ISSN:0003-2654    

DOI:10.1039/AN97500FX045

出版商:RSC    

年代:1975

数据来源: RSC

ISSN:0003-2654    

DOI:10.1039/AN97500BX047

出版商:RSC    

年代:1975

数据来源: RSC

ISSN:0003-2654    

DOI:10.1039/AN97500FP121

出版商:RSC    

年代:1975

数据来源: RSC

ISSN:0003-2654    

DOI:10.1039/AN97500BP125

出版商:RSC    

年代:1975

数据来源: RSC

摘要:

DECEMBER 1975 The Analyst Vol. 100 No. 11 97 An Evaluation of the 3-Methyl-2-benzothiazolinone Hydrazone Method for the Determination of Phenols in Water and Waste Waters Morris E. Gales, Jun. Methods Development and Quality A ssurance Research Laboratory, National Environmental Research Center, OBce of Research and Development, U.S. Environmental Protection Agency, Cincinnati, Ohio 46268, U.S.A . Friestad’s method for the determination of phenol with 3-methyl-2-benzo- thiazolinone hydrazone (MBTH) has been evaluated. The Friestad manual and automated MBTH methods were modified to lower the detection limit from 1 mg 1-1 to 1 pg 1-l. The values for phenol obtained by this method on river water and industrial waste samples were equal to or greater than those obtained by the 4-aminophenazone method.The trend towards more stringent pollution abatement practices has heightened the need for a method that can be used for determining a greater number of phenolic compounds than can be measured by the 4-aminophenazone (4AP) method. These compounds are considered to be pollutants because they cause an unpleasarit taste and odour in potable water supplies that have been chlorinated. In addition, phenols are often considered to be an indication of man-made organic contamination. The most widely used method for determining phenols in aqueous samples is the 4AP method.lS2 In this procedure the phenol is distilled and subsequently made to react with hexacyanoferrate(II1) and 4AP to form a red complex that is measured colorimetrically. The reaction with 4AP has several deficiencies, the most significant of which is its inability to measure certain para-substituted phenols3 Friestad et aL4 have described both a manual and an automated method, each of which determines phenols by oxidative coupling with 3-methyl-2-benzothiazolinone hydrazone (MBTH) .Friestad et aZ. have not only shown that the MBTH method is more universal in reactivity to phenols than the 4AP method, but also that it leads to higher molar absorptivities. The MBTH method is based on the coupling of phenol with MBTH in an acidic medium using ammonium cerium(1V) sulphate as an oxidant. The coupling takes place in the para position; if this position is occupied, the MBTH reagent will react at a free ortho position. Therefore, the reaction of MBTH with phenolic compounds is less dependent on the position of the substituent group than is that with 4AP.One disadvantage of the MBTH method, however, is that colours produced by different phenolic compounds range from red to violet and thus do not have their absorbance maxima at the same wavelength. The colours obtained have maxima ranging from 460 to 695 nm. Goulden, Brooksbank and Day5 have modified Friestad et aZ.’s automated method for use on relatively clean waters and have extended the detection limit from 10 to 0.2 pg 1-l. Their method consists basically of the automated distillation and condensation of a large sample (at a rate of 6.06 ml min-l). After the colour formation step the product is concentrated by extraction into a solvent. In the course of their work, 60 Lake Ontario samples, containing 0-15 pg1-l of phenol, were analysed by both the automated 4AP and the automated MBTH methods.The results of this study showed no statistically significant difference between the two methods. Goulden et aZ.’s automated method was not evaluated in this study because the distillation equipment and separator required are not commercially available. Rather, this paper is con- cerned with the evaluation of Friestad et aZ.’s manual and automated MBTH methods to establish their applicability for U.S. Environmental Protection Agency use by accumulating comparative data for the two methods on phenolic compounds and on samples of refinery 841 This method can be used to analyse 10 samples per hour.842 GALES EVALUATION OF THE 3-METHYL-2-BENZOTHIAZOLINONE Analyst, VOZ.IOO waste water, surface water and domestic waste. In the majority of determinations the MBTH method gave higher results, even though readings were made at a fixed wavelength. Minor changes were made in the sample volume, the reagents, the volume of reagents and the manifold. With these modifications, phenol can be determined by using both of these methods over a range from 1 to 1000pgl-l. Manual Method Apparatus Samples were measured on a spectrophotometer at a wavelength of 520 nm and with a light path of 5 cm. A 750- or 1OOO-ml separating funnel was used for the extraction with chloroform, Reagents MBTH solution, 0.05 per cent. Dissolve 0.1 g of 3-methyl-2-benzothiazolinone hydrazone hydrochloride in 200 ml of distilled water.Ammonium cerium(1V) sulphate solution. Add 2.0 g of ammonium cerium(1V) sulphate [Ce(S0,),.2(NH4),S04. 2H,O] and 1-5 ml of concentrated sulphuric acid to 150 ml of distilled water. Dissolve, in the following order, 8 g of sodium hydroxide, 2 g of EDTA (disodium salt) and 8 g of boric acid in 200 ml of distilled water. Dilute to 250 ml with dis- tilled water. Using this stock solution, make a working solution by mixing an appropriate volume with an equal volume of ethanol. After the solid has dissolved, dilute to 200 ml with distilled water. Bufer solution. Procedure Distillation If the sample has not been preserved, add 0.5 g of copper sulphate and lower the pH to approximately 4 with concentrated sulphuric acid. If the sample has been preserved with orthophosphoric acid and copper sulphate at the time of collection, add 0.5 ml of sulphuric acid.Distil 450ml of sample, add 50ml of warm distilled water to the flask and resume distillation until 500 ml have been collected. Colour develofiment To 100 ml of distillate, or an aliquot diluted to 100 ml, add 4 ml of MBTH solution. After 5 min add 2.5 ml of ammonium cerium(1V) sulphate solution. Leave the solution for another 5 min, add 7 ml of buffer and then, after 15 min, read the absorbance a t 520 nm against a reagent blank. The colour is stable for 4 h. To 500 ml of distillate add 4 ml of MBTH solution. After 5 min add 2.5 ml of ammonium cerium(1V) sulphate solution and after an additional 5 min add 7 ml of buffer. Then, after 15 min, add 25 ml of chloroform and extract the coloured product, shaking the separating funnel at least 20 times.Allow the layers to separate and pass the chloroform layer through filter-paper in order to remove any water. Read the absorbance at 490 nm against a reagent blank. Concentrations above 50 pg 1-l. Concentrations below 50 pg 1-I. Discussion Friestad et al.’s manual method has a detection limit of 1 mg 1-1 of phenol. The colour is developed by the addition of 1 ml of MBTH solution, 1 ml of a 0.2 per cent. ammonium cerium(1V) sulphate solution and 2 ml of buffer solution to 1 ml of sample. For this study a sensitivity of 1 pg 1-1 was desired, therefore the following changes were made: the sample volume was increased from 1 to 100 ml and the same ratio of sample to reagent used in the auto- mated method was used for colour development. This procedure resulted in a detection limit of 50 pg 1-1.Samples that contain less than this concentration are extracted with chloroform following development of the colour, and in this way a detection limit of 1 pg 1-1 is achieved. Automated Method Apparatus The Technicon AutoAnalyzer that was used consisted of a sampler, a manifold, a proportion- ing pump 111, a heating bath with distillation coil, a distillation head, a colorimeter equipped with 50-mm flow cell and 520-nm filter, and a recorder.December, 1975 HYDRAZONE METHOD FOR DETERMINING PHENOLS IN WATERS 843 Reagents MBTH solution, 0.05 per cent. Stock ammonium cerium(IV) sulphate solution. As described under Manual Method. Dissolve 1-Og of the solid in 150ml of distilled water, add 3.0 ml of concentrated sulphuric acid and dilute to 200 ml with dis- tilled water.Working ammonium cerium(1V) sulphate solution. Dilute 25 ml of stock ammonium cerium(1V) sulphate solution to 100 ml with distilled water. Bufler solution. Dissolve, in the following order, S g of sodium hydroxide, 2 g of EDTA (&sodium salt) and S g of boric acid in 200 ml of distilled water. Dilute the solution to 250 ml. Sodium hydroxide solution, 1 N. Dissolve 40 g of sodium hydroxide in 800 ml of distilled water. Dilute the solution to 11. Sodium hydroxide soZution, 0.01 N. Dilute 2 ml of 1 N sodium hydroxide solution to 200 rnl with distilled water. Distillation solutio?z, 10 per cent. sulphuric acid. Add 100 ml of concentrated sulphuric acid slowly to 800 ml of distilled water.Sulphztric acid, ap$roximateZy 1 per cent. Add 1 ml of concentrated sulphuric acid to 100 ml of distilled water. Wash water. Add 20 ml of 1 N sodium hydroxide solution to 20 1 of distilled water. Cool the solution and dilute it to 1 1. * * Procedure tubing and a fast flow (1 1 h-l). Set up the manifold as shown in Fig. 1. Fill the wash receptacle by syphoning, using Kel-F Pump 0.01 N Use polyethylene tubing for the sample line. R - R 0.8 Waste from still Grey - Grey 1.0 - - c-1 Mixing coil 10 turns I n To waste A-ionnnn 0 - G 0.10 0.5 per cent. MBTH O A - 0 - Y 0.16 Buffer Y - Y 1.2 Waste from flow cell G 0.10 Ammonium cerium(lV) sulphate mI min-' + I -Waste to Dump UBlack - BlackU0-3 . , 2 Air -0 Mixing coil G - G 2.0Sample 1 II - t 110 - 01(0-42 Distilling solution1 10 turns Heating bath with distillation coil Mixing coils 10 turns 10 turns 10 turns nnnn A-IO Fig.1. AutoAnalyzer I1 circuit for the determination of phenols. Sample rate, 20 per hour. Colour code: 0 = orange, G = green, R = red and Y = yellow.844 GALES : EVALUATION OF THE 3-METHYL-2-BENZOTHIAZOLINONE Analyst, Vol. 100 TABLE I COMPARISON OF RESULTS OBTAINED FOR PHENOLS IN DISTILLED WATER BY THE MANUAL MBTH AND 4AP METHODS Compound o-Cresol m-Cresol p-Cresol o-Chlorophenol m-Chlorophenol 2,3-Dimethylphenol 3,4-Dimethylphenol 3,6-Dimethylphenol 2,8Dimethylphenol p-n-Butoupphenol 2-Naphthol Concentration found (calculated as phenol)/ Concentration added rg I-' (calculated as phenol)/ I \ Pg I-' 4AP MBTH 140 100 100 100 100 100 100 100 100 100 100 100 70 <1 97 92 45 12 23 32 19 <1 130 95 66 95 97 46 60 60 53 45 60 sodium hydroxide solution through new tubing for 30 min in order to remove phenolic material from the walls.Allow the colorimeter and recorder to warm up for 30 min. Next, run a base- line with all of the reagents present, feeding distilled water through the sample line. Place appropriate standards in the sampler tray in order of decreasing concentration, then complete the loading of the sampler tray with unknown samples, using glass tubes. If samples have not been preserved add 0-1 g of copper sulphate and 2 drops of concentrated sulphuric acid to 100 ml of sample. Discussion Friestad et aZ.'s automated method, with a detection level of 10 pg 1-11, is not sufficiently sensitive for most surface waters.However, by increasing the sampling rate before distillation from 0.8 to 2 ml min-l and the sampling rate after distillation from 0.42 to 1.2 ml min-1, the detection limit was extended to 1 pgl-l. The concentration of the MBTH solution was not changed, but its flow-rate was reduced from 0.42 to 0.1 ml min-l. The concentration of the ammonium cerium(1V) sulphate solution was reduced from 1 to 0.25 per cent. The flow-rate of the buffer solution was also reduced from 1.2 to 0.16 ml min-l. 5 Wavelengthhm ~ 0 Fig. 2. Absorbance curves of 11 phenols analysed by use of the 3-methyl-2-benzothiazolinone hydr- azone method. p, Phenol; 1, o-cresol; 2, m-cresol; 3, p-cresol; 4, o-chlorophenol; 5. 3,4-dimethyl- phenol; 6, 2-naphthol; 7, P-n-butoxyphenol; 8, 2,6-dimethylphenol; 9, m-chlorophenol ; 10, 2,3- dimethylpheno1; 1 1, 3,6-dimethylphenol.December, 1975 HYDRAZONE METHOD FOR DETERMINING PHENOLS I N WATERS TABLE I1 845 COMPARISON OF RESULTS OBTAINED FOR PHENOLS IN DISTILLED WATER BY THE AUTOMATED MBTH AND 4AP METHODS Compound o-Cresol m-Cresol p-Cresol o-Chlorophenol p-Chlorophenol 2-Naphthol 3,IDimethylphenol 9-n-Butoxyphenol Concentration added (calculated as phenol)/ 87 87 87 73 73 66 Pt? I-' Concentration found (calculated as phenol)/ 4AP MBTH 65 84 69 86 2 62 60 44 60 53 6 22 Pg 1-' I 3 77 4 81 23 30 31 Results Results obtained by the MBTH and 4AP methods with and without distillation were com- pared.Table I lists the results obtained from a series of phenolic compounds added to distilled water and analysed by the manual MBTH and 4AP methods (distillation omitted). Except for o-chlorophenol, m-chlorophenol and 2,3-dimethylphenol, which gave comparable results, higher results were obtained by use of the MBTH method. Fig.2 shows the absorbance curves of these compounds with MBTH. The maxima for these compounds appear between 495 and 550 nm and a mixture of thesz 11 compounds showed a maximum absorbance at 520 nm. TABLE I11 COMPARISON OF RESULTS OBTAINED FOR PHENOLS IN OHIO RIVER WATER BY THE AUTOMATED MBTH AND 4AP METHODS Concentration found (calculated as phenol) / Pg 1-' Compound o-Chlorophenol + phenol o-Chlorophenol + p-cresol o-Cresol + m-cresol p-Chlorophenol 2-Naphthol 3,4-Dimethylphenol p-n-Butoxyphenol 4AP 158 89 61 61 10 9 21 MBTH 128 80 87 47 20 31 32 Table I1 lists the results obtained from eight phenols added to distilled water and analysed by the automated MBTH and 4AP methods with distillation. Higher results were obtained by the MBTH method for all but two of the compounds, o-chlorophenol and 9-chlorophenol.Subse- quently, samples of Ohio River water were spiked with a series of phenols. As shown in Table 111, these results confirmed the findings obtained with the distilled water samples. The recovery of o-chlorophenol was determined by analysing a series of concentrations of this compound by the automated and manual MBTH methods. As shown in Table IV, the TABLE IV RECOVERY OF O-CHLOROPHENOL BY THE AUTOMATED MBTH METHOD Concentration added Concentration found (calculated as phenol)/ (calculated as phenol)/ Pg I-' Pg I-' Recovery, per cent.15 14 93 37 28 76 73 52 71 143 99 67846 GALES : EVALUATION OF THE 3-METHYL-2-BENZOTHIAZOLINONE Analyst, VOZ. I U O TABLE V RECOVERY OF O-CHLOROPHENOL BY THE MANUAL 4AP AND MBTH METHODS WITH SOLVENT EXTRACTION Sample Sewage Sewage Sewage Sewage Distilled water Distilled water Distilled water Distilled water Concentration added (calculated as phenol)/ P.lg I-' 29 58 146 292 29 58 146 292 Concentration found (calculated as phenol) / 4AP MBTH 25 33 51 57 151 146 259 254 34 28 46 47 241 129 309 296 Pg I-' - I- Recovery, per cent. 4AP 86 88 103 89 117 79 96 106 MBTS I14 98 100 87 96 81 88 101 96 96 - - -_ Mean recovery of o-chlorophenol with the automated method decreased as the concentration in- creased.Table V gives the recovery of o-chlorophenol by the manual 4AP and MBTH methods from distilled water and sewage; unlike the results obtained by the automated method, the results obtained by the two manual methods were comparable. The two manual methods were compared by analysing a petroleum waste and a raw sewage sample after spiking them with $-cresol. Significantly higher results were obtained for both samples with the MBTH method. The recovery of 9-cresol with the MBTH method was 95 per cent. for the petroleum waste and 99 per cent. for the raw sewage, and less than 1 per cent. for both samples with the 4AP method. In earlier studies on the determination of phenol by using the 4AP method, results obtained with the automated method were equal to those obtained with the manual method.Therefore, automated 4AP and MBTH methods were used to determine the reliability of the MBTH methods. Table VI shows the results obtained on oil refinery and lumber mill wastes by the two automated methods. Table VII lists the results obtained by use of the manual 4AP and MBTH methods on industrial wastes. In all instances except one the MBTH method gave higher results. The precision of the automated MBTH method was determined at four separate concentra- tion levels over two working ranges (242 pg 1-1 and 20-134 pg I-'). They included a concen- tration near to the detection limit of the method, two concentrations at intermediate levels and one near the upper limit. Seven replicate determinations were made for each concen- tration tested.For phenol concentrations of 2.1, 5.7, 20 and 42 pg l-l, standard deviations were k0.7, &O-5, &1.2 and 50.7 pg l-l, respectively. A t concentrations of 2, 11, 53 and 134 pg 1-1, standard deviations were k0.9, &O-4, &l-1 and h1.3 pg 1-1, respectively. The TABLE VI COMPARISON OF RESULTS OBTAINED FOR OIL REFINERY AND LUMBER MILL WASTE BY THE AUTOMATED MBTH AND 4AP METHODS Concentration found (calculated as phenol)! PJZ 1-' Type of waste Oil refinery waste Oil refinery waste Oil refinery waste Oil refinery waste Lumber mill waste 5 Plywood 5 Log pond 0.5 Hardwood Sample dilution ratio 1:lOO 1 :60 3:lOO 1 :20 1 :20 1 :20 1 :200 ' 4AP MBTH7 24 24 47 48 71 70 120 124 104 98 9 10 66 76December, 1975 HYDRAZONE METHOD FOR DETERMINING PHENOLS IN WATERS 847 percentage recovery with the automated method was determined at two levels by spiking Little Miami River water with o-cresol.At concentrations of 2.7 and 65 pg l-l, the recoveries were 83 and 98 per cent., respectively. TABLE VII COMPARISON OF RESULTS OBTAINED BY THE MANUAL 4AP AND MBTH METHODS ON VARIOUS INDUSTRIAL WASTES Concentration found (calculated as phenol) /pg 1-l Sample 1 2 3 4 6 6 7 8 4AP 10 12 16 14 33 29 2060 253 MBTH 20 24 27 17 35 16 2900 396 Interferences MBTH is also used to determine aliphatic aldehydess and aromatic amines.‘ Thus these materials, if present, could cause interference. According to the literature,* however, distilla- tion removes aromatic amines from the samples, whereas aliphatic aldehydes remain in the distilled sample.The effect of the presence of these compounds was determined by analysing a sample containing n-butyraldehyde, glyoxal and formaldehyde at concentrations of 100 pg 1-1. With this procedure a green dye was formed when these compounds were made to react with MBTH; however, the dye was destroyed when the buffer solution was added. n-Butyralde- hyde did give a response equivalent to 3 pg 1-1 of phenol at this level, but glyoxal and formal- dehyde gave no response. Conclusion The results obtained with the MBTH method are higher than those obtained by use of the 4AP method when cresol, naphthol and para-substituted phenols are the dominant phenols in the sample. As with the 4AP method, the MBTH method will not give 100 per cent. recovery of all phenolic compounds when phenol is the basis for the standard curve.The detection limit and precision of the method are satisfactory for the determination of phenol in surface waters, domestic wastes and industrial wastes. The 4AP method results in a highly coloured blank, making it difficult to detect 5 pg 1-1 but the MBTH method, with extraction, has a very low blank and 1 pg 1-1 can easily be detected. Although the MBTH method appears to offer certain advantages, the usefulness and applicability of the method can be determined only after exhaustive analyses on a large variety of sample types. 1. 2. 3. 4. 5. 6. 7. References American Public Health Association, American Water Works Association and Water Pollution Control Federation, “Standard Methods for the Examination of Water and Wastewater,” 13th Edition, American Public Health Association, New York, 1971. Ettinger, M. B., Ruchhoft, C . C., and Lishka, R. J., Analyt. Chem., 1951, 23, 1783. Mohler, E. F., and Jacob, L. N., Analyt. Chem., 1957, 29, 1369. Friestad, H. O., Ott, E. E., and Gunther, F. A., “Automated Colorimetric Micro Determination of Phenol by Oxidative Coupling with 3-Methyl-2-benzothiazolinone Hydrazone,” Technicon Inter- national Congress, 1969. Goulden, P. D., Brooksbank, P., and Day, M. B., Analyt. Chem., 1973, 45, 2430. Sawicki, E., Hauser, T. R., Stanley, T. W., and Elbert, W., Analyt. Chem., 1961, 33, 93. Sawicki, E., Stanley, T. W., Hauser, T. R., Elbert, W., and Noe, J. L., Analyt. Chem., 1961, 33, 722. Received April 4th, 1976 Accepted June 23vd. 1976

ISSN:0003-2654    

DOI:10.1039/AN9750000841

出版商:RSC    

年代:1975

数据来源: RSC

摘要:

848 Analyst, December, 1975, VoZ. 100, pp. 848-853 A Rapid Method for the Simultaneous Determination of Paraquat and Diquat in Pond and River Waters by Pyrolysis and Gas Chromatography A. J. Cannard and W. J. Criddle Department of Chemistry, University of Wales Institute of Science and Technology, King Edward V I I Avenue, Cardan, CFL 3NU A rapid method is described for the determination, in aqueous systems, of two widely used herbicides known commercially as paraquat and diquat. Pyrolysis of these herbicides under carefully controlled conditions, followed by gas-chromatographic analysis of the pyrolysate, allows detection of the herbicides down to 0.01 p.p.m. A t the lowest concentration levels (0.01- 0.1 p.p.m.), there is some loss of linearity of response, possibly as a result of adsorption of the herbicides on the glass surfaces of the vessels used.This effect may have resulted in substantial errors in previously reported low levels of paraquat and diquat and may also occur in the determination of other ionic herbicides. In recent years, the commercial production and use of herbicides has increased rapidly. Of these, l,l’-dimethyl-4,4’-bipyridylium chloride (paraquat) and to a lesser degree 1,l’-ethylene- 2,2’-bipyridylium bromide (diquat) are of particular imp0rtance.l Both are quaternary ammonium salts and their structural formulae are given below. r 72’ L -I 2+ 2Br- Paraquat Oiquat As a result of their inherent stability and toxicity there has been increasing interest in them, not least from analytical chemists, with regard to environmental pollution.Previous methods of analysis have depended either on the reduction of the herbicide with sodium dithionite2 to form a free radical that absorbs in the ultraviolet region or on reduction to the corresponding piperidine followed by gas-chromatographic analysis3 Additionally, a bioassay technique has been reported4 but this procedure, although extremely sensitive, is very time consuming and is of little use if a rapid result is required. A method for the analysis of soils ,E involving catalytic hydrogenation of the herbicide followed by gas chromatography, has recently been reported, and also a pyrolytic method,8 but the latter method incorporates a preliminary ion-exchange procedure. The gas-chromatographic procedure described here requires no sample pre-treatment and can normally be completed in 15-20 min.The method has been developed with regard to the affinity of the herbicides for glass surfaces, a factor not previously reported and apparently not taken into consideration in current analytical procedures. Experimental Reagents Paraquat. Methyl viologen hydrate, obtainable from Aldrich Chemical Co. Ltd. Diquat. Obtained by crystallisation from Reglone A (Plant Protection Ltd., ICI). Standard solutions of the herbicides were prepared in de-ionised water. Preparation of standard solzttions In order to mitigate errors due to possible differential adsorption of paraquat and diquat on glass, a single calibrated flask was used for the initial calibration. After containing herbicide solutions the flask was thoroughly cleaned by repeated washing with de-ionised water.It was thenCANNARD AND CRIDDLE 849 filled with de-ionised water and allowed to stand for 1 h before checking the contents for the absence of herbicide by the proposed gas-chromatographic method. In order to examine the affinity of the herbicides for glass, standard solutions were prepared with added glass beads (1 g of BDH 100-mesh glass beads for gas - liquid chromatography) having a surface area of about 0.04 m2, i.e., about ten times that of the vessel used. Deter- minations of the herbicide content were made before and after addition of the beads. Apparatus A Perkin-Elmer F30 gas chromatograph having a standard injection-port modification for a Chemical Data Systems (CDS) Pyroprobe 190 was used throughout. Silica pyrolysis tubes were as supplied with the Pyroprobe.Studies were carried out in order to establish optimum conditions for the pyrolysis, viz., pyrolysis temperature and time, heating rate (ramp) and probe insertion distance. The data (Table I, Figs. 1 and 2) show that best results were obtained under the following conditions: temperature, 1000 "C; pyrolysis time, 5 s; ramp, 2.0 "C ms-l; and probe insertion distance, maximum. The significance of these values is discussed in detail below. TABLE I EFFECT OF PROBE INSERTION DISTANCE ON PEAK AREA AND RETENTION TIME FOR 2,2'-BIPYRIDYL Withdrawal from maximum insertion distancelmm 0 3 8 14 19 23 29 38 43 48 58 65 2,2'-Bipyridyl peak arealcounts x 10-3 99.9 96.5 99.6 94.4 99.8 97.5 94.2 95.6 98.7 94.2 92-0 85.9 Retention time/s 264 264 266 266 268 268 270 270 272 273 278 282 Gas-chromatographic conditions were as follows : column, 10 per cent.Carbowax 20M and 2 per cent. potassium hydroxide on Celite (80-100 mesh) in a 600 x 3-5 mm i.d. glass column; column temperature, 190 "C; injection-port temperature, 110 "C; carrier gas, nitrogen at a flow-rate of 40 cm3 min-l. A flame-ionisation detector was used. Measurements of peak area were made electronically by using an Infotronics CRS 208 digital integrator fitted with an angular base-line corrector. It was necessary to allow for slight day-to-day variations in detector response. In the 220 7 200 180 160 8 100 y" 60 20 ll 0.1 0.2 0.5 1.0 2.0 5.0 10.0 20.0 Pyrolysis tirne/s Fig. 1. Variation of yield of bipyridyls with pyrolysis time at different final pyrolysis temperatures: A, 1000; B, 800; and C, 600 "C.850 CANNARD AND CRIDDLE: RAPID SIMULTANEOUS DETERMINATION OF Analyst, vd.100 5 140- 100- f 80. % 60. 2 40. 8 120. \ 20 " ' I I 1 I , , I I I d / I I absence of an internal standard, a standard herbicide solution (100 pl, 1 p.p.m.) was run twice daily in order to obtain a response factor. Procedure Standard volumes of herbicide solution (up to lOOp1) were injected carefully from a 100-pl syringe [Scientific Glass Engineering Pty. Ltd. (Australia)] into the centre of a silica tube and the tube was heated continuously at 100-110 "C in an air stream from a hot-air blower. The best results were obtained when the solution was injected continuously so that it was not allowed to evaporate to dryness until all of the solution had been introduced into the tube.Care should also be taken not to allow the solution to fill completely the bore of the tube. The tube was then inserted into the coil probe of the CDS Pyroprobe 190 and the sample pyrolysed under optimum conditions (see below). Analysis of the pyrolysate was carried out by use of the Perkin-Elmer F30 gas chromatograph. The silica tubes used in this procedure were permanently kept in a furnace at 800 "C and were always handled with stainless-steel forceps. Results and Discussion The method described in this paper for the determination of paraquat and diquat is based on the following reactions. Paraquat - A N W N + 2CH3CI Although other reactions occur that give smaller fragments, it will be apparent [Fig.3 (b) and (c)] that the pyrolysis of both compounds produces few products with relative molecular masses comparable to those.of the free bases, a feature which renders the method particularly suitable for both quantitative and qualitative analysis. Soderquist and CrosbyS referred to the thermal breakdown of paraquat during direct injection at high injection-port tempera- tures, but did not attempt to develop an analytical procedure based on this effect. One of the main problems encountered in pyrolysis studies has been lack of reproducibility but use of the CDS Pyroprobe 190 minimises errors of this type and successive pyrolyses now give an acceptably reproducible pattern (Table 11). However, for best results, the pro- cedure described in this paper must be strictly adhered to.In particular, pyrolysis of samples applied directly to the coil probe or ribbon probe gave extremely poor sensitivity, owing toDecembey, 1975 PARAQUAT AND DIQUAT IN WATERS BY PYROLYSIS AND GC 42 .. W d .- 0 a -0, -03 -(o - d --hl -0 -. N c .L C .- E \ .- El I- k c, a, a h v 851852 CANNARD AND CRIDDLE: RAPID SIMULTANEOUS DETERMINATION OF Analyst, vd. 100 low yields of the bipyridyls, and the results obtained were not satisfactorily reproducible. A possible explanation is that the rapid rise in the temperature of the probe surface gives a thermal shock to the sample and solid material is thus ejected from the surface, which would result at most in partial pyrolysis of the sample with resulting low yields of base.Reproducibility would thus be a function of sample film thickness, a parameter virtually impossible to control at this level of sample size. These problems do not arise when the pyrolyses occur in the silica tube. The confined nature of the system allows time for com- pletion of the pyrolysis before the volatile compounds emerge from the tube, as unpyrolysed or partially pyrolysed material cannot easily escape from the tube. Other advantages of the method are the ease of sample application and the fact that several samples can be pre- pared prior to pyrolysis, which would not be possible if the probe alone were used. TABLE I1 VARIATION OF BIPYRIDYL PEAK AREA WITH CONCENTRATION OF PARAQUAT OR DIQUAT Bipyridyl peak area/ Relative standard Concentration, p .p .m.counts x deviation, per cent.* 0.01 0.6 20.0 0.02 1.6 13.0 0.05 4.4 11.4 0.08 7.8 10.2 0.10 9.3 9.9 0.20 19-4 7.2 0.60 51.3 (12.4)t 3.8 0.80 83-3 2.6 1.00 104.4 (12.9)t 2.1 1.00 125.0 2.0 1-40 146.6 2.0 1.60 167.6 2.0 1.80 188-8 2.0 2.00 2 10.0 2.0 * Values calculated on the basis of a minimum of five determinations. t Values for the bipyridyl peak area obtained in the presence of added glass beads. The effects of the various pyrolysis parameters are shown in Table I and in Figs. 1 and 2. The greatest sensitivity is obtained with the highest available working temperature, ie., 1000 "C, but increasing the pyrolysis time above 5 s does not significantly improve the method. In fact, it is desirable to keep the pyrolysis time to a minimum in order to reduce the ageing effect on the column produced by the pulse of hot carrier gas, which results in stationary-phase material being stripped from the column immediately beyond the probe [Fig.3 (a)]. This effect can result in adsorption of the bases on the exposed support material, giving low results. Further, it is essential that glass columns be used, as stainless-steel columns give substantially reduced yields of the bipyridyls, particularly at low concentrations of herbicide. The heating rate (ramp) of the probe is critical. Fig. 2 shows that ramps not less than 2-0 "C ms-pare necessary for highest sensitivity, but for maximum column life this value should not be significantly exceeded. The position of the probe in the injection port is not critical to the sensitivity except at withdrawal distances near to the maximum (Table I), but there is a steady increase in retention time as the probe is withdrawn.This increase is accompanied by a substantial loss in resolution (probably owing to an increase in the injection-port dead volume), which would be undesirable if pyrolysis products with similar retention characteristics to that of the bipyridyls were present. The detection limits for the method as applied to pond and river waters are governed by two main factors: the size of sample that can conveniently be introduced into the pyrolysis tube; and the ability of the column to resolve the bipyridyl peaks from those due to other pyrolysis products. Of a total of nine samples of local pond and riverwaters most gave a simple pyrolysis pattern [Fig.3 (41. The most complex pattern so far obtained [Fig. 3 (e)] shows that no interference with paraquat will occur, and that only slight interference with diquat is likely. However, a small diquat pyrolysis peak [Fig. 3 ( b ) ] can interfere to a slight extent with the 4,4'-bipyridyl peak derived from paraquat but the value for paraquat may be simply corrected when appropriate, as the size of the interfering peak is proportional to the size of the 2,2'-bipyridyl peak derived from diquat.December, 1975 PARAQUAT AND DIQUAT IN WATERS BY PYROLYSIS AND GC 853 While many references have been made to the adsorption of herbicides on various materials,l none appears to have been made to the adsorption of herbicides on glass surfaces. It will be apparent from the work described above, in which solutions were prepared with added glass beads (Table 11) that significant errors can occur when herbicide solutions are trans- ferred from one vessel to another before analytical determinations are carried out. Tfiis observation is particularly important when solutions that have low herbicide concentrations are being studied. It is recommended, therefore, that field sampling be carried out in vessels that are subsequently used for the analytical determination and the samples should be treated identically with standard solutions used for calibration purposes. The authors thank the National Environmental Research Council for financial support and Dr. J. D. R. Thomas for helpful discussions. References 1. 2. 3. 4. 5. 6. Calderbank, A., Adv. Pest Control lies., 1968, 8, 127. Calderbank, A., and Yuen, S. H., Analyst, 1966, 90, 99. Soderquist, C. J., and Crosby, D. G., Bull. Envir. Contam. Toxic., 1972, 8, 363. Funderburk, H. H., jun., and Lawrence, J . M., Nature, Lond., 1963, 199, 1011. Khan, S. U., J. Agric. F d Chem., 1974, 22, 863. Martens, M. A., and Heyndrickx, A., J . Pharm. Belg., 1974, 29, 449. Received May 13th, 1975 Accepted July 23rd, 1976

ISSN:0003-2654    

DOI:10.1039/AN9750000848

出版商:RSC    

年代:1975

数据来源: RSC

摘要:

854 Analyst, December, 1975, Vol. 100, j@. 854-856 Determination of Dimetridazole in Feedstuffs and Pre-mixes by High-speed Liquid Chromatography F. G. Buizer and M. Severijnen Rijjkslandbozswproe fstation, Kruisherengang 2 1, Maastricht, The Netherlands A method is described for the determination of dimetridazole (1,2-dimethyl- 5-nitroimidazole) in feedstuffs and premixes by high-speed liquid chromato- graphy. Dimetridazole is extracted from the sample with methanol - water (1 + 2). After liquid - liquid extraction into dichloromethane, an aliquot of the solution is injected into a high-speed liquid chromatograph. Dimetridazole (1,2-dimethy1-5-nitroimidazole) is a feed additive used for the control of blackhead in turkeys and for the treatment of haemorrhagic dysentery in pigs.The usual content in feedstuffs is 125-150 mg kg-l. Dimetridazole can be assayed by a number of method~.l-~ We normally use a thin-layer chromatographic method and a modification of the polarographic method published by the Analytical Methods Committee.5 Liquid chro- matography is an increasingly applied technique, and it was of interest to ascertain whether this technique could be applied in feedstuffs analysis. Experimental Reagents All reagents should be of analytical-reagent grade. Methanol - water (1 + 2 V/V). Hydrochloric acid, 5 mol I-' Dichloromethane. Sodium sulphate, anhydrous. Eluting solvent: chloroform - methanol (99.5 + 0.5 V/V).* Dimetridazole reference standard. Apparatus The high-speed liquid chromatographic apparatus was assembled from the following parts : a high-pressure pump, Model 6000 (Waters Associates Inc., Milford, Mass., USA), and a universal injection system, Model U6K, from the same suppliers.A stainless-steel column (600 x 3 mm), packed with Lichrosorb SI 60 slurry of particle size 10 pm, was used. A smaller column (250 x 3 mm), packed with alumina (Merck, Darmstadt, 1097) sieved to give a fraction greater than 200 mesh, was used as a pre-cleaning column directly coupled to the injector. These columns have been in use for almost 1 year. So far, there has been no necessity to change packings. Eluted components were measured with a spectrophotometer (Beckmann, Model 25, Fullerton, Calif., USA) equipped with a flow-through cell of path length 18 pm specially designed for this model (available from Waters Associates Inc.).Measurements were recorded on the recorder of the spectrophotometer. The operating conditions used were as follows : flow-rate, 0.5 ml min-l; monochromator set at 308 nm (measurements were made against eluting solvent in the reference cell); paper speed, 0.1 in min-1; and measuring range, 0-1 or 0-2 absorbance full scale. Procedure Weigh accurately 2-20 g of sample containing 0.240 mg of dimetridazole into a 200-ml glass-stoppered flask. (For feeds containing 125 mg kg-l of dimetridazole, a 10-g sample is usually taken.) Add 100.0 ml of methanol - water and agitate the mixture mechanically for 30 min. (For feeds containing 125 mg k g l of dimetridazole, 50.0 ml of methanol - water Other equipment included a rotary evaporator and normal laboratory glassware.* The chloroform used contained about 1 per cent. of ethanol.BUIZER AND SEVER1 JNEN 855 is usually added.) Centrifuge it for 5 min at 3000 rev min-1 and filter it on a filter-paper or on cotton-wool if necessary. From this filtrate, prepare a solution in methanol - water containing 2 4 0 mg 1-1 of dimetridazole (dilution factor F = final volume divided by the initial volume). Transfer 25-0ml of this solution into a 100-ml separating funnel and adjust the pH to about 1-5 with hydrochloric acid (5 mol 1-l), checking with test paper so as to ensure that the pH is in the range 0-6. Add 25 ml of dichloromethane and invert the funnel ten times. After the phases have separated, run off the dichloromethane layer through a funnel fitted with a cotton-wool plug, and containing about 5 g of anhydrous sodium sulphate, into a 250-ml evaporating flask.Extract the aqueous layer three more times with 25-ml portions of dichloromethane, shaking the funnel well for 1 min each time. Collect all of the dichloro- methane extracts in the same evaporating flask. Finally, rinse the funnel containing anhydrous sodium sulphate twice with 5-ml portions of dichloromethane. Evaporate the combined extracts just to dryness at room temperature, ensuring that the temperature does not rise above 30 “C, and dissolve the residue in 2.0 ml of eluting solvent. Operate the chromatograph under the conditions described above and inject 0.02 ml of the solution. A 0.02-ml aliquot of a standard solution of dimetridazole (500 mg 1-l) should be injected with each run of samples.Calculation Compare the peak height of the sample with the peak height of a standard. When 50.0 ml of extraction solvent have been used, the content of dimetridazole is given by: and when 100.0ml of extraction solvent have been used, by w = 40 x lo4 FrnA/m8 where w mg k g l is the mass fraction of dimetridazole in the sample, F is the dilution factor as defined above, mA mg is the mass of dimetridazole injected on to the column and ma g is the mass of sample. Results and Discussion When developing an assay method with high-speed liquid chromatography, one can work by analogy with experience gained with thin-layer chromatography. Our experience with the latter method with dimetridazole proved to be a good starting point.In our thin-layer chromatographic method for the determination of dimetridazole, chloroform - methanol (95 + 5 ) was used as eluting solvent and silica gel as the adsorbent. The RP value for dimetridazole with this system is about 0.6, and with alumina as adsorbent 1.0. For our purpose, a less polar solvent than that required for thin-layer chromatography should be used, and the solvent chloroform - methanol (99.5 + 0-5) was found to be suitable. With this solvent, the chromatograms shown in Fig. 1 were obtained. No interference was found from : ipronidazole, amprolium, ethopabate, dinitolmide, bu- quinolate, decoquinate, methyl benzoquate, acetyl enheptin, nitrofurazone, furnicozone, nicarbazin, nitrovin, carbadox, robenidine, pyrimethamine, ronidazole, monensin, sulpha- quinoxaline, sulphamezathine, sulphacetamide, tetracycline, oxytetracycline, penicillin, streptomycin, zinc bacitracin, tylosin, oleandomycin, virginiamycin and spiramycin.Meticlorpindol, when present in the usual concentration (125 mg kg-l), gave a “dimetri- dazole recovery” of 3 mg kg-I-. Furazolidone also interfered, when present in the usual concentration (50 mg k g l ) , and in this instance a “dimetridazole recovery” of 12 mg k g l was found. This interference could be eliminated by choosing a different eluting solvent : n-hexane - methanol - ethanol (65 + 15 + 20 V / V ) . The retention time of dimetridazole was sub- stantially increased by the use of this solvent. In practice, however, combinations of dimetridazole with meticlorpindol or furazolidone seldom occur, and we were therefore able to continue with our original eluting solvent.Concentrations of dimetridazole down to 10 mg k g l can be determined by the above856 BUIZER AND SEVER1 JNEN P 7s 8: Y- 7 10 0 10 0 10 0 10 0 Tirnehin Fig. 1. Typical chromatogram obtained by high- speed liquid chromatography of feedstuffs medicated with dimetridazole. method, and a lower level of detection can be achieved by making a few modifications to the method. Dimetridazole reference standard taken through the method in different amounts wils recovered completely by this method, with a standard deviation of 2 per cent. A satisfac- torily linear relationship between peak height and mass of dimetridazole injected was found, and no increase in the width of the band was observed when injecting volumes of up to 0.1 d.Dimetridazole was incorporated, at a concentration of 150 mg k g l , into a poultry feed containing 2 per cent. of grass meal and 5 per cent. of fish meal and into a pig feed containing 4 per cent. of grass meal and 2 per cent. of fish meal. The blank value for both feeds was less than 1 mg k g l . The dimetridazole was recovered completely from both feeds, with a standard deviation of 7 mg k g l . The limit of error for a 95 per cent, probability level was 4mgkg-l for the poultry feed and 5mg k g 1 for the pig feed. The results of 15 deter- minations for the recovery of 150 mg kg-1 of dimetridazole in the poultry feed were 162, 153, 142, 150, 159, 147, 155, 147, 147, 154, 145, 154, 143, 160 and 139mgkg-l; and of nine determinations in the pig feed were 139, 144,156,145, 155,157, 142,152 and 155 mg k g l . Results by this method have been continuously compared with those obtained by our current assay methods ; no significant differences between the results of the three methods were found. The presence of bentonite, sometimes used as an aid in the pelleting of feeds, caused low recoveries of dimetridazole. When 2 per cent. of bentonite was added to a feed medicated with dimetridazole about 80 per cent. of the dimetridazole was recovered. References 1. 2. 3. 4. 5. Daftsios, A. C., J . Ass. 08. Agric. Chem., 1964, 47, 231. Analytical Methods Committee, Analyst, 1969, 94, 925. Daftsios, A. C., J. Ass. Off. Agric. Chem.. 1965, 48, 301. Stone, L. R., and Hobson, D. L., J. Ass. 08. Analyt. Chem., 1974, 57, 343. Analytical Methods Committee, Analyst. 1971, 96, 746. Received July 4th.. 1976 Accepted August llth, 1976

ISSN:0003-2654    

DOI:10.1039/AN9750000854

出版商:RSC    

年代:1975

数据来源: RSC

摘要:

Analyst, December, 1975, Vol. 100, fie, 857-861 857 A Rapid Method for Monitoring Low Levels of DiD(2=ethylhexyl) Phthalate in Solutions E. Weisenberg, Y. Schoenberg and N. Ayalon Institute of Control and Standardization of Drugs, Ministry of Health, P.O. Box 1467, Jerusalem, Israel A simple gas - liquid chromatographic method using a nickel-63 electron- capture detector for the determination of di-(2-ethylhexyl) phthalate (DEHP) in blood and aqueous solutions is described. The DEHP is extracted into n-hexane in a single-step extraction and is injected directly, using 5 per cent. SE-30 as liquid phase. The detection limit was 2 ng and a linear detector response was found in the range 2-20 ng. The method was used to monitor DEHP in blood and aqueous or lipophilic solutions. The widespread use of plastic materials in medicine has been accompanied by a rapid increase in the use of phthalate esters as plasticisers.In particular, the ester di-(2-ethylhexyl) phthalate (DEHP) is added in high concentration (2040 per cent.) to poly(viny1 chloride) in order to increase the softness and flexibility of the plastic material. However, recent reports have described toxic and teratogenic effects of phthalate esters in laboratory animals.lS2 DEHP is also known to be an environmental pollutant3 and a contaminant of many products that are packed in plastic4; the attention of investigators has therefore been focused on this new potential health hazard.5 Residues of DEHP have been reported in human blood stored in poly(viny1 chloride) bags,6 in patients receiving such blood,' in milk,* in soya oil: in waterlo and in the mitochondria of the heart muscle of different animals.1l Analytical procedures have been developed for the micro-determination of DEHP ; sensitive and selective techniques are based on gas chromatography using a flame-ionisation detector12 or combined gas chromatography - mass ~pectrometry.1~ The latter method is time consum- ing, requ'ires sophisticated instruments and is therefore not suitable for screening large numbers of samples.A simple and rapid method that possesses high sensitivity, specificity and repro- ducibility is required, so that the levels of DEHP accumulating in drugs, foodstuffs, blood, tissues and the environment can easily be monitored. The most practical methods for moni- toring programmes are based on gas - liquid chromatography using electron-capture de- tectors.Lee et ~ 1 . 1 ~ employed gas - liquid chromatography with an electron-capture detector equipped with a tritium foil to detect low concentrations of phthalate esters extracted into n-hexane. This method was subsequently applied to the quantitative determination of phthalate esters in waterl5; a graph of peak area newus concentration was found to be linear over the range 100-1000 ng, the lower limit of detection being 10 ng. This paper describes a sensitive and simple method for the quantitative determination of DEHP using a gas chromatograph equipped with a nickel-63 electron-capture detector. The method was used to monitor DEHP residues in solutions (especially aqueous solutions and blood) that had been packed in poly(viny1 chloride) bags.For many years phthalate esters were assumed to be of low toxicity. Materials and Methods Reagents n-Hexane. Di-(2-ethylhexyZ) phthalate, 100.5 per cent. A cetonitrile. All other reagents were of analytical-reagent grade and were tested for freedom from DEHP. This solvent was purified by distillation or by the method of Williams.lS This material was supplied by Travmol, Ashdod, Israel. Apparatus A Packard, Model 7400, dual-column gas chromatograph, equipped with a nickel-63 electron- capture detector and a coiled glass column, 6 ft x + in i.d., was packed with 5 per cent. SE-30 on Gas-Chrom Q, 80-100 mesh. Nitrogen was used as the carrier gas at a flow-rate of858 WEISENBERG et al.: RAPID METHOD FOR MONITORING LOW Andyst, YoZ.100 20 ml min-l, and the sensitivity of the electrometer was 1 x A. The temperatures of the injection port, detector and column were set at 270, 260 and 240 "C, respectively. A 10-pl Hamilton syringe was used to inject the solutions. Method Aqueous solutions To 50 ml of the aqueous solution in a 100-ml glass-stoppered graduated cylinder, 6 ml of acetonitrile and 10 ml of n-hexane were added. The mixture was agitated for 90 min on a wrist-action shaker. After complete separation between the two phases had been obtained on standing the mixture, the n-hexane phase was collected. A series of aliquots (2-8 pl) of the n-hexane phase were then injected into the gas chromatograph. The heights of the peaks were measured and compared with those obtained from standard dilutions of DEHP run under the same conditions.mood Absorbent cotton-wool (25 mg) was introduced into the bottom of a threaded test-tube. Acetonitrile (0.2 ml) and plasma (0.5ml) were added. After the fluids were completely absorbed into the cotton-wool, 5 ml of n-hexane were added. The test-tube was next tightly stoppered with an ordinary cork (not with a screw cap) and agitated in a wrist-action shaker for 45 min. Then the n-hexane phase was separated and 2-8-pl aliquots were injected into the gas chromatograph. Results A nickel-63 radioactive source, rather than a tritium foil, was used as the detector because better resolution was obtained with the former. This results from the ability of the nickel detector to withstand high temperatures, which serve to keep the detector clean, a prerequisite for obtaining good analytical results with electron-capture detection.Different stationary phases have been recommended for the determination of DEHP by use of gas chromatography. SE-30, which was used by other investigators for the determination of DEHP by gas chro- matography with a flame-i~nkation~ or electron-capture detector,l7 was employed in this study. Our method, using electron-capture detection, was 100 times more sensitive than that employing flame-ionisation detection. This high sensitivity can be explained by the type of so-called electrophores present in the molecules. The electrophore, CO-CH : CH-CO, present in phthalate esters has a high electron absorption similar to that of polychloro derivatives.'* It has been observedlg that the electron affinity of the phthalate esters decreases from the low to the high homologues.Dimethyl, diethyl and dibutyl phthalates can be determined at levels of 100-200 pg, whereas DEHP can be determined only in the nanogram range. A typical chromatograph of a mixture of DEHP and aldrin, used as internal standard, is shown in Fig. 1. The peak is symmetrical and Gaussian in shape and can be used for quantita- tive analysis in the nanogram range (2-20 ng). Detector responses could, therefore, be meas- ured by reference to the height of the peak and not to its area; the latter measurement is time consuming or requires an expensive electronic integrator (Table I). However, the day-to-day TABLE I CORRELATION BETWEEN PEAK HEIGHT AND PEAK AREA OF DEHP AND RATIO OF ALDRIN (INTERNAL STANDARD) TO DEHP Amount Peak height Ratio, KldrinjDEHPj - aldrin: Peak area for Pg ng cm cm DEHP DEHP*/cma 110 8 3.7 3.20 f 0.14 1-16 1-54 f 0.06 55 4 2.0 1.70 1-18 0.9 (10) (10) (10) 165 12 5.4 gy 1.16 - 220 16 7 4 6.2 & 0.34 1.18 2-97 f 0.17 * Values in parentheses denote the number of determinations.December, 1975 LEVELS OF DI-(2-ETHYLHEXYL) PHTHALATE IN SOLUTIONS 859 variations in the sensitivity of the electron-capture detector must be taken into consideration and therefore daily standardisation with DEHP is recommended.Under the conditions described above, the minimum amount measured was 2 ng. Smaller amounts of DEHP could be detected at a lower electrometer range although this is limited by the background and base- line noise. In order to study the linearity of the detector response, a series of standard solutions of DEHP in n-hexane were prepared.Each dilution was mixed with an equal volume of standard aldrin solution and injected into the chromatograph. From the results shown in Fig, 2 it can be seen that a linear relationship was found between the response of the electron-capture detector and the concentration of DEHP in the 4-18-ng range, with a straight correlation ratio to the aldrin. Good reproducibility of results was established by repeated determinations of DEHP that contained aldrin as a comparative standard. The response of the electron- capture detector was found to be independent of the volume injected; identical responses were obtained when the same amount of DEHP was introduced in volumes of n-hexane.1 E Et 2 P a L a, g Retention time/min Fig. 1. Chromatograms of aldrin (A) (internal standard) and DEHP (B). For instrument parameters, see text. DEHP in 2 1.11 of mixturehg Fig. 2. Calibration graph for DEHP. of 2-lop1 1 The efficiency of this simple extraction system was verified by running blanks consisting of water fortified with DEHP. n-Hexane was found to be an effective solvent and the addition of acetonitrile optimised the extraction of the DEHP (Table 11). n-Hexane extracted more than 90 per cent. of the DEHP from the aqueous solution on the first extraction and less than 5 per cent. on the second. We therefore carried out only one extraction, which we considered to be sufficiently accurate for a rapid method.TABLE I1 RESULTS OF REPLICATE ANALYSES OF PURIFIED WATER CONTAINING ADDED DEHP Three samples were analysed a t each concentration. Amount Of DEHP DEHP foundjmg Recovery, per cent. added/mg 12.5 11.8 91.2 f 5.2 25-0 22.7 90-8 f 4.1 31.26 29.9 95.7 f 2.9860 WEISENBERG et aZ.: RAPID METHOD FOR MONITORING LOW Analyst, VoZ. 100 A series of samples of water and saline and glucose solutions packed in poly(viny1 chloride) were tested. The solutions were found to be free from DEHP [the sensitivity limit was 50 P.P.b. (parts per lo9)]. By increasing the volumes of the solutions used a sensitivity of 10 p.p.b. could be obtained. The results obtained for aqueous solutions stored in plastic bags show that the above residues of DEHP fall within a satisfactory margin of safety (Table 111).TABLE I11 DEHP RESIDUES I N AQUEOUS SOLUTIONS Source DEHP found, p.p.m. Distilled water (glass bottle) None Saline solution (glass bottle) None Saline solution - glucose solution 0.05-0*08 (0*13)* (in plastic bags) equipped with plastic tube and stored for 10 years) * Sample of saline kept for 6 months a t 37 OC. Saline solution (glass bottle with rubber stopper, 0.07-0.8 A sample of normal saline in poly(viny1 chloride) bags, kept in our laboratory for 6 months at 37 "C, was found to contain 80 p.p.b. of DEHP. Higher concentrations were found in saline solutions that had been stored for 10 years in glass bottles that were stoppered with a rubber cork fitted with a plastic tube. It is therefore worthwhile to test for DEHP contamina- tion of similar preparations used in infusions, such as protein hydrolysates, or solutions contain- ing lipophilic substances. When using gas - liquid chromatography with flame-ionisat ion detection, Rubin20 was unable to detect DEHP in a solution stored in a Travenol Viaflex con- tainer (formula PL 146) for more than 1 year over a wide range of ambient temperatures. In his experiments the blanks gave high values (0.24 p.p.m.) for DEHP; these high values can be related to the fact that chemicals are usually stored in plastic containers and can consequently be contaminated with DEHP.We therefore stress the importance not only of examining all chemicals for the presence of DEHP but also of checking all possible sources of contamination, such as water, rubber stoppers, plastic tubing and even the septum used in the chromatograph. This aspect, the appearance of interfering peaks from materials and chemicals, has recently been discussed by Levi and Nowick21 with reference to the determination of organochlorine pesticides.The migration of DEHP into blood storeg in plastic containers presents a very serious problem. Blood stored in plastic at 4 "C can extract 0-25 mg of DEHP per 100 ml per day; after 21 days the concentration of DEHP in whole blood was found to be within the limit of 50 p.p.m. Our method is suitable for the rapid determination of DEHP in large numbers of plasma samples; it is not time consuming and requires a minimum of blood, glassware and re- agents. We have established that more than 85 per cent.of the DEHP can be extracted from the plasma with reproducible results (Table IV). We used plasma rather than whole blood as many investigators have established that all of the DEHP extracted by blood is con- centrated in the plasma.22 TABLE IV RESULTS OF REPLICATE GAS - LIQUID CHROMATOGRAPHIC ANALYSES O F HUMAN PLASMA Each sample was analysed five times, Sample DEHP, p.p.m. I 70 f 4.4 I1 82 f 4.7 I11 86 f 4.3 The above procedure has also been used to extract pesticides. However, under the experimental conditions employed, no interfering peaks appeared near the retention timeDecember, I975 LEVELS OF DI- (2-ETHYLHEXYL) PHTHALATE IN SOLUTIONS 861 of DEHP. The peaks of the pesticides would interfere in the determination of phthalate esters of low relative molecular mass.Our method has also been applied to the rapid determination of DEHP in milk, the results of which will be published separately. Conclusion We are confident that our simple, rapid procedure will be of practical value for monitoring not only the migration of DEHP from plastic containers into blood and aqueous or lipophilic solutions but also the pollution produced by phthalate esters in the environment. 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. References Rubin, R. J., and Jaeger, R. J., Envir. Hlth Perspect., 1973, 3, 53. Singh, A. R., Lawrence, H. H., and Autian, J., J . Pharm. Sci., 1973, 61, 51. Rall, D. P., New Engl. J . Med., 1972, 287, 1146. Shibko, S. I., Envir. Hlth Perspect., 1973, 3, 131, Huges, H. J., FDA By-line, 1972, 3, 81 and 127. Marcel, Y. L., and Noel, S. P., Lancet, 1970, i, 35. Jaeger, R. J., and Rubin, R. J., New Engl. J . Med., 1972, 287, 1114. Cerbulis, J., and Ard, J. S., J . Ass. Off. Analyt. Chem., 1967, 50, 646. Williams, D. T., J . Ass. Off. Analyt. Chem., 1973, 56, 181. Hites, R. A,, Envir. Hlth Perspect., 1973, 3, 17. Nazir, D. J., Beroza, M., and Nair, P. P., Envir. Hltk Perspect., 1973, 3, 141. Godly, E. W., and Mortlock, A. E., Analyst, 1973, 98, 493. Hites, R. A., J . Chromat. Sci., 1973, 11, 570. Lee, F. D., Britton, J., Jeffcoat, B., and Mitchell, R. F., Nature, Lond., 1966, 211, 521. Bunting, W., and Walker, E. A., Analyst, 1967, 92, 575. Williams, I. H., J . Chromat. Sci., 1973, 11, 593. Thomas, G. H., Envir. Hlth Perspect., 1973, 3, 23. Krejci, M., and Dressler, M., Chromat. Rev., 1970, 13, 1. Weisenberg, E. , Schoenberg, Y . , and Ayalon, N., unpublished work. Rubin, R. J., Lancet, 1972, i, 965. Levi, J., and Nowick, T. W., Bull. Envir. Contam. Toxicol., 1972, 7, 193. Marcel, Y . L., Envir. Hlth Perspect., 1973, 3, 119. Received April 3rd, 1976 Accepted June 27th, 1975

ISSN:0003-2654    

DOI:10.1039/AN9750000857

出版商:RSC    

年代:1975

数据来源: RSC

摘要:

862 Analyst, December, 1975, Vol. 100, pp, 862-864 The Determination of Oxygen48 to Oxygen46 Ratios in Inorganic Phosphates by Gas - Liquid Chromatographic - Mass Spectrometric Examination of the Tri-n-butyl Derivative D. Barltrop and P. A. Lewis Paediatric Unit, St. Mary's Hospital Medical School, London, W.2 Aqueous solutions containing inorganic orthophosphates from biological materials were purified on ion-exchange columns and the phosphate was pre- cipitated as its silver salt. The dried silver phosphate was made to react with 1-bromobutane in dimethylformamide, yielding tri-n-butyl phosphate. This solution was suitable for injection directly into a combined gas chro- matograph - mass spectrometer for analysis, enabling the oxygen-18 to oxygen- 16 ratio of the original inorganic orthophosphate to be determined.Previous methods for the determination of oxygen-18 in enriched phosphates have involved its conversion into gaseous carbon dioxide or oxygen for examination by mass spectrometry.1 These methods require specialised apparatus and handling techniques. The conversion of inorganic anions into their trimethylsilyl derivatives,2 which have volatilities suitable for gas - liquid chromatography, and the use of these derivatives for determinations by mass spectrometry of oxygen-18 labelled phosphate3 have been described. In this paper an alternative method, which has been found convenient for the determination of inorganic phosphates in low concentrations in solutions derived from biological sources, is described. This procedure was developed for the determination of inorganic phosphate tracer in homo- genates of milk and faeces derived from metabolic balance studies in the newly born. Experimental Reagents These were of analytical-reagent grade unless otherwise stated.Dowex 50 W-X8 ion-exchange resin, 200-400 mesh. Dowex 2-X8 ion-exchange resin, 200400 mesh. Nitric acid, 0.1 and 1 N. Sodium hydroxide solutions, 0.2, 0.3, 0-4 and 0.5 N. Silver nitrate solution, 0-05 g m F . 1-Bromobutane. Dimethyljormamide. Potassium dihydrogen orthophosphate (oxygen-18, 85 per cent. abundance) solution, 1 mg mi-f. Apparatus The apparatus consisted of a Varian Aerograph 1700 gas chromatograph coupled via a helium separator to a Varian CH5 mass spectrometer. The chromatographic column wits a 1.54 m long x 7 mm 0.d. glass column packed with 3 per cent.silicone OV-101 on Celite AW-DMCS, 100-120 mesh (obtained from Phase Separations Ltd.) . The operating conditions were as follows: carrier gas, helium at a flow-rate of 30 ml min-l; injection temperature, 210 "C; column temperature, 170 "C; source temperature, 150 "C; ionisation voltage, 70 eV; and ionisation current, 300 PA. Procedure Acidify samples (2-20 ml) that contain at least 1 mg of phosphorus as inorganic phosphate with 1 N nitric acid in order to ensure the complete dissolution of the inorganic phosphate, and then ultracentrifuge at 90 000 g to obtain clear solutions. Determine the total free inorganic phosphate concentrations by use of the colorimetric method of Delsal and Manhouri4 on 0-1-ml aliquots.BARLTROP AND LEWIS 863 Purify the solutions by passing them through two ion-exchange columns. The first column removes cations that are likely to cause precipitation of phosphate under the alkaline condi- tions used in the second column, which is required to remove interfering anions.The first column is 1 cm in diameter by 4 cm long and is packed with Dowex 50W-X8 in the H+ form. Elute the column with 0-1 N nitric acid, then collect the first fraction, which is equivalent to the volume applied plus 5 ml, neutralise it with 5 N sodium hydroxide solution and add further hydroxide solution to give a final concentration of 0-2 N. Apply this alkaline fraction to the second column, which is 1 cm in diameter and 12 cm in length and is packed with Dowex 2-X8 in the OH- form.Elute this column with successive 50-ml amounts of 0.2, 0.3 and 0-4 N solutions of sodium hydroxide prepared in carbon dioxide free water. Collect 2-ml fractions of the eluate and locate the phosphate-containing fractions by carrying out colorimetric determinations on 0-1-ml aliquots. Next neutralise the combined phosphate fractions with nitric acid and precipitate the phosphate by the addition of 5 per cent. silver nitrate solution. Collect the yellow - green precipitate by centrifugation and wash it three times with 2ml of distilled water, then dry it at 105 "C, weigh and store it in a desiccator over phosphorus(V) oxide. Convert the inorganic silver phosphate into tri-n-butyl phosphate by reaction with l-bromo- butane, using a modification of the method of Baldwin and Higgin~.~ In the original method silver phosphate and a two-fold excess of I-bromobutane were refluxed at 103 "C for 8 h and were reported to give a 60 per cent.yield of tri-n-butyl phosphate. Ag,PO, + 3C,H,Br -+ 3AgBr + (C,H,),PO, Increased yields and a less viscous reaction medium are obtained by the addition of dimethylformamide in a volume twice that of the volume of 1-bromobutane used. In this method, heat the reaction mixture under reflux in an oil-bath at 100-110 "C for 2 h to give 90-100 per cent. yields of tri-n-butyl phosphate from silver phosphate. Finally, dilute a portion of the supernatant liquid approximately 100-fold with dimethyl- fonnamide and inject 2 pl of this solution directly into the combined gas - liquid chromato- graph - mass spectrometer.Results Tri-n-butyl phosphate gave the fragmentation pattern illustrated in Fig. 1, which shows the relative abundances of the fragments with their probable identities. Fig. 2 shows a sample enriched with 20 per cent. of oxygen-18. The P1804 to P160, ratios were thus determined by comparison of the 99 and corresponding 107, 105, 103 and 101 mass peaks. As different batches of labelled phosphate might vary in the proportions of the various phosphate species Pf60,180, P160,180,, P160180, and Pl8O4, the relative abundance of the oxygen-18 was deter- mined by calculating the ratio of the sum of the intensities of the 101-107 ions to that of the 99-107 ions. Relating this ratio to the total phosphate concentrations that had been determined colorimetrically allowed the concentration of oxygen-1 8 labelled phosphate to be ascertained.B 50 100 150 200 m/e Fig. 1. Mass spectrum of (C4H,),Pla0,. A, C4H,+; B, P1604H,; C, C4HoPls04H,+ ; and D, (C4Ho),P1604H,+.864 BARLTROP AND LEWIS m/e Fig. 2. Mass spectrum of (C4Hg)3P160p - (C4Hg)3PleOo,. The precision and accuracy of the method are indicated in Table I, in which labelled potassium dihydrogen orthophosphate solution was added in different amounts to unlabelled inorganic phosphate, milk and faeces. TABLE I RECOVERY OF ADDED PHOSPHATE LABELLED WITH OXYGEN-18 Relative abundance of leO, per cent. Observed r h \ Sample Added Observed* minus blank Aqueous orthophosphate solution 0.0 0.91 0.00 0.085 1.46 0.65 0.425 1-65 0.64 0.85 1.90 0.99 4-25 5-23 4.32 8.5 9-74 8-83 Milk .... .. . . 8.84 9.57 8.66 Faeces . . .. . . . . 19.7 21-25 20.35 * Mean of ten results. Standard deviation 0.07 0.27 0.12 0.06 0.35 0.28 0.72 1.02 Coefficient of variation, per cent. 8-1 48.6 18-6 5.9 8.0 3.2 9.5 4.8 Although concentrations of 1 per cent. of oxygen-18 in oxygen-16 could be detected, for the accurate determination of their ratios a level of at least 5 per cent. was desirable for the instrument described. The possibility of oxygen-18 enrichment in the helium separator did not appear to have any significant effect on the results. The research costs of this work were defrayed by a grant from Glaxo Laboratories Limited. Professor D. Bryce-Smith gave invaluable advice concerning the butylation procedure. 1. 2. 3. 4. 5. References Boyer, P. D., and Bryan, D. M., “Methods in Enzymology,” Volume 10, Academic Press, London, Butts, W. C., and Rainey, W. T., Analyt. Chem., 1971, 43, 538. Bar-Tana, J., Ben-Zeev, O., Rose, G., and Deutsh, J., Biochim. Biophys. Ada, 1972, 264, 124. Wootton, I. D. P., “Microanalysis in Biochemistry,” Churchill, London, 1964, p. 77. Baldwin, W. H., and Higgins, C . E., J . Am. Chem. SOG., 1952, 74, 2431. 1967, p. 60. Received May 6th, 1976 Accepted August llth, 1976

ISSN:0003-2654    

DOI:10.1039/AN9750000862

出版商:RSC    

年代:1975

数据来源: RSC

摘要:

Analyst, December, 1975, Vol. 100, $9. 865-872 865 Procedures for the Deoxygenation of Liquids J. Homer and A. Coupiand Department of Chemistry, University of Aston in Birmingham, Gosta Green, Birmingham, B4 YET Procedures are described for the deoxygenation of pure liquids, liquid mixtures and solutions of solids in liquids using tris (2,2'-bipyridy1)cobalt (11) perchlorate { [Co(bpy),] (ClO,), } and sodium tetrahydroborate (NaBH,) in a colour-indi- cating reaction. The efficiency of the procedures is assessed by reference to the lH spin - lattice relaxation times of a variety of materials. The values obtained show that the procedures are efficient, reproducible and time saving. It has been suggested1 that a mixture of tris(2,2'-bipyridyl)cobalt (11) perchlorate ([Co(bpy),] (C1OJ2'j and sodium tetrahydroborate (NaBH,) is effective for the removal of oxygen from samples when its presence is detrimental, e.g., when basic experiments on nuclear magnetic resonance are to be carried out.This paper reports a more definitive examination of the procedure, and additional techniques developed recently for use with the procedure are described. The methods developed permit the deoxygenation of several types of samples, viz., pure liquids, liquid mixtures and solutions of solids in liquids. Comments on the Detection of Oxygen While commercial equipment specifically designed for the determination of dissolved oxygen, particularly in aqueous media, is readily available, it is well known that the nuclear magnetic resonance spin - lattice relaxation time (T,) is extremely sensitive to the presence of oxygen and so this parameter can be used as a more extensive guide to the concentration of oxygen in solutions. For this reason, and because of the importance of obtaining T , data for their intrinsic value, this parameter has been studied for a representative selection of samples in order to confirm the validity of the procedures used to deoxygenate them.Fundamentally, nuclear magnetic resonance spectra derive from the quantised energy changes of individual nuclei when subject to a strong homogeneous static magnetic field (B,) and a weaker rotating field (B,) produced by a radiofrequency signal. However, the detection of the spectra depends on changes in the macroscopic magnetisation vector (M,) in the direc- tion of B,.Away from resonance and at thermal equilibrium the magnetisation vector has the value M,. After perturbation due to nuclear resonance the transient value M , returns to Mo in the characteristic spin - lattice relaxation time. The mechanism of the relaxation process can be simply considered, at a nuclear level, to be due to a net transfer of energy from the nuclei to their environment, i.e., to the lattice, which transfer is brought about by the time-dependent magnetic fields produced at the nuclei by various effects of the atoms and molecules constituting the lattice. Paramagnetic entities such as oxygen produce large fields that enhance the relaxation process and significantly reduce the value of TI. For two nuclei A and B the rate equations can be written2 as where TIaa is the total of the inter- and intramolecular contributions made by nuclei of type A to the relaxation time of A and TIBA is the specific contribution to the relaxation time of A arising from the interaction of A with B.Inspection of either equation (1) or (2) reveals that for one nucleus alone, say A, if M , is changed by a factor a at time t = 0 with a strong radiofrequency field, the instantaneous values of MZ, detected by using a low value of B,, follow an exponential decay from aM, to M,. -t Tl In (M, - M,) = - + In (M, - aM,) . . .. - (3)866 HOMER AND COUPLAND : PROCEDURES FOR Analyst, Vol. 100 Consequently, values of T, can be deduced from this so-called adiabatic rapid passage with repetitive sampling (ARPS)3 technique by plotting In (M, - M,) zleysm t.A similar simple exponential decay situation can be achieved for two nuclei provided that saturation of one nucleus [e.g., MZB = 0 in equation (l)] is achieved by double irradiation. When the resonance of one nucleus is saturated in this way an additional effect occurs. For example, when MzB = 0 and the A system regains equilibrium, dMZA/dt = 0, and equation (1) reduces to so that the intensity of the A signal increases, owing to what is called the nuclear Overhauser effect (NOE).495 Because the NOE enhancement is related [to the sixth power of the distance between A and B it has obvious immense potential in the determination of molecular structures. However, it can be seen from equation (4) that the NOE depends on T, values, which are sensitive to the presence of oxygen, and before the effect can be used definitively all of the dissolved oxygen must be removed from the samples studied.Principles of Deoxygenation In solution [Co(bpy),] (ClO,), and sodium tetrahydroborate yield a deep blue colour when all of the free oxygen has been removed. Should this solution be exposed to additional oxygen it reverts to its original brown colour. The presence of excess of sodium tetrahydroborate may induce the original blue colour, but the cycle cannot be repeated indefinitely. The detailed mechanism of the deoxygenation reaction has not been elucidated fully but evidence exists that points to two likely alternatives. The first depends on the fact that it is possible to isolate from the blue solution tris(2,2’-bipyridyl)cobalt (I) perchlorate6 and an outer-sphere reduction of oxygen by this complex may occur to give a cobalt(II1) complex.The latter, in the presence of excess of tetrahydroborate, can produce a cobalt(1) complex, which is characterised by a blue colour due to the strong charge-transfer band of the 2,2’-bipyridyl- cobalt(1) complex. The second possible mechanism depends on the fact that in the presence of phosphine ligands [Co(bpy)H,(PR,),]ClO, can be isolated.’ This suggests that the tris complex dissociates to give a monobipyridyl complex that, by analogy with work on the related rhodium system,* may take up hydrogen reversibly or oxygen irreversibly such that cobalt(II1) complexes can be formed. In the presence of excess of tetrahydroborate the cobalt(II1) complexes yield a cobalt (I) complex with its characteristic blue colour.Despite the fact that the detailed mechanism of the above chemical method for deoxy- genating liquids is uncertain, there can be little doubt that it is most efficient. Nevertheless, it is necessary to compare this method with more conventional procedures such as those based on freeze - thawing or displacement of oxygen by inert gases. The TI data for benzene given in Table I facilitate this comparison. These data show that while each procedure results in the removal of oxygen the chemical method is by far the most efficient and repro- ducible. An estimate of the extent of deoxygenation achieved can be made by reference to T , data for water. Because a saturated solution of sodium sulphite in water is commonly used as a zero point for the calibration of commercial oxygen detectors, the T, value of water distilled from such a solution under vacuum serves as a useful standard.The value of 4.0 s so obtained is higher than that of 3.4 s for air-saturated water but less than that of 4.4s obtained for a chemically deoxygenated sample. Although these values are very similar the determinations were repeated many times and in no instance was the relative order of these values different, 0r.wa.s there any overlap in the small range of values obtained for each system. The small differences in these data thus indicate a very slight residual oxygen content in the sodium sulphite treated sample, which would give rise to a small but significant zero error.Experimental Procedures A. General The recognised method of preparation,, of [Co(bpy),](ClO,), has been superseded by the following simpler method of Bhuyat.13 Add a solution of hydrated cobalt (11) perchlorate (0.01 mol, 3.67 g, in 10 ml of ethanol) to a solution of 2,2’-bipyridyl (0-03 mol, 4.60 g, in 20 ml of ethanol), Recrystallise the resulting precipitate from ethanol - water (1 + l), washDecember, 1975 THE DEOXYGENATION OF LIQUIDS 867 with ice-cold ethanol and dry over phosphorus(V) oxide so as to obtain golden brown crystals of the cobalt complex. The amounts of the two compounds used for deoxygenation should be kept as small as possible in order to minimise reduction of the sample. It is possible to deoxygenate success- fully samples of total volume 20 ml by using only 1 mg of the cobalt complex; with smaller amounts, the colour change is indistinct.The amount of tetrahydroborate used depends upon the compounds being deoxygenated. If the compounds are inert to tetrahydroborate, then about 15-20 mg of it will suffice, but if any of them react in some way 50-60 mg may be required. It should be noted that care must be exercised when handling the compounds firstly because little is known about the physical and toxicological properties of the cobalt compound and secondly because of possible side-reactions. B. Basic Deoxygenating Procedures The prime prerequisite for deoxygenation is that the cobalt complex and the sodium tetrahydroborate should both come into intimate contact with the material being deoxy- genated.This is not possible directly in all instances and an intermediate solvent may be required in order to form a link between the deoxygenating compounds and the sample. It may be miscible with the sample, thus forming a homogeneous mixture, or it may be immiscible, forming a heterogeneous mixture. In the latter instance prolonged vigorous stirring is necessary to produce a coarse emulsion so that the oxygen can diffuse into the intermediate solvent from which it is removed. The main requirements of the intermediate solvent are that it must dissolve the cobalt compound and the tetrahydroborate, preferably without reacting with either of them, and that it should have a low vapour pressure at the working temperature so that subject materials can easily be removed from it by distillation.Some suitable solvents for the active com- pounds are water, acetone, dichloromethane, dimethyl sulphoxide and dimethylformamide. C. Deoxygenation of Pure Liquids and Liquid Mixtures Using Distillation for Sample Deoxygenation is most simply carried out when the material under study dissolves the deoxygenating compounds. In this instance, and when the sample forms the top layer of a heterogeneous mixture, the following procedure should be adopted. Introduce about 10ml of sample into a suitable flask together with a magnetic follower and the appropriate amounts of the deoxygenating compounds ; when an intermediate solvent is used it should be added first together with the active compounds. Rapidly attach the flask to a vacuum line (at about 10-3 torr) and freeze the contents of the flask.Remove air via the vacuum system and allow a good vacuum to develop. Isolate the flask from the vacuum system, permit it to warm up and then stir the contents of the flask in order to allow the reaction to proceed to completion. The reaction results in the evolution of gas (presumably mainly hydrogen) and therefore distillation of the sample into the nuclear magnetic resonance tube can be hindered. When the production of hydrogen has slowed down, freeze the contents of the flask again and open the flask to the vacuum line in order to remove hydrogen. Isolate the flask and allow the contents to melt prior to distilling the required amount of sample into a pre-cooled nuclear magnetic resonance tube. After isolating the flask, freeze the contents of the resonance tube and re-open the vacuum line before sealing the tube under vacuum.In certain instances these techniques must be modified in order to avoid violent side- reactions, because, particularly with some chlorinated samples such as tri- and tetrachloro- methane, when the deoxygenating compounds are added to a heterogeneous mixture of the sample and, for example, water, an exothermic reaction occurs to give solutions with colours ranging from orange to green. The required orderly deoxygenation should be obtained by first deoxygenating the intermediate solvent layer, then freezing it, and subsequently adding the sample to it, thus freezing the latter. Finally, the trapped air should be removed and the mixture melted so as to allow the two liquids to come into contact.Transfer D. Deoxygenation of Liquid Mixtures Using Syphoning for Sample Transfer An apparatus which is suitable for the transfer of deoxygenated samples that occupy the868 HOMER AND COUPLAND : PROCEDURES FOR Analyst, Vol. 100 lower layer of a heterogeneous mixture to nuclear magnetic resonance tubes, and which is particularly useful for transferring samples of known composition, is shown in Fig. 1. Attach the apparatus to a vacuum line and allow a vacuum to develop. Rotate the three- way tap and close the Rotaflo taps in order to isolate the flask from the remainder of the apparatus. Add to the flask about 5 ml of a heterogeneous intermediate solvent together with the appropriate amounts of the deoxygenating compounds, place a magnetic follower in the flask and replace the stopper.Open the flask to the isolated manifold and draw o f f the trapped air plus evolved hydrogen. Isolate the flask, re-evacuate the manifold and repeat the process so that any remaining hydrogen is removed from the heterogeneous intermediate solvent. Vacuum line Fig. 1. Apparatus for deoxygenation using a syphoning process to transfer the sample. Freeze the solvent and add the sample material in order to freeze it. It should be noted that the volumes of each solvent must be such that the capillary tube by which the transfer of sample is made has its end well away from the unwanted layer, in order to prevent the intro- duction of unwanted material while stirring the contents of the flask.Evacuate and isolate the flask. At this stage the region between the flask and the Rotaflo taps must also be evacuated. Allow the contents of the flask to warm up while the mixture is stirred vigor- ously. The sample should now be drawn up the appropriate capillary tube by cooling the region near the Rotaflo tap (continued production of hydrogen in the main body of the flask facilitates this step). This capillary tube should be washed out by repeatedly cooling and gently warming the region of the tube near the Rotaflo tap, the liquid pushed back into the flask being allowed to equilibrate before sucking more liquid back up the tube. On the satisfactory completion of this operation set the three-way tap to the all-isolated position and very slowly open the pertinent Rotaflo tap so as to allow the sample to pass through to the nuclear magnetic resonance tube.At first it may distil but when the saturated vapourDecember, 1975 THE DEOXYGENATION OF LIQUIDS 869 pressure is reached the sample will flow down the side of the glass tube into the resonance tube. Having collected sufficient sample, shut the Rotaflo tap, freeze the sample and evacuate and seal the resonance tube. E. Deoxygenation of Solutions of Solids in Liquids The procedure devised for deoxygenation of solutions of solids in liquids is basically the same as for liquid mixtures although there are additional problems, vix., (i) cooling may induce precipitation of the solid, which may not all be redissolved, (ii) initially, after passing through the Rotaflo tap, the solvent may evaporate, leaving a deposit of solid on the walls of the tube.This deposit may be taken up by the solution after equilibrium vapour pressurehas been reached. The first problem should be minimised by avoiding saturated solutions, keeping careful control of the magnetic stirrer and allowing the flask to warm to room temperature after each cooling. The second problem can be remedied by the addition of a small glass-finger tube on the side of the Rotaflo taps near to the nuclear magnetic resonance tube. When preparing to draw off the sample, open the tap nearest to the finger and allow a flow of liquid through it. The design of the apparatus is such that the liquid will run down into the finger and not into the resonance tube. When the saturated vapour pressure of the solvent is reached in the vicinity of the resonance tube, open the other Rotaflo tap and allow the sample liquid to flow down into the resonance tube.The sample in this tube should then be carefully frozen and the tube evacuated and sealed. Prepare sufficient sample to cover both of the capillary tube ends in the flask. Results and Discussion The procedures for deoxygenating pure liquids and liquid mixtures were tested using compounds for which accepted literature values of T, exist. TABLE I SPIN - LATTICE RELAXATION TIMES OF BENZENE MEASURED BY THE ARPS TECHNIQUE AFTER DE-GASSING BY CONVENTIONAL AND CHEMICAL METHODS Values other than the literature values were measured a t about 309 K. De-gassing technique No de-gassing .. .. .... .. .. .. .. a t ton, 4 cycles . . .. .. .. . . .. .. a t ton, 1 cyclet . . .. .. .. .. .. . . 2 cycles . . .. .. .. . . .. .. 3 cycles . . .. .. .. .. .. .. 4 cycles . . .. .. .. .. .. .. 5 cycles . . .. .. .. .. .. .. 6 cycles . . .. .. .. . . .. .. Freeze - thaw procedure : Bubbling oxygen-free nitrogen for : 15min .. .. .. .. . . .. .. .. . . 30min . . .. .. * . * . .. .. .. . . 60 min .. .. .. * . .. .. .. .. . . homogeneous mixture containing dimethylformamide . . . . Chemical de-gassing : homogeneous mixture containing dimethyl sulphoxide . . . . heterogeneous mixture with water .. .. . . .. Reported in reference 9 (temperature, 298 K) . . . . .. . . Reported in reference 10 (temperature, 303 K) . . .. .. . . Reported in reference 11 (temperature, 305.1 K) .... .. * Average of four measurements. .. .. . . . . .. . . .. .. .. . . .. .. .. .. .. .. .. . . .. .. .. . . .. . . .. . . .. .. .. .. .. . . .. .. T,ls 4*8* 5.5* 23.6 21-8 2 1 4 20.8 17.7 22.1 20.9 20.9 13.1 22.7 22.8 23-1 23.4 23.0 24.8 19.3:: 18.4: 22-0$ t The f&eze -thaw operations were continuous; samples were removed after each cycle into $ Highest reported values. nuclear magnetic resonance tubes on a specially made all-glass manifold.870 HOMER AND COUPLAND : PROCEDURES FOR Analyst, Vol. 100 Nuclear magnetic resonance tubes were attached to the vacuum manifold via a double O-ring glass to metal connection. Before use each tube was heated under vacuum to dispel oxygen from the walls and the deoxygenated sample was transferred into the resonance tube by distillation or syphoning.In preliminary experiments it was found in some instances that, despite critical inspection, the seal of the resonance tube was defective in some way, for example, during 1 d the TI value for a sample of benzene would fall from about 20 s to about 6 s. It is thought that with the thin-walled resonance tubes normally used, cracks occurred on cooling after sealing. The initial method of sealing was to heat the tube under vacuum at about six places around its circumference with a pencil flame, allowing the tube to collapse gradually. As this proved inadequate, each resonance tube was subsequently heated by an all-encompassing flame near the open end, so that the tube was restricted and thickened prior to installation on the vacuum system. Under vacuum it was then possible to obtain a good seal by using a pencil flame.Values for the spin - lattice relaxation times were obtained, using the ARPS method,3 on a Varian HAlOOD nuclear magnetic resonance spectrometer connected to a Hellige He-lt fast-response recorder. An Airmec 422 signal generator was used for double resonance in order to produce effective single-spin systems where necessary. The relaxation times were calculated by using a computer analysis of In (Mo - M,) versus t data. Some typical values obtained for single compounds are given in Tables I and 11. It should particularly be noted that the T, values obtained for both the ring and the methyl protons in samples of the methylbenzenes are significantly higher than the literature values, even after taking into account the slight difference in experimental temperatures. The relaxation times of benzene (discussed earlier) and cyclohexane, which were obtained by using both the distillation and syphoning procedures, are also higher than the literature values.TABLE I1 COMPARISON OF THE IH-T, VALUES OF PHYSICALLY AND CHEMICALLY Sample material Toluenet (ring protons) . . (methyl protons) (ring protons) . . (methyl protons) (ring protons) . . (methyl protons) Acetone . . .. Dichloromethane Cyclohexane . . p-Xylene t Mesitylenet .. .. .. . . .. .. . . . . . . DE-GASSED COMPOUNDS Tl Is I \ After After No freeze - thaw chemical Literature procedure de-gassing* values de-gassing 3-76 4-18 21.3: 1 6.OS 3-29 3.50 11.6$ 9-0s 3.43 4.36 16.7: 1 4.09 3.32 3-88 7*3$ 7-59 * Values obtained (except for cyclohexane) after distillation from the pure liquid containing t Ring and methyl lH-Tl values were measured separately by saturating out the unwanted $ Values reported are the average of several measurements at about 309 K.5 Values reported are the average of several measurements at about 306.4 K. 7 The estimated error for 13 determinations is f 0 - 2 s. dimethyl sulphoxide, and after syphoning from heterogeneous mixtures with water. the deoxygenating compounds. resonance. * * Average of values obtained after distillation from a heterogeneous mixture containing The procedure for the deoxygenation of liquid mixtures was assessed, using benzene and carbon tetrachloride, as described earlier in section D. This procedure was used to obtain deoxygenated samples of benzene - carbon tetrachloride mixtures over the whole range of benzene molar fractions.The molar fractions of the components of each sample before they were deoxygenated were calculated by mass. As a check, when samples had been used inDecem bev, 1975 THE DEOXYGENATION OF LIQUIDS 87 1 the nuclear magnetic resonance experiments, the tubes were opened and the molar fractions found by measuring the refractive indices and referring to a calibration graph. For 24 pre- pared samples the difference between the molar fractions before and after deoxygenation was less than 0.02 and in 19 instances was below 0.005, which is within the experimental error for the refractive index measurements. The graphs of T, against the molar fraction of benzene (Fig.2) are nearly identical in shape with those obtained by Mitchell and Eisner.lSJ6 except that the relaxation times are slightly longer, indicating an increase in the efficiency of removing oxygen in the present work. The syphoning principle was also applied to solutions of solids in liquids. However, values of the spin - lattice relaxation times for these solutions could not be found in the literature and thus no reference was available with which to assess the merit of this procedure. Never- theless, it was decided to investigate the dependence of the spin - lattice relaxation time of the protons of 1,2,4,5-tetramethylbenzene on the concentration of this compound in carbon tetrachloride. Fig. 3 shows the values obtained. The deviations of the experimental points from the lines drawn are within experimental error, thus indicating that the required consistency in the experimental procedures has been achieved.15 14 0 0.5 1 -0 Molar fraction of benzene Fig. 2. A comparison between the 'H-T, values of physically and chemically de- gassed samples of benzene - carbon tetra- chloride: points with full lines, this work; broken lines, references 16 and 16. - - Fig. 3. Spin - lattice relaxation times of 1,2,4,5-tetramethylbenzene in carbon tetrachloride as a function of concentration, 0, T, for ring protons; X, T, for methyl protons. It has to be stressed that the benefits of the above procedures for removing free oxygen, in a colour-indicating method, from samples in which its presence is not wanted, lie in the efficiency, ease and speed with which they can be carried out. While the procedures have been used to remove oxygen from samples that were then analysed by using nuclear magnetic resonance spectroscopy, their use is not restricted to this field. For example, it is possible that they may find applications in fluorescence spectroscopy, electron spin resonance spectroscopy and in other physical analytical techniques. References 1. 2 . 3. 4. Homer, J., Dudley, A. R., and McWhinnie, W. R., J . Chem. SOC., Chem. Commun., 1973, 893. Solomon, I., Phys. Rev., 1956, 99, 559. Heatley, F., J . Chem. SOC., Faraday Trans. I I , 1973, 69, 831, and references therein. Kaiser, R., J . Chem. Phys., 1965, 42, 1838.872 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 16. 16. HOMER AND COUPLAND Anet, F. A. L., and Bourn, A. J. R., J . A m . Chem. SOC., 1965, 87, 5250. VlCek, A. A., Nature, Lond., 1957, 180, 754. Camus, A., Cocevar, C., and Mestroni, G., J . Organometall. Chem., 1972, 39, 355. Bhuyat, I. I., and McWhinnie, W. R., J . Organometall. Chem., 1972, 46, 159. Nederbragt, G. W., and Reily, C. A., J . Chem. Phys., 1956, 24, 1110. Powles, J. G., Ber. Bunsenges. Phys. Chem., 1963, 67, 328. Nolle, A. W., and Mahendroo, P. P., J . Chem. Phys., 1960, 33, 863. Burstall, F. H., and Nyholm, R. S., J . Chem. SOL, 1952, 3670. Bhuyat, I. I., Ph.D. Thesis, University of Aston in Birmingham, 1972. Reeves, L. W., and Yue, C. P., Can. J . Chem., 1970, 48, 3307. Mitchell, R. W., and Eisner, M., J . Chem. Phys., 1960, 33, 86. Mitchell, R. W., and Eisner, M., J . Chem. Phys., 1961, 34, 651. Received January 20th, 1976 Accepted August l l t h , 1975

ISSN:0003-2654    

DOI:10.1039/AN9750000865

出版商:RSC    

年代:1975

数据来源: RSC