tric crystal for the determination of paraoxon. The results of this work are shown in Figure 6. The sensitivity is not so great as the DIMP system, but nevertheless a lower concentration limit of 10 ppm can be attained. Interferences. Before considering using this detector for the detection of pesticides in the atmosphere, other common atmospheric constituents and pollutants must be tested as possible interferences. The results of this study are summarized in Table 11. The only gas that presents even a slight interference is SOz, and this could be removed if need be, or tested separately and the frequency change due to it subtracted from the total. Because of the specificity of the substrates used, interferences should present no problem unless a reaction with the substrate should occur. In this case, the interference would have to be removed prior to sampling.
CONCLUSIONS
By complexing the organophosphorus pesticide to be determined with an inorganic salt, and then using this complex as a substrate on a piezoelectric crystal detector, good specificity and sensitivity can be attained. Direct monitoring of concentrations below 10 ppm can be accomplished in less than 15 minutes. Many different organophosphorus pesticides can be determined by building detectors with various substrates. RECEIVEDfor review January 12, 1972. Accepted April 18,1972. The financial support of the Army Research Office (Grant No. DA-ARO-D-31-124-70-G69) and the National Institutes of Health (Grant No. 8-R01-OH-00345-02) is gratefully acknowledged.
Apparatus for Automated Gel Permeation Cleanup for Pesticide Residue Analysis Applications to Fish Lipids Roger C. Tindle and David L. Stalling Fish-Pesticide Research Laboratory, Columbia,Mo. The gel permeation cleanup procedure for fish and other tissue extracts in pesticide residue analysis, previously reported by Stalling, Tindle, and Johnson, has been automated. The automated system allows unattended operation while processing up to 23 samples with the system as described. Reproducibility of recoveries were quite good (coefficient of variation 5%) and cross-contamination was estimated at less than 1% The chromatography system was constructed from commercially available components so that other investigators may easily duplicate the device without the necessity for fabrication of special components. SEPARATION OF LIPIDS from pesticides is often the most time consuming process in the analysis of sample extracts for pesticide residues. Most previously reported techniques for achieving these separations can be grouped into four categories: adsorption chromatography (Florid), liquid/liquid partition (acetonitrile/hexane), forced volatilization (sweep co-distillation), and low-temperature precipitation of the lipids. None of these techniques is readily amenable to automation. Stalling, Tindle, and Johnson ( I ) reported a cleanup technique which may be readily automated. The advantages of an automated pesticide cleanup procedure include improved analytical precision and accuracy, decreased manipulative sample losses, and a significant saving of labor. Automated chromatography is widely accepted in amino acid analysis by ion exchange chromatography. In addition to commercial equipment to perform these analyses, noncommercial equipment for various types of automated aqueous chromatography systems has been reported. (1) D. L. Stalling, R. Tindle, and J. Johnson, J. Ass. Ofic.Anal. Cliem., 55, 32-8 (1972); presented at the 161st ACS National Meeting, Los Angeles, 1971. 1768
Roubal and Tappel (2) reported an automated gel permeation system for the separation and molecular weight determination of proteins. This system, however, was automated only in the sense that a commercial pump system and associated apparatus was used to quantitatively determine the protein fractions eluting from the gel permeation column. Sample loading was performed manually, and eluted samples were not collected. Hicks and Nalevac (3) made a device for the sequential addition of aqueous buffer solutions to a gel permeation column. This system used a “column follower” to sense when the solution level reached the top of the gel bed. When this occurred, the conductance between the electrodes of the column follower decreased, causing a stepper relay to advance and open a solenoid valve to admit the next solution. Again, there were no provisions for automatic sample introduction or collection. Dus, Lindroth, Pabst, and Smith ( 4 ) automated amino acid analysis by use of rotary valves to select buffers and to direct the selected buffer to one of six ion exchange columns. Samples were loaded in the tubes which connected the six output ports of the second rotary valve to the corresponding columns. A cam-type programmer was used to operate this analytical system but there were no provisions for sample collection. Many other authors have reported automation techniques for use with Chromatography columns. These include Tschida and Markowitz ( 5 ) who added automatic regeneration ~~~
~~
(2) W. T. Roubal and A. L. Tappel, Anal. Biochem., 9, 211-16 (1964). (3) G. P. Hicks and G . N. Nalevac, ibid., 12,603-12 (1965). (4) K. Dus S. Lindroth, R. Pabst, and R. M. Smith; ibid., 14,
41-52 (1966). ( 5 ) A. R. Tschida and H. Markowitz, ibid., 26,337-40 (1968).
ANALYTICAL CHEMISTRY, VOL. 44, NO. 11, SEPTEMBER 1972
CONTROLLER
W A F E R V A L V E ASSY.
Figure 1. Automated gel permeation system Components and operation of the system are described in the text. See Table I for a summary of the operation sequence
and shutdown features to a conventional Technicon chromatography system. Dymond (6)used a separate pump to withdraw samples from special sample cups and inject them into an ion exchange system. Benjamin, Pourfar, and Schneck (7) used photocells and a stepper relay system to control the sequential addition of different solvents for the separation of complex lipids on an unspecified type of column material. Eluates were fractionated by use of individual pneumatically controlled valves for each fraction desired. Manual sample introduction was used. Greater or lesser degrees of automation of commercial equipment have been reported by several authors. These reports include the work of Ertingshausen, Adler, and Reichler (8), Heumer and Lee (9), and Bennett and Creaser (IO) with amino acid analysis, and the work of Hori (11) with steroids. We have developed an automated system for the cleanup of sample extracts by gel permeation chromatography according to the technique of Stalling, Tindle, and Johnson ( I ) . This device consists primarily of commercially available components with appropriate interconnections. A unique aspect of this system is the use of a rotary valve to direct the column effluents to various receivers. Construction of this system, except for the digital controller, is easily accomplished with a minimum of shop facilities. The following description presents the necessary details for the contruction of this device with the exception of the digital controller. Construction details for the controller are presented by Tindle (12). (6) B. Dymond, ANAL,CHEM., 40,919-23 (1968). (7) A. Benjamin, M. Pourfar, and L. Schneck; J . Lipid Res., 10, 616-18 (1969). (8) G . Ertingshausen, H. J. Adler, and A. S . Reichler, J. Chromatogr., 42,355-66 (1969). (9) R . P. Huemer and K. Lee; Anal. Biochem., 37,149-53 (1970). (10) D. J. Bennett and E. H. Creaser, ibid.,pp 191-4. (11) M. Hori, Steroids, 14,33-46 (1969). (12) R. Tindle, ANAL.CHEM., in press.
MATERIALS AND EQUIPMENT
The various components in the system (Figure 1) are interconnected by means of TFE tubing and appropriate Swagelok (Crawford Brass Co., Solon, Ohio) or Chromatronix (Chromatronix, Inc., Berkeley, Calif.) fittings. Sample loops were fabricated from 17-gauge TFE spaghetti tubing (Polymer Corp., Polypenco Div., Reading, Pa.). The sample loops were calibrated by coiling the tubing to appropriate dimensions and injecting 5.0 ml of a dye solution into the tubing. The ends were then cut off inch beyond the dye mixture. This extra length allowed for the volume loss due to attachment of the sample loops to the wafer valve tubulations. The accuracy of the loop-valve arrangement was within 1 of the nominal 5.0-ml volume as disclosed by dye marker calibration after assembly. The “load” line of the Sample Introduction Valve (SIV) (Chromatronix Model SV-8031) was fitted with a Luer adaptor to allow attachment of a hypodermic syringe. The drain line passes to a suiiable waste receiver. The column was a 2.5-cm i d . x 40-cm glass column (No. 3400-D-25x60) with two adjustable end plungers (No. 3500A-25) (Glenco Scientific, Inc., Houston, Texas). The dump collect valve was a 3-way Cheminert Valve (No. CAV-3060) with 0.060-in. passages, and two pneumatic actuators (No. PA-875). Solenoid air valves (No. SOL-3-1100 DC) were used to control the pneumatic actuators (all components were Chromatronix, Inc.). The valve control box is described schematically in Figure 2. Gel permeation materials, solvents, etc., are as described by Stalling, Tindle, and Johnson ( I ) . The digital controller shown in Figure 1 served as a threesequence, multi-cycle timer to control the various switching operations of the valve control box. The controller used in this study was a rather complex solid state digital controller which is to be the subject of a later paper. However, any
ANALYTICAL CHEMISTRY, VOL. 44, NO. 11, SEPTEMBER 1972
1769
7
. . . . . .
56
?
Figure 2. Valve control box schematic Parts list Lamps LllLza
Capacitors 10 pF electrolytic, 50 WVDC C1 10 uF electrolstic. 150 WVDC C? ~, Fuse Fl type 3AG, 3A Rectifiers
B1
full wave bridge rectifier assy. 50 PI V, 4 A. D1-Da rectifier diodes, 1N5059 Connector
Jl
8 pin receptacle (Amphenol 126-221)
(Motorola MDA 1591-1)
type 24ESB, 24 V
Relays RL1-RLz SPST NO, 3-A contacts, 24 V, dc SPST NO, 10-A contacts, 24 V, dc RL3 Switches SI SPST, toggle, 5A contacts S2-S5 SPST, NO, momentary contact, 5A SS 24-position rotary, 360" rotation Air Valves SLl-SL* Solenoid air valves (described in text)
timer which allows multi-cycle operation and at least three independently timed sequences may be adapted to the control of this system. Table I lists the operational sequence. Description of the System. Figure 1 illustrates the major components of the automated chromatography system. The system consists of a solvent source and pump, a manual loading valve, a remotely controllable sample introduction system, a chromatography column, and a remotely controllable sample collection system. The valve control box allows either manual or automatic control of the sample introduction/ sample collection valve system. The outboard digital controller allows automatic unattended operation of the system. The pump used in this system is adjustable to give flow rates from 0.5 ml/min to 10-3m, the emf became more negutioe with increasing [H+] at a given constant [Ca2+]. The magnitude of this change did not appear to be reproducible and might depend upon the particular electrode construction (6). At lower pH, the emf increased in an approximately Nernstian manner with [H+], as might be expected from Equation 2 when S H>> al. ~ ~ Widely ~ varying selectivity parameters of S H = 2 x lo5(6) and lo7(5)were reported. The only other liquid-membrane electrode which has been studied in solutions of varying pH appears to be the leadselective electrode (7); however, no negative change was detected in solutions of [Pb2+], 10-2-10-4m,pH 2.5-8.0. It is of interest to note that a solid-state electrode selective toward calcium has been reported to show a negative change with pH in a solution [Caz+l = 1.4 x lo-%, pH 3-5, but not at lower [Ca2+] (8). The exact composition of this electrode has not been revealed but may be an organophosphorus compound immobilized in a thin matrix. At present the only explanation offered for the negative change is by Orme, who suggested that some impurity in the liquid-ion exchanger might be responsible (6). In view of the expanding applications of the calcium-selective electrode for measurements in multi-ionic solutions, more information about the influence of pH is clearly desirable. The aim of ( 5 ) J. W. Ross, Jr., ibid., p 57. (6) F. W. Orme, “Glass Micro-Electrodes,” M. Lavallee, Ed., John Wiley & Sons, New York, N.Y., 1969, p 376. (7) S. La1 and G. D. Christian, Anal. Chin?.Acra., 52,41(1970). (8) G. A. Rechnitz and T. M. Hseu, ANAL.CHEM., 41,111 (1969).
ANALYTICAL CHEMISTRY, VOL. 44, NO. 11, SEPTEMBER 1972
1773
