摘要:

ANALYST, OCTOBER 1986, VOL. 111 1215 Automated Procedure for pH Measurement Using a Flow Cell Mark R. Howson and William A. House* Freshwater Biological Association, East Stoke, Wareham, Dorset BH20 SSB, UK and Alan D. Pethybridge Department of Chemistry, University of Reading, Whiteknights, Reading, Berkshire RG6 ZAD, UK A flow cell for use with small glass electrodes has been designed and constructed from Perspex. The operation of the cell and the pulsing of the potassium chloride solution to the junction have been automated. The stability of the electrode potential is satisfactory with flow-rates in the range 0.2-17 ml min-1. Work carried out with two different glass electrodes in dilute buffers has indicated that it is important to check the performance of the glass electrode in both concentrated and dilute buffers, even though the slope of the electrode response in the former often appears to be acceptable.Keywords: pH measurement; flow cell; automation; liquid junction; natural waters A major source of error in pH measurement is associated with the liquid junction potential that develops across the porous ceramic plug of the reference electrode. The operational definition of pH assumes that the liquid junction potential remains constant in different solutions.1 It has been demon- strated that this is not true for many electrode pairs with constrained junctions.2 However, the widespread use of standard buffers of similar ionic strength obscures the liquid junction error because the liquid junction potentials with two such standard buffers, when compared under the same stirring conditions, are of comparable magnitude .2 As an alternative to the constrained liquid junction, Covington et aZ.3 have investigated the use of a glass flow cell, following the design proposed by Culberson.4 The advantages of this cell have recently been discussed in detail.5 The flow cell has a well defined renewable liquid junction and so overcomes some of the limitations of the constrained junction.The pH values of IUPAC buffers measured with this flow cell are in good agreement with those derivable from cells without liquid junctions.3.4 The glass flow celW suffers a number of disadvantages: (a) bubbles can become trapped near the tip of the glass electrode, necessitating the removal of the electrode; (b) the glass electrode is fitted with a ground-glass joint and is not interchangeable without regrinding the joint; (c) a sharp pinhole T-junction is difficult to make using glass; and (d) the design is not sufficiently rugged for field use.To overcome all of these limitations, a robust and inexpensive flow cell with improved flushing characteristics was designed and construc- ted from Perspex. Experimental Flow Cell Previous work by House and Donaldson6 on phosphate adsorption by calcite used a glass flow cell of the design of Covington et aZ.3 This cell was later replaced by a Perspex flow cell of similar design, but both cells suffered from poor flushing characteristics. As a consequence, the Perspex cell has been re-designed and is shown in Fig. 1. There are several advantages in using Perspex rather than glass as a construction medium, including ease of precision boring of the flow tubes and the glass electrode chamber.Moreover, the distance between the chamber and the T-junction can be accurately fixed during construction at any pre-determined distance. The re-designed Perspex flow cell consists of two parts: (a) the * To whom correspondence should be addressed. main body consisting of the glass electrode chamber and the T-junction to the reference electrode and ( b ) the glass electrode sleeve. It is designed for use with small glass electrodes having a tip diameter of ca. 5 mm. The clearance between the external surface of the glass electrode and the internal walls of the electrode chamber is 1.5 mm. The mixing in the electrode chamber has been improved by introducing the added solution off-centre at the bottom of the chamber.The exit port at the top ensures that the solution swirls around the glass H+-responsive surface, preventing any air bubbles becoming trapped. The glass electrode is held loosely in position within the electrode sleeve by a nitrile rubber O-ring. When the sleeve is screwed into position the O-ring is compressed, forming water-tight seals between the electrode and the electrode sleeve, and between the electrode and the main body. The T-junction was constructed 2 mm downstream from the glass electrode chamber, hence reducing the d.c. resistance of the system and the signal noise associated with a.c. pickup. Even so, it was found to be necessary to screen the glass electrode by encasing the flow cell in a metal container connected to the mains earth.The T-junction was formed from a 1 mm diameter conduit from the potassium chloride solution, connected to the 1.5 mm i.d. tube from the electrode chamber. No attempt was made to examine the changes in the flow cell performance associated with different geometry junctions. Fig. 1. and (c) electrode and sleeve Schematic diagram of flow cell. (a) Side view; (b) plan view;1216 v !I \I ANALYST, OCTOBER 1986, VOL. 111 Turn valve I The flow cell T-junction was connected to the potassium chloride reservoir containing the reference electrode. The glass electrode chamber was connected to one of four solutions via a valve with four inlet ports as shown in Fig. 2. The valve was rotated using a stepper motor incorporating a gearbox.All the connections were made using 2 mm i.d. PTFE tubing. The whole system was housed in an air thermostat at 25.0 k 0.2 "C. 11 rnV reading Input pH of standard buffers I r Fig. 2. Schematic diagram of apparatus. B1-B4, pH buffers; FC, screened flow cell; R, 3.5 M potassium chloride reservoir with reference electrode; ST, stepper motor, Im ex Electrical, Model ID35; L, digital linear actuater, Airpax; &, 5-ml syringe; V, miniature valve, Hamilton HVX; W, waste The performance of the flow cell was assessed with four buffers and two new glass electrodes: a Radiometer G2222C and a Russell pH SWRM/25/757. A calomel reference electrode, Radiometer K401, was used for all measurements. The electrodes were connected to a Radiometer PHM64 meter, which was interfaced to an Apple I1 microcomputer via a binary coded decimal channel. The buffers used are listed in Table 1.Measurement Procedure A computer program was developed to automate the experi- ment and to allow flexibility during the testing procedure. The flow chart is given in Fig. 3. The program centres around a keyboard scan, which automatically initiates one of the four cycles depending on the key pressed. The flow of potassium chloride solution to the T-junction was controlled with a syringe connected to a second stepper motor, modified to incorporate an internally threaded rotor fitted with a leadscrew shaft. Each pulse of this motor displaced approximately 4 p1 of potassium chloride solution. A potential drift is observed if the liquid junction is not flushed with potassium chloride solution after turning the valve.Experiments with dyes have indicated that this is because the junction retreats down the potassium chloride line. An automatic flush, dispelling approximately 0.4 ml of potassium SCAN KEYBOARD I I 'T' junction flushed? 1 pulse 1"" Rapid flush t voltage to B1,62, 63 or 64 Calculate pH of 63 or 64 U u Fig. 3. Flow diagram of control program. Program monitors the keyboard, continuously branching when specific keys are operated. J and T loops continue until another key is pressed Table 1. Values of pH at 25 "C for the buffers used in this study Molalityl Buffer number Composition mol kg-1 pH Reference Buffer 1 . . . . . . KH2P04 - Na2HP04 (1 + 3.5) 0.008695 7.413 8 Buffer 2 .. . . . . Na2B407. 10H20 0.01 9.180 8 0.03043 Buffer3 . . . . . . Dilute (1 + 9) KH2P04 - Na2HP04 (1 + 3.5) 0.0008695 7.605 3 Buffer 4 . . . . . . Dilute (1 + 9) KH2P04 - Na2HP04 (1 + 1) 0.0025 7.068 3 0.003043 0.0025 7.065 2ANALYST, OCTOBER 1986, VOL. 111 170 > 150 E . - m .- 130 c. 0 Q &lo 2 c 0 al - 9 0 - 70 1217 - - - - - chloride solution in 11 s, is written into the operating program in order to produce a sharp junction at the T. During, measurement, the flow of solution was maintained by gravity feed and the rate of flow adjusted by varying the height of W (Fig. 2). While the rate of flow of solution was changing, the liquid junction was continuously renewed by pulsing the potassium chloride line at a rate of 15 pulses min-1.The two stepper motors were controlled using two inte- grated-circuit four-phase translators (IC SAA1027, supplied by RS Components). The interface board was designed to allow connection to any TTL-compatible parallel port, such as those supplied for the BBC Model B, RML 380 or Apple I1 machines. In this application, an Apple I1 fitted with a 6522 versatile interface adaptor board (U-Microcomputers) was employed. The system was evaluated by looking at (a) the relationship between response time and flow-rate and (b) the pH value of diluted buffers 3 and 4, as recommended by Illingworth.2 These buffers have been assigned pH values using cells without a liquid junction3 and are therefore suitable for examining liquid junction effects in our flow cell. The very dilute strong acid solutions employed by Davison and Woof' may be more relevant to some natural waters than the dilute buffers used in this work, but because the inadequate buffer capacity of such acid solutions renders them susceptible to contamination, they were not used in this evaluation.Results Response Times The response time of the cell is defined as the time between switching the valve and the electrode producing a voltage reading within kO.1 mV of its final value. Although the flow-rate affected the flow cell response time, flow-rates in the range 0.2-17 ml min-1 had no effect on the final electrode output (to within kO.l mV) once a steady signal was obtained. Even at extremely low flow-rates (0.2 ml min-1) the potential remained steady once a stable junction had been formed.When the flow was stopped completely, the potential remained constant for approxi- mately 20 s before showing signs of drift (>0.2 mV). The electrode drift reported by Covington et aZ.5 for slow flow-rates was not observed here. The response of the Radiometer G2222C electrode to changing buffer solutions in the flow cell is shown in Fig. 4. 100 200 300 400 500 Timeis Fig. 4. Response of the flow cell to changing from buffer 1 to buffer 2 at different flow-rates: A, 1.9 cm3 min-1; B, 3.1 cm3 min-1; and C, 8.6 cm3 min-1 The electrode potential after the valve was switched between buffers depended on the flow-rate. At the lowest flow-rate shown, 1.9 ml min-1, the flushing was slow and produced a noisy signal until flushing was complete.At the higher flow-rates, the response time was mainly influenced by the speed of the electrode response. The response time at a flow-rate of 2.9 k 0.3 ml min-1 for a change from buffer 1 to buffer 2 was 150 s, and similarly for a change from buffer 2 to 1. A longer response time, ca. 250 s, was observed for changes between buffer 2 and the diluted buffer 3, and vice versa. The initial responses of the electrode on changing from buffer 1 or 3 at similar flow-rates were similar. However, the final stable potential was reached very quickly with buffer 1 , whereas the last 0.5 mV of change in potential occurred more slowly with buffer 3. This is attributed to the slow diffusion of ions outwards from the gel layer of the glass electrode after the ionic strength of the solution has been reduced.The minimum volume needed to obtain a steady potential depended on the flow-rate and the response time. However, in the worst instance (ie., the change from buffer 2 to 3), and with a flow-rate of 3 ml min-1, the minimum volume was 12 ml. Diluted Buffers The pH values of diluted buffers 3 and 4 were obtained using each of the glass electrodes in the flow cell in turn. Each determination was immediately preceded by calibration of the electrode using buffers 1 and 2. The results are listed in Table 2 together with the slope of the electrode response obtained using buffers 1 and 2. The pH values of these dilute buffers at 25 "C are given in Table 1. The pH values obtained for the diluted buffers 3 and 4 using the SWRM/25/757 electrode were consistently 0.03 pH unit higher than the assigned values.The measurements reported in Table 2 were made over a period of 10 weeks; no change in the performance was noticed over this period. Throughout this period, however, the second elec- trode (G2222C) gave excellent agreement with the assigned values for both diluted buffers. Conclusion It is important to have good control of the liquid junction during measurements with different buffers. Changes in flow-rate produce pressure changes. near to the junction that adversely affect the potential. The automated procedure described here eliminates errors associated with the collapse of the junction. The design of our flow cell produces excellent mixing in the electrode chamber. The stability of the electrode potential has been extended from previous reports3 to flow-rates in the range 0.2-17 ml min-1.Our experience with the glass electrode SWRM/25/757 suggests that it is important to check electrode performance in both concentrated and diluted buffers, even though the slope Table 2. Results of pH measurements with buffers pH from Number of Glass electrode Buffer flow cell* determinations SWRMI251757f. . . 3 7.627 (0.008) 12 4 7.095 (0.001) 3 G2222CJ: . . . . 3 7.601 (0.005) 12 4 7.069 (0,001) 12 * Standard deviations in parentheses. t Nernstian slope = 58.65 k 0.50 mV over 16 measurements. J: Nernstian slope = 59.21 k 0.06 mV over 26 measurements.1218 ANALYST, OCTOBER 1986, VOL. 111 of the electrode response in the former is acceptable. We feel that the SWRM/25/757 electrode tested here is not reliable for use with solutions of low ionic strength. The technique is suitable for further development for routine industrial and field applications. We thank NERC and the Freshwater Biological Association for a CASE studentship for M. R. H. References 1. Bates, R. G . , “Determination of pH, Theory and Practice,” Second Edition, Wiley, New York, 1973, p. 29. 2. Illingworth, J. A., Biochem. J., 1981, 195, 263. 3. 4. 5. 6. 7. 8. Covington, A. K., Whalley, P. D., and Davison, W., Analyst, 1983, 108, 1528. Culberson, C., in Whitfield, M., and Jagner, D., Editors, “Marine Electrochemistry,” Wiley, New York, 1981, p. 200. Covington, A. K., Whalley, P. D., and Davison, W., Pure Appl. Chem., 1985, 57, 877. House, W. A., and Donaldson, L., J. Colloid Interface Sci., in the press. Davison, W., and Woof, C., Anal. Chem., 1985, 57, 2567. Whitfield, M., “Ion-selective Electrodes for the Analysis of Natural Waters,” Australian Marine Science Association, Sydney, 1971, p. 76. Paper A6156 Received February 20th, 1986 Accepted April 25th, 1986

ISSN:0003-2654    

DOI:10.1039/AN9861101215

出版商:RSC    

年代:1986

数据来源: RSC

摘要:

ANALYST, OCTOBER 1986, VOL. 111 1219 SHORT PAPERS Flow Injection Determination of Nitrite by Amperometric Detection at a Modified Electrode James A. Cox and Krishnaji R. Kulkarni Department of Chemistry and Biochemistry, Southern Illinois University, Carbondale, IL 6290 1, USA The modification of a glassy carbon electrode by coating it with poly(4-vinylpyridine), which is subsequently impregnated with lrC164-#3-, yields an indicator electrode that permits the detection of 0.07 p.p.m. (mg 1-1) of nitrite in 20 pI of solution in a flow injection experiment. The slope of the calibration graph in the range 1 X 10-5-1 X M nitrite is only half of that obtained for a bare glassy carbon electrode, but the anion-exchange polymer coating attenuates the interference of cations such as Pb", MnIi and Fell.The coating also prevents the surface poisoning effect of thiocyanate, which has previously limited the use of bare glassy carbon and platinum electrodes for this determination. Keywords: Flow injection analysis; modified electrodes; electrochemical detection; nitrite determination The use of amperometric detection at bare electrodes in flow injection determinations has been demonstrated for analytes such as ascorbic acid,l dopamine,l nitroprusside,2 substituted phenothiazines3 and arsenite.4 Nitrite has also been deter- mined in this manner at a bare glassy carbon electrode, but the results are subject to interference by several cations and thiocyanate .5 Thiocyanate gives a particularly strong interfer- ence; the electrode becomes passivated after a single injection of a thiocyanate-containing sample.It has been demonstrated that the modification of electrode surfaces can alleviate surface poisoning in some instances and can also diminish the number of interferences in electrolytic techniques in static solutions. For nitrite determination, treating a Pt electrode by the adsorption of an iodine film eliminates the passivation caused by the accumulation of both nitrite and thiocyanate oxidation side-products.6 Coating platinum and glassy carbon electrodes with anion-exchange polymers also decreases the interferences.7 The mechanism of the oxidation of nitrite at an IrC164- J--impregnated, polymer- coated electrode has also been determined.7 In the work reported here, a modified electrode was used with a flow injection system for the determination of nitrite.The results are compared with those obtained at a bare glassy carbon electrode. Experimental The flow injection apparatus consisted of a carrier solution reservoir, a Cole Parmer MasterFlex peristaltic pump, oper- ated at 1 ml min-1, a Rheodyne 7010 injection valve with a 20-pl sample loop, a Bioanalytical Systems TL-5A electro- chemical flow cell, an IBM Instruments EC/230 potentiostat designed for use with an amperometric liquid chromatography detector and an Esterline Angus Speed Servo I1 strip-chart recorder. A pulse damper constructed in-house was also used. The carrier solution was a pH 4.6 phosphate buffer (0.2 M). The indicator electrode in the flow cell was a 7.5 mm2 glassy carbon disc and the reference electrode, which was mounted downstream from the indicator, was Ag - AgC1.The stainless- steel block of the TL-5A cell served as the counter electrode. All flow injection experiments were performed at a potential The glassy carbon electrode was polished with 0.1 pm mesh alumina as reported elsewhere .7 Electrochemical pre- treatment, which typically improves the performance of this electrode,8 was not found to be necessary when the glassy carbon served only as the modified electrode substrate, but it was used with experiments at a bare glassy carbon electrode. Of 0.85 V V S . Ag - AgCl. The pre-treatment consisted in a 5-min anodisation of the polished electrode at 1.5 V vs. Ag - AgC1, followed by a cathodisation for 1 min at - 1.2 V; this step was performed in the pH 4.6 phosphate buffer.The electrode was modified by coating it with quaternised poly(4-~inylpyridine), qPVP, and impregnating the film with a redox mediator. The PVP (Aldrich) was quaternised, as previously described,7 by the overnight reaction of a solution containing 200 mg of PVP, 50 ml of methanol and 5 ml of benzyl chloride. Three drops of the resulting solution were placed on the glassy carbon disc. After each addition, the solution was evaporated to dryness. The total deposit was about 0.12 mg of qPVP. The redox mediator was introduced by pumping a freshly prepared solution of 2 mM K21rC16 (Aldrich) in the pH 4.6 phosphate buffer into a flow cell that contained the coated glassy carbon electrode. The potential was cycled at 50 mV s-1 between 0.40 and 0.85 V vs.Ag - AgCl for 1 h using a Princeton Applied Research Model 170 Electrochemistry System to apply the potential and to monitor the process. After the impregnation of the anion-exchange polymer film, the cell was rinsed with phosphate buffer. The modified electrode was stored in the buffer when not in use. All other solutions were prepared by dissolving analytical- reagent grade salts in house-distilled water, which was further purified by passing it through a Barnstead NanoPure I1 system. A 1.00 mM stock solution of nitrite was prepared from NaN02 in the pH 4.6 phosphate buffer. The 0.010 M stock solutions for the interference study were prepared from Pb(N03)2, MnS04.H20, FeS04.7H20 and KSCN. Results and Discussion Fig. 1 shows the calibration graphs for nitrite in the 10-5- M range in pH 4.6 buffer, obtained by the flow injection method with bare and modified glassy carbon indicator electrodes.Consistent with a reported study5 that used nitrite in the range 2 x 10-6-2 x 10-5 M, a linear response was observed at the glassy carbon electrode. The non-linear response at the modified electrode was expected because the current-limiting step is not simply transport of the analyte to the surface of the electrode. Instead, diffusion of the nitrite into the polymer occurs and the actual oxidation is by the immobilised IrClb4- ; the over-all process involves the continu- ous electrochemical regeneration of IrIV in the film.7 A lower effective diffusion coefficient, due to the incursion of the nitrite into the polymer and the charge transport needed to regenerate the I P , causes the lower sensitivity obtained with the modified electrode.1220 ANALYST, OCTOBER 1986, VOL.111 1000 800 2 600 F 5 400 . 4-4 200 0 3 6 9 [NO2-l/mol I-' x lo4 Fig. 1. Calibration graphs for the flow injection determination of nitrite based on its oxidation at A, bare and B, modified glassy carbon electrodes with an applied potential of 0.85 V vs. Ag - AgCl The factors that cause the non-linearity and the lower sensitivity of the modified surface suggest that the interfer- ences that are observed at a bare electrode can be attenuated. For example, several cations are reported to interfere with the flow injection determination of nitrite at bare glassy carbon.5 The anion-exchange property of qPVP in pH 4.6 buffer solution will preclude the transport of cations to the site of the electrolysis of nitrite.Moreover, any interference by complex- ation of the nitrite can potentially be decreased owing to the affinity of nitrite for the pyridinium sites, which essentially compete with metals for nitrite ions. The above hypothesis is supported by the results of the determination of 10-4 M NO2- alone and in mixtures with equimolar concentrations of Pb", Mn" or Fell. The respective currents were 123 k 1, 118 _+ 3, 122 _+ 2 and 102 2 1 nA. The relative standard deviations were calculated from six trials for the nitrite alone and in the presence of iron; five trials were used for the other two examples. High concentration levels were selected in order to maximise any complexation and any interference due to accumulation of the electrolysis products, so these data should represent the worst example. Thiocyanate is known to interfere with electrolyses at bare metal electrodes because the product of its oxidation passi- vates the surface.In flow injection determinations at a bare glassy carbon electrode, one trial with a SCN--containing solution passivated the surface. With the present modified electrode, five consecutive injections of a sample containing 1 X 10-4 M SCN- and N02- yielded a current for the oxidation of nitrite (the current for SCN- oxidation is subtracted) of 120 f 1 nA with a range of only 4 nA. The primary limitation of the modified electrode described here is that it must be frequently reconditioned or recali- brated.9 Hence, the bare glassy carbon surface is superior for such applications as flow injection analysis, in which it is inconvenient to restore the surface, providing that the samples under investigation do not contain interfering species. This work was supported by the National Science Foundation (USA) under grant CHE-8215371. References 1. Fogg, A. G., Summan, A. M., and Fernandez-Arciniega, M. A., Analyst, 1985,110, 341. 2. Fogg, A. G., Fernandez-Arciniega, M. A., and Alonso, R. M., Analyst, 1985, 110, 345. 3. Belal, F., and Anderson, J . L., Analyst, 1985, 110, 1493. 4. Lown, J. A., and Johnson, D. C., Anal. Chirn. Acta, 1980,116, 41. 5. Newbery, J. E., and Lopez de Haddad, M. P., Analyst, 1985, 110, 81. 6. Cox, J . A., and Kulesza, P. J . , Anal. Chim. Acta, 1984, 158, 335. 7. Cox, J . A., and Kulesza, P. J., J . Electroanal. Chem., 1984, 175, 105. 8. Engstrom, R. C . , Anal. Chem., 1982, 54; 2310. 9. Cox, J. A., and Kulkarni, K. R., Idalanta, in the press. Paper A6170 Received March 3rd, 1986 Accepted May 19th, 1986 f,'

ISSN:0003-2654    

DOI:10.1039/AN9861101219

出版商:RSC    

年代:1986

数据来源: RSC

摘要:

ANALYST, OCTOBER 1986, VOL. 111 1221 Determination of Ethylene Dibromide in Aquatic Environments Arthur J. Libbey, Jr.* US Coast Guard Research and Development Center, and University of Connecticut, Marine Sciences Institute, Avery Point, Groton, CT 06340, USA Amberlite XAD-4 resin was employed for the extraction of ethylene dibromide (EDB) from water and the ED6 was separated and determined by gas chromatography using an electron-capture detector. Samples spiked with ED6 at the 1 pg 1-1 level exhibited a mean of 0.98 pg 1-1, a standard deviation of 0.105 pg I-1 and a 95% confidence level of the mean of 0.98 k 0.04 pg I-’. The average recovery was 90%. Analyses of samples spiked with EDB at the 50 ng 1-1 level exhibited a mean of 52.9 ng 1-1, a standard deviation of 12.5 ng I-’ and a 95% confidence level of the mean of 52.9 +_ 12.5 ng 1-1, with a recovery of 97%.A real sample gave a mean ED6 concentration of 6.8 pg 1-1, a standard deviation of 0.34 pg 1-1 and a 95% confidence level of the mean of 6.8 k 0.3 pg 1-1. This method represents the first reported use of an Amberlite resin for the determination of ED6 in water. Keywords: Ethylene dibromide determination; water analysis; Amberlite resin Riley and Taylor1 reported the use of Amberlite XAD-1 resin for concentrating dissolved organics in sea water. Junk et a1.2 presented a detailed study of Amberlite XAD-2 and XAD-4 resins for the extraction and separation of organics by class, employing laboratory-spiked water samples. Morrison et aZ.3 published a method using XAD-2 and XAD-7 resins, but did not include other Amberlite resins such as XAD-1, XAD-4, XAD-8 or XE340.The EPA is at present completing a “Master Scheme for the Analysis of Organic Compounds in Water,”4 but makes very limited use of XAD-2, XAD-4 and XE340 resins and employs only one solvent in the separation procedures. Less general methods employing Amberlite XAD-2,5-11 XAD-4,12-14 XAD-78 and XAD-815 have been reported, but in all instances only one or two solvents have been used, almost exclusively diethyl ether and/or methanol. As Amberlite resins are available with a variety of polarities (XAD-1 , XAD-2, XAD-4, XAD-7 and XAD-8) and numer- ous solvents of different polarities are available, the potential of these resins for the extraction and separation of organic compounds from complex aquatic environmental samples has yet to be fully realised.The purpose of this study was to initiate the development of a general method for the determination of trace organics in aquatic environments employing Amberlite resins in combina- tion with gas chromatography with electron-capture detec- tion. A literature method using XAD-2 resin2 was modified, developed and applied to the extraction, separation, identifi- cation and quantification of ethylene dibromide (EDB). EDB was selected for study because of widespread current interest in it as an environmental pollutant and the lack of analytical methods capable of determining it at levels below 50 ng 1-l. l6 Current methods involve liquid - liquid extraction and purge and trap procedures. Although Amberlite resins have been used for some halogenated hydrocarbons,5y7J7 no method involving their application to EDB appears to have been reported.EDB was first produced in the mid-1920s and the present annual production in the USA is 3 X 108 lb.14 EDB is used as a scavenger in leaded fuels, as a soil, grain and fruit fumigant, as an intermediate in the synthesis of dyes and pharmaceuticals and as a solvent for resins, gums and waxes.5 The usage of EDB for controlling tobacco rootworms in Connecticut was extensive in past years. EDB is mobile in air and water and it is expected to be mobile in soil and in the subsurface environ- * Address for correspondence: 64 Milford Road, Manchester, CT 06040, USA. ment.18 Such mobility is indicated by the finding of EDB in over 240 wells in Connecticut, all in areas used for growing tobacco.EDB has a half-life in water of about 14 years at neutral pH.19 Experimental All solvents used were nanograde materials (Mallinckrodt) and the water employed was distilled through a glass still and passed through two mixed bed ion-exchange columns and a column of charcoal. Amberlite XAD-4 resin (20-50 mesh) was purified by three successive 4-h Soxhlet extractions with methanol, then aceto- nitrile and finally diethyl ether. The purified resin was stored under methanol . Ethylene dibromide (EDB) was of analytical-reagent grade (purity >!BOLO). Apparatus A Hewlett Packard 7610A gas chromatograph with an electron-capture detector was used. The chromatograph was fitted with a 5 ft x a in 0.d.glass column packed with 3% OV-1 on 100-120-mesh Gas-Chrom Q. The liquid chromatographic column was a 15 x 0.6 cm i.d. glass tube fitted at one end with a Teflon stopcock and at the other end with a 24/40 female ground-glass joint. A 1-1 round-bottomed flask fitted with a 24/50 ground-glass stopper on one end and a 24/40 male joint on the bottom fitted on top of the glass column and acted as a removable reservoir. A small plug of silanised glass-wool was added to the column and a methanol slurry of Amberlite XAD-4 resin was slowly added to the column containing methanol, in order to prepare a 7-cm resin column. A small plug of silanised glass-wool was placed on top of the resin and the methanol was then run out to the top of the glass-wool plug. The column was flushed with 150 ml of distilled, de-ionised water.All glassware was rinsed three times with hexane, washed with detergent and rinsed with tap water, distilled, de-ionised water, acetone, hexane and finally light petroleum. All glassware openings were covered with aluminium foil after the final rinse until taken for use. Procedure Add 1 1 of a spiked water sample or well water sample to the column reservoir. Adjust the flow-rate of the sample to 8-10 ml min-1 and allow the sample to flow through the column.1222 ANALYST, OCTOBER 1986, VOL. 111 Table 1. Results for the determination of EDB in spiked water samples (1 yg I-’ and 50 ng I-’ of EDB added) and in a well water sample Sample EDB concentration 95 Yo Standard confidence level No. of Mean deviation of the mean Recovery, YO replicates High-level spiked .. . . 0.98 yg 1-1 0.105 pg 1-’ 0.98 k 0.04 pg I-’ 90 10 Well water . . . . . . 6.8 pg 1- 0.34 pg 1- 6.8 k 0.3 pg I-’ - Low-level spiked . . . . 52.9 ng 1-1 12.5 ng 1-l 52.9 k 12.5 ng 1- 97 6 4 Allow the column to run dry. Attach pressure tubing to the column delivery tip and with a vacuum pump draw a vacuum of about 23 inHg for 40 min.14 The recoveries in Table 1 indicate that no loss of EDB occurred on applying a vacuum to the column. Add 50 ml of hexane to the column and collect 25 ml in a calibrated flask, diluting to the mark with hexane eluate from the XAD-4 column. It was shown that all the EDB was removed from the column in the first 25 ml of eluate. Samples for GC were taken from the calibrated flask.The hexane was allowed to run out of the column, then the column was flushed out with 150 ml of methanol followed by 150 ml of distilled, de-ionised water until the water level was just at the top of the glass-wool. The column was allowed to stand in water until it was used again. Samples of 3 pl were injected into the gas chromatograph, the GC conditions being injection port temperature 250 “C, electron-capture detector 300 “C, oven temperature 60 “C, carrier gas [argon - methane (95 + 5 ) ] flow-rate 25 ml min-1, range 10 and attenuation 8, and the sample was run isothermally. Results and Discussion The results are summarised in Table 1. The limit of detection was calculated20 to be 37.5 ng 1-1 and found by replicate analysis of blank water spiked to 52.9 ng 1-1 to be 43.6 ng 1-1.The unspiked blank level was less than 27 ng 1-1. The results indicate that Amberlite XAD-4 resin is satisfac- tory for determining EDB in spiked samples at both the pg 1-1 and ng 1-1 levels and in real samples at the pg 1-1 level. The precision of the determination of EDB in the pg 1-1 range was considerably better than that for the ng 1-1 range. Methanol and dichloromethane were tried as eluting solvents but both eluted at the same time as EDB and thus interfered in the determination. Tetrachloroethane, trichloroethylene and 1,1,l-trichloroethane interfered when using the above GC column; however, a GC column containing a 2 + 1 mixture of 20% OV-22 - 20% OV-17 as used by Daft21 does separate all of these halogenated hydrocarbons including EDB .A column elution rate in excess of 8-10 ml min-1 led to low results. After passage of a spiked water or a well water sample the column must be dried prior to eluting the EDB with hexane. It was found that applying a vacuum to the column at the delivery tip of 23 inHg for 40 min yielded a dry column through which the hexane passed at a satisfactory rate. The method is suitable for use with larger sample volumes, e.g., 10 1, and offers the possibility of determining EDB in the 5-10 ng 1-1 range. Also, owing to the variety of polarities of the available Amberlite resins and the large number of solvents of different polarities, the general method is applic- able to a wide variety of organics at trace concentrations in water samples.22-25 Thanks are extended to Dr.Gerd A. Kleineberg for providing access to Coast Guard Chemistry facilities and Dr. Sung Y. Feng (Marine Sciences Institute) for generously providing the use of laboratory facilities, including the use of the HP-7610A gas chromatograph. Also, Dr. Alan Bentz is thanked for several helpful technical discussions. This work was supported in part by a grant from the University of Connecticut Research Foundation. 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. References Riley, J. P., and Taylor, D., Anal. Chim. Acta, 1969, 46, 307. Junk, G. A., Richard, J. J., Grieser, M. D., Witiak, D., Witiak, J. L., Arguello, M., Vilk, R., Svec, H. J., Fritz, J. S . , and Calder, G. V., J . Chromatogr., 1974, 99, 745. Morrison, J. D., Smith, J.F., and Stephan, S. F., Australian Water Resources Council, Technical Paper No. 28, Depart- ment of National Resources, Australian Government Publish- ing Service, Canberra, 1977, pp. 1-53. Garrison, A. W., Project Officer, “Preliminary Draft Report- Master Scheme for the Analysis of Organic Compounds in Water,’’ Analytical Chemistry Branch, Environmental Research Laboratory, Environmental Protection Agency, Athens, GA, 1985. Junk, G. A., Chriswell, R. C., Chang, R. C., Kissinger, L. D., Richard, J. J., Fritz, J. S . , and Svec, H. J., Fresenius 2. Anal. Chem., 1976,282,331. Coburn, J. A., Valdmanis, I. A., and Chau, A. S. Y., J. Assoc. Off. Anal. Chem., 1977, 60, 224. Grieser, M. D., and Pietrzyk, D. J., Anal. Chem., 1973, 45, 1348. Burnham, A. K., Calder, G.V., Fritz, J. S., Junk, G. A., Svec, H. J., and Wills, R., Anal. Chem., 1972, 44, 139. Suffet, I. H., Brenner, L., and Silver, B., Environ. Sci. Technol., 1976, 10, 1273. James, H. A., Steel, C. P., and Wilson, I., J . Chromatogr., 1981, 208, 89. Ishiwataro, R., Harnana, H., and Machihara, T., Water Res., 1980, 14, 1257. Ryan, J. P., and Fritz, J. S . , J. Chromatogr. Sci., 1978,16,488. Renberg, L., Anal. Chem., 1978, 50, 1836. Musty, P. R., and Nickless, G., J . Chromatogr., 1974,89,185. Stuber, H. A., and Leenheer, J. A., Anal. Chem., 1983, 55, 111. Latnieri, M., State of Connecticut Department of Health Laboratories Division, personal communication. Kissinger, L. D., and Fritz, J. S., J. Am. Water Works Assoc., 1976, 68, 435. EPA Office of Pesticides Program, “Ethylene Dibromide (EDB) Document 4, September 27, 1983,” NTIS Publication PB85-238004, National Technical Information Service, Spring- field, VA. “Ethylene Dibromide (EDB) Fact Sheet ,” American Chemical Society, Washington, DC, February 6th, 1984. Keith, L. H., Anal. Chem., 1983,55, 2210. Daft, J. L., Bull. Environ. Contam. Toxicol., 1983, 30, 492. Strachman, M. M. J., and Huneault, H . , Environ. Sci. Technol., 1984, 18, 127. Schnou, D. W., J . Water Pollut. Control Fed., 1979, 51,2467. Grabow, W. 0. K., Burger, J. S., and Hilner, C. A., Bull. Environ. Contam. Toxicol., 1981, 27, 442. Schaeffer, D. J., Schaeffer, D. I . , Tigwell, D. C., Soman, S. M., and Jonardan, K. G. Bull. Environ. Contam. Toxicol., 1973, 25, 569. Paper A51270 Received July 22nd, 1985 Accepted May 14th, 1986

ISSN:0003-2654    

DOI:10.1039/AN9861101221

出版商:RSC    

年代:1986

数据来源: RSC

摘要:

ANALYST, OCTOBER 1986, VOL. 111 1223 BOOK REVIEWS Mass Spectrometry in the Health and Life Sciences. Proceedings of an International Symposium, San Fran- cisco, California, U.S.A., September 9-13, 1984 Edited by A. L. Burlingame and Neal Castagnoli, Jr. Analytical Chemistry S-ym-posium Series, Volume 24. Pp. xxiv + 638. Elsevier. 1985. Price $133.25, Dfl 360. ISBN 0 444 42562 4 (Series). This book contains 32 papers and the titles of 48 posters presented at an International Symposium held to mark the inauguration of the new University of California Mass Spectrometry Facility. Each of the papers is followed by a discussion and both subject and author indices are included. Most of the contributors to this book are authorities in their particular fields and their papers are up-to-date accounts of the state of biomedical mass spectrometry in 1984.Much emphasis is given to tandem mass spectrometry and to the new ionisation techniques such as fast atom bombardment (FAB) and californium-252 plasma desorption, which have recently enabled mass spectrometry to be applied to large involatile molecules of biomedical interest. Peptide sequencing is a popular application of these new methods and is represented by papers on structure elucidation of tryptic peptides from bovine serum albumin (V. Ling and A. L. Burlingame) and the use of MS - MS techniques for sequencing large (8000 daltons) peptides (W. J. Richter et al.). Protein sequencing techniques are covered by both K. Biemann et al. and P. Poepstorff. A literature survey presented in the latter paper shows that the new methods have now been successfully applied to proteins with molecular weights in excess of 23 000 daltons.Other major classes of compound amenable to mass spectrometric analysis and discussed in this book are oligosac- charides (H. R. Morris et al., M. Suzuki and K.-I. Harada and H. Egge and J. Peter-Katalinic), fossil porphorins (G. Eglinton), organometallic compounds (R. T. Aplin), nucleo- tides (J. A. McCloskey), glycolipids (M. Iwamori et al.), steroids (P. V. Fennessey et al.) and steroid conjugates (C. H. L. Shackleton). Drug studies are represented by the use of condensed-phase ionisation techniques for the analysis of glucuronides (C. Fenselau) and the use of stable isotopes to study phenytoin bioavailability (Y. Kasuya et al.) and the mechanism of toxicity of both acetaminophen and valproic acid (T.A. Baillie). A carnitine conjugate of valproic acid, identified by FAB mass spectrometry (D. J. Liberato et al.) represents a new class of drug conjugate discovered by use of the new techniques. Among papers concentrating on these techniques are those on FAB ionisation (K. Rinehart, Jr.), thermospray LC - MS (M. L. Vestal), Fourier transform ion cyclotron resonance (A. G. Marshall), tandem mass spec- trometry (M. L. Gross et al.) and magnet design (L. C. E. Taylor). The emphasis of this book is very much on the use of mass spectrometry for structure determination. Its main failing is the shortage of information on quantitative aspects and, in particular, on the use of the new mass spectrometric tech- niques for trace analysis.MS - MS techniques and negative electron-capture mass spectrometry are making major contri- butions to these fields and should have been given more prominence. Apart from this, there is little to criticise. Although the presentation is “camera-ready,” the format is clear and relatively free from errors. The high standard of the work presented and the inclusion of so much state-of-the-art material makes this book a good introduction to modern mass spectrometry and its role in the structural analysis of com- pounds of biomedical interest. D. J . Harvey Practical Absorption Spectrometry: UV Spectrometry Group Edited by A. Knowles and C. Burgess. Techniques in Visible and Ultraviolet Spectrometry, Volume 3. Pp. xxii + 234.Chapman and Hall. Price f18. ISBN 0 412 24390 3. The Editors of this volume have assembled a knowledgeable and well known team of authors who have presented a basic and introductory text. Each of the thirteen chapters is self-contained and includes a few carefully chosen references. The book opens with a useful glossary, but unfortunately no indication of correct and incorrect nomenclature is given and it is not always consistent. Knowles has contributed a well written, simple and unam- bitious introduction to the theory and nature of spectra, while Tyson does the same for instrumentation. Tyson also takes a classical approach to the discussion of light sources and optical components, avoiding recent developments such as lasers. The pace of the book increases when Irish deals with monochromators and here he shows his understanding and familiarity with the field.Treherne, in dealing with detectors, gives a very brief introduction to photomultipliers; time will tell whether he has underestimated the impact of Vidicon detectors. The chapter on instrument signal processing by Ford is much more detailed; it is up to date and includes amplifiers, microprocessor-containing circuits and recorders. Tranter contributes a useful elementary introduction to computer interfacing that is clearly written. Russell and Knowles in two chapters deal with cells and measurement of spectra, they share invaluable experience and, while the chapters represent distilled wisdom, they also demonstrate that it is difficult to make these areas sound exciting.Maddams describes numerical methods such as base-line correction, multi-component mixtures and spectral stripping and here more examples might have been used to amplify an interesting account. Special techniques such as derivative and difference spectroscopy, high-precision work and densi- tometry are covered by Fell, Chadburn and Knowles in a chapter by its nature less well defined. The chapter on automated sample handling by Baber is a good introduction to the advantages of automation and the types of system available. The discussion is particularly orientated to the Skeggs-type system with disappointingly little on centrifugal analysers. Brichell’s account of instrument maintenance is most useful and should help quality control and quality assurance in routine laboratories.Useful appendices on solvent characteristics, transmission of window materials and wavelength standards contributed by Knowles and Russell complete the book. As an introductory text for the beginner this book, despite a few errors, is highly recommended. L. Ebdon Quantitative Analysis. Fifth Edition R. A. Day, Jr., and A. L. Underwood. Pp. x + 774. Prentice-Hall. 1986. Price €45.95. ISBN 0 13 746728 1. This very well written and scholarly book was produced for the North American market to cover basic reaction and equilib- rium chemistry for first-year students, introductory instrumental methods of analysis for pre-medical and pre- dental students and to provide the basic laboratory courses.1224 ANALYST, OCTOBER 1986, VOL. 111 As is common with most general analytical texts from the USA, classical chemistry is dealt with in considerable detail, taking up 50% of the space; herein is presented an excellent, clear exposition of the main aspects of the subject.The instrumental analysis sections take up a third of the text and the rest is devoted to practical work. The book does not fit UK courses, which do not, I regret to say, have as much time made available for classical material as is common in the USA or indeed in East Europe. The instrumental section does not give sufficient cover or detail for UK courses which deal seriously with the subject, if included at all. However, third-level teachers of analytical chemistry will find the book of consider- able use as a source of both lecture material and tutorial problems.D. Thorburn Burns ~ _ _ _ _ ~ ~~ ~~~ The Practice of Quantitative Gel Electrophoresis Andreas Chrambach. Pp. xvi + 265. VCH Verlagsgesell- schaft. 1985. Price DM110; $43.60. ISBN 3 527 26039 0. This book, in the series Advanced Methods in the Biological Sciences, covers the practical aspects of gel electrophoresis. The term “quantitative” gel electrophoresis is used by the author to contrast the methods described in the book, which are defined as the most efficient based on theoretical considerations, with the conventional use of gel electro- phoresis. Polyacrylamide gel electrophoresis, agarose electro- phoresis, electrofocusing and isoelectrophoresis are all described with a wealth of practical detail, from optimising buffer systems to buying and building electrophoretic equip- ment.Bravely, sources of equipment and reagents, including some catalogue numbers, are given, which, although useful now, one fears will date the book rapidly. The Appendices give further details that would otherwise disrupt the flow of the book. The text is easy to read, although with a few idiosyncrasies of style, and is clearly written by someone of great practical experience in the field of electrophoresis. References are detailed at the end and are not that numerous. The book is a specialised one but will be of use to the scientist new to electrophoresis and as a source of reference for those with more experience in this area. It is a useful book for those who are involved in the practice of gel electrophoresis. P.M. S . Clark UV-VIS-Spektroskopie und ihre Anwendungen Heinz-Helmut Perkampus. Pp. viii + 208. Springer-Verlag. 1986. Price DM148. ISBN 3 540 15467 1 ; 0 387 15467 1. This little book provides information on a range of topics likely to be of use to an analyst needing a fuller understanding of the nature and applications of ultraviolet and visible spectrophotometry. The emphasis is on the theory of the subject rather than on practical hints and the selection of topics includes most of those to be expected. Diode array detectors and the impact of microprocessors and computers are scarcely mentioned and only a handful of the 600 references are to publications later than 1982. After a brief introductory chapter and one on the principles of spectrophotometry there is a chapter on instruments, much of which is a detailed account of stray light errors.Chapter 4 deals with analytical applications and its balance seems odd. There are tables extending over several pages that list chromogenic reagents for inorganic ions and that give the A, and E values together with literature references. These tables relate both to direct spectrophotometry in solution and to extraction and flotation spectrophotometry. In contrast, the section on the photometric determination of organic com- pounds concentrates on the 2,4-dinitrophenylhydrazones of aldehydes and ketones and refers the reader elsewhere for information about other groups of compounds. A fuller treatment of organic compounds in line with that for inor- ganics would have been of great value. There is a useful section on enzymatic analysis and enzyme kinetics and then an uncritical description of multi-wavelength multi-component analysis.The chapter ends with a summary of one particular approach to the correlation of structures of organic com- pounds with values of h,,,, and E,,.,,,,. Chapter 5 is on “special methods,’’ namely two-wavelength spectrophotometry, derivative spectroscopy, reflection spec- troscopy, photoacoustic spectroscopy and luminescence exci- tation spectroscopy. These are generally clear and concise introductions for those who wish to grasp the essentials before moving on to monographs and review articles. The next two chapters cover equilibria and kinetics, respectively. The mathematical depth is uneven. For example, the kinetics of various categories of second-order chemical reactions are set out in some detail, whereas the reader is expected to be able to understand Job’s method of continuous variations without the help of a graph and to cope with the determination of the rank order of a matrix. Little appreci- ation is shown of the difficulties of attempting to solve simultaneous linear equations by inverting ill-conditioned matrices. I cannot understand why the author has devoted so much space to equilibria. Certainly spectrophotometry is an important technique for determining pK, values and the stability constants of complexes, but the spectrophotometric problems are generally the same as for the analysis of other multi-component systems and too much of the chapter relates to equilibrium theory. Chapter 7 is closer in content to the rest of the book, with sections on the stopped flow technique, relaxation spectro- scopy and photoreactions, for example. The final chapter brings together a few pages each on oscillator strengths, Gaussian and Lorentzian band shapes and the vibrational fine structure of certain electronic spectra. A . R. Rogers

ISSN:0003-2654    

DOI:10.1039/AN9861101223

出版商:RSC    

年代:1986

数据来源: RSC