摘要:

766 Analyst, August, 1979, Vol. 104, $$I. 766-770 1001 0 Studies in Gas Chromatography - Chemical lonisation Mass Spectrometry of Some Silicate Anions -1 I SiO,(TMS)Pr, I I . 1 L U I. II I 295 , , , , , 1 , 1 ' 1 1 1 1 1 1 1 J. B. Addison Atlantic Regional Laboratory, National Research Couiacil of Canada, Halifax, Nova Scotia, Canada, B3H 321 lo" 8 a- c 0- ; 100- C -0 a .- c m al - l r Chemical ionisation mass spectrometry of trimet hylsilyl derivatives of the anions sio44-, Si20,6--, Si,0Q6- and Si3010*- has been studied using isobutane as the reagent gas. The protonated molecular ion was observed in all recorded spectra. Also, the chemical ionisation spectra showed many fewer fragment peaks when compared with the corresponding electron impact spectra. - Si0,(TMS),Pr2 1 I' !' ' 8 ' 1 ' I ' I' 1 ' 8 SiO,(TMS),Pr c Keywords : Chemical ionisation mass spectronzetry ; silicate anions A method for the gas-chromatographic separation of silicate anions was first described by Lentz,ls2 and involved the formation of volatile trimethylsilyl (TMS) derivatives of the silicate anions.This TMS method was improved upon by Wu et aZ.3 in order to obtain total separation of sioq4-, Si,0,6- and Si,0,6- anions and partial separation of higher anions such as Si,O,,8- and Si,0,,8-. High-resolution mass spectrometry has been used to identify the chromatographically separated TMS derivatives,s but the molecular ion was either of very low abundance or not detected. This work reports the chemical ionisation mass spectrometric analysis of the TMS derivatives of silicate anions to yield protonated molecular ions in very high abundance and to facilitate easy identification of such anions.I I 325 I i , - - I 355 1 I , 1 Fig 1. CI mass spectra of TMS derivatives of anions.ADDISON 767 Experimental The gas chromatographic - mass spectrometric analyses were performed on a Finnigan 4000 mass spectrometer, interfaced via an all-glass jet separator to a Finnigan 9610 gas chromatograph. Chromatographic separation was made on a 12 ft x Q in stainless-steel column, packed with ultraphase 3% SE-30 on SO-100-mesh Gas-Chrom W. The gas- chromatographic conditions were as follows : column temperature, programmed from 60 to 240 "C at the rate of 6 "C min-l; injection port temperature, 200 "C; helium flow-rate, 20mlmin-l; and the separator and transfer line temperatures, maintained at 280 and 230" C, respectively. The mass spectrometer parameters were : electron energy, 70 eV; ion source temperature, 270 "C; source pressure, 8 x Torr; emission current, 0.35 mA; and scan speed, 3 s per scan.Automatic repetitive scanning was applied in the recording of mass spectrometer data and a Finnigan Incos data system was subsequently used to process the data. Isobutane was used as the reagent gas for gas chromatography - chemical ionisation mass spectrometry and was used as a make-up gas with helium as the carrier gas; the isobutane source pressure used was 0.45Torr. An enhancement technique4 that performs an auto- matic background subtraction was applied to all chemical ionisation (CI) spectra reported.Perfluorokerosene was the calibration substance on which the mass assignments were based. For the detailed preparation of the TMS derivatives, see reference 3. 100 0 100 s o . 100 0 m TJ - 2 - m o m 0 .- r 100 0 100 0 i - - - r r l t , t 200 300 400 5 00 600 171 k Fig. 2. CI mass spectra of TMS derivatives of Si20,+ anions.768 ADDISON: GAS CHROMATOGRAPHY - CHEMICAL IONISATION Analyst, VOZ. 104 Results and Discussion Figs. 1-4 illustrate the CI and electron impact (EI) mass spectra of the derivatives that were clearly separated by gas chromatography. Fig. 5 shows a typical total ion current o l ~ - l ~ ~ , 1 ' 1 * 1 ~ ~ - 1 1 ' - ' , 1' I , I I , I , I l l b l I 400 500 I I 600 ' 680 200 300 m/e Fig. 3. CI mass spectra of TMS derivatives of Si,0Q6- and Si,O,,,s- anions. Si,O,(TMS), Pr4 471 (X 20) 100 - Si,O,(TMS), Pr, % 100 - Si,O,(TMS),Pr, 531 (X20) 100 - Si,O,(TMS),Pr 561 (X20) 200 300 400 500 m/e 600 Fig.4. EI mass spectra of TMS derivatives of SG0,6- anions.August, 19 79 0 MASS SPECTROMETRY OF SOME SILICATE ANIONS 769 5 4 1 5 .O 10.0 15.0 T ime/m in Fig. 5. Typical total ion current chromatogram of TMS derivatives of silicate anions. Peaks: 1, Si,O,(TMS),Pr,; 2, Si,O,(TMS),Pr,; 3, SizO,(TMS),Prz; 4, Si,O,(TMS),Pr; 5, Si,O,- (TMS),; 6, Si,O,(TMS),Pr; and 7, Si,O,(TMS),. chromatogram, which is a plot of the accumulated ion current from the mass spectrometer source of the gas chromatograph effluent 'oemus time. The [M + I]+ ions, the protonated molecular ions, are present in all CI spectra. Com- parisons of CI and EI spectra indicate that CI spectra have fewer fragment peaks, higher relative abundances and contain sufficient information for the identification of silicate anions present. Table I lists the compounds identified by mass spectrometry and the relative abundances of the protonated molecular ions.The CI spectra show an enhancement of abundances of ions in the molecular ion region; no molecular ions were present in the EI spectra for identification, instead the [M - 15]+ ions seem to be the major structurally significant ions for identification, although they display low abundances. The EI spectra also show the usual characteristic peaks obtained by Butt and Rainey5 using gas chromatography - mass spectrometry, i.e., m/e 73 [Si(CH,),]+, 75, 147, 207 and the [M - 15]+ peak.The reported spectra represent the mass range from 200 to 680 AMU. The [M - 15]+ ions were due to the loss of a methyl radical from the TMS derivatives of TABLE I COMPARISON OF RELATIVE INTENSITIES OF [M + I]+ AND [M - 15]+ IONS I N CI AND El SPECTRA, RESPECTIVELY Relative intensity, 74 Compound 62, [M + 134- EI, [M - 15:+ m/e, [M + 1]+ A SiO,(TMS)Pr, . . .. 4.25 22.78 295 SiO,(TMS),Pr, . . . . 53.13 13.73 325 SiO,(TMS), . . . . 11.78 1.48 385 Si,O, (TMS) ,Pr, . . 58.70 1.11 487 Si,O,(TMS) 3Pr3 . . 100.00 1.51 517 si,O,(TMs), . . . . 64.00 1.57 607 Si,O,(TMS),Pr . . . . 32.26 0.99 637 Si,O,(TMS)s . . . . 100.00 2.08 667 Si,O,,(TMS),Pr, . . 100.00 - 649 SiO,(TMS),Pr . . . . 100.00 4.17 355 Si,O,(TMS) 4Pr, . . 73.00 1.27 547 Si,O, (TMS) ,Pr .. . . 100.00 1.76 577770 ADDISON the corresponding silicate anions, as a result of direct electron-impact ionisation. On the other hand, the more gentle ionisation of these derivatives yielded the protonated molecular ions; for example, CI involved electron impact of the isobutane reagent gas, resulting in the ionisation of the reagent gas and followed by transfer of a proton to a silicate anion. All of the TMS derivatives studied, except SiO,(TMS)Pr,, gave intense [M + 1]+ peaks. The reagent used in this work was a good Bronsted acid with respect to the above silicate anions, and transfer of a proton to the anions was very efficient. The low relative abundance of the [M + 1]+ ion for the SiO,(TMS)Pr, anion may be due to the greater difference in proton affinities between this anion and the reagent gas, with the result that there is a transfer of a greater amount of energy to this anion from the reagent gas, followed by a high degree of fragmentation of the anion and less efficient protonation.The high degree of fragmentation is supported by the high relative abundance of the [M - 15]+ ion present in the EI spectrum of this particular anion. It is worth mentioning that apart from the formation of completely derivatised TMS products, incomplete TMS derivatisation yielded tetramethylsilylpropyl [(TMS).Pr,]+ products. These products are formed by the replacement of SiMe, groups with propyl (Pr) groups during derivatisation in propan-2-01 and have been reported elsewhere.6-* The CI mass spectra of these [(TMS),Pr,]+ anions also show a very high relative abundance of the protonated molecular ion [M + 13’ and identification is fairly straightforward. Conclusion Isobutane chemical ionisation mass spectrometry offers a very useful approach to the identification of silicate ions.The production of the protonated molecular ions of very high relative abundance and the formation of fewer fragment peaks with enough chemical formation for characterisation make this CI technique far superior to the EI technique, which gives the [M - 15]+ ion as the most significant peak for identification and no mole- cular ion. The author is grateful to the National Research Council of Canada, Atlantic Regional Laboratory, for financial support, and to Dr. Harry Calhoun of this laboratory for providing the silicate anions. References 1. 2. 3. 4. 5. 6. Lentz, C. W., Inorg. Chem., 1964, 3, 574. Lentz, C. W., Special Report, Highway Research Board, Washington, D.C., 1966, 90, 269. Wu, F. F. H., Gotz, J., Jamieson, W. D., and Masson, C. R., J . Chromat., 1970, 48, 515. Incos Data Operational Manual, Finnigan, Sunnyvale, Calif. , 1978. Butt, W. C., and Rainey, W. T., Analyt. Chern., 1971, 43, 538. Masson, C. R., Jamieson, W. D., and Mason, F., in Jeffes, J. H. E., and Tait, R. J., Editors, “Physical Chemistry of Process Metallurgy : The Richardson Conference,” Institute of Mining and Metallurgy, London, 1974, p. 223. Eglinton, G., Firth, J. N. M., and Welters, B. L., Chem. Geol., 1974, 13, 125. Iler, R. K., “The Colloid Chemistry of Silica and Silicates,” Cornell University Press, Ithaca, N.Y., Received Febrztary 13th. 1979 Accepted March 14th, 1979 7. 8. 1955.

ISSN:0003-2654    

DOI:10.1039/AN9790400766

出版商:RSC    

年代:1979

数据来源: RSC

摘要:

Analyst, August, 1979, Vol. 104, pp. 771-777 771 Internal Standardisation and its Value in the Assessment of the Suitability of the Column for Quantitative High-performance Liquid Chromatography R. A. Moore Lilly Research Centre Ltd., Ed Wood Manor, Windlesham, Surrey, G U20 6PH The requirement that the best precision be obtained for a pharmaceutical raw material assay using automated high-performance liquid chromatography, with valve injection, over long runs, suggested that the inclusion of an internal standard would be beneficial in allowing an additional method for the calcula- tion of results. The accurate volume injection of the valve and loop system thus permitted standard calibration and subsequent sample evaluation by the peak-height ratio method as well as from absolute peak heights alone.After a number of experiments, column deterioration rendered the assays invalid, and retrospective analysis of the data demonstrated the merits of internal standard related parameters in showing the decline in performance of the column. The major benefits derived from the approach adopted were criteria for j udging the acceptability of the assay results, without excessive replication for a particular sample, and prior warning of the need to re-pack the column. Keywords : Quantitative high-perfomance liquid chromatography ; internal standardisation ; column assessment ; pharmaceutical analysis ; automatic injection In contrast to the development of gas - liquid chromatography (GLC), the more recent technique of high-performance liquid chromatography (HPLC) has grown in an environment where the options for quantitation, via a particular means of detection, are fairly numerous.Sample valves for HPLC enable accurate volumes of liquid to be injected and therefore an internal standard may appear to be of much less value than in GLC procedures. Peak heights or areas can be measured, either manually or by other means such a s with an integrator or computer . In the study described here, to aid the appraisal of quantitative HPLC, using an automatic injector with a sample valve, for the routine assay of high-purity samples (>95%) of a pharmaceutical raw material, an internal standard was also included to allow an alternative method of calculation, and manual peak-height measurements were made. The column had previously been used in several similar applications.The method resolved the drug compound of interest, an aromatic piperazine derivative with only tertiary nitrogen atoms, from potential process impurities; in the absence of a suitable related compound phenyl l-adamantyl ketone was chosen as the internal standard, eluting in a typically “clean” region of the chromatogram. Previous reportsl*2 have established the effects of variation of the major factors, namely mobile-phase composition and flow-rate, sample load and temperature, on the precision of HPLC. In the work reported here, temperature was the only factor not routinely controlled, such that a significant effect on the chromatography was to be expected (during a 12-h overnight run a variation from 23 to 12 to 15 “C might be obtained), and frequent standard injections were therefore made. One of the five experiments was also performed under temperature-controlled conditions. Experimental Reagents phosphate dihydrate were of AnalaR quality.The water used was previously distilled, and methanol and sodium dihydrogen ortho- Drug standard. The standard used was a specially purified batch characterised at 99.7%772 MOORE: INTERNAL STANDARDISATION IN THE ASSESSMENT OF THE ArtaZyst, VoZ. 104 purity (by thin-layer chromatography, differential scanning calorimetry, normalised GLC, normalised HPLC and water and solvent level determinations), and this value was used in assay calculations. A 10 mg ml-l solution of phenyl 1-adamantyl ketone (one peak only by HPLC) in methanol.A 0.05 M solution of sodium dihydrogen orthophosphate dihydrate in methanol - water (70 + 30), de-gassed under vacuum for about 1 min before use. Internal standard solution. Eluent. Apparatus Solutions were sampled by a microprocessor-operated automatic injector with a motorised sample valve and a 10-pl sample loop (Micromeritics, Model 725). A dual-piston recipro- cating pump with a solvent inlet filter (Milton Roy, Constametric 11) delivered a constant flow of 1.0 ml min-1 (requiring approximately 950 lb in-2) to a 125 x 4.9 mm i.d. stainless- steel column packed with 5-pm octadecylsilyl-bonded silica (Spherisorb SODS, Phase Separations Ltd.). The column was prepared by upwards packing from a propan-2-01 slurry using a stirred-slurry column packer (Micromeritics, Model 705) and a pneumatic amplifier pump (Haskel MCP-110) at a pressure of 4500 lb in-2. The column end-fittings included stainless-steel mesh discs of 8-pm pore diameter (two discs at the outlet end) to retain the packing.Detection was effected by ultraviolet absorption using a variable-wavelength detector with a 10-p1 flow cell, set at 240nm and 0.5 absorbance unit full-scale deflection (f.s.d.) (Cecil 212). Chromatograms were recorded at 10 mV f.s.d. (Servoscribe RE 541.20). Procedure General 9rocedure Duplicate sample masses (90 or 45 mg for experiment 1 or 3, respectively) were transferred into calibrated flasks, dissolved in a pipetted amount of internal standard solution and diluted to volume with methanol - water (70 + 30) to give solutions with nominal concentra- tions of 0.45 mg ml-l of drug and 0.50 mg ml-l of internal standard.Three standard masses, corresponding approximately to the sample mass (m), 1.1 m and 0.9 m, were treated in a similar manner to the samples. Before each run, repeated injections of a standard solution at 12-min intervals were used to check for system equilibration; during the run standards were injected frequently between samples, triplicate injections being taken from each vial. All experiments were allowed to run overnight, about 12 h being necessary. For each vial the three peak heights of drug and internal standard were measured and the values averaged to give a mean value for both components. The mean peak-height values were subsequently used to construct standard calibrations for both peak height and peak-height ratio, and to calculate sample assays.Over the course of each experiment a series of the three standard solutions was injected four or five times to allow adequate standardisation of all test solutions. A calibration graph was drawn for each series of standards, resulting in four or five graphs for peak height and an identical number for peak-height ratio, Sample assays were calculated by reference to the relevant graphs for peak height and peak-height ratio. Experiment 1 Duplicate masses from each of four sample batches gave eight test solutions, T1 to T8, which were arranged with the three standard solutions (STD1 to STD3) in the following injection sequence : STD1, STD2, STD3, T1, T2, STD1, STD2, STD3, T3, T4, STD1, etc.Y Y Series 1 Series 2 Assays were evaluated based on weighted averages from the surrounding calibration graphs, eg., the mean peak-height value for T1 was converted into an active content from the peak- height calibrations from both the Series 1 and Series 2 standards. The correct intermediate value was then calculated on the assumption that the peak height for each standard solution had changed linearly with time over the period includingAugust, 1979 SUITABILITY OF THE COLUMN FOR QUANTITATIVE HPLC 773 both series of standards. The weighted value was used for calculation of the T1 assay via peak height. A similar procedure was then followed to obtain the peak-height ratio assay. However, the time difference between the first injection of STD 1 in Series 1 and the third injection of STD 3 in Series 2 was nearly 5 h and a significant drift resulted between calibra- tion graphs ; therefore, an alternative arrangement of standard and test solutions was con- sidered to be desirable in successive experiments.Experiment 2 The solutions from experiment 1 were re-run with a new vial sequence: Series 2 Series 4 7 r STD1, T1, STD2, T2, STD3, T3, STD2, T4, STD1, T5, STD2, T6, STD3, T7, STD2, T8, STDl Y- Y Series 1 Series 3 Assays were calculated using the particular standard series bracketing the sample, e.g., the active contents of T2 and T3 were read from Series 1 and Series 2 standard calibration graphs, respectively. Experiment 3 One of the previously used sample batches and three new batches were assayed, duplicate masses being taken through the procedure as in experiment 2, except that the chromato- graphy was performed in a room controlled at 25 4 1 "C.Fresh standard solutions were also prepared. Experiment 4 ture. The solutions from experiment 3 were re-injected without control of the laboratory tempera- Experiment 5 After re-packing the front end of the column, the procedure in experiment 4 was repeated. Results and Discussion In each experiment, for each sample batch, the mean assay result and standard error of the mean (S.E.M.) were calculated from the four constituent assay values (peak height and peak-height ratio methods of calculation on the duplicate sample masses). The results are given in Table I. TABLE I MEAN SAMPLE ASSAYS AND S.E.M. (yo) Sample batch Expt.1 A 97.6 (f0.2) B 98.7 (h0.4) C 97.9 (f0.4) D* 98.3 (f0.5) E - Expt. 2 Expt. 3 Expt. 4 Expt. 5 - - - 97.0 (10.3) 97.9 (f0.5) 96.3 (f0.2) 96.9 (k0.5) 98.6 (f0.8) 100.8 (f1.4) 98.4 (f0.4) 96.2 (fO.6) 94.7 (f0.8) 96.8 (hO.9) - 97.0 (k0.7) 96.9 (f0.9) 96.0 (kO.1) 97.4 (kO.8) 96.2 (& 1.1) 96.0 (1 soln. only) - - - - - - - * Sample D is re-mixed sample B. Close inspection of the chromatograms revealed a decline in resolution of the two peaks from almost base-line separation in experiment 1 to the situation existing in experiment 4,774 MOORE: ‘INTERNAL STANDARDISATION IN THE ASSESSMENT OF THE Analyst, VoZ. 104 Internal standar 9.05 absorbance unit I )rug I 4 R 4.1 R = 3.2 R = 2.4 R = 3.3 Fig. 1. Typical chromatograms obtained in (a) experiment 1, (b) experiment 3, (c) experiment 4 (equilibration injection) and (d) experiment 4 (end of run).where the first few equilibration injections resulted in distorted, split or shouldering peaks and later equilibration injections yielded “normal” peaks with increased inter-peak valley heights above the base line. Fig. 1 illustrates some typical chromatograms obtained during the lifetime of the column, together with the resolution values (R) for the two peaks. At the end of experiment 4, removal of the column front-end fitting revealed an approxi- mately 1 mm deep depression in the packing, and this was re-filled manually with a fresh, stiff slurry of the packing material. Subsequent equilibration injections showed that the column was again giving good resolution and peak shape and experiment 5 was performed.The column settling effect, as observed above and previously in this laboratory, is likely to be attributable to one or more of three factors: (i) poor initial packing; (ii) chemical attack on the bonded phase or the silica matrix of the packing; and (iii) expulsion of packing from the column or fracture of the particles, or a combined effect. A previous exercise to compare the packing of reversed-phase columns from propan-2-01 or aqueous - methanolic sodium acetate slurries produced columns of similar efficiency and with similar settling rates in use; the latter method was recommended by the packing manufacturer to overcome electrostatic attraction between particles to aid dispersion. In the work described here, the pressure used to pack the column was far greater than that used in assay work.Consequently, the first factor above was not considered to contribute towards the settling. Chemical attack should be considered unlikely with an eluent of pH approximately 5.4, although other workers have reported settling of reversed-phase columns with phosphate buffers. According to the supplier’s instructions, two 8 pm mesh discs were used for retention of the packing, but may have allowed gradual loss of particles, especially any “fines,” and the third factor is expected to yield the major contribution to column settling. In a current investigation, 2 pm mesh discs are being used for retention of the column packing in place of 8pm mesh discs to compare column lifetimes. Comparative results should shed light on the processes responsible for column settling, including any contribution from the second factor mentioned above.Table I1 lists the parameters that best illustrate the over-all quality of the chromatography.August, 1979 SUITABILITY OF THE COLUMN FOR QUANTITATIVE HPLC 775 Mean range of triplicates The values should show little temperature effect, being dependent on measurements made over only a 0.5-h period. The results reflect the reproducibility of volume delivery of the automatic injector as well as chromatographic variation. The peak-height measurements for the drug show that very good precision is obtainable; the corresponding precision for the internal standard was not usually as good, as can be seen from the peak-height ratio values.In experiment 4, where the quality of the chromatography was low, agreement between triplicate results for peak-height measurements was poor. The less precise peak-height results for triplicate determinations in experiment 5, compared with that obtained initially, is attributed to decreasing performance of the automatic injector, as a short time later the sample valve had to be freed of a blockage. TABLE I1 PARAMETERS DEPENDENT ON CHROMATOGRAPHIC PERFORMANCE PH = peak height; PR = peak-height ratio. Mean range of triplicates (%) calibrations, % * Range of standard Expt. #-Ap, Mean difference (MD)t Mean bias (MB)$ NO. PHdrug PR PHdrug PR between methods between methods 1 0.6 1.5 3.6 1.5 0.8 - 0.4 2 1 .o 1.8 3.5 2.0 1.0 +0.05 3 0.8 0.8 1.2 3.3 2.1 -0.3 4 3.3 3.8 2.9 7.0 2.5 +1.1 5 2.3 2.3 4.5 1.4 0.3 -0.1 * Range of values measured a t standard mass m.t M D = $MI3 = number of solutions number-of solutions ’ z*e” +ve Or -ve. . ZPH assav - PR assav . Range of standard calibrations Good straight-line calibration graphs were obtained, peak-height ratio graphs giving better fits to the points than peak-height graphs. The peak-height graphs displayed an irregular sequence with time, whereas those for peak-height ratio were in regular ascending or descending order. Thus, while peak-height values must be affected by fluctuating temperature, which would determine the actual masses of solutes being injected, peak-height ratio values for two adjacent resolved peaks should be much less temperature dependent. The range of peak-height values therefore illustrates “chromatographic drift” and “temperature drift” over the duration of an experiment, while the range of peak-height ratio values reflects largely a “chromatographic drift” factor that increases as the column deteriorates.It is therefore seen that : In experiment 3 control of the temperature results in a 1.2% peak-height range compared with 3% and above for the other experiments, while the peak-height ratio range continues the increasing trend shown from experiments 1 and 2. In experiment 4, where the peak-height range is comparable to those of experiments 1 and 2, the peak-height ratio range is in keeping with the degraded state of the column. Renewal of the column front end yields a peak-height range again of the expected order, with a low peak-height ratio range typical of a stable column.(i) (ii) (iii) Mean diflerence (MD) and mean bias (MB) These terms are determined solely from assay differences between the two methods of calculation on a particular solution (either ignoring or using positive and negative signs), and are therefore virtually independent of sample and standard weighing errors. For a particular experiment these terms will provide equivalent information on the chromatography irrespective of the combination of samples and batches used. The two parameters in combination readily demonstrate the acceptability of results from a given experiment. Thus, the MD for experiment 3 suggests a wider assay variation than is desirable, although776 MOORE: INTERNAL STANDARDISATION IN THE ASSESSMENT OF THE Analyst, VoZ.104 no strong bias between methods is shown. Further column deterioration in experiment 4 leads to unacceptably high values for MD and MB, correlating with the unreliable nature of the assays. Examination of assay S.E.M. values for each experiment (Table I) reveals the same trend as with MD values, although the former will also include variation due to weighing operations. It is useful to note that column efficiency measurements, with respect to the two com- ponents of interest, are of limited worth in the evaluation of the column for quantitation purposes. The column efficiencies (plate numbers), calculated from peak width at half- height measurements, are given in Table 111. A gradual decline in column performance was found between experiments, but only in experiment 4 was a significant change shown during the run, allowing for an estimated measurement error of &loyo.TABLE 111 COLUMN EFFICIENCIES Plate number Expt. I A I No. Drug Internal standard 1 1600 4 000 2 1400 3 300 3 1400 2 800 4 (a) First equilibration injections < 1000 < 1000 (b) Start of run proper 1200 1900 (c) End of run 1400 2 600 5 1400 3 200 The efficiency for the internal standard is much more sensitive to chromatographic changes than for the drug and approximately 3000 plate numbers for the internal standard might appear to be a criterion for acceptability of the column for quantitative analysis. By this method of assessment the column is of marginal acceptability for quantitation purposes in experiment 3, a little better for experiments 2 and 5 and much better for experiment 1.However, the previously described methods of evaluation demonstrate the significantly better chromatographic stability and precision of results in experiments 2 and 5, compared with experiment 3, and even suggest superior stability for experiment 5 compared with experiment 1. Column efficiency for the internal standard is therefore an insensitive gauge for the measurement of the suitability of columns for quantitation purposes. Excluding the results from experiment 4, which are shown to be invalid, over-all sample assays with corresponding values via peak height and peak-height ratio are given in Table TTT TABLE IV OVER-ALL MEAN SAMPLE ASSAYS AND S.E.M. (yo) PH = peak height; PR = peak-height ratio.Sample batch A B C D" E F G Grand mean 98.3 (f0.3) 97.1 (f0.3) 98.1 (f0.3) 96.5 (&0.5) 96.5 (f0.4) 96.9 (&0.6) 97.3 (f0.2) Mean via PH 97.2 ( i 0 . 2 ) 98.7 ( i 0 . 4 ) 96.8 (10.5) 97.5 (k0.4) 96.0 ( 5 0 . 6 ) 96.6 ( i 0 . 7 ) 97.7 (k0.8) * Sample D is re-mixed sample B. Mean via PR 97.3 (f0.4) 97.8 ( 5 0 . 5 ) 97.4 ( 5 0 . 5 ) 98.6 (f0.4) 96.9 (f0.7) 96.4 (k0.4) 96.1 (50.8) Application of Snedecor's F-test and Student's t-test shows that for each sample the individual peak height and peak-height ratio assays belong to the same population. In each instance the grand mean assays are within 1% of the sample purity predicted by the other techniques used to characterise the standard.Aagust, 1979 SUITABILITY OF THE COLUMN FOR QUANTITATIVE HPLC 777 Chromatographic precision was also assessed from repeated equilibration injections of a standard solution, yielding typical coefficients of variation of peak height and peak-height ratio of 1.4% and 0.6%, respectively, from six injections in experiment 2, and 0.4% and 0.4% from nine injections in the temperature-controlled experiment 3.These results compare f avourably with literature ~ a l u e s . l , ~ With the exception of one sample (D), all assay results were obtained from only two weighings per sample, and therefore any statistical treatment of the results must be viewed with caution. However, the results obtained suggest that in a simple analytical procedure in which the error is mainly chromatographic, a high precision of results should be readily achieved in a single run with a good column, and adequate control as above using three or four weighings per sample; the use of an internal standard permits the assessment of the column and evaluation of the chromatographic variation of assays. Complete automation of the procedure, including measurement and calculation , would minimise the tedium associated with the regular standardisation necessary in the absence of adequate temperature control.Conclusions An internal standard is not essential for precise quantitation in HPLC, corroborating the findings of other worker^.^,^ When the greatest accuracy and precision are required, how- ever, and for repetitive assays, particularly over long runs, the inclusion of an internal standard is recommended, in order to allow calculation via both peak-height ratio and peak height alone, because of the major advantages accrued.Unless strict temperature control is observed, the variation of peak-height values with time for a given solution may be erratic, while the corresponding variation of peak-height ratio measurements gives an estimate of “chromatographic drift. ” When experimental parameters other than temperature are controlled, the “chromatographic drift” evaluates the column stability, which is the important factor governing the precision of quantitative analyses. For a routine assay a maximum variation of peak-height ratio with time may be established for acceptable column stability. Likewise, the “mean difference” and “mean bias” factors based on comparative peak height and peak-height ratio assays for each sample solution (preferably eight or more) also assess the acceptability of assay results and give an objective prior warning of the need to re-pack a column.Thus, values of MD up to 1.5% and MB up to 0.5% might be regarded as indicating acceptable assays. Until column packing methodology, materials and equipment are much more standardised, such a means of following column lifetimes should prove of great help in quantitative HPLC. It is important to monitor the complete duration of an experiment for over-all appraisal of assay precision; calculations based on initial repeated injections of a solution are insuffi- cient, as can be seen in experiment 3 where very precise equilibration injections were observed, although the peak-height ratio range and MD values showed continuing deterioration of the column and borderline acceptability of results. A mean assay difference of about 1.5% between duplicate weighings of the four samples by peak height or peak-height ratio calcula- tion methods also indicated the poor precision. The application of both peak height and peak-height ratio calculations therefore provides more meaningful information on column behaviour and assay precision than can otherwise be obtained by either technique alone on a given number of samples. Mr. G. F. Snook is thanked for helpful discussions. References 1. 2. 3. 4. Scott, R. P. W., and Reese, C. E., J . Chromat., 1977, 138, 283. Bakalyar, S. R., and Henry, R. A., J . Chromat., 1976, 126, 327. Beyer, W. F., and Gleason, D. D., J . Pharm. Sci., 1975, 64, 1567. Kern, H., and Imhof, K., Int. Lab., 1978, Jan/Feb., 65. Received October 17th, 1978 Accepted February 28th, 1979

ISSN:0003-2654    

DOI:10.1039/AN9790400771

出版商:RSC    

年代:1979

数据来源: RSC

摘要:

775 Analyst, August, 1979, Vol. 104, pp. 778-787 Analytical Methods Committee REPORT PREPARED BY THE METALLIC IMPURITIES IN ORGANIC MATTER SUB-COMMITTEE Determination of Small Amounts of Selenium in Organic Matter Methods for determining small amounts of selenium in organic matter have been examined. Determination first requires oxidative destruction of organic matter. The usual oxidising systems employing sulphuric acid initially, with hydrogen peroxide, nitric and/or perchloric acids, gave very low recoveries of selenium when high proportions of fats were present in the samples. Con- tinuous combustion and oxygen flask methods were also unsatisfactory. A wet-oxidation procedure in which nitric and perchloric acids (5 + 1) were followed by nitric and sulphuric acids gave satisfactory results.It was important to ensure that selenium was present as selenium(1V) by boiling the solution with hydrochloric acid after the oxidation. Colorimetric, gas - liquid chromatographic, fluorimetric and atomic spectroscopic methods were con- sidered for the selenium determination; of these, the last two were selected for collaborative trials. For the fluorimetric finish, the solution, after oxida- tion, was treated with 2,S-diaminonaphthalene reagent, the complex was extracted with cyclohexane or dekalin and the fluorescence in the organic phase was measured in a spectrofluorimeter with excitation at 369 nm and emission reading at 625 nm. Atomic-absorption or fluorescence measure- ments were made after hydrogen selenide generation and atomisation.The precisions of analyses for samples with selenium contents from less than 0.1 to 10 pgg-1 by the fluorimetric and hydride generation methods are illus- trated. Keywords : Selenium determination ; wet oxidation ; fluorimetry ; atomic spectroscopy The Analytical Methods Committee has received and approved for publication the following Report from its Metallic Impurities in Organic Matter Sub-committee. Report The constitution of the Sub-committee responsible for the preparation of this report was : Dr. L. E. Coles (Chairman until 1978), Mr. C. A. Watson (Chairman from 1978), Mr. W. Cassidy, Mr. P. N. Coleman (Sub-committee Secretary from 1977), Mr. R. E. Collier, Dr. W. H. Evans, Mr. M. T. Friend (from 1978), Mr. S. Greenfield, Mr. W. H. Hill, Mr. B. E. Pearce, Mr.W. L. Sheppard (resigned 1977) and Dr. J. M. Skinner, with Mr. P. W. Shallis as Secretary (until 1977) and Dr. N. W. Hanson as Secretary (from 1977). Introduction Preliminary investigations by the Sub-committee indicated that of a number of methods that have been described for the determination of small amounts of selenium, two methods in particular merited further investigation. The fluorimetric method as described by Hall and Guptal appeared to possess sufficient sensitivity, while the numerous methods based on reduction to the hydride followed by its decomposition to selenium atoms and atomic spectroscopic measurements were also thought to cover the range of interest. Before any of these methods of determination could be employed it was necessary to find a method of sample preparation that would remove the organic matrix and leave the selenium in a suitable state for fluorimetric or atomic spectroscopic determination.ANALYTICAL METHODS COMMITTEE 779 Experimental Sample Preparation It is usually convenient to wet oxidise organic matter prior to the determination of trace metals and several Sub-committee members undertook preliminary studies using dried milk as the substrate.Early results using previously recommended procedures2 based on sulphuric acid - hydrogen peroxide, sulphuric acid - nitric acid and sulphuric acid - nitric acid - perchloric acid were not encouraging, very low recoveries being obtained in the presence of samples that contained high proportions of fats. The situation was not improved by the substitution of a Bethge apparatus for the Kjeldahl flasks that were originally used.A radiochemical study by one Sub-committee member demonstrated that the selenium had not been lost from the system, but was present in a state in which it was not determined by the fluori- metric method, Similar results were also obtained by Levesque and Vendette.3 TABLE I DETERMINATION OF SELENIUM AFTER SAMPLE OXIDATION IN THE OXYGEN FLASK Test sample Milk + 10 p.p.m. of Se (0.1-g sample) . . .. .. Condensed milk + 1 p.p.m. of Se (l-g sample) . . 10 pg of Se . . .. .. .. .. .. ,. 10 pg of Se .. .. .. .. .. .. .. Wheat + 40 p.p.m. of Se (0.25-g sample) Wheat + 40 p.p.m. of Se (0.25-g sample) Milk powder + 40 p.p.m. of Se (0.25-g sample) Milk powder + 40 p.p.m. of Se (0.25-g sample) Milk powder + 20 p.p.m. of Se (0.5-g sample) ... . .. . . .. .. .. . . .. Condensed milk + 1 p.p.m. of Se (0.3-g sample) . . Final measurement method* H ydride/AAS Automatic hydride/AAS H ydride/AAS Fluorimetric H ydride/AAS Fluorimetric H ydride/AAS Fluorimetric Automatic hydride/AAS Flu orime tric Se recovery, 60-80 20-40 4 1-68 86-95 48-62 95 48-78 96 75-92 71 * AAS = atomic-absorption spectrometry. t In Tables 1-111, a single figure is the result obtained by one laboratory. A range represents the extreme results obtained by the collaborating laboratories. One of the collaborating laboratories investigated the use of a continuous combustion apparatus, but the results were still unsatisfactory. Three laboratories investigated the use of oxygen flask combustion with various samples and the results from these investigations are shown in Table I.As can be seen, these results were encouraging but still erratic and the Sub-committee was of the opinion that, as no obvious advantage had been demonstrated by these other methods of sample preparation and as they were not so widely available as wet-oxidation techniques, further efforts should be devoted to producing a satisfactory method employing the wet-oxidation technique. A sample of condensed milk that had been spiked with 1 p.p.m. of selenium was analysed by using a number of sample preparation techniques and an automated hydride generation - atomic-absorption finish. The results of this investigation are reported in Table 11. TABLE I1 DETERMINATION OF SELENIUM AFTER VARIOUS METHODS OF SAMPLE OXIDATION All determinations were carried out on a l-g condensed milk sample containing 1 p.p.m. of selenium.Method of oxidation Oxygen flask . . . . .. .. .. .. .. .. .. Oxygen bomb . . .. .. .. .. .. .. .. .. Wet oxidation - hydrogen peroxide/sulphuric acid .. .. .. Wet oxidation - nitric acid/sulphuric acid . . .. .. .. .. Wet oxidation - perchloric acid/nitric acid . . .. .. .. .. Wet oxidation - perchloric acid/fuming nitric acid .. .. .. Wet oxidation - nitric acid/perchloric acid/sulphuric acid, Kjeldahl flask Wet oxidation - nitric acid/perchloric acid/sulphuric acid, block digestion Se recovery, % .. 20-40 .. <20 .. <8 .. 4-16 .. 91-96* .. 86-9 1 * .. 75-80 .. 85-88 * These results are artificially high as the presence of small but variable amounts of nitric acid in the digest causes an enhancement of up to 30%.780 ANALYTICAL METHODS COMMITTEE : DETERMINATION OF Analyst, VoZ.104 Interpretation of Table I1 requires care, as the results obtained when using perchloric - nitric acid or perchloric - fuming nitric acid are artifically high by about 30% owing to an enhancement effect that occurs with the subsequent selenium determination when some nitric acid remains in the digest. The results obtained when using the nitric - perchloric - sulphuric acid digestion were thought to be sufficiently good to warrant a collaborative trial and a procedure based on that used by the Association of Official Analytical Chemists in Washingt~n,~~~ with the subsequent modification,6 was employed.The particular feature of this method, compared with those previously tried, was that sulphuric acid was added only after most of the sample had been oxidised by treatment with nitric and perchloric acids. The recovery of selenium from condensed milk is reported in Table 111; some of the results were obtained after the procedure had been modified to include a stage in which the digest was boiled with hydrochloric acid to ensure that all selenium was present as selenium(1V). This modification was used throughout the subsequent exercise, although examination of the results in Table I11 fails to show any statistically significant improvement. It should be noted that the hydrochloric acid concentration must not exceed 6 N, as selenium is volatilised as selenium( IV) chloride from boiling concentrated hydrochloric acid solutions.7 Recent work by Bunton et aZ.* has confirmed the importance of the selenium being present as selenium(1V) and they ensured this by treating the digest with hydrogen peroxide after removal of the perchloric acid.TABLE I11 DETERMINATION OF SELENIUM I N CONDENSED MILK SAMPLES USING A PROCEDURE BASED ON THAT GIVEN IN REFERENCES 4-6 All determinations were carried out on a condensed milk sample Laboratory Finish* Se recovery, % A Fluorimetric 75-83 B Fluorimetric 91-1 11 C Hydride/A AS 90-10s Dt Et Bt containing 1 p.p.m. of selenium. Automatic hydride/AAS 80-88 H ydride/ AFS 87-98 Fluorimetric 102-1 10 * AAS = atomic-absorption spectrometry; AFS = atomic- t Digests boiled with 4 N hydrochloric acid to ensure conversion fluorescence spectrometry.of all selenium into selenium( IV) . Selenium Determination The Sub-committee examined several methods for selenium determination, including a colorimetric procedureg and a gas - liquid chromatographic (GLC) procedure.lOJ1 It became clear, however, that only two techniques would provide the sensitivity required for the analysis of foodstuffs and additives, combined with the required degree of reliability. These were the fluorimetric method described by Hall and Guptal and methods based on reduction to the hydride followed by decomposition to selenium atoms in either a heated tube or a flame and measurement of the atomic absorption or fluorescence produced by suitable irradi- ation of the atom cloud. Because of the wide range of sensitivities of fluorimeters and spectrofluorimeters it was not possible to recommend precise conditions for making a given determination, as the calibration range and hence the mass of sample required vary from instrument to instru- ment.For methods based on hydride generation it was even more difficult to provide a recommended procedure, as not only do the sensitivities of flame spectrometers vary greatly, but the type of atom cloud produced also depends on the design of the reduction vessel, carrier gas flow-rate, flame temperature or tube temperature and the method of sample introduction. Although in the Sub-committee’s experience single-injection atomic- absorption measurements were difficult to perform reliably, the use of instruments with more sophisticated data handling than were available during this work might enable these methods to yield satisfactory results routinely.August, 1979 SMALL AMOUNTS OF SELENIUM I N ORGANIC MATTER 781 One member of the Sub-committee used an automated procedure based on that proposed by Vijan and Wood,12 which has the advantage of producing steady rather than transient signals, leading to good detection limits and better precision than could be obtained by the various “single-injection” methods when using a simple atomic-absorption spectrometer. Another member of the Sub-committee found that the non-atomic absorption signal pro- duced, caused by the large volume of hydrogen and consequent acid carry-over that occurred owing to the high acidity necessary to reduce selenium successfully, was difficult to correct for and he obtained better results by the use of atomic fluorescence.The hydride was decomposed in an argon - hydrogen diffusion flame with irradiation from an electrodeless discharge lamp operated at 40 W in a Broida-type cavity. The relative detection limits obtained by various Sub-committee members using a variety of techniques are reported in Table IV and may be useful as a guide when deciding calibration ranges and sample masses. Laboratory A . . . . B .. .. c . . .. c . . .. D . . .. E .. .. F .. .. G .. .. TABLE IV RELATIVE DETECTION LIMITS FOR SELENIUM BY VARIOUS METHODS Fluorimetric method with single-beam spectrofluorimeter . . 10 1 Fluorimetric method with single-beam spectrofluorimeter . . 20 1 Fluorimetric method with filter fluorimeter .. . . .. 100 1 Hydride generation/AAS in flame-heated silica tube . . .. 10000 2 Automatic hydride generation/AAS in electrically heated silica tube . . . . .. . . .. .. . . 2 3 Hydride generation/AFS in argon - hydrogen diffusion flame 2 2 Fluorimetric method with single-beam spectrofluorimeter . . 7 1 Fluorimetric method with single-beam spectrofluorimeter . . 30 1 Method Detection limiting * Notes t * Detection limits are calculated from the standard deviation of the lowest levels determined by each laboratory and for the purposes of this comparison are defined as twice the standard deviation of these determinations. t Notes- With fluorimetric methods it is possible to use the whole of the digest, hence for a l-g sample mass these figures represent the detection limit in parts per billion (loB).With hydride generation methods the detection limit for a sample depends on the mass of sample that i t is possible to take, the final volume of the digest and the proportion of the digest used in the analysis. The automatic method uses a pumped flow of the digest a t a rate of 3.9 ml min-l. The detection limit reported is the concentration of this sample solution in ng ml-l. 1. 2. 3. The procedures recommended in the Appendix are of necessity very general and the analyst must select sample masses and final volumes of digest to suit the measuring system. Using the recommended sample preparation procedures and making measurements by the various methods outlined above, a number of samples of organic material were analysed. The results of these experiments are given in Tables V-X.In each instance the mean is reported, together with the range, although the latter is replaced by the standard deviation where sufficient results are available. Examination of Tables V-X indicates that similar results are obtained regardless of the method of end determination. In no instance is the mean of the fluorimetric results statistically significantly different from the results obtained by hydride generation. In general, the agreement between laboratories using hydride generation was slightly better than those using the fluorimetric finish, although the number of results is insufficient to allow any conclusions to be drawn regarding the superior precision of either method.How- ever, if Fig. l, which illustrates the relationship of relative standard deviation to selenium content of the various samples, is examined, it would appear that for levels below 100 ng g-1 there may be some advantage in the use of the hydride generation procedures. (The results illustrated in Fig. 1 are based on an inter-laboratory comparison of the data and it must be emphasised that the number of laboratories using each technique was small, particularly for the lower levels of selenium when hydride generation was used.) It is worth noting that although these results were obtained with a wide range of sample matrices, the precision appears to be concentration dependent rather than sample dependent, particularly for the fluorimetric method, where all the results fall close to a single smooth curve. The precision - concentration relationship €or the hydride generation method is less well defined than that782 ANALYTICAL METHODS COMMITTEE : DETERMINATION OF Analyst, V d .104 TABLE V DETERMINATION OF SELENIUM IN SELENIUM YEAST Laboratory Finish A .. . . Fluorimetric B . . . . Fluorimetric C .. . . HydrideIAAS D .. . . HydrideIAAS E .. . . HydrideIAFSt Hydride/AFS$ G .. . . Fluorimetric Selenium content, p.p.m.* 135 (&20%, 3 results) 124 (f4Y0, 3 results) 144 (RSD 3%. 6 results) 149 (f3%, 3 results) 147 (f6%, 3 results) 137 (&ti%, 3 results) 126 (RSD lo%, 5 results) * In Tables V-X, figures in parentheses represent the total range found; where no range is given only a single result was reported.In most instances insufficient results were available t o calculate the standard deviation although this is reported when four or more results were available. Pooling the data for the various methods did not indicate a significant statistical difference between the hydride and fluorimetric methods. RSD = relative standard deviation. t Calibration graph method. $ Standard additions method. TABLE VI DETERMINATION OF SELENIUM IN CATTLE NUTS Laboratory Finish Selenium content, p.p.b. A .. . . Fluorimetric 183 (&lo%, 3 results) B .. . . Fluorimetric 237 (RSD 3.3%, 4 results) C .. . . Fluorimetric 247 (f 13%, 2 results) D .. . . HydrideIAAS 180 (&Ox, 2 results) E .. . . HydrideIAFS 230 ( & S%, 3 results) G .. . . Fluorimetric 200 (1 result) TABLE VII DETERMINATION OF SELENIUM IN KALE Laboratory Finish Selenium content, p.p.b.* A ... . Fluorimetric 110 (fox, 4 results) B .. . . Fluorimetric 125 (&Is%, 3 results) C .. . . Fluorimetric 150 (1 result) D .. . . HydrideIAAS 125 (f4%, 3 results) E ,. . . HydridelAFS 106 (&6%, 3 results) G .. . . Fluorimetric 100 (1 result) * The mean value reported for this sample by other workers is 121 p.p.b. TABLE VIII DETERMINATION OF SELENIUM IN DRIED MILK, SAMPLE A Laboratory Finish Selenium content, p.p.b.* A .. . . Fluorimetric 50 (&lo%, 2 results) B .. . . Fluorimetric 73 (RSD 14%, 4 results) C .. . . Fluorimetric 121 (f lo%, 2 results) D .. . . HydrideIAAS 80 (1 result) E .. . . HydrideIAFS 91 (RSD 5y0, 4 results) F .. . . Fluorimetric 70 (1 result) a standard deviation of 30 p.p.b.* Result obtained by neutron-activation analysis was 150 p.p.b. withAugust, 1979 50 40 30 a? s- 20 .- +J .- > 10: P - z 5 - u - ([I 0 ) - .- SMALL AMOUNTS OF SELENIUM IN ORGANIC MATTER TABLE IX DETERMINATION OF SELENIUM IN DRIED MILK, SAMPLE B - - - - 783 Laboratory Finish Selenium content, p.p.b.* A .. . . Fluorimetric 70 (-J=7%, 2 results) B .. . . Fluorimetric 99 (RSD IS%, 4 results) C .. . . Fluorimetric 125 (f.5%, 2 results) D .. . . HydrideIAAS 90 (1 result) E .. . . HydrideIAFS 103 (f6%, 3 results) F .. . . Fluorimetric 90 (1 result) a standard deviation of 30 p.p.b. * Result obtained by neutron-activation analysis was 110 p.p.b. with TABLE X DETERMINATION OF SELENIUM IN DRIED MILK, SAMPLE C Laboratory Finish Selenium content, p.p.b.* A ... . Fluorimetric 30 (&OX, 2 results) B .. . . Fluorimetric 75 (RSD 17.3%, 4 results) C .. . . Fluorimetric 100 (f3%, 2 results) D .. . . Hydride/AAS 60 (1 result) E .. . . Hydride/AFS 75 (&lo%, 2 results) F .. . . Fluorimetric 90 (1 result) precision was not stated for this sample. * Result obtained by neutron-activation analysis was 50 p.p.b.; the for the fluorimetric method, but has been represented as a single curve in Fig. 1. Neverthe- less it is still apparent that selenium concentration, rather than sample type, is the pre- dominant factor in determining the precision. This suggests that the recommended procedures will produce satisfactory results for the range of samples examined, provided that samples of suitable size are taken. Dried milk t I 10 1 0= 1 o3 1 o4 Selenium content of sampiehg g-’ Fig.1. Precision of analysis for samples with various selenium contents : + , fluori- metric method; and 0, hydride generation method.784 ANALYTICAL METHODS COMMITTEE : DETERMINATION OF Analyst, Vol. I04 Recommendation The Sub-committee recommends the methods described in the Appendix as suitable for the determination of selenium in the range 0.1-10 pg 8-l in samples of organic matter. APPENDIX Recommended Methods for the Determination of Selenium The organic matter is destroyed by wet oxidation with nitric, perchloric and sulphuric acids. Wet-oxidation methods can be hazardous, particularly when digestions of unfamiliar materials are carried out for the first time, and users are urged to familiarise themselves with previous reports on this subject.2 The selenjmm is converted into selenium(1V) by boiling with 4~ hydrochloric acid and determined either by the fluorimetric method, using 2,3- diaminonaphthalene (DAN) as reagent, or by hydride generation, atomisation and atomic spectroscopic measurement.Reagent Blank used in the test and omitting only the sample. Carry out a blank test by the entire procedure, using the exact amounts of reagents as Procedure for Oxidation of the Sample and Reduction of Selenium Reagents When possible “low in metals” grade reagents should be used. Nitric acid, sp. gr. 1.42 (71% mim). Hydrochloric acid, sp. gr. 1.18 (36% mlm). Sulphuric acid, sp. gr. 1.84 (98% m/m). Perchloric acid, sp. gr. 1.70 (72% mlm). Nitric acid - perchloric acid solution (5 + 1 V / V ) .Nitric acid - sulphuric acid solution (1 + 1 V / V ) . Apparatus Kjeldahl flasks (1OOml) were used, with means of heating at a controlled rate and with facilities for the extraction of perchloric acid fumes. Alternatively, a programmable “block digestor” may be found to be advantageous, particularly when large numbers of samples have to be processed. Destruction of Organic Matter Place a sample of suitable size (see Note 1) with several acid-washed glass beads intoa 100-ml Kjeldahl flask. Add 30 ml of nitric acid - perchloric acid (5 + 1) solution and either allow the sample to digest overnight at room temperature or heat cautiously until the initial foaming subsides and the sample is solubilised. Gradually increase the temperature until steady boiling is achieved, taking care to avoid loss of sample by bumping or excessive frothing. When the volume has been reduced by approximately half, cool and add 10ml of nitric acid - sulphuric acid (1 + 1) solution.Return the flask to the heater and continue to heat through the perchloric acid oxidation stage, characterised by vigorous surface reaction and evolution of white fumes, and ensuring that, if the yellow digest begins to darken, charring i s Prevented by the cautious addition of 1-ml increments of nitric acid. Continue to heat past the stage at which the digest develops an intense greenish yellow colour, until a clear colourless or pale yellow solution and dense fumes of sulphur trioxide are produced, and then allow to cool. Reduction of Selenium Dilute the digest with water and add sufficient hydrochloric acid to produce a solution 4~ in hydrochloric acid and boil gently for 5min; allow to cool, and dilute to a suitable known volume with water or hydrochloric acid (see Note 2).August, 1979 SMALL AMOUNTS OF SELENIUM IN ORGANIC MATTER 785 NOTES- The mass of sample chosen is determined by two factors, the first being the composition of the material and the second the expected selenium content of the sample relative to the sensitivity of the measuring system.For the volumes of reagents given above, samples low in fats and sugars and containing 50% or more of water can be digested in amounts up to 10 g, while amounts up to 6 g are suitable for samples containing 1040% of water and 3-g amounts of most drier materials can be handled. Samples with a high fat or sugar content should be limited to 1-2 g only, owing to the danger of violent reaction with perchloric acid.Larger amounts of suitable samples can be handled if small amounts of additional nitric acid are added to prevent any charring during the initial boiling stage. The volumes of water and hydrochloric acid used must be chosen such that the final volume and acidity are consistent with the measuring system being employed. When deciding on these volumes, both the effect on sensitivity and the optimum acidity for a given hydride generation system (if used) must be considered. 1. 2. Fluorimetric Measurement of Selenium Reagents 0.880) to 1 1 with water. Ammonia solution, approximately 17% m/m.Dilute 500ml of ammonia solution (sp. gr. Formic acid, 50% VlV. Cyclohexane or dekalin. Hydrochloric acid, approximately 4 M. Dilute 360 ml of hydrochloric acid (sp. gr. 1.18) to 1 1 with water. Hydrochloric acid, approximately 0.1 M. Dilute 10 ml of hydrochloric acid (sp. gr. 1.18) to 1 1 with water. 2,3-DiaminonaphthaZene (DAN) reagent. Add a few drops of 0.1 M hydrochloric acid to 1.0 g of DAN and stir to form a paste. Transfer into a 250-ml separating funnel with 200 ml of 0.1 M hydrochloric acid and shake for approximately 30 s. Extract the solution three times with 20-ml portions of cyclohexane or dekalin, dilute to 1 1 with 0.1 M hydrochloric acid and filter through a wetted filter-paper into an amber-glass bottle. Add sufficient cyclohexane or dekalin to give a protective layer 2-3 mm in depth and store in a cool, dark place (see Note 3).This solution is stable for 2-3 weeks. Hydrox.ylammonium chloride - EDTA solution. Dissolve 0.3 g of the disodium salt of EDTA and 25 g of hydroxylammonium chloride in 1 1 of water and mix well. Standard selenium solution A . Add 10 ml of nitric acid (sp. gr. 1.42) to 1.000 g of selenium and warm to dissolve. Dilute to 1 1 with water and mix well. 1 ml of solution = 1 mg of Se. Standard selenium solution 23. Dilute 10.0 ml of standard selenium solution A to 1 1 with 0.1 M hydrochloric acid. 1 ml of solution = 10.0 pg of Se. Standard selenizm solution C. Dilute 10.0 ml of standard selenium solution B to 1 1 with 0.1 M hydrochloric acid. 1 ml of solution E 0.10 pg of Se. Preparation of Calibration Graph Transfer into a series of 150-ml beakers 0.0, 2.0, 4.0 and 6.0 ml of standard selenium solution C (see Note 4).Add 1 ml of perchloric acid to the contents of each beaker and heat on a hot-plate until the first appearance in each beaker of fumes of perchloric acid. Cool, add 5 ml of 4 M hydrochloric acid to each and boil gently for 5 min. Remove the beakers from the hot-plate, cool and add to each in turn 20 ml of water, 5 ml of formic acid and 10 ml of hydroxylammonium chloride - EDTA solution. With the use of a pH meter adjust the pH of each solution to 1.8 with ammonia solution, add 5 ml of DAN solution and transfer the beakers to a water-bath maintained at 50 "C. After 30min, remove the beakers from the water-bath, cool and transfer the solutions with a minimum of water washings to separate 250-ml separating funnels fitted with PTFE786 ANALYTICAL METHODS COMMITTEE : DETERMINATION OF AnaZyst, VoZ.104 stopcocks. Adjust the volumes to about 70 ml with water and extract each with 10 ml of cyclohexane or dekalin. Allow the layers to separate, discard the lower aqueous layer and wash the organic layers each with 25 ml of 0.1 M hydrochloric acid. Allow the layers to separate and discard the lower acid layer. Measure the fluorescence of each cyclohexane or dekalin phase in a spectrofluorimeter with excitation at 369 nm and reading the emission at 525 nm (see Note 5). Correct the readings for the blank containing no added selenium and plot a calibration graph from the net values. Preparation and Measurement of the Sample Solution Take the sample solution from the Destruction of Organic Matter above, or a suitable aliquot containing not more than 0.6pg of selenium, and transfer it into a 150-ml beaker with the aid of 20ml of water and continue as outlined under Preparation of Calibration Graph from “.. . 5 ml of formic acid and 10ml of hydroxylammonium chloride- EDTA solution . . .” to “. . . with excitation at 369 nm and reading the emission at 525 nm (see Note ti).’’ Correct the fluorescence reading for the reagent blank, and determine the selenium content of the sample by reference to the calibration graph. NOTES- The procedure involving the formation of the DAN - selenium complex and subsequent handling of the solutions should be carried out in diffused light.The calibration range can be adjusted to suit the particular fluorimeter or spectrofluorimeter being used, provided that it is compatible with the sample mass or the aliquot of i t taken for analysis. When using instruments that do not produce corrected spectra, the apparent excitation and emission wavelengths are influenced by the instrument optics and the detector response; it is therefore necessary to scan around the quoted wavelengths to find the maxima for a given instrument. 3. 4. 5. Measurement of Selenium by Hydride Generation, Atomisation and Atomic Spectroscopy Principle of Method The diluted acidic digest is reduced by sodium tetrahydroborate( 111) solution to yield hydrogen selenide, which is carried by a stream of carrier gas to a heated area where it is atomised and the concentration of atoms produced is measured by atomic spectroscopy. Because of the wide range of apparatus available for applying this family of techniques, it is not possible to give instructions as to how the determination should be performed.The following comments may, however, be helpful. 1. The detection limits for selenium obtained by these methods vaned by at least three orders of magnitude according to the technique employed by Sub-committee members. The choice of sample mass and final digest dilution volume must be compatible with the useful working range of the given generation - spectrometer apparatus available. If a single injection system is used for atomic-absorption measurements] the detection limit is very poor unless simultaneous background correction is employed.An automatic evolution procedure such as that employed by Vijan and Wood12 requires the minimum of labour and produces good detection limits and precision. For every individual apparatus, the optimum acidity, carrier gas flow-rate, sample volume, sodium tetrahydroborate(II1) volume and concentration will vary. These para- meters must be optimised before attempting to obtain useful measurements. The atom reservoir for making atomic-absorption measurements is preferably an electrically heated silica tube furnace operating at a temperature of at least 800 “C. Satis- factory results may be obtained by using a silica tube of approximately 10mm i.d. with a central side-arm for “hydride” introduction, heated with an air - acetylene flame. An alternative method of making atomic spectral measurements is to use the technique of atomic fluorescence. When using this technique an argon - hydrogen-entrained air flame is a suitable atom reservoir. When discrete sample injection is used, the experience of the Sub-committee is that only syringes or injectors of all-plastic construction are suitable for sample handling. It should be noted that optimum performance in the absence of interfering ions may 2. 3. 4. 5. 6. 7. 8.August, 19 79 SMALL AMOUNTS OF SELENIUM I N ORGANIC MATTER 787 be obtained either by injection of the sample to the generator already containing the sodium tetrahydroborate( 111) or vice versa, according to the system employed. This situation may be modified in the presence of certain ions that inhibit hydride generation.13 Optimum performance of a given system may be found to vary according to whether or not the carrier gas inlet to the generator is above or below the aqueous reactants. 9. 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. References Hall, R. J., and Gupta, P. L., Analyst, 1969, 94, 292. Analytical Methods Committee, Analyst, 1960, 85, 643. Levesque, M., and Vendette, E. D., Can. J . Soil Sci., 1971, 51, 85. Ihnat, M., J . Ass. Off. Analyt. Chem., 1974, 57, 368. Ihnat, M., J. Ass. Off. Analyt. Chem., 1974, 57, 373. Ihnat, M., and Miller, H. J., J . Ass. Ofl. Analyt. Chem., 1977, 60, 813. Vogel, A. I., “A Text-book of Macro and Semimicro Qualitative Inorganic Analysis,” Fourth Bunton, N. G., Dixon, E. J., and Michie, N. D., J . Ass. Off. Analyt. Chem., 1978, 61, 48. Cresser, M. S., and West, T. S., Analyst, 1968, 93, 595. Shimoissi, Y., Analytica Chim. Acta, 1973, 64, 465. Christian, G. D., and Young, J. W., Analytica Chinz. Ada, 1973, 65, 127. Vijan, P. N., and Wood, G. R., Atom. Absorption Newsl., 1974, 13, 33. Pierce, F. D., and Brown, H. R., Analyt. Chem., 1976, 48, 693. Edition, Longmans, London and New York, 1954, p. 584.

ISSN:0003-2654    

DOI:10.1039/AN9790400778

出版商:RSC    

年代:1979

数据来源: RSC

摘要:

788 SHORT PAPERS Analyst, August, 1979 Hydrocarbon Analysis of Naphtha Using Capillary-column Gas Chromatography Under Isothermal Conditions Pradeep Kumar, S. L. S. Sarowha and P. L. Gupta Indian Institute of Petroleum, Dehra Dun-248005, India Keywords ; Hydrocarbon analysis ; naphtha ; capillary-column gas chromatography ; Kovdt's retention indices Several attempts have been made,l-6 with varying degrees of success, to carry out a detailed individual hydrocarbon analysis of petroleum naphtha and its fractions, which find extensive use as feedstocks for various petrochemical processes and products. Because of the com- plexity and occurrence of a large number of hydrocarbon isomers, gas chromatography with long capillary columns and special facilities such as temperature and pressure programming has been tried.Such procedures are difficult to apply to routine analyses. This paper reports a method of detailed hydrocarbon analysis using a squalane capillary column (150 ft x 0.01 in) without prior separations or temperature or pressure programming. At room temperature fairly good resolution of various isomers in naphtha fractions boiling up to 150 "C is achieved. For the separation of some selected compounds, however, running of the chromatogram at another temperature is necessary. From a study of the variation of Kov&t's retention indices with temperature for different hydrocarbon types, it is shown that a reasonably good separation of various hydrocarbons can be achieved at 55 "C. Data from two chromatograms, at room temperature and at 55 O C , are combined to give the final results, showing good repeatability and therefore suggesting that the procedure is promising for the routine analysis of naphtha samples.Experimental The gas chromatograph used for this study was basically a Philips Instrument, Model PV 4000 series, with a flame-ionisation detector modified to accommodate capillary columns by incorporating a Hamilton inlet splitter. The capillary column was a stainless-steel (150 f t x 0.01 in) WCOT type, coated with squalane and supplied by Perkin-Elmer. Peaks in the chromatogram are identified by measuring Kov&t's retention indices, con- firming selected hydrocarbons with the help of available reference compounds or using literature data whenever available (assuming that a reproducibility of 1 for the retention index could be achieved in practice).Air-peak distances (calculated by using Peterson and Hirsch's methods) have been subtracted from retention distances for the calculation of Kov&t's indices. Results and Discussion The chromatogram of a typical naphtha (boiling range 40-150 "C) run at room tempera- ture is shown in Fig. 1. From Fig. 1 and Table I it is evident that peaks differing by two or more retention index units are generally well separated, and that for the same difference in retention index values, separation appears to improve as the retention index values increase, Thus, at the beginning of the chromatogram components differing by two or more retention index units are not well resolved, e.g., cyclopentane, 2,3-dimethylbutane and 2-methylpentane, but towards the end of the chromatogram even components differing by less than two retention index units show good separation.Hence, it appears that the peak width (measured in terms of retention indices) increases at a slower rate than the increase in retention index.SHORT PAPERS 789 TABLE I KOVAT'S RETENTION INDICES OF IDENTIFIED HYDROCARBONS USING A SQUALANE CAPILLARY COLUMN Measured Kovat's retention AI Concen- Peak indices of 10 - tration chromatograms from typical 55 "C atures-la yo mlm number peaks from At in a -*-, liter- naphtha, from Figs. 1 Hydrocarbon and 3 Butane .. .. 2-Methylbutane . . Pentane 2,2-Dimethylbutane' ' Cyclopentane 2,3 -Di met h y 1 butane' ' 2-Methylpentane . . 3-Methylpentane . . Hexane .... 2,2-Dimethyl- pentane .. Methylcyclopentand ' 2,4-Dimethylpentane Benzene .. .. 2,2,3-Trimethyl- butane 3,3-Dimethyi-' * * pentane .. .. Cyclohexane . . 2-Methylhexane . . 1,l-Dimethyl- cyclopentane . . 2,3-Dimethyl- pentane .. .. 3-Methylhexane . . 1-cis-3-Dimethyl- cyclopentane . . 1-trans-3-Dimethyl- cyclopen tane . . 5-Ethylpentane . . 1-fruns-2-Dimethyl- cyclopentane . . 2,2,4-Trimethyl- pentane .. Heptane (n-C,i . . 1-cis-2-Dimethyl- cyclopen tane 2,t-Dimethylhexand ' 1.1.3-Trimethyl- (l-c-2-DMCP) (2,2-DMCn) . . . . cyclopen tane 2,2,3,3-Tetramethylbu- tane (2,2,3,3-Tet.MC, Methylcyclo- hexane (MCH) . . 2,bDimethylhexane Ethilcycloientane 2,4-Dimethyihexand ' 2.2.3-Trimethyl- (1,1,3-TMCP) (2 6-DMC ) . . (ECP) (2,4-DMCe) .. . pentane (2,2,3- TMC6) .. ..l-truns-2-cis-4- Trimethylcyclo- pentane (1-t-2-c- Dimethylhexand ' Toluene .. .. 4-TMCP) .. (3.3-DMCJ . . .. 3,3- (3.3-DMCa) l-tr~ns-2&-3-Tri- methylcyclo- pentane (l-t-2-c-3- TMCP) 2,3,4-Trimeth;l- ' * pentane (2,3,4- TMC ) 2,3,3-Thnethyl- ' pentane .. .. l,l,2-Trimethyl- cyclopentane . . 2-Methyl-3-ethyl- pentane 2,3-Dimethylhexand ' 2-Methylheptane . . 4-Methylheptane . . 3,4-Dime thylhexane 3 4 6 6 7 8 9 10 11 12 12 13 14 16 16 17 18 19 19 20 21 22 23 23 - 24 26 26 26 I) 26 26 27 28 29 30 31 32 32 33 34 36 36 36 36 37 38 39 $6 "C 400.0 473.0 500.0 534.6 562.0 564.6 568.3 582.6 600.0 624.7 624.7 629.0 633.3 636.3 665.4 667.6 665.4 669.1 669.1 674.8 678.9 682.9 685.3 685.3 - 700.0 716.4 721.3 721.3 721.3 721.3 727.0 730.0 730.0 - 737.0 740.6 740.6 744.1 748.8 764.9 768.1 758.1 768.1 764.2 765.0 768.1 400.0 500.0 569.9 569.9 569.9 686.0 600.0 626.1 626.1 626.1 641.2 - - - 660.6 662.3 664.8 675.6 672.6 675.6 683.6 686.8 686.8 689.6 - 700.0 722.8 720.4 724.7 726.6 726.6 726.6 735.5 732.9 737.6 742.6 743.7 747.6 749.1 - 759.9 764.6 759.9 764.6 766.8 - - 0.00 0.07 0.00 1.38 0.90 0.17 0.45 0.00 0.89 E} 0.34 2.47 1.44 1.32 2.22 0.17 1.96 0.86 0.31 1.63 1.65 0.64 1.69 1.08 0.00 2.10 0.62 1.91 0.8 1.6 1.9 0.1 0.6 0.1 2.0 1.3 2.2 3.7 0.1 2.5 Trace 0.1 6.4 1.4 0.6 0.6 1.4 1.0 0.9 0.3 1.4 Trace 2.2 0.3 0.1 0.2 2m3g} 11.4 2.43 0.22 0.4 1.83 0.5 0.42 0.4 1.52 Trace 1.63 0.4 1.32 0.2 2.42 5.6 1.65 0.6 1.27 0.2 2.02 0.1 2.32 0.4 '0::: Oe2 0.16 1.2 0.16 0.1 1.00 0.1 Measured KovPt's reten tion Peak indices of number peaks from from chromatograms Figs.1 ,-*-, Hydrocarbon and 3 25°C 55'C 1 -cis2 -trans-4 -Tri- methylcyclo- pen tane ..3-Methyl-3-ethyl- pentane .. .. 2,2,4,4-Tetramethyl- pentane . . . . 3-Ethylhexane . . 3-Methylheptane . . methylcyclo- pentane .. .. 2,2,5-Trimethyl- hexane 1-cis-3-Dimethyl- * * cyclohexane . . 1-fruns-4-Dimethyl- cyclohexane . . 1,l-Dimethyl- cyclohexane . . 1-Methyl-cis-3-ethyl- cyclopentane . . Cycloheptane . . l-Methyl-truns-3- ethvlc yclopen tane 2,2,4-Trimethyl- hexane .. .. l-Methyl-frans-2- eth ylcyclopentane 1-Methyl-1-ethyl- cyclopentane . . 1-trans-2-Dimethyl- cyclohexane . . l-cis-2-czs-3-Tri- methylcyclo- pentane .. .. 1-cis-4-Dimethyl- cyclohexane . . I-trans-3-Dimethyl- cyclohexane . . Octane .. .. Isopropylcyclo- pentane 2,2-Dimethy&ptane 1-Methyl-cis-%ethyl- cyclopentane .. 2,2,3-Trimethyl- hexane .. 2,4-Dimethylheptane 1-cis-2-Dimethyl- cyclohexane . . 2-Methyl-4-ethyl- hexane . . .. Prop ylcyclopen tane 2.6-Dimethylheptane Ethylcyclohexane . . E thylbenzene 2,b-Dimeth ylheptane 3,6-Dimethylheptane 3,3-Dimethylheptane 1,1,3-Trimethyl- cyclohexane . . 2,3,3-Trimethyl- hexane 1,1,4-Trimethyl- . cyclohexane . . l-cis-3-cis-6-Tri- meth ylcyclohexane fi-Xylene .. 2,3,4-Trimethyl- hexane .. .. m-Xylene 2,S-Dimethyiheptank' 4-Ethylheptane . . 1-cis-2-cis-3-trans-4- Tetramethyl- cyclopentane . . 3,4-Dimethylheptane 4-Methyloctane . . 2-Methyloctane . . l-cis-2-tr~ns-3-Tri- 39 39 39 40 40 41 41 42 42 42 43 44 44 44 44 45 46 46 47 47 47 48 49 60 61 - 62 62 63 63 64 64 56 66 - 56 57 58 59 60 60 61 62 63 64 66 66 67 768.1 768.1 768.1 771.0 771.0 774.1 774.1 779.9 779.9 779.9 783.2 786.3 786.3 786.3 786.3 788.4 796.5 796.6 800.0 800.0 800.0 806.8 813.7 816.8 819.0 - 822.6 822.6 826.6 826.6 829.3 829.3 832.0 832.0 - 834.7 835.6 837.2 840.4 843.9 843.9 846.3 - - - - - - 772.1 - - 772.1 772.1 779.8 776.6 786.4 788.3 786.4 788.3 794.7 791.6 - 791.6 794.7 803.3 803.3 806.6 806.6 800.0 - - - - - 829.3 830.1 834.6 834.6 - - - - - 840.3 841.3 843.8 847.0 849.5 - 852.4 853.5 856.7 859.5 860.2 861.5 863.3 AI Concen- 10 - tration At in a from typical liter- naphtha, aturea-1' % mlm 2.18 0.1 1.95 - 1.90 - g; } 1.0 2*27} Traces 0.65 :::} 2.9 2.91 1.93 0.2 t:::} 0.6 2.14 3.64 - - 0.1 2*87} 1.2 2.77 2*96 } 0.7 2.61 0.00 1.7 2.61 0.2 0.45 0.1 2.64 0.2 1.38 } 0.3 0.20 3.07 0.2 0.66 0.9 2.03 0.9 2.69 2.3 2.80 1.3 g:;: } 0.6 0.25 - 1.15 - 2.82 - 1.86 - - - 2.68 1.4 1.60 0.1 - .~ ~.~ ~ 2.64 4.2 0.65 0.9 0.46 0.1 - - 1.06 0.1 0.30 0.8 0.16 0.6790 SHORT PAPERS Analyst, Vol.104 Peak number from Figs. 1 Hydrocarbon and 3 3-Ethylheptane .. 68 3-Methyloctane . . 69 l-Methvl-truns-3- isoprbp ylcyclo- pentane .. 69 1-Methyl-cis-i:iso- propylcyclo- pentane .. .. o-Xylene l-trum-3-Dimdthyl-' * cis-2-ethylcyclo- pentane .. .. 3,3-Diethylpentane . . 1-Methyl-1-propyl- cyclopentane . . l-Methyl-t~ufis-2- 69 70 71 72 73 propylcyclo- pentane .. 73 propylcyclo- pentane .. .. 74 1-Methyl-cis-3: TABLE I (contimed) Measured Kovgt's retention ~I Concen- indices of tration peaks from lo hi i n a chromatograms from typical ,--A-t liter- naphtha, 26OC 55OC atureo-18 %m/m - 866.7 0.45 0.2 - 869.3 0.35 - 869.3 - 1 0.7 - 870.9 3.10 1.5 - 869.3 - - - 877.0 - - 880.6 2.90 0.1 - 882.3 - - - 882.3 - - - 883.7 - - Peak number from Figs.1 Hydrocarbon and 3 l-Methyl-tram-3- propylcyclopentane 75 l-cis-2-cis-4-Tri- methylcyclo- hexane .. .. 76 l-cis-2-truns-4-Tri- me thylcyclo- hexane ,. .. 77 1-trans-2-Diethyl- c yclopen tane 78 Isobutylcyclopentak 78 1-Methyl-1-ethyl- cyclohexane .. 79 1-cis-3-Diethyl- cyclopentane .. 79 1-trans-3-Diethyl- cyclopentane .. 79 Isopropylbenzene . . 80 Nonane .. .. 81 Measured KovPt's retention AI Concen- indices of 10 - tration peaks from At in a ch;omatograms from typical c---~---, liter- naphtha 25 oc 55 oc ature*-'* % mlm' 885.1 - - - - 886.5 - - - 889.6 - - 891.7 - - - 891.7 - - 894.3 - - 894.3 - - 894.3 - - 888.6 895.3 2.50 - 900.0 900.0 0.00 - - - - - Table I shows retention index values measured from the chromatograms.For overlapping peaks a slight error in retention index values is possible owing to difficulties in correctly locating individual peak maxima. Temperature coefficients of the indices (AI/A.t) have also been reported in the literature.*12 By increasing the column temperature naphthenes and aromatics are retained longer than alkanes ; hence, preferential separation of naphthenes and aromatics from alkanes could be achieved. With small changes in the column tempera- ture, the separation of naphthenes from aromatics is difficult, as the AI/At values are only marginally different for these two types of hydrocarbons. For very large temperature changes, however, the retarded naphthenes rind aromatics may start interfering with the alkanes of higher carbon number.As an example, 3,3-dimethylhexane (3,3-DMC6) and toluene elute together at room temperature; at 55 "C they are separated fully, but if the temperature is increased to 75 O C , it is found that 3,3-dimethylhexane merges with 1-trans- 2-cis-4-trimethylcyclopentane (l-t-2-c-4-TMCP) and toluene starts to interfere with 1-trans- 'I- - Timdmin Fig. 2. Section of a chromatogram at a column temperature of 55 "C.August, 19 79 SHORT PAPERS 791 A uqasuodsa t f ZL - J PL - m 1' -2 l 9 - 3 s9p9 I 1792 SHORT PAPERS Analyst, Vol. 104 2-cis-3-trimethylcyclopentane (l-t-2-c-3-TMCP). With the aim of separating the maximum number of pairs (that elute together at room temperature) and avoiding interference from neighbouring components, a compromise temperature of 55 "C has been selected.The separations possible by judicious selection of column temperature are illustrated further in Fig. 2, where a section of the chromatogram run at 55 "C is shown. At 25 "C methylcyclohexane elutes with 2,2-dimethylhexane (2,2-DMC6), 1,1,3-trimethylcyclopentane (1,1,3-TMCP) and 2,2,3,3-tetramethylbutane (2,2,3,3-Tet.MC4). At a column temperature of 55 "C, 2,2-dimethylhexane and 1,1,3-trimethylcyclopentane are separated, as shown in Fig. 2, but methylcyclohexane (MCH) and 2,2,3,3-tetramethylbutane are still not separated from each other as the AI/At values for these compounds differ only slightly. Some typical separations are given in Fig.3, which shows a complete chromatogram of a naphtha sample at 55 "C. Components not separated from others at room temperature but separated at 55 "C are marked in the chromatogram with circles. KovAt's indices of the hydrocarbons at 55 "C are also shown in Table I for ready identification. Separated components are measured from this chromatogram ; the other compounds interfering at room temperature cah be estimated by subtraction. The composition of a typical straight-run naphtha fraction (40-150 "C) up to C, hydrocarbons is shown in Table I (last column) as an illustration. The concentrations reported were calculated from normalisation of peak areas (measured by triangulation). The only significant pair remaining unresolved is ethylcyclohexane and ethylbenzene. If need be, ethylbenzene can be determined by analysing the sample on a polar column, such as 7,s-benzoquinoline. Hence, the proposed method, utilising isothermal chromatography at two temperatures, shows reasonably good separation. 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. References Polgar, A. G., Holst, J. J., and Gorennings, S., Analyt. Chem., 1962, 34, 1226. Gupta, P. L., and Kumar, P., unpublished work. Merchant, P., Jr., Analyt. Chem., 1968, 40, 2153, "Annual Book of ASTM Standards," American Society for Testing and Materials, Philadelphia, Egiazarov, Yu. G., Kulikov, V. I., Kuzyaeva, V. V., Smol'skii, A. M., Kozlova, T. I., Barkovskaya, Sanders, W. N., and Maynard, J. B., Analyt. Chem., 1968, 40, 527. Schomburg, G., and Dielmann, G., J . Chromat. Sci., 1973, 11, 151. Peterson, M. L., and Hirsch, J., J . Lipid Res., 1959, 1, 132. Hively, R. A., and Hinton, R. E., J . Gas Ckromat., 1968, 6, 203. Tourres, D. A.. J . Chromat., 1967, 30, 357. Soj'ak, L., and Bucinska, A., J . Chromat., 1970, 51, 75. Bricteux, J., and Duyckaerts, G., J . Chromat., 1966, 22, 221. 1974, Part 25, p. 671. E. B., and Kozlov, N. S., Chem. Technol. Fuels Oils, 1974, 10, 179. Received August 15th, 1978 Accepted January lst, 1979

ISSN:0003-2654    

DOI:10.1039/AN9790400788

出版商:RSC    

年代:1979

数据来源: RSC

摘要:

792 SHORT PAPERS Analyst, Vol. 104 The impact Cup: A Simple Aid in Flame Spectrometric Analysis at High Analyte Concentrations Malcolm S. Cresser Department of Soil Science, Uwiuersity of Aberdeen, Meston Walk, Aberdeen, AB9 2 UE Keywords ; Impact cup ; flame spectrometry ; high analyte concentrations ; pneumatic nebuliser The upper useful concentration limit in any flame spectrometric method of analysis is imposed by the onset of a high degree of curvature in the calibration graph. A number of approaches may be adopted in order to extend the working range. If an absorption, emission or fluorescence line is available that exhibits poorer sensitivity, this line can be used in place of the more sensitive line. In some instances useful alternative wavelengths are not available,Azlgust, 1979 SHORT PAPERS 793 however, or the change in sensitivity may be so drastic that the alternative wavelength is of limited value.In atomic-absorption spectrometry burner rotation or a separate, short-path burner can be used to increase the sensitivity for 1% absorption and thus extend the working range. In the author’s experience, this approach often leads to poorer precision, because of the increase in background noise caused by the greater relative importance of fluctuations in molecular absorption and/or emission at the edges of the flame. Moreover, both of the above approaches suffer from the serious drawback that the incidence and extent of inter- ferences increase significantly when solutions with high concentrations of analyte element are nebuilised, because of the larger size of the residual solid particles after evaporation of the solvent .l+ An alternative approach to the problem is to take steps to decrease either the qebulisation rate or the nebulisation efficiency. The nebulisation rate can be reduced by res‘ricting the nebuliser capillary diameter or by increasing its length.Unfortunately, this may lead to a greatly increased tendency towards nebuliser clogging with only small changes in sensitivity, The nebulisation rate can also be reduced by the addition of a miscible, high-viscosity additive, which involves a time-consuming sample preparation step, or by employing a peristaltic pump to reduce the rate of solution uptake. It has also been demonstrated5 that the useful working range can be extended by nebulisation of discrete aliquots of sample that are sufficiently small for the nebuliser not to reach its nonnal equilibrium working conditions.All of these approaches considerably decrease the throughput of samples for a given instru- ment. This short paper reports a novel approach to the problem of increasing the useful working range of flame spectrometric determinations. The impact bead of a conventional atomic- absorption spectrometer is replaced with a PTFE impact cup, which serves to reduce considerably the nebulisation efficiency. Experimental Reagents Stock solutions containing 1000 pg ml-l of analyte or interfering element were prepared from analytical-reagent grade salts, and diluted as appropriate. The stock magnesium solutions (MgCl,.GH,O and MgS0,.7H20) were standardised by complexometric titration.Apparatus Measurements were made using a Baird A3400 atomic-absorption spectrometer operated under normal conditions for routine analysis. When appropriate the impact bead was replaced with a PTFE cup made as shown in Fig. 1 to fit on the manufacturer’s impact bead support arm. The interior of the cup was made with sloping sides and PTFE was chosen as the construction material to minimise the possibility of sample solution being retained inside the cup, which might have produced memory effects. Impact cup Nebuliser \ u- Impact bead support Locking nut Fig. 1. Design of the PTFE impact cup.794 SHORT PAPERS Analyst, Vol. 104 Results and Discussion It was found that the use of the impact cup in place of the usual impact bead caused a 12-fold change in nebulisation efficiency, which in turn caused an approximately 12-fold change in sensitivity in both atomic-absorption and -emission spectrometry for readily atomised elements.Thus the absorbances from 1 pg ml-l solutions of cadmium, copper, magnesium and zinc nebulised into an air - acetylene flame were reduced 115, 12.0-, 12.5- and 12.2-fold, respectively. The emission signals from 1 pg ml-l solutions of potassium and sodium were reduced 13.8- and 14.8-fold, respectively. The differences between these factors are attributable to curvature in the calibration graphs, particularly that due to ionisation in the instances of sodium and potassium. To investigate the effect of use of the impact cup on precision, 0.2 and 0.5 pg ml-1 mag- nesium solutions were nebulised repeatedly into a 130-mm air - acetylene flame, using the impact bead, and the absorbances measured.A 1 pg ml-l magnesium solution was nebulised periodically and the scale expansion adjusted if necessary to give an arbitrary reading of 0.8 for this solution. The coefficients of variation were found to be 0.028 and 0.012 at 0.2 and 0.5 pg ml-1 of magnesium, respectively. This procedure was then repeated using the impact cup, but for 2.4 and 6.0pgml-l magnesium solutions, using a 12 pgml-1 standard solution to adjust the scale expansion. The coefficients of variation were found to be 0.035 and 0.013, respectively. Thus the precision when the impact cup is used is comparable to that when the impact bead is used at 12 times lower concentrations. As the use of the impact cup appears to be similar in effect to a 12-fold dilution,significant improvements should be observed in the linearities of calibration graphs, Typical graphs for the determination of sodium in an air - acetylene flame by emission spectrometry at 589 nm are shown in Fig.2, using 130- and 10-mm flames for the impact bead and a 130-mm flame Sodium concentration/pg ml-’ Fig. 2. Flame emission calibration graphs for sodium a t 589nm. A, Impact bead, 130-mm flame; B, impact bead, 10-mm flame: and C , impact cup, 130-mm flame. for the impact cup. The photomultiplier voltage was adjusted so that identical arbitrary readings were obtained from a 20 pg ml-l sodium solution in each instance. The curvature of the calibration graph is eliminated over the range 6-20pgml-l when the cup is used, although curvature due to ionisation is observed over the range 0-6 pgml-l.Similar ionisation-induced curvature is normally observed with the impact bead over the range 0-0.5 pg ml-l for the 130-mm flame if a calibration graph is plotted over the range 0-1 pg ml-1. It can be eliminated by the addition of an ionisation buffer. The calibration graph for the range 0-240 pg ml-l of sodium using the impact cup and a 130-mm flame is identical, within the limits of experimental error, with the graph for the range 0-20 pg ml-l using the impact bead. If the impact cup is used in conjunction with the small flame, the useful working range extends to more than 1000 pg ml-l. No significant memory effects were observed.August, 1979 SHORT PAPERS 795 Typical atomic-absorption calibration graphs for magnesium, as both the chloride and the sulphate, are shown in Fig.3. The degree of scale expansion was adjusted to give an arbitrary reading of 0.8 for 20 pg ml-l of magnesium as the chloride in each instance. A slight curva- ture is observed over the range 0-8 pg ml-l when the impact cup is used with the 130-mm flame. This curvature is normally observed over the range 0-0.7 pg ml-l using the impact bead and the 130-mm flame. Using the impact cup and the small flame, the calibration graph for magnesium, as the chloride, over the range 0-240 pg ml-l was similar to that obtained using the small flame and impact bead over the range 0-20 pg ml-l.It is interesting to note that, whereas magnesium as the chloride and as the sulphate gave identical calibra- tion graphs over the range 0-20 pug ml-l when the impact cup and the 130-mrn flame were used, the calibration graphs for the two magnesium species were different (see Fig. 3) when the impact bead and the 10-mm flame were used. The elimination of this interference over the range 0-20 pg ml-l is almost certainly attributable to the fact that only extremely fine droplets reach the flame when the impact cup is utilised, so that the smaller residual solid particles are completely ~olatilised.~,~ r Magnesium concentration/pg ml-' Fig. 3. Flame atomic-absorption calibration graphs for magnesium a t 286nm. A, MgCl,, impact bead, 10-mm flame; B, MgCl, and MgSO,, impact cup, 130-mm flame; and C, MgSO,, impact bead, IO-mm flame.To see if the reduction in interference was general, the effect of various amounts of phosphorus on the determination of 100 pg ml-l of calcium in a stoicheiornetric air - acetylene flame was investigated. The results, shown in Fig. 4, clearly demonstrate that a substantial 0.9 1 1 0 1 20 40 60 80 100 Phosphorus concentrationlpg ml- ' Fig. 4. Effect of phosphorus on the absorbance from 100 pg ml-l of calcium. A, Impact cup, 130-mm flame; and B, impact bead, 130-mm flame.796 SHORT PAPERS reduction in the degree of interference is observed when the impact cup is used, providing further support for the hypothesis that the fraction of spray reaching the flame when the impact cup is used consists only of very fine droplets.A similar reduction in the degree of interference may be achieved by 12-fold dilution using the impact bead. Conclusions The use of an impact cup in conjunction with a pneumatic nebuliser produces an effect that in every respect is equivalent to 12-fold dilution for flame spectrometric analysis on the system used. Use of the impact cup therefore increases the useful working range by a factor of 12 without the usual problems associated with flame spectrometric determinations at high analyte concentrations such as increased incidence and extent of interferences and/or reduced sample throughput. If the impact cup is used in association with the techniques usually employed to extend the working range, much greater extensions become possible. If the impact cup and impact bead are fitted side-by-side on two arms of an inverted Y, it is possible to change from one to the other in a few seconds, without even extinguishing the flame. The author is indebted to George Wilson for the manufacture of the impact cup and to Fiona Mitchell for assistance with the experimental work. References 1. 2. 3. 4. 6. Kirkbright, G. F., and Sargent, M., “Atomic Absorption and Fluorescence Spectroscopy,’’ Academic Price, W. J., “Analytical Atomic Absorption Spectrometry,” Heyden, London, 1972, p. 91. Cresser, M. S., and MacLeod, D. A., Analyst, 1976, 101, 86. Cresser, M. S., Lab. Pract., 1977, 171. Cresser, M. S., Anulyticu Chim. Ada, 1976, 80, 170. Press, London, 1974, p. 617. Received February 12th, 1979 Accepted March 20th, 1979

ISSN:0003-2654    

DOI:10.1039/AN9790400792

出版商:RSC    

年代:1979

数据来源: RSC

摘要:

Analyst, August, 1979 Book Reviews 797 ANNUAL REPORTS ON ANALYTICAL ATOMIC SPECTROSCOPY, REVIEWING 1977. Volume 7. Edited by J. B. DAWSON. Pp. x + 291. London: The Chemical Society. 1978. Price A17.50; $35 (CS Members L13). Under the editorship of J. B. Dawson the reviewing team have once again lived up to the expectations created by the earlier volumes in this series. In his foreword the editor comments that the first aim of the reviewers is to provide the reader with up to date information on world wide developments in the field of analytical atomic spectroscopy. This is nicely demonstrated by the observation that over 20% of the references are from papers delivered at conferences and that are likely to be overlooked by many of those not in the conference jet-set. However, this volume provides an even greater service at this particular time by its reporting of developments in inductively coupled plasmas and related techniques, which with the advent of commercial instruments are rapidly demonstrating their potential in the trace-element field.Personally, I appreciate this publication very much and I recommend this particular volume to all practising analysts who use atomic-spectroscopic methods. JOHN AGGETT ESTIMATING THE HAZARD OF CHEMICAL SUBSTANCES TO AQUATIC LIFE. Edited by JOHN CAIRNS, JR., K. L. DICKSON and A. W. MAKI. Pp. viii + 278. Philadelphia, Pa.: American Society for Testing and Materials. 1978. Price $19.50. A workshop was held in July 1977 to discuss the “Application of aquatic toxicity testing methods as predictive tools for aquatic hazard evaluation.” Originally proposed by A.W. Maki it was called together as a result of the US Toxic Substances Control Act 1976. The workshop members included chemists, toxicologists, ecologists and aquatic biologists who were organised into specialist groups. Each group was assigned one of the following topics: (a) toxicological effects, (b) environmental concentration, (c) environmental fate, (d) hazard assessment and (e) procedures for estimating hazards to aquatic life, and these topics formed the basis of five one-day workshop sessions. Each session was opened with the reading of a plenary paper by a member of the specialist group. The workshop discussion that followed was summarised by the specialist group for that session and later approved by the other members of the workshop.This book is the publication of the plenary papers and the discussion reports for each of the workshop sessions. It is inevitable, with a workshop of this type, that much overlap and repetition of ideas and subject matter occur between sessions. In fact, the plenary papers for sessions (b) and (c) cover much of the same ground and the two specialist session groups report a single discussion. Such repetition can be useful as it often provides an alternative viewpoint. However, stricter directives to authors concerning their subject matter could have been useful in providing more of the relevant subject material presented for some sessions. The plenary papers vary a great deal in their length, breadth of coverage and their citation of scientific literature, though the discussion reports do attempt to balance the individual papers.Nevertheless, the workshop acknowledges that many of its findings are related to organic chemicals but suggests that most aspects of hazard assessment apply equally well to the inorganics. Much of the content of the book relates to two documents that are conveniently reproduced in the appendices. One is a “Proposed working document for the development of an ASTM draft standard on standard practice for a laboratory testing scheme to evaluate hazard to non-target aquatic organisms, ” by the ASTM sub-committee E35.21 on Safety to Man and Environment. The other is the “Criteria and rationale for decision making in aquatic hazard evaluation (third draft),” by the Aquatic hazards of pesticides task group of the American Institute of Biological Sciences. The applicability of these documents and a third alternative proposed by R.A. Kimerle, a member of the workshop, was discussed. In its conclusion the workshop lists several areas of scientific research where future work would aid the assessment of the hazards of chemical substances to aquatic life. In general, this book is well written and presented and would be a useful addition to the library of any scientist or administrator working in this field, particularly as it gives reasonably up to date information on hazard evaluation in aquatic systems as interpreted by a number of different individuals and committees. ASTM Special Technical Publication 657. K. R.BULL798 BOOK REVIEWS Analyst, Vol. 104 LIQUID SCINTILLATION COUNTING. Volume 5. Edited by M. A. CROOK and P. JOHNSON. Proceedings of a Symposium on Scintillation Counting organised by the Radiochemical Methods Group (Analytical Division, The Chemical Society), Bath, England, September 13-1 6, 1977. Pp. x + 222. London, Philadelphia and Rheine: Heyden. 1978. Price $28; L14; DM64.50. The change of title of the Symposium from Liquid Scintillation Counting was justified by the contents, which incorporated whole body counting and radioimmunoassay. There was little that was novel, either at the meeting or in this volume except that it is very useful to have the cross- fertilisation of ideas between disciplines. The quality of the writing is good, but the use of single spacing on a typewriter to ease reproduction causes references and superscripts to run into the previous line and results in some confusion in places.The twenty-six chapters are divided into five parts and I have chosen for comment what I considered to be outstanding in each section. The plenary lecture on sample preparation by Bransome gives timely warning and advice, which is of value to us all. Immunoassay is a powerful method, in that it can identify several hundred constituents in a biological fluid, but unfortunately the plenary lecturer does not include any references with which to follow up the technique. I would like to see the technique of Ayrey, for using heavy metals loaded into scintillators, developed further as it seems to be a powerful method for counting electron-capture nuclides by measuring a higher proportion of K X-rays.Whole body monitoring for clinical applications is well served by two chapters by Burkinshaw and his colleagues from Leeds, with Boddy providing the cost effectiveness to persuade the clinician. The advances in instrumentation and technique were in no way new but the chapter by Horrocks is masterly in its presentation, if over simplistic in representing the liquid scintillation detector as linear down to energies below 100keV. The criticisms by Bransome of the misuse of automatic counters is answered by Johnson in chapter 20 in a well written chapter on computing methods which warn you when your data are meaningless or of little value. The final two papers on measuring small amounts of radioactivity in minute precious samples by Burleigh and Otlet demonstrate what meticulous attention to detail can produce when using carbon-14 dating techniques.This is the last symposium on liquid scintillation counting to be organised by the Radiochemical Methods Group and in the future the emphasis should be upon practical workshops to improve the use of the technique. This opinion reflects the well developed state of the method of liquid scintillation counting for which this five-volume series will provide a convenient reference library for all users. J . A. B. GIBSON THIN-LAYER CHROMATOGRAPHY. Second Edition. By JUSTUS G. KIRCHNER. Techniques of Chemistry, Volume X I V . Pp. xxvii + 1137. New York, Chichester, Brisbane and Toronto: John Wiley. 1978.Price L42.50; $84. This is Volume XIV in the Techniques of Chemistry Series, and may be considered to be one of the standard authoritative handbooks available covering thin-layer chromatography. It is a weighty tome of over 1100 pages with over 6000 references. Since the first edition, several of the chapters have been revised completely and all have been modified and brought up to date. The volume retains the general format of the first edition and it is divided into two parts, the first part dealing with various techniques of thin-layer chromatography, and the second, much to the satisfaction of the practical reader, with applications. The latter includes acids, alcohols, alkaloids, amines, antibiotics, carbohydrates, dyes, hydrocarbons, lipids, nucleic acids, pesticides, pharmaceutical products, phenols, natural pigments, steroids, essential oils, vitamins and many others.In addition, there is an Appendix listing compounds and types of compounds to be found in the book together with the test number of the reagent(s) to be used for their detection. There is both a subject and a compound index, which have been included in preference to an author index. To complete the various lists the addresses are given (mainly in the USA) of commercial firms mentioned in the text. There is always something to criticise in any book if one is determined to do so and this is no exception; for example, plastics analysts might wish for more details of separations of additives other than plasticisers/stabilisers (perhaps more in the ultraviolet absorber range), and specialistsAztgust, 1979 BOOK REVIEWS 799 in other fields might also feel deprived, but this is a minor carp for no single movable volume can be exhaustive when covering so many fields.My advice to the chromatographer who has not got a reference volume and can afford this one is to buy it. D. SIMPSON INSTRUMENTATION FOR HIGH-PERFORMANCE LIQUID CHROMATOGRAPHY. Edited by J. F. K. HUBER. Journal of Chromatography Library, Volume 13. Pp. xvi + 204. Amsterdam, Oxford and New York: Elsevier. 1978. Price $34.75; Dfl180. Any chemist who has been an instructor in one of the many high-performance liquid chromato- graphy (HPLC) training courses of recent years is well aware of the need for a good, comprehensive book to help in the selection of equipment for specific analytical requirements.Unfortunately, despite the impeccable qualifications of each of the authors, this volume does not provide a satisfactory outline of the state of the art in instrumentation for liquid chromatography. The systems engineering concepts outlined in the Editor’s introduction must be accepted as the methodology used by those engineers who actually design liquid chromatography apparatus for manufacture, but correlation of this mathematical approach with real performance from real hardware is not uniform throughout the book. However, individual chapters such as that on Electrochemical Detectors (H. Poppe), Sample Introduction Systems (J. C. Kraak) and Column Design (J. C. Kraak) stand out as examples of clear exposition, with good coverage of the subject and a good selection of easily understood line drawings of a range of individual components.Surprisingly, Professor Poppe’s second chapter on other detectors is much less satisfactory, as it lacks the line drawings of flow cells, etc., which are so essential to an understanding of the sources of temperature and flow sensitivity in optical detectors in particular. However, this section is better in this respect than that discussing Radiometric Detectors (P. Markl), which relies solely on the mathematical approach. The remaining chapters on Pumps and Gradient Systems (M. Martin and G. Guiochon) , Preparative Liquid Chromatography (A. Wehrli) and Mass Spectro- metry (E. Kenndler and E. R. Schmid) do not reach the standard of the three best sections, but no part of the book is less than a useful guide to an understanding of HPLC instrument com- ponents.The literature coverage of individual sections varies as much as the subject matter, but on the whole coverage is reasonably complete up to 1977, with one or two 1978 references. Within these limits, most of the major papers on each of these equipment topics do appear in the reference lists at the ends of separate sections. It is a considerable disappointment that the standard of presentation of individual chapters is so variable. The cost of the book is not unreasonable for what is essentially a reference volume on apparatus, but is too high for recommendation for individual purchase. However, until a better balanced competitor appears this volume must be a good choice for the shelves of the larger chemical libraries.R. A. WALL ENVIRONMENTAL POLLUTANTS : DETECTION AND MEASUREMENT. Edited by TAFT Y. TORIBARA, JAMES R. COLEMAN, BARTON E. DAHNEKE and ISAAC FELDMAN. Environmental Science Research, Volume 13. Pp. xii + 500. New York and London: Plenum. 1978. Price $51. This is the 13th volume in the series Environmental Science Research and it records the pro- ceedings of the 10th International Conference on Environmental Toxicity held at Rochester, New York, in May 1977. This conference, with the detection and measurement of environmental pollutants as its theme, was unusual in that there was a limited attendance of around only 30, of whom 75% were presenters of papers. The reason given in the Preface that this small gathering “should provide more meaningful discussion” is hardly convincing, nor is the international designa- tion given to this conference, only two participants apparently being from an organisation outside the USA.The book follows the pattern of the conference in that the 23 papers with their associated discussions, presented in five sessions, are published as chapters within five titled sections. The first section of four chapters deals with general aspects of the specification of analytical problems in the detection and measurement of pollutants in the environment. This is followed by a section with the somewhat misleading title of “More Familiar Principles” and containing a miscellancy of five singularly interesting but unrelated topics, as diverse as ion-selective electrodes and the800 BOOK REVIEWS Analyst, Vol.104 identification of the source and the fate of petroleum from oil spills at sea. The four chapters on methods for field use that follow strikingly illustrate the advances made in the sampling and detection of gaseous pollutants, ranging from the use of passive samplers involving gas permeation to that of lasers. The true value of this volume is probably to be found in the last two sections where some of the latest and most powerful analytical techniques for use in pollution studies are described. The first five chapters concern the application of various microprobe methods, such as transmission electron microscopy, secondary ion mass spectrometry, auger electron spectroscopy and X-ray photoelectron spectroscopy to the analysis of airborne particulates arising from or associated with pollution.Also included is an account of the development and an assessment of the potential of one of the newest microanalytical techniques, that of electron energy loss spectroscopy. In the final section, devoted to a variety of physical analytical methods, the applications of the two well established techniques of X-ray and nuclear activation analysis are reviewed. Also, a valuable re-evaluation is made of the potential use of Raman spectroscopy in the chemical identifi- cation of airborne particulates. Other topics described are the use of time-of-flight spectroscopy for measuring the size distribution of airborne particulates and of surface ionisation mass spectro- metry for the continuous analysis of both inorganic and organic pollutants in air or water. This is a generally well produced volume, most chapters being of considerable individual merit. However, the reviewer finds difficulty in recommending i t either to the general reader wishing for an overview of the present state of the art of detecting and measuring environmental pollutants, or to the instrumental specialist who has an interest in developing new and better methods of analysis. The book has something for each class of readership but possibly not sufficient to justify the high cost of around k25. Rather, this is seen as a reference book for the library shelf where all can have access. R. WOOD

ISSN:0003-2654    

DOI:10.1039/AN9790400797

出版商:RSC    

年代:1979

数据来源: RSC