ISSN:0144-557X    

DOI:10.1039/AP98825FX009

出版商:RSC    

年代:1988

数据来源: RSC

摘要:

March 1988 ANPRDI 25(3) 57-104 (1988) Analytical Proceedings Proceedings of the Analytical Division of The Royal Society of Chemistry CONTENTS 'Research and Development Topics in Analytical Chemistry' 57 Reports of Meetings 58 Summaries of Papers 58 98 Equipment News 101 SAC Silver Medal (Rules) 102 Conferences and Meetings 103 Courses 104 Analytical Division Diary ROYAL SOCIETY OF CHEMISTRY The Annual Meeting on RESEARCH AND DEVELOPMENT TOPICS IN ANALYTICAL CHEMISTRY will be held at Plymouth Polytechnic on July 18th and 19th, 1988 The 1988 Meeting will be the twenty-fifth anniversary of the initiation of this meeting. Papers and poster presentations are invited describing work carried out by postgraduate research students in universities and colleges and by young research workers in industrial and other establishments. Oral contributions are to be presented by the student or his industrial counterpart in 20-minute lectures. The number of oral contributions may be limited in order to balance the overall programme. Titles of contributions should be submitted to the Secretary of the Analytical Division at the address given below by March 6th. Brief abstracts should follow by March 31st. Those who wish to attend or who have any queries about the meeting should write to the Secretary of the Analytical Division, Royal Society of Chemistry, Burlington House, London WIV OBN. Typeset and printed by Black Bear Press Limited, Cambridge, England

ISSN:0144-557X    

DOI:10.1039/AP98825BX011

出版商:RSC    

年代:1988

数据来源: RSC

摘要:

h 57 Reports of Meetings North West Region The sixty-third Annual General Meeting of the Region was held at 6.30 p.m. on Friday, January 15th, 1988, at the College of Higher Education, Chester. The Chair was taken by the Chairman of the Region, Mr. E. R. Adlard. The following office bearers were elected for the forthcoming year: Chairman-Mr. E. R. Adlard. Vice- Chairman-Dr. C. J. Peacock. Honorary Secretary-Dr. G. Davison, 34 Beech- fields, Doctors Lane, Eccleston, Chorley , Lancashire PR7 5RE. Honorary Treasurer-Mr. B. Taylor. Honorary Assistant Secretary-Mr. E. D. Worthing- ton. Members of Committee-Mr. D. S . Corrigan, Mr. R. G. Feasey, Dr. P. Fielden (co-opted), Dr. L. A. Gifford, Mr. P. Morries (ex officio), Dr. J . Owen, Dr. P. A. Sewell and Dr. P. J. Whittle. Mr. G.W. Earnshaw and Mr. P. A. Moorcroft were re-appointed as Honor- ary Auditors. Midlands Region The thirty-third Annual General Meeting of the Region was held at 6 p.m. on Tuesday, December 8th, 1987, at the University, of Technology, Loughbor- ough. The Chair was taken by the Chair- man of the Region, Mr. J. E. W. Tillman. The following office bearers were elected for the forthcoming year: Chairman-Dr. A. Brai thwaite. Vice- Chairman-Dr . C. L. Graham. Honorary Secretary-Mr. H. E. Brookes, 35 Dunster Road, West Bridgford, Nottingham NG2 6JE. Honorary Treasurer-Mr. D. M. Peake. Honorary Assistant Secretary-Dr. T. E. Edmonds. Members of Committee-Dr. B. J . Birch, Dr. A. M. G. Macdonald, Professor J. N. Miller, Dr. R. M. Smith, Dr. J. M. Thompson and Mr. J . E.W. Tillman (ex officio). Mr. W. G. Harris and Mr. H. Pugh were re-appointed as Honorary Auditors. East Anglia Region The twentieth Annual General Meeting of the Region was held at 1.45 p.m. on Tuesday, November 17th, 1987, at Uni- lever Research, Sharnbrook. The Chair was taken by the Chairman of the Region, Mr. A. M. C. Davies. The following office bearers were elected for the forth- coming year: Chairman-Mr. A. M. C. Davies. Vice-Chairman-Mr. R. P. Mun- den. Honorary Secretary and Treasurer- Mr. P. R. Brawn, Unilever Ltd., Col- worth House, Sharnbrook, Bedfordshire MK44 1LQ. Honorary Assistant Secre- tary-Mr. P. Snowdon. Members of Com- mittee-Mr. A. Anderson, Dr. A. Brown, Dr. C. Creaser, Mr. A. Henderson, Dr. M. Kibblewhite and Dr. H. Moss. Mr. A. G. Croft and Mr.C. E. Waterhouse were re-appointed as Honorary Auditors. South East Region The thirteenth Annual General Meeting of the Region was held at 2 p.m. on Thursday, December loth, 1987, at the Scientific Societies Lecture Theatre, Sav- ile Row, London, W.l. The Chair was taken by the Chairman of the Region, Dr. J. G. Firth. The following office bearers were elected for the forthcoming year: Chairman-Mr. G. F. Phillips. Vice- Chairman-Mr . D . W. Houghton. Honorary Secretary-Dr. A. H. Andrews, Beecham Pharmaceuticals, Clarendon Road, Worthing, Sussex BN14 8QH. Honorary Treasurer-Mr. D. Blair. Honorary Assistant Secretary- Mr. P. J . O’Neil. Members of Committee -Dr. R. M. Belchamber, Mr. W. B. Chapman, Dr. J. E. Davies, Dr. J. G. Firth (ex officio), Mr. R. Goulden (co- opted) and Mr. H.I. Shalgosky (co- opted) Dr. J. E. Newbery and Mr. H. I. Shalgosky (co-opted). Dr. J . E. Page and Mr. D. C. M. Squirrel1 were re-appointed as Honorary Auditors. Northern Ireland Region The seventh Annual General Meeting of the Region was held at 5.15 p.m. on Wednesday, November l l t h , 1987, at the Queen’s University of Belfast. The Chair was taken by the Chairman of the Region, Professor D. T. Burns. The following office bearers were elected for the forth- coming year: Chairman-Professor D. T. Burns. Vice-Chairman-Mr. R. A. Hall. Honorary Secretary and Treasurer-Mr. W. J. Swindall, Department of Che- mistry, David Keir Building, Queen’s University, Belfast BT9 5AG. Members of Committee-Mr. E. L. Donaldson, Dr. M. Harriott and Mr. A. J.McCormack. Mr. V. Beavis and Mr. C. Wilson were re-appointed as Honorary Auditors. Micro and Chemical Methods Group The forty-fourth Annual General Meet- ing of the Group was held at 1.50 p.m. on Thursday, December loth, 1987, at the Scientific Societies Lecture Theatre, Sav- ile Row, London, W.1. The Chair was taken by the Chairman of the Group, Mr. C. A. Watson. The following office bear- ers were elected for the forthcoming year: Chairman-Mr. C. A. Watson. Vice- Chairman-Mr. M. J. Graham. Honor- ary Secretary-Mr. P. R. W. Baker, 55 Braemar Gardens, West Wickham, Kent BR4 OJN. Honorary Treasurer-Mr. M. R. Cottrell. Members of Committee- Mrs. D. E. Butterworth, Miss L. Dixon, Mr. M. A. Russell, Mr. W. J . Swindall, Dr. E. Vidgeon and Mr. M. J. West. Mr. S. Bance and Mr.H. I. Shalgosky were re-appointed as Honorary Auditors. Thermal Methods Group The twenty-third Annual General Meet- ing of the Group was held at 5 p.m. on Wednesday, January 6th, 1988, at the Polytechnic, Hatfield. The Chair was taken by the Chairman of the Group, Professor D. V. Nowell. The following office bearers were elected for the forth- coming year: Chairman-Mr. P. J. Haines. Vice-Chairman-Mrs. J. A. Hider. Honorary Secretary-Dr. C. J. Keattch, Industrial and Laboratory Ser- vices, P.O. Box 9, Lyme Regis, Dorset DT7 3BT. Honorary Treasurer-Dr. R. H. Still. Members of Committee-Dr. G. M. Clark (co-opted), Miss J. Griffiths, Mr. C. Jones, Dr. R. C. Mackenzie (co-opted), Professor D. V. Nowell (ex officio), Dr. J. T. Pearson, Dr. F. R. Sale, Dr. A. Smith and Dr. R. S. Whitehouse. Dr. A. Dyer and Mr. D. Griffiths were re-appointed as Honorary Auditors. Electroanalytical Group The eighteenth Annual General Meeting of the Group was held at 12.15 p.m. on Thursday, December loth, 1987, at the Scientific Societies Lecture Theatre, Sav- ile Row, London W.l. The Chair was taken by the Chairman of the Group, Professor M. L. Hitchman. The following office bearers were elected for the forth- coming year: Chairman-Professor M. L. Hitchman. Vice-Chairman-Professor A. K. Covington. Honorary Secretary-Mr . A. E. Bottom, Kent Industrial Measure- ments Ltd., Oldends Lane, Stonehouse, Gloucestershire GLlO 3TA. Honorary Treasurer-Dr. B. J. Birch. Honorary Assistant Secretary-Dr. J . P. Hart. Members of Committee-Dr. P. N. Bart- lett, Dr. A. G. Fogg (ex officio), Dr. E. A. H. Hall, Dr. D. Midgley, Dr. S. M. Slater, Mr. D. Thomason and Dr. C. M. G. van den Berg. Dr. R. M. Smith and Mr. J. E. W. Tillman were re-appointed as Honorary Auditors.

ISSN:0144-557X    

DOI:10.1039/AP9882500057

出版商:RSC    

年代:1988

数据来源: RSC

摘要:

58 ANALYTICAL PROCEEDINGS, MARCH 1988, VOL 25 Research and Development Topics in Analytical Chemistry The following are summaries of twenty of the papers and posters presented at a Meeting of the Analytical Division held on July 8th-9tht 1987, in the University of Strathclyde, Glasgow. Application of Derivative Spectroscopy to the Alizarin Fluorine Blue - Lanthanum Determination of Fluoride Marie M. Ferris, Brian Bingham and Michael A. Leonard Department of Pure and Applied Chemistry, Queen's University, Belfast BT9 5AG In dilute solution at pH 4.5, alizarin fluorine blue (1,2- dihydroxyanthraquinon-3-ylmethylamine-N,~-diacetic acid; AFB) is yellow (Amax. 423 nm), its lanthanum complex is red (Amax, 495 nm) and with fluoride ions this forms a blue ternary (AFB)2La2F2 complex (Amax, 567 nm) instead of disintegrating to form AFB and LaF3.1.2 The spectral changes which occur on adding fluoride to (AFB)2La2 are illustrated in Fig.1. This reaction has been used since 1958 as the basis of a specific absorptiometric analysis method for fluoride at about the 0.05 ri----300Fg F- I 0.420 0.315 V (0 + z n Q: 0.210 Wavelengthhm Fig. 1. Absorption spectra of AFB - La3+ solutions containing increasing amounts of fluoride. AFB = 5 x 10- 5 M; La3+ = 5 >( 10 5 M; pH = 4.6 (total acetate concentration about 0.06 M); solvent, 25% acetone; F- concentration, 0.05-0.4 p.p.m. (also 3 p.p.m.); 1 .O-m cell; slit width, 2.0 nm; scan speed, 960 nm min-1. Blank is 25% acetone, pH 4.6 to 0.4 p.p.m. level. Fig. 2 shows spectra obtained when reagent - fluoride solutions are examined against a blank containing everything except fluoride; analytical measure- ments are usually made using the peak at 616 nm.In purely aqueous solution the calibration graph shows poorish linearity close to the origin which, when linear regression analysis is applied, yields an elevated detection limit. However, between 0.32 0.24 al 0 t 0.16 z s 0.08 0 increments I 500 600 700 8 Wavelengthhm 0 Fig. 2. Absorption spectra of AFB - La3+ solutions containing increasing amounts of fluoride measured against the full blank (all comwnents except fluoride). Other conditions as for Fig. 1 Table 1. Calibration graph linear regression results: aqueous solutions A or dAldA dAldpg or d'AldA2 Intercept. Slope per 100 cm3, etc. Absorption (low blank) 2.47 x 10 2.47 x 10 3 Absorption (high blank) - 6 .8 6 X 10 ' 2.47 x 10-3 First derivative (low blank) - 1 . 3 0 ~ 10 I 7.89 x 10- 3 First derivative (high blank) -3.20 x 10-2 9.18 x 10 3 Second derivative (low blank) -2.61 x 10 ' 4.22 x 10-4 Second derivative (high blank) - 1 . 5 2 ~ 10 ' 7.05 x 10-J Correlation coefficient 0.9944 0.9944 0.9916 0.9910 0.99 10 0.9950 No. of points 9 9 6 9 8 9 S.d. of pts about line A . dAIdh. etc. 3.85 x 10--2 3.85 x 10 3 1.08 x 10 1 1.81 x 10 2 7.54 x 10 J 1.04 x 10- 3 Lowest deter- minable X (pg per 100 cm3 of F) 3.81 3.80 8.65 1.71 5.68 3.60ANALYTICAL PROCEEDINGS. MARCH 1988, VOL 25 59 Table 2. Calibration graph linear regression results: 25% acetone solutions Intercept. Slope A or dAldh dA /dug Correlation No. of or dzAldh2 per 100 cm3, etc.coefficient points 2.83 x 1 0 - 3 Absorption (low blank) 7.60 x lo--’ 6.43 x 10- 3 0.9989 8 7.39 x 10 ~ 0.9992 8 First derivative (low blank) -1.21 x 10-1 4.10 x 1 0 - 2 0.9945 7 First derivative (high blank) -1.083 X 4.66 x 10 - 2 0.9991 8 Second derivative (low blank) -1.28 X lo--’ 3.38 x 1 0 - 3 0.9970 7 Absorption (high blank) Second derivative (high blank) 1.89 x lop3 3.11 x 1 0 - 3 0.9984 8 10 and 30 pg of fluoride in 100 cm3 linearity is good. Addition of acetone to about 25% improves both near-origin linearity and sensitivity but separation between the (AFB),La2 and (AFB)2La2F2 spectra diminishes. 1’40 A 0.88 0.36 2 I$ -0.16 -0.68 - 1.20 500 40 pg F- 600 700 Wavelengthhm 800 Fig. 3. First derivative spectra of AFB - La7+ solutions containing increasing amounts of fluoride measured against the full blank.Conditions as for Fig. 2 except for a scan speed of 120 nm min I; Ak = 10 nm The availability of a Perkin-Elmer Lambda 9 ultraviolet - visible - near infrared spectrophotometer with excellent deriv- atising facilities led us to investigate whether working with derivative spectra (dAldh and d2Aldh2) would improve the fluoride analysis. The main feature of the first derivative of a Gaussian absorption band is a strong maximum followed by a strong minimum, with zero crossing coincident with A,;,, . The vertical distance between the peak and the trough is the amplitude and this is proportional to concentration provided that Beer’s Law is obeyed in zero derivative. The second derivative spectrum is inverted with respect t o the zero derivative and it features two satellite peaks flanking a strong minimum which corresponds to h,;,, .The vertical distance between the minimum and either the long or the short wavelength wing may be related to concentration. The valuable features of derivative spectra are the improvement of resolution between overlapping bands and the accentuation of S.d. of pts about line A, dAldh. etc. 4.00 x 10 3 3.91 x 10 5.06 X 10-2 2.68 x 10-2 3.13 X 10- 2.36 X 10-3 Lowest deter- minable X (pg per 100 cm3 of F) 1.72 1.47 4.55 1.59 3.50 2.08 “sharp” spectral features.7.8 Fig. 2 shows the band width of the fluoride complex spectra to be comparatively small; hence, a derivative-based improvement seemed a reasonable possi- bility. 0.060 I N x W -0.036 -0.068 0.028 0.004 I I--- 5 pg F- 15-pg increments 40 pg F- -0.100 I t Wavelengthhm Fig.4. Second derivative spectra of AFB - La3+ solutions containing increasing amounts of fluoride measured against the full blank. Conditions a\ for Fig. 3 Experimental Stock solutions ( 5 X 1 0 - 4 ~ ) of AFB and lanthanum nitrate were made up as described in references 1-6. A pH 4.6 buffer solution was prepared containing 150 g of hydrated sodium acetate and 75 cm3 of glacial acetic acid per litre. Standard fluoride solution ( 5 pg cm-3) was prepared from AnalaR or electronic grade sodium fluoride. Solutions for study contained 10 cm? of AFB, 10 cm3 of L$+, 3 cm3 of buffer solution and 0-8 cm3 of fluoride solution, made up to 100 cm3 with the lanthanum added last.A 25-cm3 volume of acetone can be included. More complete details can be found in references 1,3 and 6. Purely aqueous solutions require 1 h to reach equilib- rium, but the reaction in 25% acetone is much faster. Spectra were measured against either a blank containing buffer (and acetone) only Q’. a full blank containing everything except fluoride.60 ANALYTICAL PROCEEDINGS, MARCH 1988. VOL 25 Results Fig. 3 shows the first derivative spectra of a series of solutions in 25% acetone plotted against the full blank. There is a clear isosbestic point at 615 nm, a positive peak at 602 nm and a negative peak at 635 nm; response was taken as the ordinate difference between the 602- and 635-nm peaks. The isosbestic point corresponds well with the higher A,;,, of Fig.2. Fig. 4 shows the 25% acetone full-blank second derivative spectra; these show good negative peaks at 565 and 615 nm. Measure- ments were taken between the 542 nm positive peak and the 565 nm negative peak. Table 1 shows linear regression results for the purely aqueous system and Table 2 those for the 25% acetone system. Points at high fluoride levels which obviously deviated from the line due to lack of reagent excess were excluded. Conclusion In purely aqueous solution there is little to choose between the various spectral derivatives; the lowest determinable mass of fluoride is disappointing in all instances. In 25% acetone solution "high blank" zero and first derivative analyses give equally good high-quality results: the second derivative is slightly inferior.Derivative "low blank" analyses are distinctly inferior. It is difficult to compare sensitivities when the ordinates are different but we feel that results from conven- tional 25% acetone high blank absorbance t'ersus concentra- tion graphs are not appreciably improved by derivatisation. 1. 7 -. 3. 1. 3 . 6. 7. 8. References Leonard. M. A , . and West. T. S.. J . Clirm. Soc., 1959, 3577. Langmyhr. F. J . , Klausen. K. S., and Nouri-Nekoui. M. H.. A n d . Chirn. Acru, 1971. 57. 341. Belcher, R.. and West, T. S . . Tolarira. 1961. 8. 853. Yamamura, S. S.. Wade. M. A.. and Sikes. J. H.. Aiirrl. C h e i ~ i . . 1962. 34, 1308. Greenhalgh. R.. and Riley. J . P.. And. Chirri. Acta. 1961. 25. 179. Leonard. M. A.. in Johnson. W. C.. Ediror. "Organic Rcagents for Metals and for Certain Radicals." Volume 2.Hopkin and Williams. Chadwell Heath. 1963. p. 1 . Cahill. J . E . , Am. Luh.. 1079, 11. 79. Fell. A. F., PTOL.. A m l . Dill. Chrm. SOC.. 1978. 15. 260. Selective Fluorogenic Flow-injection Procedures for Primary, Secondary and Tertiary Amines in Non-aqueous Media 1. R. C. Whiteside and P. J. Worsfold Department of Chemistry, University of Hull, Hull HU6 7RX A. Lynes and E. H. McKerrell Shell Research Ltd., Thornton Research Centre, P.O. Box 7, Chester CHI 3SH Many compounds containing amine groups have useful industrial applications, e.g., corrosion inhibitors, surfactants 1. Primary amine acHo + R - NH2 + HS.CH2.CH2.0H CHO o-Phthalaldehyde (OPA) 2-Mercaptoethanol @ME) and biocides, and they are often formulated in oil-based matrices.Three non-aqueous flow-injection procedures with Methanol 2. Secondary amine CI + R’R - NH R ’ R acetate Ethyl @l>O+HCl ‘N’ NO, NO, 7-Chloro-4-nitrobenzo-2-oxa-l,3-diazole (N BD-CI 1 3. Tertiary amine AcO AcO c)- 1 \ Id -0Ac HBC C-OAC I 1 AcO Tertiary catalysed amine condensation b A C O O ! - O / ~ ~ - OAc Fig. 1. Fluorogenic reactions 0 Exc = 340 nm Em = 431 nm Exc = 480 nm Em = 525 nm Exc = 400 nm Em = 441 nm - ~ ~~~~~ ~ ~~ Table 1. Calibration data for hexylamine. dihexylamine and triethylamine Detection Relative standard limit deviation ( 1 1 = 5 ) . Linear range! Functional group Compound (20.)/m~ mM Prim a ry am i n e Hex y 1 a rn i n e 0.03 1 .o 0 4 . 8 Secondary amine Dihexylarnine 0.12 1.3 0-13 Tertiary amine Triethylnmine 0 .005 2.1 0-2.5ANALYTICAL PROCEEDINGS, MARCH 1988.VOL 25 OPA/2-ME in ethyl acetate 61 - 340 nm 45 "C w Waste 1. Primary amine determination 20-pi injection I U 2. Secondary arnine determination (primary amines masked) 20-11 I i n j ect i o n 0.5 ml min-1 3 100 cm, 20 cm, OPN2-ME in 45 "C ethyl acetate I 45°C NBD- CI in ethyl acetate u 3. Tertiary amine determination 20 pI injection 100 cm, 45 "C 20 cm, 45 "C Waste I 441 nm U u . " " ' " ' * soiunon Waste - Fig. 2. Flow injection manifolds fluorescence detection, selective for each of the amine functional groups (i. e. primary, secondary and tertiary) have been developed by using the reactions shown in Fig. 1. Primary amines can be determined directly by derivatisation with o-phthalaldehyde(0PA) and 2-mercaptoethanol (2-ME) to form isoindole derivatives.' Secondary amines can be determined by derivatisation with 7-chloro-4-nitrobenzo-2- oxa- 1.3-diazole (NBD-CI) after on-line masking of primary amines by reaction with OPA/2-ME.' Tertiary amines can be determined by their ability to catalyse the cyclic condensation of malonic acid and acetic anhydride to form a fully acylated phloroglucinol carboxylic acid3; primary and secondary arnines do not interfere because they form non-fluorescent N-substi- tuted and N.N-disubstituted amides.Experimental The flow-injection manifolds used for the three procedures are shown in Fig. 2. For the determination of primary amines standards made up in ethyl acetate were injected into a single line carrier stream of OPA (58 mM) and 2-ME (58 mM) in ethyl acetate. For the determination of secondary amines standards made up in ethyl acetate were injected into a stream of OPAI2-ME (58 mM) and merged 100 cm downstream with NBD-C1 (28 mM) in ethyl acetate.For the determination of tertiary amines standards made up in tetrahydrofuran were injected into a single line carrier stream of malonic acid (4% m/V) in acetic anhydride. In all instances a 100 cm reaction coil at 45°C was used and the detector was a fluorimeter (Perkin-Elmer LS-2) fitted with a 7 y1 flow cell. Results and Discussion Calibration data for hexylamine, dihexylamine and tri- ethylamine are given in Table 1, the results being typical of those obtained using low relative molecular mass aliphatic amines. The procedure for primary amines is specific, but the procedure for secondary amines is dependent on the capacity of the OPA/2-ME stream to remove primary amines prior to derivatisation with NBD-C1, apd this is affected by reagent concentration, reaction time and temperature.The procedure for tertiary amines is specific but primary and secondary amines can cause quenching of the fluorescence signals. These procedures could therefore be used to determine the total primary, secondary or tertiary amine content of a sample, but would need to be used in conjunction with HPLC for complete characterisation of a physical and chemical mixture of amines, and work is currently being done in this area. The authors thank Thornton Research Centre, Shell Research Limited, for an Extra-Mural Research Grant in support of this work.References 1. 2. 3 . Simons. S. S., and Johnson, D. F..J. Am. Chem. Soc.. 1976.98. 7098. Whiteside, I . R. C., Worsfold. P. J . . and McKerrell. E., And. Chim. Acra, in the press. Whiteside. 1. R . C., Worsfold. P. J . . and Lynes. A . . Anal. Chim. Arm, 1987. 192, 77.62 ANALYTICAL PROCEEDINGS, MARCH 1988, VOL 25 Retention Modification in HPLC Using Metal Ions Simon J. Bale and Roger M. Smith Department of Chemistry, University of Technology, Loughborough, Leicestershire LE 7 7 3TU Stephen G. Westcott and M. Martin Smith Glaxo Group Research, Ware, Hertfordshire SG 72 ODJ Metals are being used increasingly in HPLC as mobile phase additives to alter the selectivity of a separation. A number of methods have appeared in which a metal compound is incorporated in the mobile phase and is then adsorbed on to the silica surface of a reversed-phase packing.,., This alters the selectivity by changing the polar, steric and solvation proper- ties of the stationary phase.Another technique has involved the addition of chiral metal compounds, usually amino acid complexes, to the mobile phase to facilitate enantiomeric separations.3-4 A method which has seen little utilisation to date is the use of simple metal salts as mobile phase modifiers, such as zinc chloride5 and nickel acetate.6 In many instances the role played by the metal ions in these separations has not been fully established. We have recently investigated the effect of a range of metal ions on the retention of 2-aminophenol.7 This was used as a model compound, to give a more detailed understanding of the retention modifica- tion caused by simple transition metal ions in the mobile phase.It was found that all of the metals studied caused a reduction in retention, owing to the formation of a more polar species with the metal ions, and that the size of the reduction was dependent on the metal ion concentration. The pH of the mobile phase was also shown to be an important factor, and in general, increasing the pH increased the size of the effect of the metal ions. The metals studied had effect in the order Cu > Ni > Co > Zn > Cd > Mn This work has now been extended to take into account the various equilibria present in solution and a theoretical model has been developed to explain the variation in retention of 2-aminophenol with metal ions in the mobile phase.Experimental The chromatography was performed by using a Pye Unicam PU4010 pump, a Rheodyne 7125 injection valve fitted with a 20 p1 sample loop and a Pye Unicam PU4020 variable wavelength detector set at 280 nm. The retention of 2-amino- phenol (20-p1 injections, 1 0 - 3 ~ ) was studied using a porous polystyrenedivinylbenzene polymer column (PLRP-S, 100 mm x 5 mm) thermostatted at 30 “C using a water jacket. Aqueous anion (A-). The relative proportions of these species are pH dependent, and the equilibria between the three forms are governed by the pK, values of the amine and phenol groups, K,, and Ka2. In the presence of metal ions the phenolate anion forms chelates and two further species may be introduced into the equilibrium, the 1 : 1 (MA+) and 2 : 1 (MA,) chelates.The proportions of these two species are dependent on their formation constants K1 and K2. When 2-aminophenol is injected on to an HPLC column and eluted with an eluent containing metal ions the same equilibria as above are set up, and these are, in turn, in equilibrium with the stationary phase: The capacity factor ( k ’ ) for the over-all system can be defined as Amount in stationary phase Amount in mobile phase k’ = - q[AH+,]s + [AHIS + [A-IS + [MA+]\ + [MA215 [AH+,], + [AH]m + [A-lm + [MA+], + [MA21m - where q = phase ratio. Each of the five species has a capacity factor: k’b for the protonated amine, k ’ , the neutral molecule, k ’ , the phenolate anion, k ’ c , the 1 : 1 chelate and k‘c2 the 2 : 1 chelate.If substitutions are made for the equilibrium concentrations and stationary phase terms, the following equation is obtained: mobile phases (10, 20 and 30% methanol), containing 0.26 M ammonium acetate and low levels of transition metal acetates. were used at a flow-rate of 1 ml min-1. The pH of the mobile phase (6, 7.24 or 8) was adjusted by use of concentrated hydrochloric acid or sodium hydroxide solution. Discussion In aqueous solution, and in the absence of metal ions, 2-aminophenol is present in three forms: the protonated amino cation (AH,+), the neutral molecule (AH) and the phenolate This equation contains a term for 2 : 1 chelate formation involving the free ligand concentration [A-1. This suggests that the capacity factor of 2-aminophenol, and in general the analyte,, is concentration dependent.However, because the metal ions present in the mobile phase are in vast excess (between 0.02 and 0.4 M compared with 0.000 1 M analyte at pH 7.24), and 2 : 1 chelate formation depends overall on the free ligand concentration [A-12. only very small amounts of the 2 : 1 chelate are formed with the injection concentrations used in HPLC. Chelate formation tends to the formation of MA+ asANALYTICAL PROCEEDINGS, MARCH 1988, VOL 25 63 the metal ion concentration increases and the equation for capacity factor can be reduced to At very low concentrations, when the ligand is in excess, competition with protons dominates the expression and insufficient ligand becomes available for 2 : 1 chelate forma- tion.In the absence of metal ions this equation reduces still further to give an equation describing the variation of the capacity factor of an ampholyte with pH; this form contains the two expressions derived by Miyakex et al. for the capacity factor of an acidic and a basic compound. The variation of capacity factor with metal ion concentration at constant pH was tested using the expression a + b[M2+] 1 + c[M2+] k’ = where a , b and c are constants related to those above. If the pK, values of the groups are known then all the other capacity factor terms and formation constants can be determined from a, b and c. 121 I 0.00 0.05 0.10 0.15 0.20 0.25 0.30 0.35 0.40 Nickel ion concentrationh Fig. 1. Graphs showing the correlation for the effect of the presence of nickel ions on the capacity factors of 2-aminophenol, between the experimental values at pH 6 (a), pH 7.24 (B) and pH 8 (O), and the corresponding calculated values depicted by the solid lines This expression was fitted to experimental data by using a non-linear least squares computer program (Genstat 4.03).High coefficients of correlation have been obtained and the same constants have been obtained under a variety of experimental conditions. The formation constant of the 1 : 1 chelate of 2-aminophenol with nickel ions was determined by using the equations derived above with experimental data obtained at pH 6, pH 7.24 and pH 8 (Fig. 1). The formation constants found in this way were logloK1 4.2 (pH 6), 4.67 (pH 7.24) and 4.85 (pH 8). These constants were calculated by using a pK, value of 10.43 for the phenol group, which had been previously determined by spectroscopy, and by HPLC using the relationship between capacity factor and pH.While very good correlations have been obtained for fitted and experimental data the accuracy of the calculated formation constants requires an accurate determination of k ’ c l , the capacity factor of the 1 : 1 chelate. This often requires the use of very high concentrations of metal ions in the mobile phase, more than 0.5h1, which poses problems for the detection background, and adjustment of the pH. Conclusion The addition of metal ions in the mobile phase is a technique which may be used to alter the selectivity of a separation. Analytes that are capable of chelation can be made to form polar species with metal ions and their retention times are reduced.The influence that a given metal may have is dependent on the magnitude of its chelate formation constant with the analyte, its concentration and the mobile phase pH. A model has been proposed for the behaviour of 2-amino- phenol in mobile phases containing metal ions which accurately describes experimental measurements. This model may be useful in predicting the effect of metal ions in the mobile phase on the chromatography of other chelating analytes. References 1. 2. 3. 4. 5 . 6. 7. 8. Lochmuller, C. H.. and Hangac, H . H., J . Chromatogr. Sci.. 1982, 20, 171. Cooke, N. H. C., Viviattene, R . L. R., Eksteen, R., Wong, W. S . . Davies, G., and Karger, B. L., J . Chromatogr., 1978, 149, 391.Le Page, J . N . , Lindner, W.. Davies, G . , Seitz, D. E.. and Karger, B. L., Anal. Chem., 1979, 51. 433. Gilon, C., Leshem, R., Tapuli, Y., and Grushka, E., J . Am. Chem. Soc., 1979, 101, 7612. Walters, V., and Raghaven, N . V., J . Chromarogr.. 1979, 176, 470. Sternson, L. A., Dixit. A . S . , Riley, C. M.. Siegler. R. W . , and Schoech, D . , J. Pharm. Biomed. Anal., 1983, 1, 105. Smith, R. M., Bale, S . J . , Westcott, S. G., and Martin-Smith. M., Analyst, 1987, 112, 1209. Miyake, K., Okrumura, K., and Terada. H., Chem. Phurm. Bull., 1985, 33. 769. HPLC Determination of Trace Metals in Industrial and Environmental Samples Mary Meaney, Joseph Mooney, Michelle Connor,and Malcolm R. Smyth School of Chemical Sciences, NlHE Dublin, Glasnevin, Dublin 9, Ireland The determination of trace metals using HPLC-based technol- ogy has received much attention in recent years.1-h The main reasons for this stem from the possibilities firstly of having a multi-element technique using equipment that is now com- monly available in most laboratories and which need not -be dedicated solely to this application, and secondly, of obtaining information on the "speciation" of a certain element in a certain matrix.To date, we have investigated the application of reversed phase HPLC to the determination of trace concentrations of copper(I1) and iron(II1) in anaerobic adhesives,' and the levels of iron(II1) and aluminium(II1) in soil and clay samples following a variety of sample extraction procedures.x The determination of trace metal concentrations by HPLC is based on the assumed quantitative formation of a stable metal chelate in situ.The metal ion of interest can be introduced on to64 ANALYTICAL PROCEEDINGS, MARCH 1988, VOL 25 the column either by “direct injection” of the metal ion in its ionic form, or by injection of a solution of the metal chelate formed “externally” using the same ligand that is employed in the chromatographic eluent. By using the “direct injection” technique, limits of detection in the low p.p.m. range can usually be achieved. Much lower limits of detection (of the order of 50 p.p.b.) can, however, be achieved following external formation of the metal chelate, assuming that the background metal ion concentration is of the order of 10 p.p.b. This is illustrated for the case of iron(I1I) in Fig.1. In the optimisation of an HPLC method for the determina- tion of a certain metal ion, the following parameters are of importance. 1. Ligand concentration. The optimum ligand concentration should be investigated bearing in mind the range of metal ion concentrations to be determined, and should be the concentra- tion which gives rise to the largest and most stable absorbance reading (relating to the chelate) in the shortest time possible. 2. p H . The acid - base equilibria of the ligand should be investigated in order to optimise conditions for chelate formation: again, the absorbance readings obtained for the chelate under different conditions of pH should reflect the largest and most stable readings in unit time. 0.002 Absorbance I 0 5.0 10.0 Tirneh in Fig.1. Comparison of HPLC chromatograms for the determination of iron(II1) following ( a ) “in situ” formation of iron(II1) - oxine complex, and ( h ) external formation of the iron(II1) - oxine complex 3. Solvent composition. The ratio of aqueous to organic solvent should be varied to investigate the stability of the chelate under the conditions that are generally employed for HPLC separations, e.g., the copper(I1) - oxine chelate has been found to be unstable at aqueous buffer: acetonitrile ratios of greater than 50 : 50, and hence these ratios cannot be used for HPLC separations involving this metal chelate .7 4. Salt concentration. In some instances it has been reported that metal chelates can be stabilised by the addition of a salt to the mobile phase, e.g., the addition of potassium nitrate to the mobile phase was advocated by Bond and Nagaosab for stabilisation of the iron(II1) - oxine chelate.5. Detection wavelength. The detection wavelength should be optimised with respect to the absorption maximum of the chelate and the background absorbance due to the mobile phase of choice. 6. Column selection. It has been our experience that the best separations of metal chelates is achieved on columns which have been “end-capped” to prevent free SiOH groups on the stationary phase interfering in the separation. The number of metal ions which can be determined simultaneously by using this approach can be of the order of five or six, but this depends to a large extent on the choice of ligand, the operating conditions and the matrix involved.The ligands investigated to date include dithiocarbamates, P-dike- tonates, Schiff bases, dithizone, crown ethers, etc. In our studies, we have concentrated on the use of oxine as a ligand because of: firstly, the separation that one can achieve, especially with respect to copper(I1) - and iron(II1) - oxinates; and secondly, the ability to use this compound as an extractant of metal ions from complex matrices such as soils and clays. The HPLC approach to trace metal analysis can therefore be useful in situations where a large number of metal ions (particularly transition metal ions) need to be determined in a short space of time or as a complementary method to atomic absorption spectroscopy (AAS) or anodic stripping voltam- metry (ASV) when the results of two independent methods are required.The information that one obtains from HPLC determina- tions of this kind does not necessarily correlate with that obtained from AAS, however, as this approach is primarily geared to looking at the “free” or “available” metal ion concentration rather than the “total” metal ion concentration. This has been demonstrated in our studies both on the anaerobic adhesives7 and on the soil and clay samples.8 In the former study, anaerobic adhesive formulations were investi- gated which contained 300-400 p.p.m. of EDTA. In these formulations, however, EDTA is only present in the “soluble” form to the extent of approximately 20 p.p.m. and the HPLC method developed could only detect concentrations of cop- per(I1) and iron(II1) spiked into the formulations when this concentration limit was exceeded.In the instances of soil and clay samples discrepancies have been noticed in the concentra- tions of aluminium(II1) extracted under various conditions and determined separately by HPLC and AAS. The AAS results were consistently higher than those obtained by using HPLC and the difference in these results would appear to be due to “bound” aluminium(II1) species co-extracted with the “free” portion. This ability of HPLC to speciate between “free” and “bound” metal ions in environmental and industrial samples. coupled perhaps with electrochemical detection, should prove of major interest in the near future. References 1 . 2. 3 . 4. 5 . 6. 7. 8. Bond, A. M., and Wallace. G. G., And.Chem.. 1982.54. 1706. Smith, R. M.. Butt, A. M.. and Thakur, A.. Anulwt. 1985, 110. 35. Inatimi. E.. J . Chromatogr.. 1983, 256. 256. Guira, R. C., and Carr. P. W., J. Chromutogr. Sci.. 1982.20. 10. O’Laughlin, J . W.. and Hanson. R. S . , Anal. Chem., 1982,54, 178. Bond, A. M., and Nagaosa. Y., A~zal. Chim. Acta. 1985. 178. 197. Mooney, J . . Meaney, M., Leonard. R., Wallace. G. G.. and Smyth, M. R . , Analyst. 1987, 112. 1555. Meaney, M.. Connor. M.. Breen, C.. and Smyth. M. R.. in preparation.ANALYTICAL PROCEEDINGS, MARCH 1988, VOL 25 65 Ion Exchange Resin and Soil Solution Measurements of Soil Potassium and its Uptake by Ryegrass and Clover G. D. Wimaladasa" and A. H. Sinclairt Department of Soil Fertility, The Macaulay Land Use Research Institute, Craigiebuckler, Aberdeen AB9 2QJ There is no universal agreement on the best methods for extracting soil solution or non-exchangeable soil potassium.A centrifugal method for extracting soil solution for the determi- nation of potassium and a mixed ammonium chloride - ion exchange resin procedure for extracting both exchangeable and non-exchangeable potassium are described. Introduction Soil potassium occurs notionally in four different forms that are in equilibrium with each other, but differ in their availability to plants. Mineral K= Non-exchangeable K S Exchangeable K& Soil Solution K Potassium release to plants in a single growing season is dominated by equilibrium I, whereas I1 and I11 affect the long-term availability. 1 Plants obtain their nutrients largely, if not exclusively, by uptake from solution in soil.2 However, little attention has been paid to the seasonal variations in K concentration in the soil solution and its relationship with the soil solid phases.There are various reasons for this. Soil solution has been regarded as difficult to obtain in a form likely to relate closely to that present in undisturbed soil. Numerous methods have been proposed either to isolate soil solution, such as liquid displacement and lysimeters, or to estimate potassium in soil solution, such as electro-ultrafiltration (EUF) and quantityhntensity measurements, but none has proved entirely satisfactory.*-4 In this paper we describe a modified centrifugal method of isolating soil solution and the behaviour of K+ in soil solution within the rooting zone of ryegrass in the field.An assessment of the rate of replenishment of potassium in the soil solution from the solid phase is required in order to understand soil - plant - potassium interactions. Numerous methods have been used to measure the concurrent release of non-exchangeable potassium together with exchangeable potassium during the growing season.5-7 In this paper a novel method involving NH,+-ion exchange resin has been used to study the chemistry of soil potassium dynamics and the results have been compared with standard methods. 111 Materials and Methods Soil Solution Soil solutions were isolated by centrifugation of 300 g of field moist soil at 3000 x g for 60 min in a two-component cell (diameter 9 cm and total height 10 cm) separated by a perforated disc.8 Whatman No.42 filter-paper was placed on the perforated disc and disposable polythene bags were used to reduce the possible contamination of the soil solution. The solution collected in the lower part of the cell was separated and re-filtered through a Whatman No. 42 filter-paper to avoid contamination due to fine clay material. Potassium concentra- tions were determined by flame photometry. When the soil samples were too dry to allow collection of enough solution for chemical analyses, samples were re- ~~ ~~~~ ~ ~~ ~ ~ ~ ~ ~ * Home Address: Tea Research Institute of Sri Lanka. Talawakelle. Sri Lanka. + Present Address: The North of Scotland College of Agriculture. Aberdeen. moistened by adding de-ionised water up to 90% of the field capacity.Re-moistened soil samples were stored in the cold room for 3-4 d and thoroughly mixed. Forty-eight soil samples can be centrifuged per day provided that the centrifuge has a six-component head with appropriate dimensions equivalent to the cell. Mixed Ammonium-Cation Exchange Resin and Chloride-Anion Exchange Resin Preparation Four grams of Duolite 225 H resin (BDH, particle size of 0.50-1.18 mm) and 4.0 g of Amberlite C1-resin (BDH, IRA-400, particle size of 0.30-1.18 mm) were introduced into a nylon netting bag (3.5 x 6.5 cm) with a mesh size of 0.2 mm. The oblong and tetrahedron shaped bags were sealed by means of an electric sealer.9 These mixed-resin bags were equilibrated with excess of 1 M ammonium chloride and 0.0 1 M hydrochloric acid (1 : 1) solution in an orbital shaker.After the equilibration the pH of the solution was measured and a fresh solution of 1 M NH&l and 0.01 M HCl was introduced. This procedure was repeated until the pH was constant. The excess ions were washed free in the same vessel with de-ionised water until the chloride ion concentration dropped to a minimum. The final resin material contained NH4+ and C1- ions. Resin bags were stored under de-ionised water in order to avoid breakdown of the resin beads. July August 0 June Fig. 1. Potassium concentrations (p.p.m.) in soil solutions from treatment N,K,, where "C" is the time of grass cutting and "F" is the time of fertiliser addition Soil Potassium Extraction With Ammonium Chloride - Ion Exchange Resin Air-dried soil, equivalent to 5.0 g of oven-dried soil, which had been passed through a 2-mm sieve, was equilibrated with 100 ml of de-ionised water and a mixed-resin bag in a wide mouth polypropylene bottle for 16 h (overnight) held in an end-over- end shaker.After the equilibration, the ion exchange resin bags were removed from the soil - water suspension and washed thoroughly with de-ionised water. The resin bags were eluted with 100 ml of 1 M NH&l and 0.01 M HCl (1 + 1) solution for 30 min and the potassium concentration was measured by flame photometry.66 ANALYTICAL PROCEEDINGS, MARCH 1988, VOL 25 Table 1. Correlation coefficients of linear regressions of exchangeable and resin extractable soil potassium on cumulative potassium uptake and quadratic regressions on yield of ryegrass at different cropping stages K uptake at cut no.Yield at cut no. K measurement 1 1 + 2 1 + 2 + 3 1 1 + 2 1 + 2 + 3 1 M NH4 OAc (KJ 0.928 0.981 0.981 0.682 0.869 0.934 MixedNH,CI - ion exchange resin (K,) 0.943 0.982 0.980 0.642 0.845 0.917 (all P <0.001) Nitrification of NH4-Ion Exchange Resin The possibility of nitrification of NH4-ion exchange resin material during the 16-h equilibration with soil - water suspen- sion was studied by using “Nitrapyrin” [2-chloro-6-(trichloro- methy1)pyridinel as a nitrification inhibitor. 10 Three soils of different pH values (4.0, 4.8 and 5.5) were equilibrated with 4.0 g of the ammonium-cation exchange resin and 100 ml of de-ionised water for 16 h with and without the presence of 15 vg ml-1 “Nitrapyrin.” A control set of soils were equilib- rated without resin bags.After the equilibration, the soil suspensions were filtered through Whatman No. 42 filter- paper. The nitrate concentration in the filtrate was measured colorimetrically after reduction to nitrite with hydrazine (Industrial Method No: 333-86E, Technicon). Results and Discussion Soil Solution Potassium A ryegrass and clover field experiment (Countesswells soil series)” with and without fertiliser potassium was selected to monitor the concentration of potassium in soil solutions within the rooting zone soil. Soil solution potassium levels were in the range of 0.2g1.26 p.p.m. in the plots with zero potassium applied and 0.50-4.38 p.p.m. in the plots receiving 70 kg ha-1 cut-1 of potassium as potassium chloride.The nitrogen addition to all plots was 110 kg ha-1 cut-1 as ammonium nitrate. Soil samples were collected at 2-week intervals except immediately after fertiliser application, when sampling was at 2-d intervals. Fig. 1 gives the variation of soil solution potassium levels. The letter “C” indicates when the sward was cut and “F” indicates the timing of ammonium nitrate application. The increase in soil solution potassium concentration following the addition of nitrogen fertiliser alone (N2K,,) is attributed to displacement of K+ from exchange sites by NH4+ from the fertiliser. This finding has implications in fertiliser use. Mixed Ammonium Chloride - Ion Exchange Resin Potassium Potassium was extracted from 41 soil samples with the mixed ammonium chloride - ion exchange resin (K,) and the results compared with yield and crop potassium uptake from these soils by ryegrass grown in pots, and with potassium extracted by the conventional method using I M ammonium acetate solution at pH 7.0 (K,,).3 The K, results in Table 1 show similar linear correlations between soil test value and K uptake by ryegrass compared with conventional Kc,.An additional advantage of the mixed ammonium chloride - ion exchange resin is its ability to extract available soil phosphorus. 12 Loosely held, non-exchangeable K, Kint, is not extracted by 1 M ammonium acetate solution, but becomes available to ryegrass. A close correlation between Kint and K,, has been shown,’ which helps to explain the observed close correlations between K,, and ryegrass data.However, the mixed ammonium chloride - ion exchange resin does extract part of the non-exchangeable K. This ability to extract non-exchangeable K is demonstrated by the relationship K, = 0.674 K,, + 0.935, r = 0.986 (P <0.001) The equation shows that the greatest proportion of non- exchangeable K, compared with exchangeable K, is extracted by the ion exchange resin from soils of low exchangeable K. The rate of release of this non-exchangeable K is being investigated by means of sequential ion exchange resin extractions. A quadratic relationship was found between the yield of ryegrass and soil K values (Table 1). Nitrification of Ammonium-Ion Exchange Resin The nitrate concentrations in Table 2 indicate that nitrification of ammonium-ion exchange resin was not significant during the 16-h period.Table 2. Nitrate-nitrogen concentrations (p.p.m.) in aqueous extracts of soil with and without the presence of ammonium-ion exchange resin and nitrapyrin Soil pH Extractant 4.0 4.8 5.5 Soil - water 0.52 0.91 2.25 Soil - water - NH,+-ion exchange Soil - water - NH4+-ion exchange resin 0.55 0.86 1.60 0.54 0.72 1.53 resin - nitrapyrin Standard error of means = k0.047 Conclusion The proposed centrifugal and mixed ammonium chloride - ion exchange resin techniques have proved useful for measuring the K concentration in soil solution and its replenishment from exchangeable and non-exchangeable sources as part of a wider study of soil - plant - potassium dynamics. We wish to thank Dr. D . Atkinson and members of the Departments of Soil Fertility, Spectrochemistry, Statistics and Technical Services for their assistance.G. D. Wimaladasa wishes to thank the Director, Tea Research Institute of Sri Lanka, for financial assistance and study leave. 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. References Sinclair. A. H., J . Soil Sci.. 1979. 30. 757. Adams, F.. Burmester, C . , Hue. N . V . . and Long. F. L.. Soi2 Sci. SOC. Am. J., 1980, 44, 733. Sinclair, A. H., Plunt Soil, 1982, 64, 85. Ghorayshi, M., and Lotse. E. G., Swedish J . Agric. Res.. 1986. 16. 143. Sinclair, A. H . , J. Soil Sci., 1979. 30, 775. Salomon, M., and Smith. J . B., Soil Sci. SOC. Proc.. 1957. 21, 222. Arnold, P. W . , Nature, 1958, 182. 1594. Linehan, D. J . , Sinclair. A. H.. and Mitchell. M. C.. Planr Soil. 1985, 86, 147. Sibbesson, E..Plant Soil. 1977. 46. 665. Powell. S. J . . and Prosser. J . I.. A p p . Environ. Microbiology. 1986. 52. 782.ANALYTICAL PROCEEDINGS. MARCH 1988. VOL 25 67 11. Glentworth, R., and Muir, J. W . , “The Soils of the Country Around Aberdeen, Inverurie and Fraserburgh. Memoirs of the Soil Survey of Scotland. Edinburgh,” HM Stationery Office. Edinburgh, 1963. 12. “The Macaulay Institute for Soil Research, Annual Report,” Volume 56, Macaulay Institute for Soil Research, Aberdeen, 1986. p. 183. On-line Pre-concentration wit Injection - Inductively Couplec In c 3 - >. In c al c C 0 - + s . t,- .- c - .- .- I a Boron-specific Resin in Flow i Plasma Emission Spectrometry D. R. Anderson and C. W. McLeod Department of Metals and Materials Engineering, ShefYield City Polytechnic, Sheffield S I 1 WB Recent studies in our laboratories have demonstrated the usefulness of activated alumina for on-line trace enrichment of a wide range of cations and anions, including phosphate, chromate and sulphate.1 This approach, however, was rela- tively unsuccessful for borate, on account of poor analyte retention characteristics. Given the need for improving current methodology for the determination of boron in complex samples, an alternative column packing was sought. Thus, a boron-specific resin was utilised in the present study. Experimental The flow injection manifold and ICP instrumentation have been described elsewhere.2 The microcolumn (1.5 x SO mm) was packed with Amberlite XA743 (Rohm and Haas) of particle size range 180-420 pm.Experimental parameters were: carrier stream flow-rate (distilled water), 0.75 ml min-1; sample volume, 250 pl; eluent, nitric acid (1 M, 250 pl). Plasma operating conditions were: forward power, 1.1 kW; coolant argon flow, 17 1 min-1; nebuliser argon flow, 0.7 1 min-1; analytical wavelength, 249.77 nm. Standard solutions were prepared from BDH chemicals (Spectrosol). A synthetic steel sample was prepared (iron 1000 pg ml-1, boron 1 pg ml-1, pH 3.0) and EDTA added to prevent iron precipitation. In the procedure, boron was deposited on the column either by injection (250 pl), or by pumping the sample for specified time periods, e . g . , 2 min. Elution was subsequently effected by injection of nitric acid (250 yl); the signal was measured either as an emission - time profile or an integrated peak area.Results and Discussion A typical emission time response for the deposition - elution sequence is shown in Fig. 1. For the experimental parameters given, and for column lengths greater than 5 cm, it was Eluent I I I B I - I I 1 0 50 100 150 1 Time/s 3 Fig. 1. The deposition - elution sequence: sample boron, 1 pg ml - I established that more than 90% of analyte was deposited for samples of pH 3-10, and that nitric acid (1 M , 250 pl) was an effective eluent. For column lengths less than 5 cm, deposition efficiency was impaired. A scheme for the deposition - elution process, analogous to the boric acid - mannitol reaction, is: Resin - borate complex At pH values above 3 and in the presence of the resin, complexation takes place and the equilibrium is shifted to the right.Upon addition of acid, the boron species reverts to boric acid. A calibration graph, based on the injection of samples, was established over the concentration range from 10 yg 1-1 to 100 pg ml- 1 , and this revealed a significant breakthrough of analyte at concentrations above 20 pg ml-1. The breakthrough capacity for boron at this point was estimated to be 23 pg per gram of resin. The use of larger sample injection volumes gave enhanced sensitivity. The relative standard deviation for seven replicates at 100 yg 1-' was 1 .S%. 0 Tirne/s Fig. 2. iron, 1000 yg ml-1 The deposition - elution sequence: sample boron. 1 yg ml- The matrix removal capability of the flow injection manifold was investigated with respect to iron, because it is well known that the determination of boron in steels by ICP-AES is subject to spectral interference.3 A synthetic steel sample (iron 1000 pg ml-1, boron 1 pg ml-1) was injected (250 pl) into the manifold and the response is shown in Fig.2. This reveals the time resolved emission responses for iron and boron. Work is currently in progress on the determination of boron in complex matrices.68 ANALYTICAL PROCEEDINGS. MARCH 1988. VOL 25 References 1. McLeod. C. W.. J . At. Spectrom.. 1987. 2 . 539 2. 3 . McLeod. C. W., Cook, I. G.. Worsfold. P. J.. Davies, J . E.. and Queay, J . , Spectrochim. Actrr, Purl B. 1985. 40. 27. Wallace, G . F.. AI. Spectrost.. 1981. 2 . 61. Laser Ablation for the Mobilisation of Refractory Analytical Atomic Spectroscopy Simon Chenery Applied Geochemistry Research Group, Department of Geology, Imperial London S W7 2BP Michael Thompson Materials College, South in Kensing ton, Department of Chemistry, Birkbeck College, Gordon House, 29 Gordon Square, London WC1 6BT Katherine Tim m i n s DQA/TS, The Royal Arsenal East, Woolwich, London SE 18 6TD The use of laser ablation for solid sample introduction has been demonstrated by several groups’-J as an alternative to conven- tional mobilisation methods such as aqueous nebulisation.The performance of the method was satisfactory for certain specific tasks. but fell short of the standard needed for general analysis. It seemed clear that the improvement of laser ablation systems required a study of the fundamental processes involved. The study concentrated on three areas: the ablation process, the form of the ablated material and the transport of that material.Experimental The laser used was a ruby-rod type delivering approximately 1 J of energy in both normal and Q-switched mode, which enabled us to control both the number and the peak power of the laser pulses. The laser light was focused on the sample surface in the ablation chamber by using a Zeiss binocular microscope of x 200 or x 500 magnification. The glass ablation chamber had a volume of 18 cm3. The input and output tubes of the ablation chamber were tangential and there was a window made of a glass microscope cover slip for laser light entry. This is currently the best design devised so far, in terms of sample transport ability and the prevention of “fogging” of the window.The sample was carried from the chamber by a stream of argon through 5 mm i.d. polythene tubing t o an impinger. This device was used to remove material greater than 8 pm in size that might otherwise have dropped out on the way to the ICP torch. A PTFE change-over valve was incorporated into the gas line t o allow a change of sample without the entrainment of air. The ICP-AES spectrometer was used with the following conditions: coolant flow, 12 I min-1; auxilary flow, 0.3 1 min-1; injector (chamber carrier) flow. 0.5 1 min-1, all argon; power. 1.2 kW at 27 MHz; viewing window, a 4-mm square centered at 14 mm above the load coil. The laser was synchronised with the spectrometer electron- ics by an optical switch.The laser ablation events were observed by both still photography and on high-speed cine film. The single shot photographs were taken with a 140 mm lens. 160 ASA colour film and 0 . 5 s exposure, the exposure starting before and finishing after the event. These were taken both in air and argon. The cine film was shot in colour at a speed of 1000 frames s-1. The filtration of the aerosol of the ablated material was conducted at various points along the system by means of a 0.4 or 0.8 pm Nucleopore filter in a polythene holder. The form of the ablated material was studied. after coating the filters with gold, by secondary and backscattered electron microscopy. The composition of ablated material was determined using the energy dispersive X-ray analysis system and the size distribu- tions were made automatically on the same system.The mass loss on ablation was obtained from a piece of copper (weighed t o 1 X lo-’ g) before and after the event. The mass of ablated material collected on the filters at given points in the gas line was determined spectrochemically after dissolu- tion of the copper on the filter in nitric acid. Results Two processes can be observed when mild steel is ablated by a lightly Q-switched laser in air, a plume of incandescent material and luminous particles flying off at high velocity. If the same material is ablated in an argon atmosphere then the plume is still visible but the particles are not. This suggests that the plume is a plasma and that the particles are burning. Under the same conditions both ceramic materials and zirconium also show plumes.Zirconium gave particularly luminous burning particles in air, and was chosen for the cine photography at a speed of 1000 frames s-1. The cine film was shot with the same conditions as the still photography. The film showed in the first frame (l/lOOO s) the production of the plasma plume. The second frame showed the production of particles with a much reduced plasma plume. The third frame showed no plasma plume and no further particles being produced. Subsequent shots o n l y showed particles in flight. A secondary electron micrograph (Fig. 1 ) of a cross-section through an ablation crater produced by a lightly Q-switched laser shows a tear-drop shaped crater with a smooth neck and overlapping material. once apparently molten on the surface of the sample. The bottom of the crater is rough with pitting.thus suggesting a different ablation process. A further secondary electron micrograph (Fig. 2) shows the form o f typical ablated material (stainless steel) that reaches the plasma. Most of the particles collected were spherical and Fig. 1. sample. Depth of crater. 580 !(rn Cross section through laser ablation crater i n a stainlesh-steelANALYTICAL PROCEEDINGS. MARC€{ 1988. VOL 25 69 almost smooth. with a small amount of amorphous material on the surface. This amorphous material can also be seen on the filter in the background. The spherical particles had evidently been formed directly from the liquid. The amorphous material is thought to be condensate from the plasma plume.The particle size distribution (Fig. 3) of this spherical material shows that a majority of particles are between 0 and 2 pm, but with a tail up to 8 pm. However. the mass distribution shows the 4-6 ym sizes to be most important in terms of instrument response. Fig. 2. Diameter, 3.7 pin A sphere of stainless steel produced by laser ablation. An inspection of material collected in the impinger showed spheres of assorted sizes greater than 1 ym in groups of up t o 20. welded together by a meniscus. Thus. the spheres must have been molten when they came into contact or were separating out from a larger droplet. When materials such as ceramic that are more crystalline or brittle are ablated. the finer material leaving the torch is similarly spherical. but larger angular fragments that do not appear to ha1.e been melted can be collected from the impinger. Mass loss experiments with copper showed that at maximum laser power approximately 20 pg were ablated at each shot. €lowever. only 30% of this leaves the chamber (the rest appears as quenched molten material on the chamber walls) and only 5% passes through the torch. Of the material leaving the torch. it is estimated that only 1% of that mass is of a size less than (1.8 pm. Conclusions It seems clear that two separate processes contribute to the over-all ablation event and to the material that is eventually analysed. The plasma plume is formed initially from material directly volatilised from the sample surface. This gives rise to the intense emission above the ablation site and to the fine amorphous material that is condensed from the vapour phase. Subsequently, a pool of liquid is formed by interaction of the solid sample with the plasma plume. and this is explosively ejected from the crater to form the spherical particles. This double process has relevance to the question of selective volatilisation, even though the spherical matter predominates in the material that is analysed. - 50 - ," 40 - c + 1 2 3 4 5 6 7 8 9 3 DiameterlKm Fig. 3. Particle size distribution and mass distribution of material from the ablation of a stainless-steel sample. The solid blocks represent the number of particles and the dotted blocks the relative mass Only 30% of the mobilised material leaves the ablation chamber and only a fraction of this reaches the torch for analysis. There is clearly scope for improving the efficiency of the transport process, and possibly some benefit to be gained by separating the condensate from the spherical particles for selective analysis. References 1. Ahercrombie. F. N., Silvester. M. D.. blurry. A. D., and I3arringer. A. R.. in Barnes, R. M.. Editor. "Applications of I n d uc t i ve 1 y Coup led PI asm as to Emission Spectrometry , " Franklin Institute Press. Philadelphia. 1978. p. 121. Ishizuka. I-.. and Uwarnino. Y.. Anuf. Chem.. 1980. 52, 125. Thompson. M., Goulter. J . E.. and Sieper. F.. Anu!)'Sf. 1981, 106. 31. Carr. J . W . . and €lorlick. G . , Spectrochirn. Acta. Purl R , 1982, 37. 1. 2. 3 . 1.

ISSN:0144-557X    

DOI:10.1039/AP9882500058

出版商:RSC    

年代:1988

数据来源: RSC

摘要:

ANALYTICAL PROCEEDINGS. MARC€{ 1988. VOL 25 69 Coal Analysis by Analytical Atomic Spectrometry (ICP-AES and ICP-MS) Without Sample Dissolution Huw G. M. Parry and Les Ebdon Department of Environmental Sciences, Plymouth Polytechnic, Drake Circus, Plymouth, Devon PL4 8AA As coal will be the major source of energy for the forseeable future. there will be an increasing demand f o r trace element information on coal because of concerns such as quality control , corrosion, contamination. cat a14 s t poisoning and environmental pollution. It has therefore become essential to develop accurate and reliable technique\. which can determine the amount of these elements present i n coal. Present techniques require coal to be in solution. Not only is this dissolution tedious and difficult. but it is alw prone to analyte loss.contamination, incomplete dissolution and hazards. r.g. , the use of perchloric and hydrofluoric acids. Therefore. the direct analysis of solid samples of coal would be advantageous. Several methods for solid sample analysis have been described in the literature.' 5 but were considered less than ideal because of matrix effects o r problems encountered in calibration. A method of solid sampling which eliminates these problems is slurry atoniisation. which involves sample intro- duction as aqueous suspensions of finely powdered material.70 ANALYTICAL PROCEEDINGS, MARCH 1988, VOL 25 The similarity of this approach to aqueous solution analysis enables the use of conventional instrumentation with only minor modifications and calibration using aqueous stan- dards."' In the work reported here techniques have been established for the determination of major, minor and trace elements in coal by the introduction of slurries into an inductively coupled plasma (ICP) for both atomic emission and mass spectrometry.Experimental Instrumentation For atomic emission spectrometry (AES), all measurements were performed using two commercially available rapid sequential instruments. In the first instrument (Plasmakon S-35, Kontron Spectroanalytik, Eching, Germany), power for the plasma was supplied from an rf generator with a maximum forward power of 3.5 kW at 27.12 MHz. A torch (H. Baumbach, Ipswich, Suffolk) of the Greenfield design, with a 3-mm bore demountable injector, was used. Samples were introduced via a double-pass Scott-type spray chamber using an unblockable V-groove nebuliser made from Kel-F (PS Analytical, Orpington, Kent).In the second instrument (ARL 3520, ARL, Luton, Bedfordshire), power was supplied by a 2.5 kW maximum, 27.12 MHz rf generator of the closed loop design. The torch was of the Fassel type and either a fixed 1.2 mm bore injector or a two-piece torch with interchangeable injectors up to 3 mm bore were used. The spoiler, in the spray chamber supplied with the instrument, was removed to allow an undisturbed flow of sample. The unblockable nebuliser referred to above was used to aspirate the slurries and calibration standards. Table 1. Slurry atomisation - inductively coupled plasma atomic emission spectrometry results for certified reference coal SARM 18 Al*O3 Fe203 CaO KzO Ti02 Cr Mn c o Sr V Zn Ni c u Slurry Certificate atomisation*, value, o/o mlrn O/O mlrn 2.440 k 0.008 2.57 0.248 k 0.005 0.29 0.145 k 0.003 0.18 0.104 f 0.005 0.145 0.090 t 0.002 0,114 w g - ' CIgg-' 13.3 k 0.25 16 19.3 _t 0.13 22 5.7 k 0.1 6.7 41.7 2 0.4 44 17.8 f 0.3 23 6.1 k 0.5 5.5 8.1 * 0.2 10.8 4.5 k 0.1 5.9 * Results k 2 standard deviations.Certificate range, o/o mlrn 2.54-2.61 0.28-0.29 0.17-0.19 0.14-0.15 0.11 1-0.116 F".gg-I 14-18 2 1-23 5 3-7.2 42-45 2 1-25 5.2-6.8 10.1-1 1.5 5.2-6.3 For mass spectrometry (MS), t'ie instrument used was a commercial instrument (PlasmaQuad, VG Isotopes, Winsford. Cheshire), with an rf generator and torch box supplying a maximum forward power of 2.0 kW at 27.12 MHz to the plasma.The torch was of a Fassel design with a 1.2 mm bore injector. Samples were introduced via a double-pass, water- cooled spray chamber. Again, the same unblockable nebuliser was used for sample introduction. Reagents The slurries were prepared by using 0.1 g per 100 ml of Aerosol OT (BDH Chemicals, Poole, Dorset) as dispersant in doubly distilled, de-ionised water. All working standards were pre- pared by taking various aliquots of the standard stock solution and diluting to the required concentration. Sample Preparation Four certified reference material coals were analysed: sub- bituminous coals NBS SRM 1632(a) and NBS SRM 1632(b) (National Bureau of Standards, Washington, DC, USA), SARM 18 and SARM 20 (South African Bureau of Standards, Pretoria, S.A.).The coal slurries were prepared by shaking 1 g of each sample with 10 g of zirconia beads (Glen Creston, Stanmore, Middlesex) and 4 ml of Aerosol OT solution (0.1 g per 100 ml) in a 30 ml Nalgene bottle on a standard laboratory flask shaker for 12 h. Particle size in the slurries was measured by using a Coulter counter (Model TAII, Coulter Electronics, Northwell Drive, Luton, Bedfordshire) and NBS 1632(a) and SARM 20 were shown to be 100% less than 2 pm in size, whilst NBS 1632(b) and SARM 18 were shown to be less than 10 pm in size. The slurries were washed from the beads and transferred to a 100-ml calibrated flask and diluted to volume with the Aerosol OT solution. For ICP-MS analysis, slurries of 0.2 g per 100 ml were prepared for minor or trace level elements, while for major elements slurries containing 0.002 g per 100 ml were prepared.With these concentrations no cone blocking effects were observed. Table 2. Slurry atomisation - inductively coupled plasma atomic emission spectrometry results for certified reference coal SARM 20. The value in parentheses is uncertified Slurry Certificate Certificate atomisation*. value, range, O/o mtm % mlm YO mfm A1203 9.88 k 0.09 11.27 11.16-1 1.73 Fe703 1.14 k 0.01 1.17 1.15-1.19 CaO 1.87 k 0.01 1.87 1.85-1.89 0.14-0.15 KZO 0.14 k 0.02 0.14 TiOz 0.61 k 0.01 0.63 0.6 1-0.65 Cr Mn c u Sr V Zn Ni c o M g - ' 69.7 t 0.7 79.1 t 0.3 17.6 k 0.1 319 k 6 46.3 k 0.05 15.3 t 0.05 14.2 k 0.1 22.3 k 0.1 CLgg-I (67) 80 18 330 47 17 25 8.3 * Results k 2 standard deviations. CIg g- I 77-82 15-19 318-338 45-50 14-18 7.6-9 23-26 Table 3.Slurry atomisation - inductively coupled plasma atomic emission spectrometry results for certified reference coal NBS SRM 1632(a). The values in parentheses are uncertified Slurry Certificate atomisation*, value, Yo mim Yo mlrn (3.1) A1 2.56 k 0.02 Fe 1.03 k 0.01 1.11 k 0.02 Ca 0.25 t 0.01 0.23 k 0.01 Ti 0.16 2 0.01 (0.175) Cr Mn c u Sr V Zn co Ni CIS g- ' 33.0 k 0.3 29.3 k 0.2 16.3 t 0.1 94.5 k 0.5 39.8 2 0.4 26.0 k 0.3 9.3 k 0.3 18.5 k 0.5 * Results k 2 standard deviations. CIgg-' 34.4 t 1.5 28.0 t 2.0 16.5 k 1.0 (88) 11.0 k 3.0 28.0 k 2.0 (6.8) 19.4 k 1.0ANALYTICAL PROCEEDINGS, MARCH 1988. VOL 25 71 Results and Discussion Tables 1-4 show the results from the analysis of the reference coals using simple aqueous calibration.Good agreement with the certified value for both major and trace elements was achieved. The values for aluminium are lower than the certified values and this was attributed to the inability of the plasma to atomise refractory aluminium oxide species completely. Table 4. Slurry atomisation - inductively coupled plasma atomic emission spectrometry results for certified reference coal NBS SRM 1632(b). The values in parentheses are uncertified Slurry Certificate atomisation*, value, '/i m/m '/o m/m Al 0.690 t 0.01 0.855 k 0.019 Fe 0.685 * 0.01 0.759 k 0.045 Ca 0.184 t 0.005 0.204 k 0.006 Ti Cr Mn c u Sr V Zn Co Ni Il.gg-I 345 k 6 12.5 t 0.7 12.1 t 0.5 6.1 k 0 . 2 99.6 t 2.5 11.0 k 0.1 10.2 t 0.1 3.0 t 0.1 7.0 t 0.2 T Results k 2 standard deviations. Il.gg- ' 454 k 17 (1) 12.4 2 1.0 6.28 t 0.3 (102) (1-1) 11.89 k 0.78 2.29 2 0.17 6.10 k 0.27 Table 5 shows the results from the analysis of reference coal SARM 18 by ICP-MS and the excellent agreement with the certified values achieved using aqueous calibration.The exception opce again is aluminium. The rare earth elements are difficult to determine by ICP-AES because of their inherently complex spectra; however, in ICP-MS this problem does not arise, and these elements can more readily be determined. Good agreement with the certified value was achieved for cerium and samarium by slurry atomisation ICP-MS. The poor result for uranium was attributed to sample contamination resulting from a previous determination of a uranium rich sample. Conclusion The above results for coal slurry analysis by ICP-AES and ICP-MS show good agreement with the certified coal values.Slurry atomisation is thus a practical method for coal analysis using both techniques. Provided that the particle size was below 8 pm, good agreement with the certified values was obtained by using aqueous calibration, with simple sample pre-treatment and calibration and without the need for modification of existing instrumentation. Table 5. Slurry atomisation - inductively coupled plasma mass spec- trometry results for certified reference coal SARM 18. The values in parentheses are uncertified Slurry Certificate atomisation*, value. % mim Yo mlm A1 0.80 k 0.05 2.57 Ti 0.095 t 0.001 0.114 Fe 0.300 k 0.007 0.290 Be Cr Mn CO Ni c u Ge Rb Ba Ce Sm U Mg Il.gg-' 3.7 k 0.22 654 k 16.4 16.1 k 0.9 23.4 t 0.8 6.7 t 0.4 10.1 2 0.3 7.5 2 0.8 8.2 k 0.8 8.1 k 0.4 74.5 k 1.1 22.3 k 1.8 2.1 k 0.2 3.4 k 0.1 * Results 2 2 standard deviations.ugg-' (4.1) 660 16.0 22.0 6.7 10.8 5.9 8.1 78.0 22.0 2.0 1.5 (8.0) Certificate range, % m/m 2.54-2.6 1 0.11 1-0.116 0.28-0.29 M g - ' 600-660 1418 2 1-23 5.5-7.2 9-13 5.2-6.4 6.7-9.5 71-78 21-24 1.9-2.2 1.5-2.0 The authors acknowledge the support of HGMP by SERC and British Coal under the CASE Studentship Scheme, and Applied Research Laboratories and VG Isotopes for the provision of instrument time and advice. 1. 2. 3. 4. 5. 6. 7. References Denton, C. L . . Himsworth, G . . and Whitehead, J . , Analysr, 1972. 97, 461. Jack\on, P. F. S . , and Whitehead. J . , Analysr, 1966, 91, 418. Ure. A . M.. and Bacon, J .R.. Analyst, 1978, 103, 807. Guidoboni, R. 3 . . Anal. Chem.. 1973, 45, 1275. Ruch. R. R . , Cahill, R . A , . Frost, J . K . , Camp. L. R.. and Gluskoter, H. J . , J . Radioanal. Chem., 1977. 38, 415. Ebdon. L., and Wilkinson, J . R . , J . Anal. A f . Specrrom.. 1987. 2, 39. Ebdon, L., and Wilkinson, J . R., J . Anal. A t . Specrrom.. 1987. 2. 325. Use of a Continuous Spectrophotometric Monitor for Nitrate Based Fertilisers in Hydroponic Cultivation J. Richard Clinch and Paul J. Worsfold School of Chemistry, University of Hull, Hull HU6 7RX Harry Casey and Sue M. Smith Freshwater Biological Association, River Laboratory, East Stoke, Wareham, Dorset BH20 6BB The on-site monitoring of chemical species in water has several advantages over laboratory-based methods, e.g., fewer prob- lems with sample stability and contamination, immediate and continuous transmission of information and savings on labour and time.The attractions of spectrophotometric detection for such monitors include good selectivity, robust instrumentation and applicability to a wide range of chemical species. There are several potential area5 of application for spectrophotometric field monitors, including water quality monitoring, water- intake protection, effluent monitoring and nutrient budget studies. This paper describes the use of a portable, fully automated nitrate selective spectrophotometric field monitor to investigate the dosing of a liquid fertiliser t o an experimen- tal, hydroponically cultivated, water cress bed.72 1 - Peristaltic - Injection - Reaction - Detector valve manifold Pumps ANALYTICAL PROCEEDINGS, MARCH 1988, VOL 25 Reagents 1 Fig.1. Block A I A I 1 1 I I I I Sample Switching valve H Standard I I I I -J -1 diagram of the automated spectrophotometric field monitor. (Reproduced, with permission, from reference 1) Experimental The water cress bed was continuously supplied with fresh spring water containing 4.50 mg 1-1 of nitrate as nitrogen, which passed through the bed and was led to waste. The spring water supply was periodically dosed with concentrated liquid fertiliser, such that a concentration of 20.0 mg 1-1 of nitrate as nitrogen was obtained at the outflow of the bed. The nitrate monitor was placed at the outflow of the bed and monitored the nitrate concentration of the effluent every 30 min, as has been described previously.1 A block diagram of the monitor is shown in Fig. 1. A single-board computer (Control Universal, Cambridge) is used to control .a 12-V, solenoid-activated, sample injection valve (Chemlab Instruments) with a sample volume of 30 pl, a 12-V two-way solenoid valve (Lee, Westbrook, USA) and two 240 V peristaltic pumps (Ismatec Mini S-820). The digitised output from the detector is sent to a miniprinter and a 24 x 2 character liquid-crystal display. The manifold is based on the reduction of nitrate to nitrite in a copperised cadmium column and derivatisation with N - l - naphthylethylenediamine dihydrochloride and sulphanil- amide. The resultant pink dye (Amax, = 542 nm) is detected by a solid-state light emitting diode - photodiode detector fitted with a flow-through ce11.2 To compensate for any drift, the sample signal is always ratioed against the signal from a 5.0 mg 1-1 nitrate as nitrogen standard.Results and Discussion Fig. 2 shows the variation of nitrate concentration at the outflow of the water cress bed over a 40-h period commencing at 4.30 p.m. The time required for liquid to pass through the bed is 1-2 h, which explains the time lag when dosing is started or stopped. During dosing the nitrate concentration at the outflow is 16-17 mg 1-1, indicating an uptake by the water cress of 3-4 mg 1-1, and when dosing is stopped the nitrate concentration at the outflow remains constant at 2.5 mg 1-1, indicating an uptake of 2.0 mg 1-1 from the spring water. The unexpected rise in signal after 26 h is probably due to a change in the flow-rate of spring water through the cress bed.This short term field trial clearly shows that the current dosing procedure should be linked to a feed back mechanism from the monitor to the dosing pump, which would result in more efficient and cost-effective dosing. The results are, however, encouraging with regard to the operation of the spectrophotometric field monitor. It is anticipated that the nitrate monitor could operate unattended for 1-2 weeks before reagents, nitrate standard and the copperised cadmium column Dosing Dosing Dosing O f f z 20 1 0 10 20 30 40 Time/h Fig. 2. over a 40-h period Nitrate concentration at the outflow of the water cress bed would need to be replaced. The monitor is also simple, cheap and portable, and the manifold can be readily modified to suit a wide range of spectrophotometric derivatisation procedures.The authors thank Wessex Water for financial support and one of us (J.R.C.) thanks the Science and Engineering Research Council for a scholarship. The Freshwater Biological Associa- tion is a grant-in-aid body of the Natural Environment Research Council. References 1. 2. Clinch, J . R.. Worsfold, P. J.. andCasey. H., Anul. Chim. Acfu. 1987, 200, 523. Worsfold, P. J . , Clinch. J . R.. and Casey. H., Anul. Chim. A m , 1987. 197. 43. A Novel Hydrothermal Treatment Process T. M. Khong and C. F. Simpson Analytical Science Group, Department of Chemistry, Birkbeck College, University of London, 20 Gordon Street, London WCl H OAJ The use of fluidised-bed technology for the production of chemically bonded phases for HPLC has already been demon- strated.1 These phases are produced by a silani~at~uii reaction of the surface silanols of silica gel and reactive organosilane.ANALYTICAL PROCEEDINGS. MARCH 1988.VOL 25 540 7.3 EJ - + This paper investigates the pre-treatment of the base silica in the fluidised-bed reactor. Generally, pre-treatment of silica gel involves treating the silica gel by boiling in acid followed by washing. This serves to purify and hydroxylate the surface fully. Finally, the silica gel is heated in an oven at 200 "C in air or under vacuum for a given duration.2.3 During initial studies into the fluidised-bed method of producing chemically bonded phases for HPLC, the viability of hydrothermally treating the silica as a means of activation prior to reaction was investigated.Hydrothermal treatment refers to the physical modification of silica gel by exposure to water or to gaseous water (superheated steam) at temperatures higher than the normal boiling point of 100°C. The autoclaving process is commonly used for such treatments in which the silica gel is held in water or steam over 100°C at super- atmospheric pressures. The net effect is a sintering of the silica gel, which results in an increase in the pore size of the silica particles with a corresponding decrease in surface area. 520 I N Cn 500 (II ? Experimental Materials Employed HPLC grade silica of particle size 20 pm, pore size 6 nm, manufacturer's surface area 540 m' g-1.A - - Equipment described1 and is shown in Fig. 1 . Hiridsing tower. The fluidising tower has previously been Hydrothermal Treatment About 60 g of silica were placed in a fluidising tower (45 mm i.d.). Steam was generated in a vaporiser with a water inlet flow of 2.5 pl min-1. The steam was carried by a flow of dry nitrogen (BOC) (-lO(LlSO ml min-1) to a heat exchanger, where it was superheated prior to passage to the appropriately heated fluidising tower. The temperature in the fluidised bed was maintained at k 2 "C of the set temperature. Samples were withdrawn from the reactor by suction and transferred to dried sample vials. where they were sealed. Fluidising gases used were dried using activated Linde SA molecular sieves (BDH). Surface Area and Particle Size Measurements The BET nitrogen adsorption was performed using a home- f At ~ O S D h e r 2 Condenser k Pre-tower (-)-& Man om et er heater I Steam Silane Flow meter - Fig.1. Schematic diagram of fluidised bed reactor 0 4 0 m A 0 20 40 60 80 100 120 Time/h Fig. 2. Effect of hydrothermal pretreatment on surface area of d i c a gel (measured by BET). Temperatures are: 0. 150 "C: +. 200 "C; A. 250 "C 420 ' Instrumentation The chromatograph employed consisted of a Shimadm LC-SA pump and a Shimadzu SPD 2AM ultraviolet detector operating at 254 nm. A Rheodyne 7133 fitted with a 5 pl internal loop was used. Columns were thermostatted in a water jacket at 20°C. Mobile phases were prepared on a volume basis using LC-grade solvents. Results and Discussion Fig.2 shows the effect of hydrothermal pre-treatment on the surface area as measured by the BET nitrogen adsorption. The results show that generally there is a decrease in surface area " 0 12 24 36 48 60 72 84 96 108 120 Timeih Fig. 3. Graph showing the effect o f hydrothermal pre-treatment duration on pore diameter at three different temperatures: 0. 150 "C: +. 200 "C; A. 250 "C74 ANALYTICAL PROCEEDINGS. MARCH 1988, VOL 25 after treatment for 108 h. The change in the physical nature is considerably less drastic than if the silica were treated by autoclaving methods. Fig. 3 shows that the radius of the pore remains relatively constant (up to 48 h) under a wide range of treatment conditions. 13.5 I Tem per at u r e/"C Fig. 4. hydrothermal pre-treatment and silanisation on carbon loading Graph showing the effect of temperature of fluidised bed In Table 1 , it is shown that hydrothermal treatment followed by reaction at 200°C yields the highest percentage carbon loading.Furthermore, Fig. 4 shows that there is a linear relationship between the temperature of hydrothermal treat- ment and the carbon loading achieved. As the surface area and pore diameter were nominally similar after the hydrothermal pre-treatment at the various temperatures, the carbon loading on the resultant bonded phase reflects the degree of surface silanol concentration. At 400 "C, surface silanols are dehy- drated, forming siloxanes, and hence fewer active silanols are available for chemical reaction. This is illustrated in Table 1 by the low carbon loading at 400°C.50 t 40 30 a- g 20 10 0 A '"'0 10 20 30 40 50 60 Sizelpm Fig. 5. Graph showing the effect of fluidised bed hydrothermal treatment at 150 "C on the distribution of particle size of silica gel (by POP) with time Particle attrition in fluidised beds caused by the vigorous agitation of the friable solid particles may produce fines.5 Consequently, solids treated in the fluidised bed may be smaller in size. However, in these experiments fines are continuously removed because of the constant upward flow of the carrier gas through the bed (Fig. 5 ) . Furthermore, the particle size distribution becomes narrower as the duration of hydrothermal treatment progresses. The narrowing of size distribution is probably due to the break up of the larger particles, giving rise to particles of smaller size.However, the mean particle size of 20 pm was unchanged after the hydrothermal treatment. The removal of fines and the narrow particle size distribution are essential for the production of efficient HPLC columns as column permeability is improved by this means, thus leading to more efficient columns. Improved column performance has previously been demonstrated. 1 Table 1. Carbon loading and coverage values for C, (uncapped) reacted at various temperatures Temperature Coverage: S u r face Pore reaction/"C C, % per nm (SSf,J/m2 g - nm None - - 689.6 6.96 200 13.07 I .93 605.2 6.96 300 11.66 1.72 628.6 6.66 400 10.28 1.52 602.3 7.10 of no. of ligands area diameter/ Chromatograms (Fig. 6) show the effect of hydrothermal pre-treatment in conjunction with fluidised bed preparation for RPLC.10 5 Ti me/m in 15 10 5 Ti me/m in 15 10 5 Ti melm in Fig. 6. Comparative separations performed on three C, reversed phases: ( a ) , separations performed on commercially available equi- valent C,; ( b ) , separations performed on C, packing without hydrothermal treatment; ( c ) , separations performed on C, packing with hydrothermal treatment. Conditions are: flow-rate. 1 ml rnin- 1; mobile phase, methanol - water ( 1 : 1 V / V ) ; temperature, 20°C. Samples are: 1. acetone; 2. phenol; 3 . p-cresol; 4. 2.5-xylenol; 5. p he n e t o I e Conclusion The hydrothermal treatment described in this paper is less vigorous than that effected by use of an autoclave and as such does not change the physical properties of silica gel signifi- cantly.However, this pre-treatment process is most suited to activating silica gel prior to reaction for the preparation of chemically bonded phases. The advantages of this process are the removal of fines, the narrowing of particle size distributionANALYTICAL PROCEEDINGS. MARCH 1988, VOL 25 75 and the maximisation of the surface silanols for subsequent reaction. It is also shown that when this process is used in conjunction with subsequent fluidised bed preparation of chemically bonded phases, the RPLC produced shows better chromatographic characteristics. The authors would like to thank Professor J. H. Knox (Edinburgh University), Mr. M. Harris (King's College, London) and Mrs. M. Brockbanks (Dysons Instrument Ltd.) for the use of their equipment.References 1 . Khong. T. M.. and Simpson. C. F . . C'lzrornutoqruphia. 1987.24. 385. 2. Hernetsherger. H.. Maasfeld. W.. and Ricken. H.. C'lzromato- graphia. 1976. 9. 303. 3. Rehak. V.. and Smolkovri, E.. J. Chromafogr.. 1980. 191. 71. 4. van Kreveld, M. E.. and van den Hoed, N . . I . ('hromnrogr.. 1973. 83, 1 1 1. 5 . Yates, J . G.. "Fundamentals of Fluidised-hed Chemical Processes." Rutterworths. Sevenoaks. 1983. Comparison of Vapour-phase Temperatures and Chemical Interferences for Wall, Platform and Probe Atomisation in Electrothermal Atomisation Atomic Absorption Spectrometry 0. 0. Ajayi and 0. Littlejohn Department of Pure and Applied Chemistry, University of Strathclyde, Cathedral Street, Glasgow GI 7XL C. B. Boss Department of Chemistry, North Carolina State University, Ra yleigh, NC 27650, USA Various procedures have been proposed for the reduction of vapour p h ase interferences in elect ro t h e r m a 1 at om isa t i on atomic absorption spectometry (ETA-AAS).The various approaches studied fall into two categories; methods that involve removal o f some of the matrix components prior to the atomisation stage, or procedures that involve atomisation into a high temperature environment. The former procedures include matrix modification , I - extraction and pre-concentra- tion. In the latter category, platform+’ and probeX I 4 atomisa- tion have been used with some success to reduce the extent of vapour phase interferences, i n comparison with conventional tube-wall atomisation. Other techniques in this category include capacitive discharge heating,l3 the side-heated iso- thermal furnace’-‘ and the double furnace. 1i This study has been undertaken to compare the chemical interferences caused by chloride salts in the determination of volatile elements, such as lead and gallium, when atomised by tube-wall, platform and probe atomisation procedures. An attempt has been made firstly to calculate the vapour phase temperature experienced by the analyte species for the three modes of atomisation, and secondly, to correlate the tempera- tures with results obtained in the interference studies.A two-line absorption procedure was applied to calculate the vapour phase temperature. using lead as the thermometric species. If’ The effect of varying the atomiser tube wall thickness at various points along the length of the tube was also investigated to establish the dependence of vapour phase temperatures and interferences on tube temperature grad- ients.1’ Experimental Pye Unicani PLJ 9000 and SP 9 atomic absorption spec- Table 1. Temperature programmec; tor Pye I’nicam S P 4 atomiser Stage Wall Platform Dry Hold time% 32 32 AAh Hold timeis 15 IS Tempera t urt9OC 160 I60 Ramp No.’ 6 6 Temperature “C YMl-900 6O(L 1 5( I( 1 Ramp No.” 3 3 Pre-herit I Ho I d time I s Temperature:”C - RampNo.’ - - Pre-heat I1 Temperaturei’C Hold timeis Ramp No.’ Atomise Tern pe ra t ure /“C Hold time14 - 3 Ramp No.’ 0 0 Clean Temperature ‘“C 2800 9800 Hold timeis - 3 - 7 Ramp No. ’ 0 0 3 ) 0 ( ~ 9 8 0 0 3 I( L98( )O 3 . Ramps 0. 3 and 6 implv >2000.200 and 20°C 4 1 . respectively Probe 500 50 h hU0- 1600 15 3 400 5 0 2800 0 7 - Functions selected Recorder Recorder Temperature control Gas stop Tern pe r at u re con t rnl Peak timer Gas stop Temper at u re con t r o I76 ANALYTICAL PROCEEDINGS, MARCH 1988. VOL 25 Table 2. Wall and vapour temperatures and corresponding percentage reco\ ery of lead in presence of sodium chloride for different tube shapes. Pye Unicam SP-9 atomiser. wall atomisation Tem pe ra t ure+/"C Lead recovery.: '%, At start of Pb peak# At Pb peak height At end of Ph peaks I(!;" I?I:V NaCl ?'Yo n~iV NaCl Wall Wall Wall Wall Wall Wall Peak Peak Peak Peak Tube" centre end Vapourq centre end Vapourq centre end VapourT height area height area Unmodified 1405- 1132- 1166- 1503- 1325- 156-C 191% 1799- 159s- S9 90 53 33 Modified 1410 1130 1187 1629 122% 15-47 1909 1384- 1505 94 68 33 36 Modified - - 1737 1357- 2021- 1919 1x97- 9941- 1635 92 108 70 89 1407 1135 1309 1645 1556 1603 19-38 1838 1700 tube I 12x6 1337 tube I1 1474 2108 1911 2501 * Pyrolytically coated tube.t Set temperature 2000 "C. 5 Defined as time during atomisation when lead absorbance is 90"L o f peak height \.slue. SO ng ml 1 of lead. 10 yl volume. Errors in calculation of vapour temperature were typically 250-200 "C. trometers were used, along with a PU 9007 data station and PU 9090 data graphics system, respectively. A PU 9095 video furnace programmer was used with an SP-9 furnace power supply as the atomiser. Two microcomputers. an IBM-AT and an Apple IIe. were used for data processing with Tempcalc, Lotus 123 and Pyro software.The latter. non-commercial. programme was written for operation with the Ircon optical pyrometer that was used t o measure the surface temperatures of the atomiser tube. Vapour temperatures were calculated by the two-line lead absorption method described by Siemer and Lewis.16 The vapour temperatures were estimated from the temporal signals of 5 pg ml-1 of lead at 280.2 nm and 368.3 nm. Absorbance data were collected at 200 data points over the duration of the transient signals. This information was used with the "Tempcalc" program t o obtain time-resolved wpour temperature profiles. The interference caused by 0.01-1% mlV concentrations of sodium chloride and magnesium chloride on 50 pg I- 1 of lead and of zinc chloride on 100 pg I-! of gallium were compared for tube-wall, platform and probe atomisation.The various atomiser temperature programmes are given in Table 1 . Results and Discussion Influence of Tube Design on Surface and Vapour Temperatures and Vapour Phase Interferences for Wall Atomisation Pyrolytic graphite coated electrographite tubes for the Pye Unicam SP 9 atomiser were modified by reducing the wall thickness at different sections along the length of the tube. Modified tube I was created by shaving off a 0.5-mm layer of graphite from a 5-6 mm mid-portion of the tube to increase the temperature gradient from the hottest section at the injection hole towards the cooler tube ends.17 The second design. modified tube 11. was prepared by removing a 0.5 mm layer of graphite from 5-6 mm sections starting 10 mm from the ends of the tube. This reversed the temperature gradients such that the ends of the tube heated faster and to a higher temperature than the unaltered central section.17 The vapour and surface temperatures achieved with the conventional and modified tubes were observed to be design dependent. Vapour temperatures were calculated from ab- sorbance measurements at the start, peak and end of the lead atomic absorption signals and the values are given in Table 2. The start and end points were defined as the times during the atomisation cycle when the lead signal had reached 20% of the peak value. Corresponding optical pyrometer temperature measurements obtained for the centre and ends of each of the tube designs are also listed in Table 2.Where two or three repeat measurements o r calculations were made, the range o f temperature values obtained is given in Table 2. There was evidence that the vapour temperature experienced by the lead atoms in the modified tube I1 was higher throughout the duration of the lead atomic vapour than for the other two tube designs. This is a result of the in\'erted temperature gradient that is created along modified tube 11. which promotes heating of the vapour by the hot end sections. prior t o \raporisation of the analyte from the tube centre. The gas temperature at the time of analyte vaporisation is therefore higher than for modified tube I and the unmodified tube. Interference studies performed with the three tube designs (Table 2) also confirmed the advantage of modified tube 11.Although there was very little difference in the performance of the tube designs with regard to the interference of a 1% i ? l i I / sodium chloride solution on 50 ng ml-1 o f lead. at 2% tTz1V of sodium chloride a higher level of lead recovery was obtained with modified tube 11. It would appear that there is some advantage to be obtained by altering the shape of an electrothermal atomiser tube to allow vaporisation into a higher temperature vapour than is currently achieved with conventional tubes of uniform wall thickness. Comparison of Tube and Vapour Temperatures and Chemical Interferences for Wall, Platform and Probe Atomisation Besides altering the temperature gradient along a graphite tube, it is known that vaporisation into a high temperature environment can be achieved through the application of platform and probe atomisation procedures.A comparison of wall and vapour temperatures for the three modes of atomisa- tion (wall, platform and probe) is given in Table 3 for measurements under conditions selected to give a final tube temperature close to 2000 "C. The vapour temperatures were calculated by the lead two-line method using absorbance \ d u e s at the start. peak and end o f the lead signals. Corresponding wall temperatures at the centre of the tube were obtained ivith an optical pyrometer. The results presented in Table 3 show that the platform has a cooling effect on the temperature of the tube, which causes a slight reduction in the apparent vapour temperature at times corresponding to the start of the lead peak, although equivalent or higher vapour temperatures are achieved in comparison to wall atomisation by the end of the lead peak. Insertion of a graphite probe into a heated atomiser tube does not greatly influence the wall temperature.which was reasonably constant throughout the duration of the lead atomic absorption signals used to calculate the \rapour tem- perature. However, the probe did have a cooling effect o n the vapour to the extent that there was a greater difference between the vapour and wall temperatures at the three times during the existence of lead atoms compared with ~vall atomisation. Even so, the average vapour temperature for probe atomisation was still higher than that obtained n,ith n-all atomisation.ANALYTICAL PROCEEDINGS.MARCH 1988, VOL 25 77 lable 3. Wall and "apour temperatures for wall. platform and probe atomisation Me asu re d:' Ca I c u I a t e d t e m pe rat u re s!"C At lead peak height At start of lead peak": .4 t o ni i sa t ion Set Wall Wall Wall At end o f lead peak ' Tube mode temperuture!"C centre Vapour.; centre Vapourt centre Vapourt Total p!.rolytic Wall 2000 1 - 1 2 1400 1827-1905 1650 190(l- 1924 1498 graphite U n ni od i fie d Platform 2000 - 1 39(& 1478 1690 1357-1398 1722-1735 1421-1623 p y ro I y t i c a I1 y (TPG 1 coated s I0 t t e ci t 0 t a I Probe 2300 1003-2044 1466- 1608 2035-3 1 9 1 1605- 164 1 2033-2 164 1548- 1764 p y ro 1 y t ic Fraphite ' Defined as time during atomisation when lead absorbance is X"il o f peak height \ d u e ..: Errors in calculation of vapour temperatures \yere typically kW-200 "C. Table 4. Comparison of wall. platform and probe atomisation interferences for magnesium chloride on lead and zinc chloride on gallium Percentage recoiwy of analyte signal, "/o Lead '< Gallium+ 0.5"/, v , ' V MgCI, 1 .O"/o miV MgCI, At om i s a t i o n Peak Peak Tube mode height Area height Area Total pyolytic Wall -11 42 36 36 graphite p yr o 1 y t ic a I I y coated U n mod i f i e d PI at fo r m 71 86 59 81 (TPG) Slotted total Probe 91 85 81 70 pyrol ytic graph it e ' 10 pl volumes of 50 ng ml 1 lead solution at 2000°C. + 10 pl volumes of 100 ng m l ~ 1 gallium solution at 2800 "C. 0.5% rniV ZnCI, Peak Peak height Area height Area 13 20 6 16 l.O'% miV ZnCI, 106 85 113 89 93 81 86 77 Table 3 gives details of an interference study conducted to investigate the effect of magnesium chloride on lead under the atomisation conditions described in Table 3.As expected from the vapour temperature data, lead is less susceptible to interferences from magnesium chloride when probe atomisa- tion is used rather than wall atomisation. What is interesting is that platform atomisation is also better than wall atomisation, even although the vapour temperature at the time of the peak lead signal is apparently lower than that achieved with wall atomisation. This suggests that the effective interference concentration in the atomiser tube at the time of lead vaporisation is lower than for wall atomisation. possibly due t o greater temporal separation of the analyte and interferent vapours with platform atomisation compared with wall atom- isation.With platform atomisation the superiority of area measurements compared with height measurements is not unexpected when the vapour temperatures given in Table 3 are taken into account. It might be expected that the higher vapour temperature during the latter period of the lead signal would cause the area interference to be lower than that of the height measurement. Results are also given, in Table 4, for an interference study to evaluate the effect o f zinc chloride on gallium. A higher atomisation temperature of 2800°C was used for these measurements and i t was expected that the cooling effect of the platform and probe would be less significant at this tempera- ture compared with measurements made at 2000°C. This indeed appears t o be the case with both the platform and probe giving much better recoveries of the gallium peak height and peak area signals in comparison with wall atomisation.I n general. there appears t o be no substantial difference between probe and platform atomisation with regard to the zinc chloride interference effect, and close t o quantitative recovery of gallium is obtained for both atomisation modes at least t o the 0.5% mlV of zinc chloride level. Conclusions The surface temperature of a tube and the temperature gradient along the tube that exists at the time of analyte vaporisation have a direct influence on the gas temperature experienced by atoms in electrothermal atomisation.By altering the wall thickness of a graphite tube so that the ends of the tube heated up faster than the centre, it was possible to increase the effective vapour temperature during the period of analyte measurement t o an extent similar to that achieved by platform and probe atomisation. The vapour temperature that exists at the time of sample vaporisation has a direct influence on the degree of vapour phase interferences that are caused by chloride salts on volatile analytes such as lead and gallium. In order obtain a complete picture as t o the influence of probe and platform atomisation on interferences, it is necessary t o consider the temporal overlap of analyte and interference species as well as the vapour temperature that exists in the tube during the measurement period.The authors are grateful for a Nigerian Government Scholar- ship (for O.O.A.) and acknowledge the loan of spectrometers and electrothermal atomisers by the Marketing and Develop- ment Departments of Pye Unicam Limited, Cambridge. References 1 . 2. 3. Slavin. W.. Manning. D. C.. and Carnrick. G. R.. At. L'vov. R. V.. Spectrochim. Actu, Purt R . 1978, 33. 153. Ediger. R. D . . Ar. Ahsorpr. Ncwsl.. 1975. 14. 137. Spiwrmc.. 1981. 2. 137.78 4. 5 . 6. 7. 8. 9. 10. Slavin. W.. Carnrick. G . R.. Manning. D. C.. and Pruzkowska. Ottaway. J . M.. Atzul. Proc.. 1984. 21. 55. 11. E.. At. Spectrosc.. 1983. 4. 69. 12. Slavin, W . . and Manning. D. C.. Spectrochini. Acia, Purr B. 1980. 35. 701. 13. Bezur. L . . Marshall. J . , Ottaway, J .M.. and Fakhrul-Aldeen. R.. Atialj,sr. 1983. 108. 553. 14. Carroll, J . . Marshall. J . . Littlejohn. D., and Ottaway, J . M.. Fresenius Z . A w l . Cheriz.. 1985. 322. 145. 15. Littlejohn. D.. Cook. S . . Durie. D.. and Ottaway. J. M.. Spectrochim. Acta, Part B . 198.1. 39. 295. 16. Ottaway. J . M.. Carroll, J . . Cook, S . . Corr. S. P.. Littlejohn. 17. D.. and Marshall. J.. Freseriius Z. Aiiul. Chein.. 1986. 323. 747. ANALYTICAL PROCEEDINGS. MARCH 1988, VOL 25 Giri. S . K.. Shields. C. K.. Littlejohn. D.. and OttaLvay. J . M., Anuljst. 1983. 108. 211. Slavin. W., and Manning, D. C.. Specrrochirn. Aciu. Part B . 1982, 37, 955. Chakrabarti. C . L.. Wan. C. C.. I-lamcd. 11. A , . Bertels. P. C.. and Grcgoire. G. C.. A m / . Chrrn.. 1980. 52. 167. Frech. W., Baxter.D.. and Hutsch. B.. Anal. Chew.. 1986. 58. 1973. Frech, W . , and Jonsson. S . . Specrrochiin. ACIN, Purr B , 1982.37, 1021. Siemer. D . D . . and Lewis. L. C.. Arid. Cliem.. 1983. 55, 99. Littlejohn. D.. and Ottaway. J . M., Anulj.si. 1979. 104. 208. Surfactant-selective Electrode for Ion-pair Titrations in Mixed Solvents C. J. Dowle and B. G. Cooksey Department of Pure and Applied Chemistry, University of Strathclyde, Cathedral Street, Glasgow G I 1XL W. C. Campbell ICI plc, Chemicals and Polymers Group, P.O. Box 90, Wilton, Middlesbrough, Cleveland TS6 8JE A commonly used procedure for the analysis of surfactants in detergent formulations involves an ion-exchange separation using ethanolic hydrochloric acid, with subsequent quantifica- tion of the anionic and cationic fractions by a two-phase titration.This procedure has several drawbacks. some of which can be overcome by using surfactant-sensitive electrodes to detect the end-point of a single-phase titration. In the two-phase titrations. the toxicity and reactivity of the organic phase presents problems. Also. the colorimetric end-point requires considerable experience. being very subjective, and the variation of partition coefficients for the range of surfactant homologues can yield inaccurate results. Surfactant ion-selective electrodes, which are particularly useful in mixed solvent systems. have been developed for the analysis of anionic and cationic surfactants. I The electodes are constructed from PVC membranes incorporating ion pairs as ion exchangers.coated upon graphite substrates. They are simple. robust and inexpensive. and respond to ionic surfactant concentrations between lo-’ and 1 0 - ( 7 M . Electrode Construction The ion-selective electrodes are constructed as illustrated in Fig. 1. The sensing membrane is composed of 59.95% mlnz tritolyl phosphate. 39.95% rnlrti PVC (molecular weight 100000). and 0.1W0 niitzi of a surfactant salt. For the cationic ion-selective electrode. the salt used was tetrabutylammonium dodecylsulphate. For the anionic ion-selective electrode. the salt used %as n-hexadecyl trimethylammonium pentane-1-sul- phonate. C A 10mm D \ I - - - - - D 30 mm =lo0 mm * \ Fig. 1. Construction of the sensor electrode: A, spectrochemical graphite (diameter 5 rnm): B. coaxial cable; C. glass tube.10 mm o.d. and 9 m m i.d.: D. cpoxj resin The reproducibility of the electrcxk manufacture was acceptable considering the simple dip-coating technique employed. For example, when the measurements were made with a series of six freshly prepared anionic surfactant ion-selective electrodes. a cell potential for a 10-4 M sodium dodecylsulphate (aqueous) solution of 199.3 k 3.88 mV was obtained. Methodology The electrochemical cell consisted of the appropriate ion- selective electrode and an Activion standard calomel reference electrode. The cell potentials were measured by an EIL 7050 potentiometer linked to a Linseis chart recorder. The analyte solutions were stirred magnetically and. for all measurements, constant electrode immersion levels were maintained.For potentiometric titrations, 4.0 -+_ 0.05 ml of each sample were taken and diluted ten-fold with water. The titrant was added at a rate of 1.25 ml min-1 from a Radiometer ABU 80 autoburette. The cationic surfactants were titrated against aqueous sodium dodecylsulphate (0.004 M ) ~ while the anionic surfactants were titrated with aqueous Hyamine 1622 Large potentiometric end-point breaks were recorded (80- 250 mV) which were extrapolated manually for quantification. Favourable comparison with standard two-phase titrations’ was observed. A wide range of surfactant types can be successfully quantified by using this procedure. These include alkylbenzene sulphonates. alkyl sulphates, alkylether sul- phates. quaternary ammonium salts and amine oxides. They operate over a wide series of homologues, exhibiting little differential selectivity. (0.004 M ) .Applications of Electrodes Particular value is seen in quantifying the anionic and cationic fractions eluted from an automatic ion-exchange system.‘ which separates commercial formulations into component fractions. i.e. , soaps, anionic, cationic and non-ionic surfac- tants. These are eluted in hydrochloric acid - ethanol ( 1 + 9 V/V). Following a ten-fold dilution of the medium with water, the electrodes described in this paper quantified surfactant concentrations very successfully in contrast with commercial surfactant electrodes, which do not operate satisfactorily in this medium. Table 1 illustrates the application of the electrodes for the quantification of commercially formulated detergents, after ion-exchange separation.Comparative concentrations determined by a standard two-phase titration are also given in Table 1. It should be noted that it was only possible to conduct a volumetric comparison as the true molecular weights of the blends were not known. Conclusions The titration technique described for locating end-points in surfactant ion-pair titrations offers improvements upon theANALYTICAL PROCEEDINGS. MARC13 1988, VOL 25 79 Table 1. Comparative analysis o f total cationic and anionic surfactants prescnt in real samples using the potentiometric and two-phase titration techniques End-point for total anionics End-point for total cationics Sample Pot en t io me t r i c 24.6 24.55 Washing-up liquid (2) 31.75 31.8 31.8 Washing-up liquid ( 1) 24.6 Washing powder Washing powder 1 ) 24.6 24.6 24.6 2) 2.0 2.05 2.0 Washing powder (3) 29.75 29.5 29.7 Two - phase 24.55 24.6 24.55 31.55 31 .6 31 .55 24.5 24.5 24.5 1.9 1.9 I .H 30.0 30.05 30.0 Potcntiometric Two-phase None present None present 4.25 4.15 4.2 0.50 0.55 0.50 1.40 1.40 1.35 4.2 4.2 4.15 0.70 0.65 0.75 1.40 1.40 1.40 existing two-phase titration.The electrodes described are also References insensitive to inorganic salts, and offer good performance in some mixed solvent systems where a Nernstian response can be observed below the critical micelle concentration over a wide range. I . 2. 3. Dowle. C. J . . Cooksey. B. G.. Ottaway. J . M.. and Campbell. W. C., Analyst, 1987, 112, 1299. Reid. W. W., Teriside Deterg.. 1967, 4. 292.MacDonald. L. S . . Cooksey. B . G . , Ottaway. J . M.. and Carnpbcll. W. C., Anal. Proc., 1986,23. 448. Atomiser, Source, Inductively Coupled Plasma in Atomic Fluorescence Spectrometry (ASIA): a Comparison of Nebulisers S. Greenfield, M. S. Salman, M. Thomsen and J. F. Tyson Department of Chemistry, Loughborouah University of Technology, Loughborough, Leicestershire LEI 1 3TU Liquid sample introduction with pneumatic nebulisation is the approach used in the vast majority of flame and plasma atomic spectroscopy determinations. It is the most popular form of sample introduction as it is simple, inexpensive and robust, without being particularly prone to memory effects. The concentric pneumatic nebuliser is probably the most popular nebuliser for use with an ICP,I-3 although cross-flow nebulisers are also widely used with the 1CP.C" Table 1.Optimum conditions fluorescence Variable Coolant gas flowil min I Plasma gas flow11 min - 1 Nebuliser gas flow11 min I Liquid flow-ratelmi rnin- 1 Observation height above the Power in plasma/kW Slit widthimrn coilicrn for lead determination in atomic Atomiser Source ICP ICP 8.5 30 25 -.- 1 3 3.5 4.2 5.3 - 10.1 1.8 0.39 5.9 3.6 A major disadvantage' of both cross-flow and concentric nebulisers is that the sample solution has to pass through a narrow capillary, so that care must be taken to ensure that the sample solution contains no small particles which could block the nebuliser. Babington# originally described a nebuliser in which the gas issued from a small orifice into a flowing solution.Nebulisers based o n this principle have recently been proposed for use with the ICP.7-".1" The use of a spray chamber appears to be imperative with these nebulisers before the aerosol is injected into the ICP. The spray chamber essentially acts as a sorting device, allowing only the droplets below a certain size to reach the atom cell.ll The two most popular types of spray chamber are the double-pass spray chamber described for ICP work by Scott et al. and the vortex-type of cloud chamber (or cyclone spray chamber) described for ICP work by Greenfield. 13 Ultrasonic nebulisation has already, since the first work on ICP-AES, been recognised as a powerful method for aerosol generation.lJ.15 Two types of ultrasonic nebulisers have been developed; in the first type the analyte solution flows over the surface of the transducer as described by Boumans and de Beer,]-' Olsen et al.16 and Taylor and Floyd." In the second type.the ultrasonic energy is focused on a Mylar film as described by Wendt and Fassel.14 Recently, Fassel er al. l X . 1 Y reported extensive studies of ultrasonic nebulisation. There have been a number of compari- sons of the performance of nebulisers used in ICP-AES. 14.20.21 In this paper a comparison of nebulisers used in the ASIA instrument is described. ~~ Table 2. Comparison of nebulisers on the source plasma under optimum conditions with the GMK nebuliser on the atomiser plasma Nebuliser on source Total fluorescence plasma signal/mV GMK 395.6 dc Galan plus cyclone 395.5 de Galan plus impact 3x5 spray chamber wall spray chamber80 11 ANALYTICAL PROCEEDINGS, M A R C H 1988, VOL 25 cm 4 Top view 10 cm + I Glass flask (500 ml) The aerosol produced is collected from the central region of the spray chamber and transported to the plasma torch Side view Aerosol to plasma torch To drain Fig.1. The cyclone spray chamber Experimental The dual-plasma atomic fluorescence - atomic emission spec- trometer (ASIA spectrometer) has been described else- where.’* A comparison was made between a commercially available GMK nebuliser and a de Galan nebuliser. The latter was used in combination with a cyclone spray chamber with central aerosol collection (Fig. 1) and impact wall spray chamber (Fig. 2). Also included in this comparison was an ultrasonic 4erosol to plasma Delivery solution \ Argon inlet 1 \de Galan nebuliser I I To drain k 5crn 4 the use of two plasmas in the ASIA instrument results in a total of 12 variables including the spectrometer slit width (Table 1).The search began with the source plasma followed by the atomiser plasma, changing each variable one at a time and keeping all others constant, thus setting each variable in turn to a value producing the maximum in the figure of merit (total fluorescence signal for lead at 280.2-283.3 nm). This process was repeated in a cyclical manner until the figure of merit used showed no significant change. The effect of using a desolvating system was also investigated on both atomiser and source plasmas. 35 cm Water out Water in To plasma Delivery solution - Water out To drain Fig.3. Ultrasonic nebuliser with desolvation system The fluorescence signal under optimised conditions was measured for an integration period of 10 s. A fixed concentra- tion of 100 p.p.m. of lead was introduced into the atomiser plasma and the source plasma was run with a 20% lead solution with recirculation (see Fig. 4). An Apple IIe computer was used for data collection. Delivery solution de Galan nebuliser Tygon tubing Sample Fig. 4. Recycling sample introduction system on source plasma Fig. 2. Impact wall spray chamber Results and Discussion Table 1 shows the optimum conditions for lead at h = 2 280.2-283.3 nm in atomic fluorescence in the best arrangement of nebulisers on the instrument, i.e.. the GMK nebuliser on the atomiser plasma and the de Galan nebuliser with cyclone spray chamber on the source plasma. nebuliser designed to focus the ultrasonic energy on a Mylar film (Fig. 3).An optimisation procedure was carried out by using an alternating variable search (AVS) method in whichANALYTICAL PROCEEDINGS. MARCH 1988. VOL 25 81 Table 2 shows the fluorescence signal obtained with the de Galan nebuliser in combination with both the cyclone and impact wall spray chambers with the GMK nebuliser on the atomiser plasma. The results show no significant difference between the commercial GMK nebuliser and the de Galan nebuliser with the spray chambers described above. A comparison was made between the GMK nebuliser, the de Galan nebuliser in combination with both cyclone and impact wall spray chambers and the ultrasonic nebuliser, while keeping the GMK nebuliser on the source plasma.The results are given in Table 3. The signal from the de Galan nebuliser with the impact wall spray chamber was lower than that obtained with the GMK nebuliser and the de Galan nebuliser used in conjunction with the cyclone spray chamber. When the ultrasonic nebuliser was used, it gave a fluorescence signal 3-4 times as great as all of the pneumatic nebulisers. but its precision was poor. Table 3. Comparison of nebulisers on the atomiser plasma under optimum conditions with the GMK nebuliser on the source plasma Nebuliser on atomiser Total fluorescence plasma signal/mV GMK 100.1 de Galan plus cyclone 91.2 de Galan plus impact 74.8 spray chamber wall spray chamber Ultrasonic 366.6 The effect of desolvation was studied with the GMK nebuliser on the atomiser plasma and the de Galan nebuliser with the cyclone spray chamber and the ultrasonic nebuliser while keeping the GMK nebuliser on the source plasma.Results indicated in all instances that the total signal was less with desolvation. This effect is probably caused by the dry a e r a o l increasing the temperature of the atomiser plasma as it was not possible t o turn the power down and still maintain a stable plasma. A study of the effect of desolvation on the source plasma was made with the GMK nebuliser and the de Galan nebuliser while keeping the GMK nebuliser on the atomiser plasma. An increase in the fluorescence signal was observed with desolvation. The best detection limit (85 p.p.b.) was obtained with the pneumatic nebulisers.The arrangement was the de Galan nebuliser plus cyclone spray chamber on the source plasma with the GMK nebuliser on the atomiser plasma. The authors thank Professor L. de Galan for the generous provision of two nebulisers, the Iraqi Government for financial support for MSS, and the SERC and BDH Ltd. for financial support for MT. I . 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. References Meinhard, J. E., ICP Inf. Newsl.. 1976. 2. 163. Ebdon. L.. Cave, M. R., and Mowthorpe. D. J . . Anal. C'him. Actu. 1980, 115. 179. Scott, R . H . , ICP Inf. Newsl.. 1978. 3. 425. Kniseley, R. N . , Amenson, H.. Butler. C. C.. and Fassel. V. A . . Appl. Spectrosc.. 1974, 28. 285. Donohue. D. L., and Carter, J .A , . Anal. C'hem., 1978,50,686. Novak, J . W., Lillie, D . E., Boom, A. W., and Browner, R. F., Anal. Chem.. 1980. 52. 576. Ebdon, L.. and Cave, M. R . , Anctlysr. 1982. 107. 172. Babington, R . S . , Pop. Sci., 1973. May, 102. Ripson, P. A. M. R . , and de Galan. L.. Speclrochim. Acra, Part B , 1981, 36, 71. van der Plas, P. S . C . , and de Galan. L.. Spectrochim. Acta, Part B , 1984. 39, 1161. Novak, J . W., and Browner. R . F., Anal. Chem.. 1980,52,792. Scott, R. H.. Fassel, V. A , , Kniseley. R. N . , and Nixon. D. E.. Anal. Chem.. 1974, 46, 75. Greenfield. S . . Ph. D. Thesis. Loughborough University of Technology, 1979. Wendt, R.. and Fassel, V. A.. Anal. Chem., 1965, 37, 920. Boumans, P. W. J. M., and de Beer, F. J . , Spectrochim. Acta, Part B , 1972, 27.391. Olson. K. W., Haas, W. J . , and Fassel, V. A , , Anal. Chem., 1977, 49, 632. Taylor. C. E., and Floyd T. L., Appl. Spectrosc.. 198 1,35.408. La Freniere, K. E.. Rice. G. W., and Fassel, V. A , , Spectro- chim. Acta, Purr B , 1985. 40. 1495. Fassel, V. A., and Bear. B. R.. Specnochim. Acta, Part B , 1986. 41, 1089. Gustavsson, A , . Spectrochim. Acta, Part B , 1984, 39, 743. Garbiano, J. R.. and Taylor. H. E.. Appl. Specrrosc.. 1980. 34, 584. Greenfield, S., and Thomsen, M.. Spectrochim. Acta. Part R. 1986. 41, 677. A Comparison of the Alternating Variable Search and Simplex Methods of Optimisation for Plasma Atomic Emission Spectrometry S. Greenfield. M. S. Salman. M. Thomsen and J. F. Tvson Department of Chemistry/ Loug h borough University b f Techno log yI L oug h boroug h, L eicestersh ire LE113TU The employment of an expensive atomic emission spec- trometer, using the inductively coupled plasma (ICP), in the most efficient and cost-effective manner requires a knowledge of the optimum operating conditions for the plasma in the single-element and simultaneous multi-element analysis rate and coolant gas flow-rate.modes. In previous studies1 it has been shown that measured emission from a plasma source is dependent upon the following parameters: height of viewing, power in plasma, and the three gas flows, namely, nebuliser gas flow-rate, plasma gas flow- The alternating variable search (AVS) method was first used Table 1. Variation in SBR during the optimisation of lead in emission at 405.78 nrn Nebuliserl Coolanti Plasma/ Order of variables I min- I I min - 1 Heighticm PoweriW 1 min- I SBR After 1st cycle 0.9 15 4.5 612 0.5 2.58 After 2nd cycle 1 .o 1 1 4.6 612 0.3 2.97 After 3rd cycle 1 .o 11 4.6 644 0.3 3.01 Starting conditions 0.5 20 3.5 485 2.0 1.0882 2.5 0 .- c.2 2.0 -0 3 2 n 1.5 m 0 c - 1.0 ANALYTICAL PROCEEDINGS, MARCH 1988, VOL 25 - - - - :I! Z 10 5 0 3.0 . 3.0 2.5 - ' 2.5 0 0 .- c .- * 3 2.0 -0 C C 3 3 e 2 0 0 5 1.5 - 5 1.5 CU m n In 0 0 2.0 - c - * m 1.0 - 1.0 - 0, UJ v) m .- .- 0.5 - 0.5 I Q - w 50-54 - - - - - - 1 Optimisation after one cycle, O/O Fig. 1. Optimisation by AVS for all 120 orders 5 10 15 20 0 0.5 1 .o 1.5 In 1960 Rosenbrock3 modified the alternating variable search method. One of his modifications was to recognise that step length in the optimisation was critical and, if chosen correctly, a great deal of time could be saved.Therefore, he suggested that if the optimisation appeared to be moving in a good direction the step length should be lengthened by a factor of three. If a worse result was obtained then a smaller step size, usually halved, should be taken in the opposite direction. Simplex optimisation has been used in analytical chemistry for many years. The original technique was derived by Box and Wilson6 in 1951, then developed by Spendley et al.7 in 1962, and later applied to analytical chemistry by Longs in 1968. Deming and Morganq-10 have helped to clarify the technique for the analyst and the use of simplex optimisation in analytical chemistry has been reviewed by Deming and Parker." A good description and historical account of simplex optimisation has been provided by Betteridge et al.12.13 Nelder and MeadIJ modified the original sequential simplex procedure of Spendley and this version has been most widely applied.The computer has played a large role in the application of this technique and BerridgeIS has reported using a simplex program (written in BASIC) for several years. Several groups of workers in the field of plasma spectrometry have reported using this simplex method, both with a calculator and with a computer program, for optimisation of an ICP. Ikm Experimental The instrument used for this work, which has been described elsewhere,21 is both an atomic fluorescence and atomic by Greenfield and Burns2 for an ICP system. This method of optimisation has been described in full and given a more theoretical treatment by several workers.ss 3.0 > 0.5 t n l 1 I I 3.0 I '0 -2.0 2 0 1.5 n 0, Y c 2 1.0 0 m 0.5 t n l 1 3.0 3.5 4.0 4.5 5.0 5.5 6.0 400 500 600 700 800 900 1000 0 0.5 1 .o 1.5 2.0 Viewing height above coil/cm Power in plasmaiW Plasma gas flow-rate/l min-1 cycle: 0.third cycleANALYTICAL PROCEEDINGS. MARCH 1988. VOL 25 200 3.0 0 5 10 15 20 25 0 83 I I 5 10 15 20 : 6.0 - 5.5 - . 6 r' m 5.0 - .- al Jz 0, 5 4.5 - 3 .- > 4.0 3.5 11 1000 900 800 5 2 700 i - 600 - L a, 3 500 I 400 300 Vertex number 5 10 15 20 25 0 Vertex number Fig. 3. Vertex number I I I Map of simplex optimisation for lead Vertex number emission spectrometer and it was used in the emission mode. The plasma generator is a Radyne SC1S model of 2.4 kW output at 36 MHz and it is free-running.A Fassel torch was used to contain the plasma. and a Meinhard nebuliser and a Scott spray chamber arrangement were used t o introduce samples into the plasma. The emission radiation was focused by means of a 2-in diameter quartz lens on t o the slits of a modified Optica monochromator. The recorded signal was collected and integrated by an Apple IIe computer, and the same computer was used t o run the simplex program. The AVS method of optimisation was carried out in the following manner: firstly. the order in which the variables were to be taken was decided upon: secondly, the boundary conditions were set. These conditions were set at the limits for easy plasma operation; for example.a plasma gas flow-rate of lower than 0.3 1 min-1 causes the plasma tube t o burn, but a plasma gas flow-rate of greater than 2 1 min-1 causes the Table 2. Optimum plasma parameters obtained with the AVS method Element Wavelengthinm I rnin 1 lmin I Pb I 40s. 78 1 .o 1 1 .o Al I 396.15 1 .o 1 1 .o Na I 589.0 1 .5 1 1 .o Ca I 422.67 1 . 2 1 1 .o Ca I1 393.36 9.0 19.0 Mn I 279.18 1 .o 16.0 Sn I 326.23 0.8 16.0 Nebuliseri Coolant! 5 10 15 20 25 Vertex number plasma to be extinguished. The initial step size was that which was easily controllable with the equipment available and caused a measurable change in the figure of merit. The figure of merit chosen for all of the optimisation was signal to background ratio (SBR). A short program was written to facilitate rapid data collection (lo# integration) and automatic calculation of the SBR.The computer program used for the simplex optimisation was kindly supplied by P. Norman and L. Ebdon. and was modified slightly t o incorporate the short program described above to record the SBR as a sub-routine t o be called when the simplex asked for a value of the response factor. The variable names were entered into the computer program with the appropriate ranges. Also entered was the required precision for the termination of the simplex. A precision of 3% was specified as this was directly comparable with the two standard deviation value that was used in the AVS method. Heighticm PoweriW 1.6 644 4.7 739 4.8 549 4.7 612 4.5 676 4.6 54') 4.4 517 Plasmai 1min-1 0.3 2.0 3.0 2.0 1.1 1 .o 2.0 SBR 3.01 4.79 56.68 6.63 3.48 6.61 4.4884 ANALYTICAL PROCEEDINGS, MARCH 1988.VOL 25 Table 3. Optimum plasma parameters obtained with the modified simplex method N e bu I i se ri Coo I an ti PI a s in a: Element Wavelengthinm I rnin 1 min - I Height’cm PoweriW I min - I SBR Pb I 405.78 1 .o 12.8 4.6 644 0.9 2.97 A1 I 396. I5 1.1 13.1 4.8 77 1 1.3 4.61 Na I 589.0 1.4 14.9 4.9 549 1.2 50.25 Ca I 422.67 1.1 13.9 3.9 612 1.1 5.62 Results and Discussion In order to establish whether the sequence in which the five variables are taken had any effect on the optimisation, all possible permutations of the five variables were taken and an optimisation carried out using the AVS method. The resulting 120 optimisations (factorial 5 ) were carried out for lead at 405.78 nm with a 10 p.p.m.solution for the signal and triply-distilled water for the background reading. The results of a typical experiment are shown in Table 1 and the order of the variables in this example was as follows: “injector,” “coolant,” “height,” “power” and “plasma.“ The AVS method was carried out as described above and 106 of the 120 optimisations were completed in three cycles, with the remaining 14 orders requiring four cycles of the optimisation; the SBR still showed a significant improvement after three cycles but only very marginally in each instance. The percentage optimisation after the first cycle against the number of orders is plotted in Fig. 1. This gives an indication of how well the optimisation procedure is doing after one cycle. For most of the experiments the optimisation was at least 65% complete after only one cycle.A complete optimisation experiment took, on average, 2 h to perform with approxi- mately 45 min being spent on the first cycle. The AVS optimisation of lead is mapped for the five variables, “height,” “power,” “plasma,” “coolant” and “injec- tor,” respectively, in Fig. 2. It can be seen that for the final variable, the nebuliser gas flow-rate, the values of the SBR in the second and third cycles overlap as the optimum is reached. These plots of the SBR against the variable for all the cycles of the optimisation give a good indication of the behaviour of the variable and how sensitive the figure of merit is to small changes in the value of each variable. A further six experiments were performed with the AVS optimisation procedure and the summary of the optimum conditions is given in Table 2.Optimisations for all the elements studied were completed within three cycles. 3.0 2.5 2.0 tx m v, $ 1.5 0 Cl al oc 1 .o 0.5 0 I 1 I 1 5 10 15 20 Vertex number Fig. 4. for lead Plot of SBR against vertex number for simplex optirnisation The modified simplex program was used to optimise for a variety of elements and the progress of the simplex is mapped out for lead in emission at 405.78 nm for comparison with the AVS method (Fig. 3). The map of the simplex procedure with respect to the figure of merit, SBR, is shown in Fig. 3 . Twenty vertices were required to reach the optimum conditions: however, several sets of parameters had to be rejected as one or more of the variables lay outside the boundary conditions. 60 50 40 (I m e Q) 0 Q) n 30 a 20 10 0 c , I 1 ~- 5 10 15 20 25 30 Vertex number Fig. 5 . Simplex optimisation for sodium A further three optimisations for different elements were performed with the modified simplex method and a summary of the optimum conditions is given in Table 3. The simplex reached the optimum conditions within a period of approxi- mately 2 h and required between 20 and 25 vertices. The map for sodium is shown in Fig. 5 . from which it can be seen that a good response factor of 52.87 was obtained for the twelfth vertex; however, the simplex went on to a total of 26 vertices before satisfying the criteria for termination achieving an SBR of 50.25 at the optimum. Financial support for MSS from the Iraqi Government and for MT from the SERC and BDH Ltd. is gratefully acknowledged. References 1. Greenfield. S . . Ph. D. Thesis. Loughborough Unixrsity o f Technology. 1979. 2. Greenfield. S . . and Burns. D. T., Aiiril. c‘him. Acra. 1980. 113. 20s. 3 . Box. M. J., Davies. D.. and Swann. W. H.. “Non-Linear Optirnisation Techniques.“ Oliver and Boyd. Edinburgh. 1969. 4. Barley, B. M.. “Studies in Optimisation.” Intertext. Leighton Buzzard, 1974, p. 31. 5 . Bunday. B. D.. “Basic Optimisation Methods.“ Edward Arnold. London. 1984. 6. Box, G . E. P.. and Wilson. K. B.. 1. Roj. Srurisr. Soc.. 1951. 13. 1. 7. Spendley, W., Hext. G . R.. and Himsworth. F. K.. T&- nometrics. 1962. 4. 441.ANALYTICAL PROCEEDINGS, MARCH 1988. VOL 25 8. 9. 10. 11. 12. 13. 14. 15. Long. D. E.. Ariirl. C'liiiti. A m . 1969. 46. 1963. 16. Deming. S. N . . and Morgan. S. L.. A i d C'liein.. 1973.15.278a. Morgan. S. L.. and Deming. S. N . . Aiiril. Cheui.. 1973. 46. 1170. 17. Deminz. S. I-., and Parker. L. R.. CRCCrif. Re\,. Aritil. C'tiem.. 1978. 7 . 1x7. 18. Betteridge. D.. Wade. A. P.. and Howard. A . G.. Trilriritri. 1985. 32. 709, 19. Retteridge. D.. Wade. A . P.. and Howard. A. G.. Tiilririlii. 1985. 32. 723. 20. Nelder. J . A , . and Mead. R.. C'onipiitcv- J . . 1965. 7. 308. Berridge. J . C.. Arid. Proc.. 1983. 20. 39. 21. 85 Ebdon. L . . Cave. M. R.. and Mowthorpe. D . J . . Aiicrl. C'him. Ac.ta. 1980. 115. 179. Tcrhlanchc. S . P.. Viser. K . . and Zeeman. P. B.. Spectrocliirii. Actii, Piirt B. 1981. 36. 293. Leary. J . J.. Brooks. A. E.. Dorrzap. A . F.. and Golightly. D. W.. Appl. Spectrmc.. 1982. 36. 37. Syen. G. S.. Long. S . . and BroLvner. R. F . . Appl. S p m r o s c . . 1986, 40. 236. Ebdon. L.. Norman. P . , and Sparkes. S . T.. Spectrochitri. ACIN. Prrrt R, 1987. 42, 619. Greenfield. S . . and Thomsen. M.. Spectrochirn. Acru, Purr R. 1985. 1369.

ISSN:0144-557X    

DOI:10.1039/AP9882500069

出版商:RSC    

年代:1988

数据来源: RSC

摘要:

ANALYTICAL PROCEEDINGS, MARCH 1988, VOL 25 85 The Effect of Electrolyte Chain Length on Electroendosmotic Flow in High Voltage Capillary Zone Electrophoresis K. D. Altria and C. F. Simpson Analytical Science Group, Department of Chemistry, Birkbeck College, University of London, 20 Gordon Street, London WCIH OAJ When a high voltage is applied across a surface which is in contact with an electrolyte solution a movement of this solution occurs. I This flow is termed electroendosmosis (EEO) and it is utilised in high voltage capillary zone electrophoresis (HVCZE)' to sweep solutes along the capillary to the detecting system. It has been reported' that solutions of the carrier electrolyte. cetyl trimethylammonium bromide (CTMAB). give a negative flow direction with respect to that "normally" obtained in HVCZE.This has been investigated by studying the EEO flow characteristics of a homologous series of CTMAB related cationic electrolytes. Here we report the findings of this study. Experimental The apparatus used has been described previously.3 A voltage of 30 kV was applied across 100 cm x 75 !mi fused silica c a pi 1 I a r i e s sup p 1 i e d by S c i e n t if i c G 1 ass E n gi n e e r i n g , M i 1 ton Keynes, Buckinghamshire. The EEO flows were measured by weighing the mass of buffer transfer. as previously described. 1 The data presented are the means of duplicate measurements. All of the reagents used were of analytical grade and supplied by British Drug Houses (BDII). Poole, Dorset. except for the propyl trimethylammonium bromide and hay1 trimethylammonium bromide, which were prepared by react- ing the appropriate alkyl halide and trimethylamine as supplied by BDH.Solutions were prepared by weight in de-gassed. near conductivity water obtained from a Elgastat Spectrum still. Model No. SC1. High Wycombe. Buckinghamshire. Results and Discussion The results obtained are given in Fig. 1. All log concentration L'er.yu.s flow-rate graphs, for the CTMAB homologues, when extrapolated back to a common point of 0.0151 cm' s-1 V-I? gave correlation coefficients of 0.99 or better. This common point had an identical EEO flow-rate with that from near conductivity water. The observed flow was found to diminish with both e 1 e c t r o 1 y t e con ce n t ratio n and i nc r e ase d h y d r o c ar bo n c h a i n lengths.Electrolytes with longer chains than C-6. passed through a point of zero EEO flow and gave negative flows (Fig. 1). The hydrated surface of a silica capillary has a negative charge, due to surface ions such as ionised silanols. It is suggested that the cationic detergent molecules adsorb on to the capillary wall and present their positive charges into the bulk solution. Molecules with a longer hydrocarbon chain protude their charge further into the solution and shield the negative surface ions more. This shielding results in a decrease in the negative charge and EEO flow velocity. If the surfactant has a sufficiently long chain, and is in sufficient concentration. then the net surface charge becomes positive and the EEO flow direction is reversed.0.002 r 0.001 - > r- w; -0.000 V w n w 2. -0.001 a -0.002 1 L I 1 I L -8 -7 -6 -5 - 4 - 3 - 2 Log concentration - 1 Fig. 1. UEEO \'er.yll.s log concentration: c). cetyl (C-16 : +. tetra (C-14): .. lauryl (C-12): 0. brnzyl (C-6): W , hexyl (C-6): 0. propyl (C-3): A. mcthql (C-1) Conclusions Previous 1 y reported re s u 1 t s h ave bee n investigated and explained. Adsorption of surfactant on to the capillary wall reduces the EEO flow-rate. and for long chain derivatives reverses flow direction. The extent of this flow alteration is dependent on both surfactant concentration and hydrocarbon chain length. Financial support from SERC and Shandon Southern Products Limited is gratefully acknowledged by one of us (K.D.A.). References 1. 2. 3 . Stevens. T. S..and Cortcs. €1. J.. Anal. C'Iwrn.. 1983. 55. 1365. Lukacs, K. D.. and Jorgenson. J . W.. Anal. Chmi.. 1981. 53. 1298. Altria. K. D.. and Simpson, C. F.. AtiuI. Proc.. 1986. 23. 453.86 ANALYTICAL PROCEEDINGS. MARCH 1988, VOL 25 Equivalent Separations on Different Reversed Phases Yook-Ming Lau and C. F. Simpson Analytical Science Group, Birkbeck College, University of London, 20 Gordon Street, London WCI H OAJ There is a wide range of reversed phase packings available in the market produced by a variety of manufacturers. They all produce octadecylsilyl (ODS) bonded phases and it would be thought that on this basis there would be no difficulty in selecting a suitable column material. Unfortunately, there is a wide variation between the separation characteristics of ODS phases produced by different manufacturers and even on a batch to batch basis from a given supplier.Ogan and Katzl have shown that even with the relatively simple separation of polyaromatic hydrocarbons (PAHs) completely different chro- matograms are obtained using nominally the same phases. Various factors can contribute to this variation in separation characteristics, e . g . , the different bonding chemistries used, whether the phase is monomeric or "bulk" modified and even the nature of the silica matrix used can have a profound effect on separations obtained. 3 2 bL 0, -1 1 I I I 1 I 1 I 1 I I 0 1 2 3 4 5 6 7 Carbon no. of benzoates between columns of a disparate nature. The present paper describes experiments designed to evaluate this proposition.Experiment a1 The apparatus consisted of a Waters M-45 solvent delivery system and a thermostatted column fitted with a Rheodyne 7413 valve having an injection volume of 1 pl, while a Cecil CE212 variable wavelength ultraviolet detector was used to monitor the effluent, its output being displayed on a Servo- scribe recorder. The column and detector were thermostatted by circulating the column jacket and detector jacket with water maintained at 20 "C. Prior to entering the Rheodyne valve, the mobile phase was pre-heated to the column operating tem- perature by pumping the supply first through a coil thermo- statted with a water jacket at 20 "C. The two reversed-phase columns employed were a Spheri- sorb S10 ODSl obtained from Phase Separations Limited, Queensferry, Clwyd, and Partisil 10 ODs-2 obtained from Whatman Limited, Kent.These were 25-cm long, 4.6 mm internal diameter, stainless-steel columns. Partisil 10 ODs-2 is 10 ym octadecylsilyl polymeric (15% carbon content) phase, which is fully end-capped, whereas Spherisorb S10 ODs1 is a 10 ym octadecylsilyl (10% carbon content) monomeric phase which is partially end-capped. Mobile phases were prepared by weighing a known mass of modifier on a Mettler AE160 balance into a three-base solvent of methanol - water (25 + 75. 50 + 50 and 75 + 25 m/m, respectively). Modifiers investigated were homologues of alcohols obtained from the Aldrich Chemical Company Limited, Dorset. Alkyl benzoates (C,-C,) were used as test solutes. Methyl and ethyl benzoates were obtained from BDH Chemicals Limited, Poole.Propyl. butyl. pentyl and hexyl benzoates were prepared via the Schotten - Bauman reaction. 0.15 I L I 1 1 I I 0 1 2 3 4 5 6 7 Carbon no. of benzoates Fig. 1. ( a ) , Butanol (g per 100 g) in 50 + 50 m/m methanol - water o n Partisil 10 ODs-2. 0, 0.0; A , 0.005; 0, 0.025; @, 0.05. ( b ) . As ( a ) , but on Spherisorb S10 ODSl In an attempt to control predictably this variation in retention, systems of modifiers in various base mobile phases have been investigated. The effect of chain length of a homologous series of alkanol modifiers, in three different base solvent mixtures of methanol and water, have been studied by using two commercially available reversed phase packings (RPC). It has been shownz-3 that modifiers in the mobile phase can be used to determine the chromatographically available surfaceholume within a column.Conversely, because the action of a modifier is to be adsorbed on the reversed-phase medium, it is possible that columns could be modified to present the same surface area/volume for interaction with solute molecules and hence to obtain equivalent separation D & A ' 4 I 1 1 0 0.01 0.02 0.03 0.04 0.05 ( Butanol in 50 + 50 m/m methanol - water, O/O m/m 06 Fig. 2. ODs-2. 0, Ethyl benzoate; A . hexyl benzoate Solute retention of ethyl and hexyl benzoates o n Partisil 10 Results and Discussion Effect of Different Modifier Systems The ternary mobile phase systems investigated consisted of various low concentrations of a modifying agent added to a base solvent. Figs. l ( u ) and l(h) show the results obtained on chromatographing a series of alkyl benzoates both with and without modifiers on Partisil 10 ODs-2 and Spherisorb S10ANALYTICAL PROCEEDINGS, MARCH 1988, VOL 25 - 87 ( iii) 0 0.01 0.02 0.03 0.04 0.05 0.06 0 0.01 0.02 0.03 0.04 0.05 0.06 "/o mim of alcohol in 25 + 75 mim methanol - water o/o mim of alcohol in 25 + 75 mim methanol -water I ( iii) I I 1 1 I I 1 1 L 1 L 1 I 0 0.01 0.02 0.03 0.04 0.05 0.06 0 0.01 0.02 0.03 0.04 0.05 0.06 o/o m/m of alcohol in 75 + 25 m/m methanol - water o/o mim of alcohol in 75 + 25 m/m methanol - water Fig.3. ( u ) . Alcohols in methanol - water on Partisil 10 ODs-2; solute hexyl benzoate. 0. Ethanol: A , propanol; 0, butanol; a, pentanol; H, hexanol. (i), Methanol - water 25 + 75 mlm; (ii), methanol - water 50 + 50 m/m; (iii).methanol - water 75 + 25 mlm. ( b ) , As ( a ) , but on Spherisorb S10 ODSl ODSl, respectively. The net effect of including the modifier in the mobile phase is to reduce the retention of the solutes by reducing the surface area or pellicular volume available1 of the reversed phase for interaction with solute molecules. It is also clear that these graphs obey the Martin equation4 for retention of homologues. Thus. solute retention can be modified by including a limited amount of a longer chain organic molecule into the mobile phase. and this is strongly adsorbed on to the reversed phase in order to provide a stable system. Further, the retention volume of the solutes is a substantially linear function3 of the modifier content of the mobile phase.Fig. 2 shows the retention of ethyl and hexyl benzoates on Partisil 10 ODs-2 in a mobile phase consisting of an increasing percentage mass of butanol in 50 + SO m/m methanol - water. As the alkyl chain length of the modifier is increased, progressively smaller concentrations of modifier are needed to achieve the same degree of solute retention. This is shown in the series of graphs in Figs. 3(a) and 3(b) by the reduction in retention of hexyl benzoate as the solute in a mobile phase consisting of various homologous alcohols in 25 + 7.5, 50 + 50 and 75 + 25 mlm, respectively, of methanol - water on Partisil 10 ODs-2 and on Spherisorb S10 ODS1. A study of Fig. 3(a,i) shows that 0.05% mlm of ethanol is required to give the same retention as 0.005% of pentanol. This is because the adsorption strength of ethanol is relatively weak compared with pentanol in the mobile phase and hence the degree and stability of the surface for solute interaction is less than with hexanol.The selection of an appropriate modifier system for equi- valent separation can potentially be achieved by including a modifier at a specific concentration which will provide equivalent surface area/volume in the two columns for the solute molecules to interact. By overlaying the individual graphs in Figs. 3(a) and ( b ) for the two columns, the points of intersection of the lines represent the concentration of modifier that is required in the methanol - water system for equivalent separation. Fig. 4. ( a ) . Column, Partisil 10 ODS2; mobile phase.methanol - water (75 + 25 rn/rn); flow-rate. 1 ml min-1; temperature, 20 "C; sample, homologous series of alkyl benzoates (Cl-C6). ( b ) , As ( a ) , but on Spherisorb S10 ODSl Fig. 4 shows the difference in separation of alkyl benzoates on the two columns using 75 + 25 mlm methanol - water. The separation obtained from the Spherisorb column is too fast.ANALYTICAL PROCEEDINGS. MARCH 1988, VOL 25 Fig. 5. 0.025% mlm hexanol ( h ) ; flow-rate, 1 ml min- I ; temperature. 20 "C; sample, homologous series of alkyl benzoates (C,-C,) Column. Spherisorb S l 0 ODSl; mobile phase, methanol - water ( 5 0 + 50 mirn) ( a ) . methanol - water ( 5 0 + 5 0 nzirn) modified with Therefore, the methanol content is initially reduced to 50 + 50 m/m, which Fig. 5(a) shows to be too slow a separation. By the appropriate addition of the type and concentration of modifier to the methanol - water system, it is possible to reduce retention in a chromatographic run to a reasonable time.On the addition of 0.025% mlm of hexanol to the 50 + 50 m/m methanol - water system, the retention of solutes is reduced to a reasonable time, as is shown in Fig. 5 ( 6 ) . Modifiers in the appropriate methanol - water content which give identical slopes for both columns in the graph of the log Vrr against the carbon number of a solute are chosen, The retention of solutes on a Partisil column in Fig. 6 is similarly reduced by the addition of 0.025% m/n7 of pentanol to the 75 + 25 m/m methanol - water system. i; Fig. 6. Column, Partisil 10 ODS2; mobile phase, methanol - water (75 + 25 mlm) ( a ) .methanol - water (75 + 25 mlm) modified with 0.025% mim pentanol ( b ) ; flow-rate, 1 ml min I : temperature, 20 "C; sample, homologous series of alkyl benzoates (C,-C,) At that particular slope, the graph of the log Vr, (Spherisorb) against log Vr, (Partisil) should give a slope close to unity. The comparison of chromatograms in Fig. 7 of the respective modifier systems on the two columns shows equivalent separations between Spherisorb and Partisil. Conclusions Solute interaction decreases with increasing modifier concen- tration. Further, solute retention is a linear function of the mobile phase composition. On increasing the alkyl chain length of the modifier, progressively smaller concentrations are needed for the same degree of interaction.a) L Fig. 7. Sample. homologous series of alkvl benzoates (C,-C,,); temperature, 20 "C; flow-rate. 1 ml min mobile phasc. methanol - water (75 + 25 1dm) modified with 0.025% tnirn pentanol ( ( 1 ) . methanol - water (50 + 50 tnim) modified with 0.025'% mim hexanol ( b ) : column. Partisil 1 0 ODS?. (0). Spheriwrb SIO ODSl ( h ) Solute interactions on octadecylsilyl (ODS) reversed phases are variable. By varying the concentration and alkyl chain length of the modifier employed in the mobile phase system. an equivalent surface arealvolume is provided in the two columns for solute molecules to interact.ANALYTICAL PROCEEDINGS, MARCH 1988, VOL 25 89 References 3. Scott, R. P. W . , and Simpson, C. F., Faraday Symp. Chem. SOC. 15, 1980, 69.4. Martin, A. J . P., Biochern. Soc. Symp., 1949, 3, 4. 1. Ogan, K.. and Katz, E.. J . Chromatogr., 1980, 188, 115. 2. Beezer, A. E., and Simpson, C. F., in the press. Factors Affecting Detection Limit in Flow Injection Solution Spectrophotometry Andrew 8. Marsden and Julian F. Tyson Department of Chemistry, Loughborough University of Technology, Loughborough, Leicestershire L E 1 1 3TU There are two factors that control the detection limit in any analytical procedure, the magnitude of the signal and the magnitude of the noise. The strategy for reducing the detection limit is, therefore, to maximise the signal and minimise the noise. As far as flow injection (FI) solution spectrophotometry (SS) is concerned, the magnitude of the signal is governed by the extent of dispersion produced by the particular manifold used.Most FI - SS procedures involve the on-line generation of an absorbing product from the reaction between the injected sample and a reagent stream. There has to be sufficient mixing between the sample and reagent to ensure complete reaction across the entire sample profile. Thus, the requirement is to design the dispersion produced by the manifold to give this degree of mixing without unnecessary dilution. Several sources of noise exist in any flow injection system. In decreasing order of magnitude these are pump roller noise, mixing noise (arising from incomplete merging of the carrier and reagent streams) and electronic noise (from the detection and recording devices). The second factor, which only arises with manifolds containing confluence points.sometimes can also be caused by the peristaltic pump. There are some minor factors which can also affect the limit of detection in flow injection analysis (FIA), such as the geometry of the confluence point and the presence of refractive index gradients. In this paper, strategies for reduction of base-line noise in the spectrophotometric determination of chloride I using a merging stream manifold are discussed. There are several benefits in using a merging stream manifold as opposed to a single-line manifold. These include the elimination of negative peaks at low sample concentrations (when the reagent absorbs at the analytical wavelength), matrix matching of the carrier stream, variation of the respective flow-rates of reagent and carrier to suit the conditions of the reaction being used and the removal of refractive-index effects by the injection of a large volume.This last strategy, as well as allowing the detection of the product while the refractive index gradients on the leading and trailing edges are outside the field of view of the detector, also maximises the sensitivity as the dispersion is governed only by the relative flow-rates of the sample carrier and reagent carrier streams. These can be optimised with regard to sample dilution at the confluence point, the requirement of excess reagent to ensure sufficient reaction and the sample throughput rate. Such a strategy is not possible with a single-line manifold, a5 injection of a large volume leads to the formation of a double peak due to lack of reagent in the centre of the sample zone.’ Experimental Apparatus The basic flow injection system used is shown schematically in Fig.1. A Cecil Instruments CE2373 ultraviolet - visible spectrophotometer fitted with an 8 PI flow cell (Hellma) was used as the detector, and the manifold consisted of 0.5 mm and 1.0 mm i.d. PTFE tubing throughout. A Microperpex peristal- tic pump (LKB) was used, fitted with a second pump head, In order to achieve different flow-rate ratios between the reagent and carrier lines, silicone rubber pump tubes of various internal diameters were used. The total flow-rate used was 2.5 ml min-1. The volume injected was 700 pl, as this eliminated troublesome refractive index effects; a Rheodyne 5020 sample injection valve was used to inject chloride samples into the carrier stream.Samples were prepared by serial dilution of a 1000 mg 1-1 stock aqueous solution of chloride (1.65 g 1 - 1 of NaC1). The chloride reagent had the following composi- tion: mercury(I1) thiocyanate, 0.625 g 1-1; iron(I1I) nitrate, 30.30 g 1-1; nitric acid, 3.3 g 1 - 1 ; and methanol, 15% (VIV). n..-- Reagent I 4 3.0 Flow cell Carrier (water) Waste - 1.3 mm i.d. Sample Fig. 1. Manifold for chloride determination Base-line Noise Reduction As indicated earlier, the base-line noise is caused by several factors, the most important of which is pump roller noise. The peristaltic pump is not entirely pulse-free, and these pulsations give rise to periodic base-line fluctuations. Simple pulse dampers were fitted to each line immediately after the pump in an attempt to reduce some of the roller noise.These pulse dampers consisted of a small glass T-piece, the longer arm of which was allowed to fill with air when connected into the manifold. As the air column was compressed, the effect was to damp down the rhythmic pulsations of the flow. As the pressure increased within the manifold the air column was compressed inside the tube. There was a slight delay before the next pump roller made contact with the tubing. thus the pressure decreased in the manifold and the previously compressed air column acted on the fluid to restore the flow. The whole cycle was then repeated. The next source of background noise in a manifold of the type investigated is the mixing noise which arises from the incomplete merging of the carrier stream with the reagent stream at the confluence point.This incomplete homogenisa- tion of the two streams is thought to be caused by the pump rollers being slightly out of phase, resulting in a type of “segmentation” of the two streams at the confluence point. The different densities of the streams meant that mixing was made more difficult. Several designs of confluence point were investigated, in an attempt to discover which geometry was most effective in reducing base-line noise.3 The designs included a Y-shaped 120” confluence, and a T-piece with 0.8 mm bore. In addition to the basic manifold (50 cm of 1.0 mm i.d. PTFE tubing), various devices were inserted into the manifold downstream of the confluence point to aid the mixing by homogenisation of the stream segments.The first of these was a small packed-bed reactor (PBR). This reactor was construc-90 ANALYTICAL PROCEEDINGS, MARCH 1988, VOL 25 ted from a length of glass tubing (4 cm), of internal diameter 1.5 mm. It was filled with glass beads 0.3-0.35 mm in diameter. This reactor was inserted into the manifold in an attempt to provide a so-called tortuous path for the streams merged at the confluence point. The second device constructed to mimimise mixing noise consisted of a tube of PTFE packed with a single line of glass beads (ballotini). The internal diameter of the tube was 0.58 mm and the diameter of the beads in the range 0.3-0.35 mm. Thus, the beads pack the tube in single file, forming a single bead string reactor (SBSR)."S The third de- vice was a tightly coiled wide bore (1 .O mm) open tube reactor placed immediately after the PBR; finally, a Visco Jet Micro Mixer (Lee products) was used.(b' 10.004 A H 60 s Fig. 2. base-line noise using a 120" angle confluence point (a) Base-line noise using a T-piece confluence point; ( b ) Results and Discussion Regarding the confluence geometry, more extensive work is warranted, but it does appear that a Y-shaped device is preferable, see Fig. 2. Future work may include an investiga- tion of confluence points with acute angles between the inlet and outlet. The mixing noise due to poor homogenisation of the samplekarrier and the reagent stream was greatly reduced by the use of the packed-bed reactor. The base line noise reduction can be seen in Fig. 3.However, the use of the PBR alone did not seem to be completely effective. As can be seen from Fig. 4, the use of the wide-bore open-tube reactor placed immediately after the PBR produced a marked reduction in the noise. The tight coil is believed to introduce secondary flow 60 s U T 0.004 P 1 Fig. 3. only Level of base-line noise obtained using a packed-bed reactor patterns within the stream, thus thoroughly mixing the effluent from the PBR. The dispersion of PBR and SBSR is not excessive, as the mixing effect is caused by the increased residence time over the usual open tube reactors.6 0.004 A I Fig. 4. Base-line noise level obtained by using a packed-bed reactor and a 50 cm length, I .0 mm i.d., coiled open tube Initially, the SBSR did not produce improved mixing over that produced by the PBR; this is most likely a result of the development of higher back pressures.The pump used is incapable of working effectively against more back pressure than that generated by the PBR described. A more robust peristaltic pump will be used in future work. Similar levels of noise reduction to those obtained with the'PBR have been reported for a 25 cm coiled SBSR.7 The Visco Jet Micro Mixer was compared with the PBR and 50-cm coil, Fig. 5. As can be seen, the performance of the mixer was disappointing. The difference in density of the two liquids may be the reason why the mixer did not work very well. The development of the manifold for the maximum signal to noise ratio by using the techniques described earlier has led to an improved detection limit, from 1 mg 1-1 initially to 1 ng 1-1 for chloride in water.I T 0.004 A II V Fig. 5. the confluence point. Base-line noise obtained using the Visco Jet Mixer in place of Conclusions It is possible to reduce the base-line noise encountered in an FI - SS procedure based on a manifold consisting of simple open-tube reactors. The most effective method of reducing the noise, and thus improving the detection limit. was found to be the use of the packed-bed reactor in conjunction with the wide-bore, coiled, open-tube reactor. Both devices are placed after the confluence point. Financial support for this work from the SERC and BDH Chemicals Ltd. (who generously supplied the chemicals and the Visco Jet Mixer) through the CASE scheme is gratefully acknowledged.References 1. RGiiCka. J . , and Hansen, E. H.. Aizal. Chirn. Acra. 1976. 87. 353. 2. Tyson. J. F.. Arial. Chim. A m . 1986. 179. 131. 3. Silfwerbrandt-Lindh. C.. Nord. L.. Danielsson. L. G.. and Ingman, F.. Anal. Chim. A m . 1983. 160. 1 1 . 3 . Reijn. J. M., and Poppe, H.. Atzal. Chim. Acta. 1983. 145. 59. 5 . van den Berg. J . H. M.. Deelder. R. S . . and Egberink. G. M.. Anal. Chim. Acrii. 1980. 114. 91. 6. Reijn, J . M.. van der Linden. W. E.. and Poppe. H.. Aiid. Chim. A m . 1981. 123. 229. 7. Patton. C. J.. and Crouch. S . R.. Aiiul. C'hhi. A m . 1986. 179. 189.ANALYTICAL PROCEEDINGS, MARCH 1988, VOL 25 91 Problems Associated with the Analysis of Hydrogen Sulphide Using the Ethylene Blue Method Christine F.Wood and lain L. Marr Department of Chemistry, University of Aberdeen, Meston Walk, Old Aberdeen AB9 ZUE The ethylene blue method, used for the colorimetric determi- nation of traces of hydrogen sulphide in air, was developed from the earlier methylene blue method by Docherty et al. in 1971.1 Ethylene blue is formed by reacting sulphide with p-diethylaminoaniline in an acidic medium, and oxidising the product with iron(II1) to form the dye. Use of the diethyl analogue has advantages over the methylene blue reaction because of its increased sensitivity and stability. However, problems began to arise in our laboratory when the method was used for an environmental study by a succession of workers over several years. It was noticed that the slopes of calibration graphs varied often by as much as a factor of two, although the individual researchers were capable of achieving good repro- ducibility after a certain amount of practice.It appeared, therefore, that accidental variation of the level of one or more of the constituents was a possible reason for these discrepan- cies. A study of the various stages of the reaction leading to the formation of ethylene blue revealed that changing the amounts of absorbant (zinc acetate solution) or oxidant [ammonium iron(II1) sulphate] had little effect on the reaction, but that variation in the concentration of sulphuric acid used as the solvent for the p-diethylaminoaniline reagent caused large changes in the final absorbance value. Having found that, in principal, the pH at the time of the reaction between the p-diethylaminoaniline and the sulphide (present as zinc sulphide) was crucial to the final absorbance reading, it was decided to extend the investigation to find the best conditions for the reaction.Experimental Instrumentation 10 mm glass cells. Spectrophotometer. Pye Unicam SP6-550 instrument with Solutions Reagent. Dissolve 0.156 g of p-diethylaminoaniline in 50 ml of sulphuric acid of varying concentration. Oxidant. Dissolve 45 g of ammonium iron(II1) sulphate in 1 1 of 1% V/V sulphuric acid. Absorbent. Dissolve 25 g of zinc acetate in 1 1 of distilled water. Dilute by a factor of 10 to give a working solution. Standard sulphide solution. Wash some crystals of sodium sulphide in ethanol and dry them carefully. Weigh out accurately 3.75 g and dissolve them in 0.1 M sodium hydroxide solution.Make up to 1 1 in a standard flask. Dilute 1 ml of the stock solution to 500 ml to give a working solution of 2 pg ml-1 of sulphide. Results and Discussion Experiment 1 Known amounts of sulphide were added to a series of 25-ml standard flasks, each containing approximately 23 ml of zinc acetate solution. To each flask was added 0.40 ml of reagent solution, every one being prepared in a different concentration of sulphuric acid. The mixtures were shaken and left to stand for 1 min. Next. 0.25 ml of oxidant was added to each flask, the mixtures were shaken again, and then left for 5 min. The absorbances of the resulting solutions were measured at 670 nm in a 10-mm cell. The results are illustrated on Fig.1. The study indicates that the maximum absorbance is obtained when the pH of the reaction mixture is 1.65. However, as the graph also indicates, the amount of sulphuric acid added is extremely critical to the final absorbance reading. To obtain an error of 5% or less in the absorbance value, it is necessary to keep the solution at pH 1.65 + 0.10. 0.5 0.4 0 C (D e 2 n a 0.3 0.2 0 I I I 0.02 0.04 0.06 0.08 0.10 Concentration of H2S04 in final solutionlhn Fig. 1. sulphide Graph of absorbance versus final acid concentration for 8 pg of 0.3 I 0.2 er t lu e n a v) 0.1 0 1 2 3 4 5 6 7 8 9 1 0 1 1 1 2 1 3 1 4 Development time/min Fig. 2. Effect of acid concentration on the rate of formation of' ethylene blue. Reagent dissolved in: 1. 50%; 2, 30°/"; 3 , 20%; and 4, 8% V/V aqueous sulphuric acid92 ANALYTICAL PROCEEDINGS, MARCH 1988.VOL 25 Experiment 2 The development of the colour of some solutions with varying sulphuric acid concentration was followed by plotting absorb- ance ver.su.y time (time = 0 when the oxidant was added to the reaction mixture). The results are illustrated on Fig. 2. Experiment 2 shows that the time required for complete colour development depends on the concentration of sulphuric acid. It appears that, as the concentration of sulphuric acid in the reaction mixture is decreased, the development time also decreases. Thus, it can be seen that the pH of the reaction mixture affects both the absorbance obtained and the time required for the colour to develop. This would explain the comparatively poor reproducibility of the method, as a slight change in the acid concentration can cause a significant change in the sensitivity.However, it is now possible to use the method with confidence, provided that great care is exercised in the preparation of reagents and their addition to the reaction mixture. Once formed, the ethylene blue will remain stable for several days. With careful addition of reagents the precision is satisfactory (RSD = 2.9% for 11 samples, with absorbances ranging from 0.391 to 0.429 for 5 pg of sulphide in a final volume of 25 ml). Attention is now being paid to the feasibility of modifying the method to include control of the pH of the reaction mixture with the use of an appropriate buffering system. References 1. Rees, T. D.. Gyllenspetz, A. B..and Docherty, A. C., Analysr. 1971. 96. 201. Ore and Mineral Analysis by Slurry Atomisation - Plasma Emission Spectrometry Michael E. Foulkes, Les Ebdon and Steve Hill Department of Environmental Sciences, Plymouth Polytechnic, Drake Circus, Plymouth PL4 8AA Trace elemental analysis of ores and minerals is often difficult to achieve. Typically. such samples are resistant to simple dissolution procedures. The necessity for using solubilisation techniques that are hazardous, time consuming and subject to problems of contamination and analyte losses has led to an interest in the direct analysis of solid samples. Plasma emission spectrometry offers a promising route for solids analysis, as analytical plasmas offer the advantages of a high-temperature source, suitable for matrix destruction, a rapid multi-element capability, long linear working ranges and detection limits in the nanogram per gram range.A number of techniques have been used to introduce solids into plasmas'-3 but many suffer problems such as the need for matrix matching of calibration standards or the need to modify instrumental hardware. Analysis using an inductively coupled plasma (ICP) is traditionally performed using a solution introduced via a nebuliser - spray chamber arrangement. One way of retaining the normal sample introduction system when analysing solid samples is to use a suspension of finely powdered solid sample, the slurry technique, and this also offers the potential for calibration using simple aqueous standards and essentially conventional instrumentation.Lx Instrumentation Two computer controlled, rapid, sequential ICP spectrometers were used in the study (S-35, Kontron Spectralanalytik.Eching, FRG, and 35000, ARL, Luton, Bedfordshire). The former was fitted with a Greenfield diameter torch (29 mm 0.d.) and the latter with a Fassel diameter torch (20 mm 0.d.). Both torches were of the demountable form with 3 mm i.d. injector tubes. All analyses were performed by using an unblockable V-groove nebuliser (PS Analytical, Orpington, Kent) and Scott-type double-pass spray chamber. Slurries were maintained in suspension by use of a magnetic stirrer and fed to the nebuliser by means of a peristaltic pump. The particle size of the sample is known to be of major importance in slurry atomisation.6.7 The time taken for a solid matrix to form atoms depends not only on the temperature of the atomising cell but also on the size of the matrix particle.When nebulisation is used to introduce the slurry particles to the atomising cell this size-related effect is compounded by the efficiency of the transport system. A suitable comminution process is therefore required for most ores and minerals. 40 1 ', s 6 E < 20 > 10 n 2 4 6 8 10 12 14 Particle sizeipm Fig. 1. The effect of grinding time on the particle size distribution of an in-house ore (bottle and bead method): --> 0; - - -. 0.5; - - - -. 1; , 3 hour . _ . _ Slurry Preparation Two wet grinding techniques were used to produce fine slurries directly from ores and minerals. Firstly. a bottle and bead method6 which used 2.5 mm zirconia beads (Glen Creston.Stanmore, Middlesex) shaken with the samples (in the ratio 10: l ) , in an aqueous solution of a suitable dispersing agent within a polythene container, using a laboratory flask shaker. Secondly, a micronising method which used cylindrical agate media vibrated with the sample in a polythene vessel using a commercial microniser (McCrone Research Associates, Lon- don). Both comminution processes were performed in a small Table 1. Operating conditions used on Kontron S-35 ICP for analysis of CANMET CRM sulphide reference ores Torch type Greenfield Gas flows/l min ~ I Coolant 16 Auxiliary 0.6 Injector 1.9 Injector dia./mm 3 Forward powerJkW 1 . S Nebuliser type V-GrooveANALYTICAL PROCEEDINGS, MARCH 1988, VOL 25 93 Table 2.Slurry atomisation results for CANMET orcs using aqueous standards. Values in parentheses are indicative. uncertified. values only CANMET CRM ore KC- 1 a CANMET CRM ore MP- l a Certified. Yo Slurry, Yo Certified, '/; Slurry. o/o c u 0.629 L 0.015 0.45 * 0.01 c u 1.44 t 0.01 0.99 k 0.01 Zn 19.02 i 0.10 16.7 2 0.20 Zn 34.65 f 0.15 33.7 k 0.50 Pb 4.33 t 0.03 3.7 k 0.05 Pb 2.24 k 0.0-3 2.27 +- 0.03 Sn 1.28 i 0.03 0.86 k 0.02 Sn 0.61 k 0.02 0.44 k 0.02 As 0.84 t 0.02 0.74 * 0.02 Ag 0.167 ? 0.002 0.1 I k 0.005 Fe (6.2) 4.9 * 0.10 Mo 0.029 k 0.001 0.004 k 0,0004 ugg- ' Fe (10.9) 8.6 t 0.20 Ag 69.7k2.2 4s t 4 volume of water containing a suitable dispersing agent, e.g., 0.1% mlV sodium hexametaphosphate, to stabilise the slurry. The particle size distribution with grinding time for a typical sulphide ore using the bottle and bead method is shown in Fig.1. The micronising method produced a similar slurry distribution t o the 1 h grinding period shown in Fig. 1 in only 10-15 min. Under these conditions greater than 95% (by volume) of the particles were under 8 pm in size. Contamina- tion was kept to a minimum by using grinding media with a measure of hardness greater than that of the sample. 15 1 1 0.5 1 .o 2.5 6.3 16 25 Particle size/prn Fig. 2. Change with injector bore of slurry particle size distribution delivered to the plasma. Constant injector gas flow-rate 1.7 1 min-1. - . _ . Starting slurry (coarse); -, 2.2 mm injector; - - - -, 3.6 mm injector Transport Studies The size distributions of slurry particles issuing from various injector tubes of different internal diameters at various gas flow-rates were measured on a test bed comprising a V-groove nebuliser, a double-pass spray chamber and an injector tube.The particle size distribution of a slurry that enters the analyte channel of a plasma clearly differs from that of the starting slurry (Fig. 2) and the distribution was found to be dependent upon the carrier flow-rate and the bore of the injector (Fig. 2). It can be seen from the particle size distribution of material which passed the injector that the maximum particle size of the starting slurry should be reduced to below 8 pm, if representa- tive transportation of the sample is to be achieved. u l I ~ 1 . 2 1.5 1.9 2.4 3.0 3.9 5.0 6.4 8.2 10.5 13.6 17.7 23.7 Particle size/pn Fig.3. In-flight particle size distribution of ore slurry and solution exiting from double pass spray chamber: - - - -. solution: -. slurry. Injector gas flow-rate. 1.5 I min I If calibration by aqueous standards is to be used the transport properties of the slurry should ideally model those of a solution. Laser diffraction studies of in-flight slurry particles (ore particles within an aqueous envelope) that exit from a V-groove nebuliser and those from a double-pass spray chamber show little difference in size distribution from those of solution aerosols under the same conditions (Fig. 3). Table 3. Plasma operating conditions used for the determination of copper and silver in ore slurries KC-la and MP-la Forward powerlk W Gas flowsil min Coolant Auxiliary Injector Injector boreimm Viewing heightimm *g c u Nebuliser type 2.0 18 0.6 2.6 3.0 18 10 V-Groove _ _ _ ~ ~ ~ - Slurry Analyses Certified reference material ores, namely CANMET sulphide reference ores KCla and MPla (CANMET project, 555 Booth Street, Ottawa, Ontario, Canada) have been analysed by slurry atomisation by using the conditions shown in Table 1.The results by slurry atomisation with simple aqueous standards are shown in Table 2. The ratio of found t o certified concentration, expressed as a percentage recovery. is also shown. These recoveries correlate well with the elemental heats of atomisa- tion, as is shown in Fig. 4. While atomisation efficiencies of I uu Zn Pb 9 0 1 : As 80 t Al 0 Fe 70 - 8 8 60 - 5 50- [r 40- a, W 30t 2o 10 i Mo 0 I 0 100 200 300 400 500 600 7 Ay,/kJ rnol-l 0 Fig.4. versus heat of atomisation Percentage recovery of elements in CANMET CRM ores several elements in these ores are lower than those recorded from solutions of the same elements, plasma operating conditions have been determined (Table 3) where the viewing94 ANALYTICAL PROCEEDINGS, MARCH 1988, VOL 25 height emission profiles of copper and silver in the ore slurries are equivalent to those in solutions (Figs. 5 and 6). Full recoveries can therefore be obtained under these modified operating conditions. 7 I m ln C c ; 104 u > ln al C . c .- c .- '0 103 E .- m .- Lu I I t Background 7 60.0 1 02 0 20.0 40.0 Viewing heightlmm Fig. 5. tions of copper in solution and slurry (KC-la) Viewing height emission profiles for equivalent concentra- Slurry particles of 8 pm and less have been shown to reach the plasma by use of the conventional transport system.Alumina slurries of known composition, with size distributions exceeding 90% (by volume) less than 8 pm, but of various size fractions, have been analysed for their aluminium content by using the normal operating conditions shown in Table 4(a). Table 4. Operating conditions used for the analysis of alumina slurries ( a ) ( b ) A1 recovery, "10 74 90 PowerIkW 1.5 1.5 Height/mm 20.0 6.4 Injector gas11 min- 2.0 1.2 Auxiliary gas11 min- I 0.6 0.6 Coolant gad1 min-' 17.0 17.0 ~~ The recoveries obtained are shown in Table 5 . The ultra-fine alumina (100% less than 2.5 pm) gives 99% recovery and indicates that the atomisation efficiencies of slurry and solution are very close, The atomisation efficiency of a much coarser alumina slurry can be improved, relative to solution, by using a medium to high power plasma (1.5 kW) and low carrier flow-rates (1.2 1 min-1) through a 3 mm bore injector.The reduced gas velocity allows an increase in residence time for slurry particles in the plasma and recoveries have been increased from 74 to 90% using the conditions shown in Table 4 w . Table 5. Variation with particle size of the ratio of aluminium signal from an alumina slurry to the signal from an aqueous aluminium solution of the same concentration ( E , YO) Percentage of particles less than 8 , 5 , 3 and 2.5 pm, respectively E , O/o 8 5 3 2.5 99 100 100 100 100 74 94 88 69 47 55 93 77 46 25 Real samples of in-house sulphide ores were analysed using the conditions shown in Table 6 and these results are shown in Table 7.Conclusions Slurry atomisation plasma emission spectrometry is a viable lo5 7 I I In ln c 5 104 s . c > ln C al C .- c .- 103 .- m ul .- ; I Background I I 102 ' I I I 0 20.0 40.0 60.0 Viewing heightlmm Fig. 6. Viewing height emission profiles for equivalent concentra- tions of silver in solution and slurry (KC-la) technique for the analysis of ores and minerals. The choice of plasma operating conditions is important, as is the particle size of the slurry, which needs to be less than 8 pm for efficient recoveries to be obtained. This particle size range can be Table 6. Plasma operating conditions used on ARL 35000 ICP for analysis of in-house sulphide ore slurries Torch type Forward powerIkW Gas flows4 rnin-1 Coolant Auxiliary Injector Injector dia./mm Viewing height/mm Nebuliser type Fassel 1 .5 16 1.5 1.7 3.0 V-Groove 15 achieved rapidly for an ore by means of the treatments outlined above.Minimal changes to conventional plasma hardware are needed for slurry analysis to be performed and it is an inexpensive technique to implement. Table 7. Slurry and solution analysis of in-house sulphide ore Slurry, Yo Solution. '/o c u 0.337 2 0.02 0.37 k 0.005 2.6 k 0.1 Zn 2.41 k 0.08 Fe 13.7 k 0.06 18.2 2 0.8 Pb 0.28 2 0.02 0.27 k 0.03 Clgg-' Zr 90 k 4 94 k 9 A!? 119k 16 119k 14 The authors gratefully acknowledge the financial support by the SERC and BP under the CASE scheme to one of us (M. E. Foulkes) which has made this work possible.References 1. 2. 3. 4. 5. 6. 7. 8. Ng. K. C., Zerezghi, M., and Caruso. J . A.. Anal. Chem.. 1984, 56, 417. Thompson, M., Goulter, J . E.. and Sieper, F., Analyst. 1981. 106, 32. McLeod, C. W., Clarke, P. A , . and Mowthorpe. D. J . , Specrrochim. Acta, Part B, 1986,41. 63. Wilkinson, J . R., Ebdon, L., and Jackson, K . W . , Anal. Proc., 1982, 19, 305. Watson, A. E., and Moore. G. L., S. A f r . J . Chem.. 1984.37, 81. Sparkes, S. T., and Ebdon, L.. Anal. Proc., 1986. 23, 410. Ebdon, L., and Wilkinson, J . R.. J . Anal. A t . Specrrom.. 1987, 2, 39. Ebdon. L., and Wilkinson. J . R.. J . Anal. A t . Specrrom.. 1987. 2, 325.ANALYTICAL PROCEEDINGS, MARCH 1988, VOL 25 95 Development and Application of Novel Electrothermal Atomisation Atomic Absorption Spectrometry Procedures of Relevance to the Electricity Supply Industry A.M. Quinn, R. Nichol, L. Miller and D. Littlejohn Department of Pure and Applied Chemistry, University of Strathclyde, 295 Cathedral Street, Glasgow G I IXL G. B. Marshall Central Electricity Generating Board, Central Electricity Research Laboratories, Kelvin Avenue, Leatherhead, Surrey, KT22 7FE The CEGB, in common with many industries, is interested in the development and application of methods for the rapid and sensitive determination of trace elements in a variety of materials, such as waters, ashes, particulate matter, soils and plants. Power-station fly ash can be used in land reclamation, so methods of analysis are required for the determination of toxic trace metals in both the ash and crops grown on the reclaimed land.In these and many other analyses, it would be desirable to minimise sample treatment to achieve almost direct determination of analyte elements. A number of electrothermal atomisation atomic absorption spectrometry (ETA-AAS) procedures have been evaluated at the University of Strathclyde to assess their suitability for industrial analysis. Three procedures for the direct analysis of solids, the cup-in-tube,‘ the ring chamber tube2 and slurry sample introduction,3.4 have been evaluated for the direct determination of lead in powdered kale by ETA-AAS. In a separate study, a heated probe atomiser has been designed and constructed for use in ETA-AAS with a view to reducing the matrix interferences frequently encountered in the analysis of foodstuffs and environmental samples by conventional pro- cedures.Comparison of Three Procedures for the Direct Analysis of Solids One of the major features of ETA-AAS is that the sample is almost always analysed in solution. If samples such as dried plant materials could be analysed in the solid state, extensive sample preparation times would be eliminated and any contamination which may occur during a dissolution process would be reduced. The sample used throughout the investiga- tion was Bowen’s kale, a certified reference material. This sample was selected because it was typical of a crop that could be grown on reclaimed land. Lead was chosen as the analyte because of its toxic nature. The selection of lead as the test analyte was also useful as it is susceptible to matrix interfer- ences caused by the chloride salts that are present in many biological samples.The study should therefore indicate how each of the methods copes with the increased matrix interfer- ences that occur when analysing plant materials without dissolution. Cup-in-tube Procedure The cup-in-tube method can be considered to be an extension of the stabilised temperature platform furnace (STPF) method. Initial work carried out by Grobenski et 01.5 illustrated the potential of STPF conditions for the direct determination of lead in rock samples. A continuation of this work by Vollkopf et al.1 resulted in the development of a specially adapted furnace for the direct determination of solids, the cup-in-tube procedure.The method involves the use of a specially slotted graphite tube into which a graphite cup is inserted. The tube and cup are made of high density graphite coated with a pyrolytic graphite layer. Specially adapted cones are required to use the cup-in-tube with a Perkin-Elmer HGA 600 atomiser. Approximately 1 mg of powdered sample is deposited into the cup, which is then transferred to the furnace for atomisation. During the atomisation stage, the cup acts like a large platform and is heated mainly by radiation. The analyte appearance time is therefore delayed relative to wall atomisation and, because the tube is heated under maximum power to the pre-set temperature, the sample should be vaporised into a high temperature environment. Under such conditions vapour phase interferences should be reduced, or even eliminated, and hence direct calibration with acidified standard solutions should become possible.However, as is indicated in the results section, interference-free analysis was not possible in the determination of lead in Bowen’s kale and the use of standard additions and matrix modification procedures was necessary. Ring Chamber Procedure The ring chamber tube was designed by Schmidt and Falk2 specifically for the direct analysis of solid samples. The specially adapted pyrolytically coated graphite tube is rotation- ally symmetrical with an external cylindrical sample chamber at the centre of the tube (see Fig. 1). The tube consists of two pyrolytically coated sections which lock together. The special tube construction allows convenient and exact placing of the solid sample, which is typically up to 15 mg in mass.During the atomisation stage of the furnace programme, the analyte atoms escape through a small orifice in the tube into the main volume of the furnace through which the hollow cathode lamp radiation is focused. A small hole is bored approximately 2 mm from the end of the tube to allow matrix particles and vapour to escape before condensing at the cooler ends of the tube. Fit .~5.85mm-8mrn-.-11 mrn-. -. -. -4- I 1 1-28 mm- Fig. 1. Schematic drawing of the ring chamber tube The design of the ring chamber tube is such that the resistivity is similar to that of a conventional atomiser tube used with an EA 3 electrothermal atomiser (Janoptik Corporation). In this study the atomiser was operated in conjuction with a Jana Model AA 3 atomic absorption spectrometer.When the ring chamber tube is heated to the atomisation temperature, the temperature towards the end of the tube is slightly greater than that at the centre because of the larger mass of the ring chamber. The heating rate of the central section is also slower than that of the ends of the tube and so vaporisation into a high temperature environment is possible.96 ANALYTICAL PROCEEDINGS, MARCH 1988, VOL 25 Slurry Sample Introduction of Solids The slurry method involves the suspension of a finely pow- dered solid in solution and the introduction of this solution directly to the atomiser. The method therefore combines the advantages of both solid and liquid sampling in that no tube modification is required and conventional autosamplers can be used to deposit the slurry volume.One of the main drawbacks of the cup-in-tube and ring chamber tube methods is the need to repeat micro-scale weighings in a reproducible manner. With the slurry procedure, only one weighing is required and the deposition of small aliquots of the sample is simplified through the use of a fully programmable autosampler. This allows reproducible mass deposition and reproducible place- ment of the sample in the tube. The use of the slurry method, however, does not overcome problems associated with sample inhomogeneity and matrix interferences and both difficulties must be taken into consideration when developing methods. Throughout this investigation a thixotropic thickening agent, Viscalex HV30,3?4 was used for the preparation of slurry solutions.Viscalex HV30 is an acrylic copolymer containing carboxyl groups. It is supplied in the form of an acidic low viscosity emulsion, but on dilution and neutralisation the emulsion produces a viscous gel over the pH range 6-10. When preparing a slurry, the aim is to produce a suspension that is stable over the period of time required for the analysis, but without making the liquid so viscous that it cannot be dispensed quantitatively by an autosampler. Slurry stability depends upon the particle size of the solid, the slurry concentration and the concentration of Viscalex in the slurry. In this work, a slurry method was developed for the determination of lead in Bowen's kale using a Perkin-Elmer HGA 500 atomiser and a Perkin-Elmer 3030 atomic absorption spectrometer with deuterium background correction. Analysis of Bowen's Kale A summary of the three procedures under consideration is given in Table 1 along with the typical sample masses analysed per determination. In each method, standard additions were required to counteract the chemical interferences caused by the kale matrix in the determination of lead. Matrix modification was also used with the cup-in-tube and slurry methods.The results obtained by the three methods are given in Table 2, and the concentrations of lead are in reasonable agreement with the certified reference value. All calculations were based on peak area absorbance measurements made at the 283.3 nm lead line, except for the slurry method, where peak height absorbance was used.The large standard deviation obtained with the cup-in-tube suggests that this procedure is unsatisfactory for precise quantitative analysis in the determination of lead in kale. In contrast, the 1Ooh precision obtained with the other two procedures was considered acceptable for solid sampling routines. Table 1. Summary of solid sampling procedures Accommodated Technique sample massimg Method Cup-in-tube <1 Standard addition calibration of lead with loo% HN03 as matrix modifier Ring chamber S1S Standard addition calibration of lead (standard addition left to absorb into the kale then dried slowly at 60 "C in an oven overnight) Slurry 0.2 Standard addition calibration of lead with 5% NH4H2P04 as matrix modifier, oxygen ashing On the basis of analytical accuracy and precision, speed of repeat analyses and scope for automation of the method using standard ETA-AAS instrumentation, the slurry sample intro- duction procedure was considered to be the most suitable for Table 2.Comparison of measured lead concentration in Bowen's kale obtained by different solid sampling procedures Pb contenti Technique Clgg-l Cup-in-tube 2.6 2 1.1 Slurry 2.55 k 0.29 Ring chamber 2.58 k 0.25 Certified content 2.4 k 0 . 5 the rapid analysis of solids. With the cup-in-tube and ring chamber tube, analysis times of about 5 min, inclusive of weighings, were required, in comparison with the 2 min cycle time for slurry analysis. Although a period of 5-10 min is required in order to prepare a slurry, the time saved in multiple atomisations of a sample, the convenience of within-tube matrix modification and the ease of automation more than compensate for the minor inconvenience of the sample preparation step. Design and Construction of an Independently Heated Probe for ETA-AAS Various procedures have been adopted in order to minimise the effect of chemical interferences in ETA-AAS analysis including the use of matrix modifiers, peak area measurement and platform atomisation.In the latter approach, the intention is to minimise the occurrence of interferences through vapori- sation of the sample into a high temperature environment. An alternative procedure to platform atomisation is probe atom- isation, which gives better temporal separation of furnace heating and sample volatilisation.6 In this procedure, the sample droplet is injected on to the head of the thin graphite probe that is positioned inside the graphite tube.The droplet can be dried and charred inside the tube, but before the atomisation stage the probe is removed and the furnace heated to the desired atomisation temperature. Once the temperature within the furnace has been stabilised, the probe is re-inserted and heated rapidly by radiation from the hot tube to give vaporisation into a high temperature environment. Under these circumstances vapour phase chemical interferences are reduced due to the increased dissociation of analyte molecules in the high temperature vapour compared with atomisation from the tube wall. Although the probe system separates the heating of the vaporisation surface from the atomisation medium, it is not possible to control the heating rate and temperature of the probe and tube independently with current designs.For any particular type of probe, the rate of heating and temperature attained depend on the temperature of the graphite tube. An attempt was therefore made to design and construct a probe device that could be heated independently of the graphite tube. 2 c m 2.5 crn 0.5 c m Ridge 1.5 rnm -500 pm Side elevation 5 c m 0.5 c m I 3.5 crn 4 Plan view Fig. 2. Schematic drawing of an independently heated graphite probeANALYTICAL PROCEEDINGS, MARCH 1988, VOL 25 97 Probe Design A variety of probes were evaluated before the selection of the design illustrated in Fig.2. A template was constructed to allow laboratory preparation of probes from electrographite bars of dimensions 1 x 0.3 x 5 cm. Other probes of a similar design were prepared by Schunk (UK) Limited. The independently heated probe assembly consisted of three components: the probe mount; the gas box for the pneumatic solenoid; and the power supply. The power supply was constructed from an adapted Pye Unicam SP 9 atomiser power supply. The stems of the probe are positioned in two graphite blocks, which act as electrical contacts. The probe mount is held on two guides which direct the movement of the probe when the pneumatic solenoid is activated. The gas box contains the solenoid valves and the electrical trigger system required for automatic movement and heating of the probe.A typical independently heated probe programmed for the SP 9 atomiser is given in Table 3. At the second dry stage (which could also be a char stage) the spectrometer is auto zeroed and the solenoid manually triggered to remove the probe from within the atomiser tube. During the pre-heat stage, the tube temperature is elevated to the required value, and at the atomisation stage the "record" function activates the pneumatic solenoid and the probe is moved back into the heated atomiser tube. As the probe mount slides along the guide rails, a small microswitch triggers the probe power supply, which causes the probe to be heated when electrical contact is made between brass contacts on the end of the probe mount and a fixed plate positioned at the end of the guide rails.The probe is heated almost instantaneously to approximately 1000 "C and after 2 s it reaches a temperature of about 1600 "C. Table 3. Typical programme for independent heating probe Step TemperaturePC Time/s Ramp Functions - Dry 450 SO 5 450 3 2 Auto-zero. Probe manually removed from furnace - Pre-heat 2500 5 0 Atomise 2500 4 0 Rec., gas stop, Clean 2700 3 2 peak timer - Fig. 3 illustrates the effect of auxiliary heating of the probe. Its temperature is compared with the temperature attained by the unheated probe when inserted into an atomiser tube heated to a set temperature of 2000°C. Independent heating of the probe significantly increases the heating rate to the equilibrium temperature, which is approximately 200 "C greater than for the unheated probe.At higher atomisation tube temperatures (e.g., 2700 "C), the effect of auxiliary heating on the probe heating rate was not as pronounced, but an increase in the final probe temperature of between 150 and 200°C was observed. Preliminary results with the independently heated probe suggest that auxiliary heating will be of most use when a comparatively low atomiser tube temperature is selected for probe atomisation. Most of the probes used in the initial stages of this project were prepared from electrographite strips. Although this form of graphite is easily shaped and filed, it is more susceptible to oxidation and thermal and mechanical stress. The lifetime of individual probes is somewhat low at present (about 20-30 firings) but should be improved by better inert gas sheathing of the probe assembly and the use of high density graphite, pyrolytic graphite or microporous glassy carbon for probe manufacture. Further evaluation is now required to assess the usefulness of the independently heated probe in removing vapour phase interferences and improving the atomisation of more refractory elements. X 9 1.65 . al 2 1.60 z % 1.55 I- 1.50 5 1.45 U 1 I 0 2 4 6 ' 1.40 Time/s 3 Fig. 3. Optical pyrometer temperatures of the heated (+) and the unheated (0) probe during an atomisation stage of set temperature 2000 "C The provision of a Research Studentship (for A.M.Q.) by the Central Electricity Generating Board, and support from the SERC (for R.N.) are gratefully acknowledged. The donation of graphite probes by Schunk UK Ltd. and the loan of an SP-9 furnace power supply and atomiser by Pye - Unicam Limited are also acknowledged. The provision of facilities at the Central Institute for Optics and Spectroscopy, Academy of Sciences of the GDR, Berlin, by Professor Heinz Falk were greatly appreciated, particularly with regard to the ring chamber tube AAS experiment. This paper is published with the permission of the Central Electricity Generating Board and the University of Strathclyde. References 1. Vollkopf, U., Grobenski, Z..Tamm, R . , and Welz, B.,Analyst, 1985, 110, 573. 2. Schmidt, K. P., and Falk, H., Spectrochim. Actu, 1987, 42B, 431. 3. Littlejohn. D., Stephen, S. C.. and Ottaway. J. M., Anal. Proc.. 1985.22.376. 4. Stephen, S. C . , Littlejohn. D.. and Ottaway, J . M., Analyst, 1985. 110, 1147. 5 . Grobenski, Z . Lehmann, R.. Tamm, R., and Welz. B. Mikrochim. Acfa, 1982, 1, 115. 6. Corr, S. P . , and Littlejohn D.. J . A n d . Atom. Spectrom., 1988, 3 . in the press. ROYAL SOCIETY OF CHEMISTRY: ANALYTICAL DIVISION ELECTROANALYTICAL GROUP The Electroanalytical Group will be organising a symposium on the subject of Electrochemical Sensors, dealing particularly with microelectronic devices and their biomedical applications. It will be held at Thorn EMI, Hayes, Middlesex, on May 4, 1988. For details contact Mr. A. E. Bottom, Kent Industrial Measurements Ltd., Oldends Lane, Stonehouse, Gloucestershire GLlO 3TA.

ISSN:0144-557X    

DOI:10.1039/AP9882500085

出版商:RSC    

年代:1988

数据来源: RSC

摘要:

98 ANALYTICAL PROCEEDINGS, MARCH 1988, VOL 25 Equipment News Accessories for Atomic Spectroscopy The HGA-700 graphite furnace and the AS-70 autosampler are new accessories for use with the makers' Model 1100 and Model 2100 atomic absorption spec- trometers. Both are fully computer con- trolled, allowing complete operation of the complete system via the spectrometer keyboard, with all parameters and analy- tical data displayed on the same video screen. Literature is available. Perkin-Elmer Ltd., Post Office Lane, Beaconsfield, Buckinghamshire HP9 1QA. Scanning Spectrophotometers The new DU-70 series of instruments follow the makers' DU-60 series. They offer capabilities ranging from sample analysis of turbid solutions to the analysis of DNA samples of less than 50 PI. A complete 16-bit multi-processor system accompanies the spectrometer so that readings can be taken rapidly, to be displayed in real time.Calculations are performed as quickly as parameters can be entered. The DU-70 provides six scanning speeds from 60 to 2400 nm min-1, and at speeds of up to 2400 nm min-1 any sample can be scanned in under 1 min. Without waiting for a screen change the user can select from ten calculation options, which include deriva- tives and smoothing for finding additional spectral information. The DU-70 system monitors kinetic reactions with real-time graphic display of data. Gel separations are simply achieved and the sample is scanned only once. Using stored data any information can be calculated within seconds, including peak migration dis- tance, areas under peaks or molecular height of bands.Beckman Ltd., Progress Road, Sands Industrial Estate, High Wycombe, Buck- inghamshire. Spectrophotometers A complete range of ultraviolet - visible instruments is offered, from the analogue- readout Chroma 252 colorimeter to the versatile Response I1 with its dual disk drives for data manipulation. The three new Chroma colorimeters offer perfor- mances to suit education, routine and complex applications. In conjunction with its water jacketted cuvette the top of the range Chroma 257 is an ideal colorimeter for a wide selection of Gilford reagents. The 258/259 spectrometers are micro- processor controlled via a four position touch pad. Bright LED displays allow an absorbance reading to be indicated to kO.001 and also prompt the user through each measurement routine.The 258 is ideal for routine colorimetric or spectro- Ciba Corning Chi photometric analysis and offers four modes of operation: absorbance, trans- mittance, concentration and routine kinetics. The 259 offers the same modes with greater versatility. The 7800 spectro- photometer is, in its standard format, a low cost scanning instrument, Through the use of plug-in modules the capability of the optical bench can be enhanced. The major options include mono or colour CRT displays, a high resolution printer - plotter, a thermal graphic printer, appli- cations credit cards and a floppy disk drive unit. The Response I1 is a high performance spectrometer equipped with two 5Y4-in disk drives and comprehensive data manipulation facilities, which allow data or analysis programs to be stored and retrieved at will and make operations such as the comparison of samples and calcula- tions of final results simple and con- venient.Ciba Corning Diagnostics Ltd., Hal- stead, Essex C 0 9 2DX. Vacuum Fourier Transform Infrared Spectrometers The FTS-40 and FTS-60 instruments are housed in heavy duty, compact, cast enclosures which can be operated under reduced pressure, vacuum or nitrogen purge. The spectrometer is divided into front and rear compartments which are separately evacuable. The rear (inter- ferometer) compartment can be main- tained at reduced pressure while samples or accessories are being changed in the front compartment. The spectrometers can be used in both the mid- and far- infrared regions. Interchange kits are available so that a far-infrared instrument can be exchanged into the mid-infrared, and vice versa.The resolution of the FTS-40 is 2 cm- I , optionally increased to 0.5 or 0.25 cm-1, and that of the FTS-60 is 0.5 cm-1, optionally increased to 0.25 or poma colorirneters 0.1 cm-1. The sample compartment will accommodate most sampling accessories. The UMA 300 microscope, photoacoustic accessory and large gas cells can be accommodated if the sample compart- ment is used in the purge mode. Polaron Equipment Ltd., Digilab Divi- sion, Bio-Rad Microscience Division, 53- 63 Greenhill Crescent, Watford Business Park, Watford, Hertfordshire WD1 SQS. Sample Handling Accessory for Infrared Spectroscopy The Precision Plastics Film Press (PPFP) can guarantee the production of plastic films to a constant and reproducible thickness and so will be of particular importance in applications which require the identification of many components, such as the analysis of polymer blends, where it is vital to use films of constant thickness.It can completely replace the hot pressed film method of sample prepa- ration. The PPFP unit includes a thermo- statically controlled oven calibrated up to 300°C, a cooler and a set of brass dies which can also be heated and cooled quickly. The dies can ensure the produc- tion of films to reproducible thicknesses of 20, 50, 100, 200 and 500 pm. Pye Unicam Ltd., York Street, Cam- bridge CB1 2PX. Multi-channel X-ray Diffractometer A new, compact bench-top XRD spec- trometer uses a curved multi-channel detector instead of conventional scanning detectors and can, therefore, reduce analytical time from hours to a few seconds.It will take all types of acces- sories, including furnaces, and it is parti- cularly suited to dynamic studies, powder diffraction, phase transitions, texture and stress analysis. Spectrolab Ltd., P.O. Box 25, New- bury, Berkshire RG16 8BQ.ANALYTICAL PROCEEDINGS, MARCH 1988, VOL 25 99 Aluminium Coated Capillary Columns The AQ5 Ultrabore series have an inter- nal diameter of 0.53 mm and are intended for use in place of conventional packed columns. The aluminium coating greatly increases the strength of the material, extending column life and allowing col- umns to be coiled to 1.50 mm diameter, thus fitting all popular makes of gas chromatograph. A full range of bonded phases is available, from low polarity BP-1 (SE-30 OV-1) to the more polar BP-21 (CW-20M FFAP) phases.The high thermal conductivity of aluminium pro- vides rapid heat transfer to the phase during temperature programming. Ultrabore columns can be operated in existing packed column instruments by the use of the makers' range of adaptors and converters. Scientific Glass Engineering (UK) Ltd., 1 Potters Lane, Kiln Farm, Milton Keynes MKll 3LA. Chromatographs Two new products are announced: the Shimadzu LC-8A preparative HPLC system and the Shimadzu GC-14A capil- lary GC system. The former is a computer controlled system which offers a wide flow-rate control range, allowing both analytical and preparative HPLC.The GC-14A is designed for capillary GC but is compatible with packed columns. A temperature control circuit ensures accu- rate maintenance of oven temperature and this, coupled with uniformity of temperature distribution within the oven, delivers the superior reproducibility of retention times required for capillary GC. Up to four different sample injectors can be fitted and FID, ECD, FTD, TCD and FPD detectors can all be fitted. For on-screen processing of chromatograms both instruments can be interfaced with the Shimadzu C-R4A Chromatopac data processor. Dyson Instruments Ltd., Hetton Lyons Industrial Estate, Hetton, Houghton le Spring, Durham DHS ORH. Refractive Index Detector for Chromatography The PU4026 is available as a module or as part of the makers' PU4100 quaternary liquid chromatograph.Accurate temper- ature control is achieved via a Peltier thermoregulator, and the Fresnel flow cell design allows complete compatibility with dynamic mixing chromatography svstems. Pye Unicam Ltd., York Street, Cam- bridge CB1 2PX. Pulseless HPLC Pumping Systems The 350 series consists of the isocratic HPLC pump, the 3.52 ternary HPLC gradient system and the new 350/04 quaternary (4-solvent) microprocessor controlled gradient system. The 350 series of HPLC pumping systems offers abso- lutely pulseless flow. Applied Chromatography Systems Ltd., The Arsenal, Heapy Street, Mac- clesfield, Cheshire SKI 1 7JB. HPLC Columns A range of totally metal free columns and cartridges is announced. The columns are made from Glass Lined Metal Tubing (GLT) using end fittings incorporating a Polyglas frit encased in rigid fluoro- polymer.They offer high performance with effective plates in excess of 75000 m-1 for spherical silica based packings. Packing options include particle sizes of 3 and 5 ym with pore sizes of 8 or 30 nm. The columns are available in 4 and 2 mm internal diameter, with lengths of 100 and 250 mm. The cartridges have a 4 mm internal diameter and a length of SO mm, with other lengths available soon. A brochure is available. Scientific Glass Engineering (UK) Ltd., 1 Potters Lane, Kiln Farm, Milton Keynes MKll 3LA. HPLC Solvents A range of high purity solvents offers very low levels of alkaline and acidic impuri- ties, with non-volatile residues below 3 p.p.m. The maximum water content is O.O5%, with the majority at 0.01%.Literature is available. May and Baker Ltd., Liverpool Road, Eccles, Manchester M30 7RT. Materials for Flash Chromatography A range of columns and accessories is available for the rapid and efficient isola- tion of organic compounds from crude reaction products. There is a choice of six column sizes, from 18 to 100 mm internal diameter. Literature is available. A range of silica for chromatography is offered under the brand name Colpak; this includes a grade, Colpak C60, specifically designed for flash chromatography. May and Baker Ltd., Liverpool Road, Eccles, Manchester M30 7RT. Analytical Chromatography Data Systems The PU6000 integration system allows even the newcomer to tackle the complete procedure of data acquisition, integration and reporting.Based on windows tech- nology, the PU6000 ofers multi-tasking and interactive data handling, so that it is possible to collect data, develop a method and produce a report at the same time. The software and data capture hardware can handle up to six asynchronous chan- nels of chromatography. Pye Unicam Ltd., York Street, Cam- bridge CB1 2PX. Fraction Collector The Model 2100 fraction collector collects fractions on either a time or volume basis. Time increments (from 0.1 to 99 min) or drops (from 1 to 990) are selected via the front panel and digital counters are used to monitor progress. The capacity for 12, 13, 16 and 18 mm diameter test tubes is 100. If used with the Model 700 or Model 402 HRLC gradient systems it can receive an external signal for fraction advance.Bio-Rad Laboratories Ltd., Caxton Way, Watford Business Park, Watford, Hertfordshire WD1 8RP. Thin-layer Chromatography Gel Scanner The Shimadzu CS-9000 is designed for use in scanning TLC plates, autoradiographs, silver stains, and polyacrylamide and protein gels. Particularly useful for elec- trophoresis, it provides 2-D scanning as standard. Resolution is high (50 pm) and a colour CRT and a printer - plotter are included as standard. Especially for TLC there is a zig-zag scanning facility for irregular spots. On-screen reintegration and manipulation of chromatograms can be performed after a scan. V. A. Howe and Co. Ltd., 12-14 St. Ann's Crescent, London SW18 2BR. Gel Dryers The ability to select pre-set programmed cycles for optimised drying of a variety of gels is one of the features of the Model 543 and Model 583 gel dryers.Others include a quick seal track for instantaneous gasket sealing and selectable temperature set- tings from 50 to 90°C. For sequencing gels, the pre-set cycle speeds up drying by eliminating pre-heating requirements. The dryer rapidly elevates the tempera- ture to its 90 "C maximum and then slowly decreases it to the final set value (usually 80 "C). The optimal performance temper- ature is therefore reached quickly and drying should take 20-25 min. Bio-Rad Laboratories Ltd., Caxton Way, Watford Business Park, Watford, Hertfordshire WD1 8RP. Electronic Controller - Switcher for Electrophoresis The Pulsewave 760 is an electronic con- troller - switcher for DNA pulsed field, field inversion (FIGE) and orthogonal field electrophoresis (OFAGE).De- signed for the separation of chromosome size DNAs, from 50 000 to 2 000 000 bases in length, it has wide applications in gene mapping and identification, and all aspects of chromosomal analysis. These would include a role in the Human Genome Sequencing Project, the first100 ANALYTICAL PROCEEDINGS, MARCH 1988, VOL 25 step of which will be the construction of large-scale restriction enzyme maps of the chromosomes. To separate these a Pulse- wave will be required to perform the current switching. Bio-Rad Laboratories Ltd., Caxton Way, Watford Business Park, Watford, Hertfordshire WD1 8RP. Sample Processing Station Sample preparation with the makers’ Bond Elut solid phase extraction columns has been made more efficient with the introduction of the Vac Elut SPS 24- sample processing station.This will accept 24 solid phase extraction columns at a time and it reduces processing times to a minimum. The SPS 24 features a waste and collect facility. All columns are easily accessible during sample applica- tion and elution, and because waste sol- vents are ported through the central funnel there is no need to clean the vessel interior and insert the sample rack prior to sample elution. Analytichem International, P.O. Box 234, Cambridge CB2 1PE. Clinical Chemistry Analysers The Vitalab 30 series incorporates multi- function video display, alphanumeric thermal printer, programmable 21-test memory, Peltier electronic temperature flow cell control and the Vitalab PTFE bellows sipper pump.The Vitalab 31 has an automatic wavelength selection. All are capable of non-linear tests using curve fitting software. The 30 series differs as follows: Vitalab 30 has plug-in filters, Vitalab 31 has a built-in 8 position auto- matic filterwheel, and the Vitalab 32 has a monochromator. The Vitalab 30, 31 and 32 can be converted into the Vitalab 102, 112 and 122 chemistry batch analysers by the addition of a sampler and program- mable diluter. Vital Scientific Ltd., Huffwood Trad- ing Estate, Partridge Green, Sussex RH13 8AU. Analyser of COSHH Measurements The Photovac 10S50E gas analyser is the latest addition to the Photovac 10s range of environmental monitoring systems. Designed to measure key gases as required by the new COSHH regulations, it features improved hardware and soft- ware.A computer version, the 10S70E is also available with modem control and an RS232 interface to an external computer remote terminal. Centronic Sales Ltd., 275 King Henry’s Drive, New Addington, Croydon , Surrey CR9 OBG. Mercury Vapour Analyser The Jerome 41 1 mercury vapour analyser uses a patented gold film sensor which overcomes specificity problems asso- ciated with atomic absorption spectro- scopy. Designed for use in all working environments, it will detect and digitally display mercury concentrations as low as 0.001 mg m-3 in 10 s. Interferences such as water vapour, sulphur dioxide, aro- matic hydrocarbons or particulate matter do not affect the accuracy of the Jerome 41 1.Also available is the Jerome gold coil mercury dosimeter, a collection device for mercury vapour which can be used for time weighted averages of contamination to individuals in the workplace. Arizona Instrument Corporation, P.O. Box 670 Cookham, Maidenhead, Berk- shire SL6 9BJ. Portable pH Meters The pH1 range of portable pH meters has been extended. The new instruments feature digital display, automatic one- button standardisation and the makers’ Auto-Read and Auto-Find facilities, which provide lock-on stable readings and instant recognition of 5 standardisation buffers, respectively. A range of acces- sories is available. A choice of three meters suits a wide variety of tasks. Start-up kits are also offered, together with a laboratory organiser and wall or shelf mounts.Beckman Ltd., Progress Road, Sands Industrial Estate, High Wycombe, Buck- inghamshire. Meat Speciation Test Kit Domino 5 offers a simple and accurate screening test for identifying the species in uncooked meat samples. The Domino 5 test kit is based on the agar gel double immunodiffusion technique, which has been found to have a screening sensitivity of 5%. The method takes only a few minutes to set up and can be incorporated into the routine quality control of any meat processing company. IDS, Usworth Hall, Washington, Tyne and Wear. Protein Purification The Rotofor cell provides a new alterna- tive for preparative scale protein purifica- tion by isoelectric focusing. By incorpor- ating a cylindrical focusing chamber that rotates around its axis the new cell over- comes traditional scale-up problems.It is capable of greater than ten-fold purifica- tion of a protein sample in a single 4 h run. In certain instances proteins of single band purity can be recovered. Bio-Rad Laboratories Ltd., Caxton Way, Watford Business Park, Watford, Hertfordshire WD 1 8RP. Automated Sequence Reader An extension of the Gene-Master product line, a new instrument is based on the work completed by Professor E. M. South- ern and John Elder at Oxford University. Incorporating a fast optical scanner, which uses a linear array detector, the automated sequence reader can read a set of four lanes in approximately 4 min and a 38 X 50 cm gel in about 2 h. With 24-h operation over 100 000 bases can be read per day. On high quality gels accuracy is 99% with the hard copy (i.e., autoradio- gram) being retained for verification.After digitising each image the system will apply pattern recognition software to correct automatically for lane wander, “smiling” and other artifacts and it will assign a sequence. Analysis is by the Gene-Master or an IBM AT or compat- ible computer. A range of DNA and protein analysis software is available for use with the Gene-Master. Bio-Rad Laboratories Ltd., Caxton Way, Watford Business Park, Watford, Hertfordshire WDl 8RP. Oven for Oil Recovery Research Redwood Corex (Services) Ltd., of Aber- deen, have taken delivery of a nitrogen purged hot air oven to perform core flooding experiments in reservoir condi- tions. The oven is used to evaluate gas injection techniques for increasing oil recovery in North Sea Reservoirs.It has a maximum operating temperature of 200 OC, with four air-circulating fans pro- viding temperature uniformity through- out the chamber of -t 1 “C. Safety features include a hydrocarbon gas analyser alarm system, which stops an experiment if the concentration of hydrocarbon gases approaches a pre-set limit. R. E. Pickstone Ltd., Fison Way, Thetford, Norfolk IP24 1HT. Blender A new 7-speed, positive push button, solid-state controlled blender is available with a 1 1 stainless steel or 1.2 1 heat resistant glass container. It has an enamelled die-cast zinc base with chrome collar. The blades are of stainless steel and speeds range from 3500 to 21000 revolutions min-1. Christison Scientific Equipment Ltd., Albany Road, East Gateshead Industrial Estate, Gateshead NE8 3AT.ANALYTICAL PROCEEDINGS, MARCH 1988, VOL 25 101 Software for Routine Formulation and Other Weighing Applications Labware Formulating is a new, user- friendly software package that improves the productivity of routine jobs involving a lot of calculating and recording with conventional methods.It is for use with the Epson PX-4 hand-held computer and either one or two of the makers’ balances in production, filling plants, stores and laboratories. Also available is XPac-M software, which can be programmed to suit the needs of specialised tasks. Mettler Instrumente AG, CH-8606 Greifensee, Switzerland. Electron Microscope The CM30/STEM is a 300 kV trans- mission electron microscope. Its high acceleration voltage permits the investi- gation of specimens of greater thickness, for example in the examination of the three dimensional structure of live or material science objects.In high resolu- tion structure analysis the point resolution of 0.23 nm can be combined with local chemical analysis by X-rays or diffraction down to an area of 2 nm2. Philips Industrial and Electro-Acoustic Systems Division, P.O. Box 218, 5600 MD Eindhoven. The Netherlands. Stereo Measurement Package A new package has been launched for use with the makers’ AN 10000 X-ray micro- analysis system. Computer based image processing systems such as the AN 10000 can determine true quantitative topo- graphical data from a flat-plane image. This is achieved by acquiring a stereo pair of images, either directly into the AN 10000 image store or by digitising TV camera photographs and comparing them.From parallax measurements rela- tive heights on the specimen’s surface can be determined. Link Analytical Ltd., High Wycombe, Buckinghamshire. Eight Colour Plotter Designed for CAD or business graphics, the SE283 incorporates two graphics lan- guages as standard: Hewlett-Packard’s HP-GL and the makers’ own BBC-GL. Together with the choice of RS232-C or IEEE 488 interface, this makes the SE283 compatible with almost all available graphics software and hardware. Its plot- ting area is 287 mm X 410 mm with margins of only 5 mm. ANSI A and B formats are also catered for. Resolution is better than 0.0s mm and vector speed is programmable from 1 to 50 cm s-1. British Brown-Boveri Ltd., Darby House, Lawn Central, Telford.Shrop- shire TF3 4JB. Literature Literature is available on the Shimadzu UV-2100 ultraviolet - visible recording spectrophotometer, the Shimadzu UV- 3100 ultraviolet - visible - near-infrared recording spectrophotometer, the RF- 5000 recording spec t ro fluo ro p h o t ome t er , and the Shimadzu AA-660 atomic absorp- tion - flame emission spectrometer. V. A. Howe and Co. Ltd., 12-14 St. Ann’s Crescent. London SW18 2LS. A colour brochure gives full details of the Kratos LIMA systems. The latest instru- ment is the LIMA-3 laser ionisation mass analyser, which enables both organics and inorganics to be analysed at sensitivities down to parts per million. Kratos Analytical, Barton Dock Road, Urmston. Manchester M31 2LD.A brochure announces the Osmo Orion laboratory reverse osmosis unit. This unit produces 600-4000 1 day-’ of purified water. Osmonics Inc.. 595 1 Clearwater Drive, Minnetonka, Minnesota 55343, USA. A catalogue contains information on a wide range of materials for research and design engineering. The range includes pure metals and alloys, polymers, ceram- ics, composites, glassy alloys, biopoly- mers, various inorganic and organic com- pounds, single crystals and L-B films. Goodfellow, Cambridge Science Park, Milton Road, Cambridge CB4 4DJ. Brochures describe the Suprafuge 22 and the Cryofuge M 7000, both from Heraeus Sepatech. The Suprafuge 22 is a refriger- ated centrifuge featuring high perfor- mance microprocessor control linked to an advanced drive system; speeds of up to 22 300 revolutions min-1 are attainable and the temperature range is from - 19 to +40°C. The Cryofuge M 7000 is a large capacity refrigerated centrifuge designed to meet the demand for increased produc- tion in the blood transfusion service and the pharmaceutical industry. V. A. Howe and Co. Ltd., 12-14 St. Ann’s Crescent, London SW18 2LS. Issue 5 , September 1987, of Laboratory Equipment Info gives details of a wide range of equipment and materials, from silica gel plates and the CD 60 Densi- tometer for quantitative TLC to the HI/ 8414 Stick pH meter and the HI18543 Oxymeter dissolved oxygen meter. Refer- ence is also made to the new Camlab catalogue, now available. Camlab Ltd., Nuffield Road, Cam- bridge CB4 1TH. The Autumn, 1987, issue of Howe Lahor- atory News gives information on a new Hamilton PRPX300 ion-exclusion col- umn, the Ambis two-dimensional beta scanner, Shimadzu spectrophotometers, the Sald-1000 laser diffraction particle size analyser, Heraeus Sepatech centri- fuges, the Bioscan QC 2000 for radioiso- tope counting and other equipment. V. A. Howe and Co. Ltd., 12-14 St. Ann’s Crescent, London SW18 2BR. Volume 1, June 1987, of Innovation, a PR magazine from Shimadzu, has a number of articles, including Scientific Instrumen- tation and Instrument Science, Gardens of Kyoto, Trends in Imaging Devices for Medical Use, and an article about the Company’s founder, Genzo Shimadzu. Shimadzu Corporation, 1 Nishinokyo- Kuwabaracho, Nakagyo-ku, Kyoto 604, Japan.

ISSN:0144-557X    

DOI:10.1039/AP9882500098

出版商:RSC    

年代:1988

数据来源: RSC

摘要:

ANALYTICAL PROCEEDINGS, MARCH 1988, VOL 25 SAC Silver Medal 101 Nominations are invited for the award of ing in any field covering the practice and is normally made annually to the candi- the SAC Silver Medal. which is for the teaching of analytical chemistry. The date who, in the opinion of the AD encouragement of young scientists work- award is accompanied by a cash prize and Council, has made the greatest contribu-102 tion and whose work has made the most significant impact on any branch of analy- tical chemistry. In addition, the future promise of the candidate is taken into consideration. It is hoped to provide an opportunity for the successful candidate to deliver a lecture to the Division on a suitable occasion subsequent to the presentation of the Medal. The rules are follows: The award of the Silver Medal will normally be considered annually by the Honours Committee, acting on behalf of the Council of the Division, but an award may not be made if it is considered that the work of no candi- date reaches the required standard.Candidates must be British subjects of 35 years of age or under at January 1st in the year in which the award is made. Evidence of age will be required.* * Although the age limit has been reduced in this Rule, for the next year, consideration will continue to be given to candidates up to 38 years old. ANALYTICAL The merits of the candidate’s work may be brought to the notice of the Council by any person (being a mem- ber of the Analytical Division of the Royal Society of Chemistry) who desires to recommend the candidate by letter addressed to The President, Analytical Division, The Royal Society of Chemistry.The letter should be accompanied by a short statement of the candidate’s career (date of birth, education and experience, degrees and other qualifi- cations, special awards, etc., with dates, and any other relevant informa- tion) and a list of titles of, and refer- ences to, papers or other works publi- shed by the candidate, independently or jointly. One reprint of each paper (or other work) for which reprints are available should be submitted. The award will be made on an over-all assessment of the candidate’s contri- bution, the impact of hidher work and hidher future promise in any field PROCEEDINGS, MARCH 1988, VOL 25 covered by the principles, teaching and practice of the analytical sciences. No restriction is placed as to where the work is conducted. The Committee assessing the applica- tions shall be at liberty to call any candidate for interview. The successful candidate will receive the sum of f l O O in addition to the Medal. The decision of the Council shall be final. Any alteration to these rules shall be subject to the approval of the Council. Recommendations for the next award should be made to The President, Analy- tical Division, The Royal Society of Che- mistry, Burlington House, London, W1V OBN, by March 31st, 1988.

ISSN:0144-557X    

DOI:10.1039/AP9882500101

出版商:RSC    

年代:1988

数据来源: RSC

摘要:

102 ANALYTICAL PROCEEDINGS, MARCH 1988, VOL 25 Conferences and Meetings 3rd CNA International Meeting on Flavourings May 29-31, 1988, Lugano, Switzerland The topics of this congress will include “The Analytical Control of Flavourings in the Food Industry as an Essential Guaran- tee of Quality and Safety.” For information on the congress con- tact Professor Fernando Tates, Cattedra di Aromatizzazione dei Prodotti Alimen- tari del Dipartimento di Scienze e Tecnol- ogle Alimentari e Microbiologiche, Uni- versita degli Studi, Via Celoria, 2 20133, Milano, Italy. 1st International Symposium on Separa- tion of Chiral Molecules May 31-June 2, 1988, Paris, France The Societe Franqaise de Chimie will be organising this symposium, which is spon- sored by E. Merck. The sessions, which will consist of plenary lectures and oral and poster contributions on the following subjects: Direct Separation: Static and Dynamic Crystallisations; Crystallisation by Diastereoisomer Formation; Chromat- ographic Methods (GC, LC) Based on ROYAL SOCIETY OF CHEMISTRY: ANALYTICAL DIV CHEMOMETRICS GROUP A meeting entitled WHAT CAN CHEMOMETRICS DO FOR YOU? will be held at the University of Edinburgh on March 29th-30th, 1988 S ON The Chemometrics Group of the Analytical Division will organise this meeting in the Department of Chemistry.The speakers will be J. Mendham (Thames Polytechnic), K. Burton (Polytechnic of Wales), M. Bracewell (Macaulay Institute), R. J. Howarth (BP Research), J. M. Thompson (Birmingham University), J. McNicol (The Scottish Agricultural Statistics Service) and P.Norman (ICI). For further information contact Dr. S . Haswell, School of Chemistry, Thames Polytechnic, Wellington Street, London SE18 6PF.ANALYTICAL PROCEEDINGS, MARCH 1988, VOL 25 103 Chiral Separative Phases (CSP) or Not; Enzymatic Separations. For further information contact Societe Fransaise de Chimie, Congress Depart- ment: G. Perreau, 250 Rue Saint- Jacques, 75005 Paris, France. Laboratory Information Management Systems Conference June 14-16, 1988, Pittsburgh, PA, USA The first international LIMS meeting was held in 1987 and was attended by 400 people. The second one will be held in the Westin William Penn Hotel, Pittsburgh, and will consist of four lecture sessions (each with one invited and four contri- buted papers) and two poster sessions (one of which will be for vendors).The address for information is LIMS Conference, P.O. Box 2166, Lower Burrell, PA 15068, USA. Nuclear and Radiochemistry July 11-15, 1988, Brighton The second international conference in this series will form part of the twenty- first anniversary celebrations of the Radiochemical Methods Group of the Analytical Division of the RSC. It will be held in the Brighton Metropole Hotel. The invited lecturers will be Dr. D. C. Hoffman (USA), Dr. D. Comar (France), Dr. J. Silvester (UK), Professor E. A. Scheikert (USA), Dr. C. Allen (UK), Professor B. F. Myasoedov (USSR), Dr. E. I. Hamilton (UK), Dr. K. Roessler (FRG) and Dr. A. Dyer For further details contact The Sec- retary, Analytical Division, The Royal Society of Chemistry, Burlington House, London W 1V OBN.International Symposium on Current Separation Techniques for Macromol- ecules August 22-24, 1988, Uppsala, Sweden The symposium will be held at the Upp- sala University Biomedical Centre. Con- (UK). tributions will be presented in the fields of Liquid Chromatography, Supercritical Fluid Chromatography, Field Flow Frac- tionation, Hydrodynamic Chromato- graphy and Electrophoresis. Poster con- tributions are still required. For details contact The Swedish Chem- ical Society, The Analytical Division, Wallingatan 26B, S-111 24 Stockholm, Sweden. 102nd Annual International Meeting and Exposition of the AOAC August 29-September 1, 1988, Palm Beach, FL, USA This meeting, which will be held at the Breakers, will consider Analytical Metho- dology, with a spotlight on Biotechnol- OgY.For information contact Margaret Ridgell, AOAC, 1111 North 19th Street, Suite 210, Arlington, VA 22209, USA. 4th Separation Science and Biotechnology Symposium August 31-September 3 , 1988, Gargnano del Carda, Italy The above symposium will be held at the Palazzo Feltrinelli. The challenges brought about by biotechnology will be discussed. In addition to analytical and preparative chromatographic techniques, others state of the art approaches such as sedimentation, field-flow fractionation, capillary to preparative gel electrophore- sis and other biomacromolecular and particle separation techniques will be included. For further details contact the confer- ence Chairman, Professor Pier Giorgio Righetti, Euro Business Centre, P.O.Box 10552, 1001 EN Amsterdam, The Netherlands. 18th International Symposium on Envi- ronmental Analytical Chemistry and 4th International Congress on Analytical Techniaues in Environmental Chemistrv September 5-8, 1988, Barcelona, Spain This symposium will include plenary lectures, invited and submitted research papers and poster presentations covering the whole field of environmental ana- lytical chemistry. Contributions will be welcomed on New Instrumentation tech- niques, On-line versus Off-line Methodol- ogies in Environmental Analysis and Sample Handling, Environmental Fate and Metabolism of Pollutants, and Distri- bution and Speciation of Metals. The address of the Congress Office is Palacio de Congresos, Dpto Conven- ciones, Avda Reina Ma. Cristina s/n, 08004 Barcelona, Spain. Fifth Annual International Pittsburgh Coal Conference September 12-16, 1988, Pittsburgh, PA, USA The University of Pittsburgh will be host- ing this conference, which will be held in the Pittsburgh Hilton Hotel. For further information please contact Pittsburgh Coal Conference, MEMS, One North- gate Square, 2 Garden Center Drive, Suite 211, P.O. Box 270, Greensburg, Pennsylvania 15601, USA. Eastern Analytical Symposium October 3-7, 1988, New York, NY, USA The twenty-seventh EAS will be held in the New York Hilton Hotel. For informa- tion contact Dr. Stephen Scypinski, Ber- lex Laboratories Inc., 110 East Hanover Avenue, Cedar Knolls, NJ 07927, USA. California Association of Toxicologists 1988 Quarterly meetings will be held on May 7, August 6 and November 5, 1988. For details contact Paul Sedgwick, Orange County Sheriff-Coroner, Forensic Science Services, P.O. Box 449, Santa Ana. CA 92702. USA.

ISSN:0144-557X    

DOI:10.1039/AP9882500102

出版商:RSC    

年代:1988

数据来源: RSC

摘要:

ANALYTICAL PROCEEDINGS, MARCH 1988, VOL 25 103 NIR Workshop March 24, 1988, Chorleywood A Near Infrared Reflectance Workshop will be organised by the Flour Milling and Baking Research Association at their new conference facility. Dr. Brian Osborne will be in the Chair, and the emphasis will be on practical rather than purely theoret- ical aspects, providing an update on NIR technology as applied to the cereals indus- tries. For further information please contact Mrs. Pat Beeston. Conference Officer, FMBRA, Chorleywood, Hertfordshire WD3 5SH. Loughborough Short Courses July, 1988, Loughborough The following Short Courses will be organised by the Department of Chem- istry of the University of Technology. “High Performance Liquid Chromato- graphy,” July 4-8, 1988. Fee f450 includ- ing residence and all meals (2420 if paid in advance); non-residence f330 (f300 if paid in advance). “Statistics for Analyti- cal Chemistry,” July 5-8, 1988. Fee f360 including residence (f330 if paid in advance); non-residence $265 (f240 if paid in advance). “Flow Injection Analy- sis,” July 6-8, 1988. Fee f305 including residence and all meals (f280 if paid in advance). “Modern Approaches to Ultraviolet and Fluorescence Spec- trometry,” July 15, 1988. Fee f450 includ- ing residence and all meals (f420 if paid in advance); non-residents f330 (f300 if paid in advance). Further details available from Mrs. J. E. Stirling, Department of Chemistry, Loughborough University of Technology, Loughborough , Leicestershire LEl 1 3TU.

ISSN:0144-557X    

DOI:10.1039/AP9882500103

出版商:RSC    

年代:1988

数据来源: RSC