ArtaZyst, April, 1973, Vol. 98, $9. 251-256 251 An Electrical Conductivity Detector for Paper, Thin-layer and Column Chromatography* BY V. DI STEFAN0 AND P. MARINI (Ccntro Spcrimentale Metallurgico, V i a di Caste1 Romano, Rome, Italy) An electrical conductivity detector has been developed for application to the paper, thin-layer and column chromatography of ionic substances. Two arms of an a.c. Wheatstone bridge comprise two pairs of electrodes applied directly on to the separating layer so that a signal ensues from the bridge when a chromatographic spot passes one electrode pair while the other pair is traversed by the eluting agent. The sensitivity is about 1 pg and the precision is about 2 to 3 per cent. THE problem involved in the quantitative determination of the components of a mixture after a chromatographic separation has been carried out is still difficult and complex.While the identification of the individual components can be simplified to measurement of R, values, the quantitative analysis is carried out by using techniques and methods that are often exclusively applicable to only one or other of the components of a mixture and great care must be exercised to avoid interference effects from the chromatographic support. The methods capable of more general use can be classified as direct and indirect methods. The former permit quantitative analysis to be carried out directly on the chromatographic medium without other manipulation, while indirect methods require a series of manipulations after the development of the chromatogram. Examples of the application of such methods can be found in the literat~re.l-~ With indirect methods, quantitative analysis needs to be performed by highly qualified personnel and by use of rather complex instrumentation, while chromatographic separations are carried out simply and often with very simple apparatus.However, there are relatively few direct methods of quantitative analysis of general applicability, especially in paper and thin-layer chromatography. The present work is intended as a contribution towards pro- viding a direct detection method that is useful in many instances. A description with illustrations is given of an electrical conductivity detector applied to paper, thin-layer and column chromatography, which is capable of detecting small amounts of any substance that is a charge carrier and that can be eluted by substances that exhibit a certain electrical conductivity .CONFIGURATION OF THE SYSTEM- The simplest form of such a detector as originally assembled consists of two parallel, narrow, thin gold plates between which is enclosed a strip of paper, thus forming a con- ductivity cell, which is used as an arm of a Wheatstone bridge; the bridge is connected to a differential amplifier whose output is fed into a chart recorder. One end of the paper strip is immersed in the eluting agent and a flow of eluting agent is established along the strip. When this flow becomes constant, the signal from the Wheatstone bridge is also constant and the bridge can be set to zero. A solution of a conducting substance in the eluting agent used is prepared and 2 to lop1 of this solution are applied with a precision microsyringe close to the immersed end of the paper strip, which results in a spot being formed which travels in the flow of eluting agent towards the conductivity cell.When the spot reaches the electrodes a conductivity peak is recorded. More elaborate arrangements were subsequently designed and the electrodes were applied to thin layers of silica, alumina and cellulose, as well as to a special type of column. In Fig. 1 is shown the best arrangement attained at present. Four electrodes are now used, forming a working cell and a reference cell. They consist of thin, narrow platinum strips, 1 mm wide x 1 mm thick and. 35 mm long, placed 1-5 mm apart, embedded in a PTFE support on which the thin layer is deposited in the usual way.The eluting agent is fed in at the upper end and the sample is applied at a suitable distance from the feeding point * Presented at the Third SAC Conference, Durham, July 12th to 16th, 1971. Q SAC and the authors.252 DI STEFAN0 AND MARINI : AN ELECTRICAL CONDUCTIVITY DETECTOR [AndySt, VOl. 98 of the eluting agent, either with a microsyringe or by using a narrow strip of polished gold on which the substances to be determined have been deposited from solution by evaporation of the solvent. The conveniently supported gold strip is brought into contact with the layer and contact maintained until a further sample is required to be applied. One end of the thin layer, which is kept in a horizontal position, is exposed to the air so as to enable the eluting agent to evaporate; the other parts of the chromatographic system are suitably sealed in order to prevent its evaporation; efforts are at present being made to embed the electrodes in a glass plate, glass being a more satisfactory material in this respect than PTFE.The above electrode arrangement yielded the most satisfactory results, but other arrange- ments were also tried, as shown in Fig. 2. In Fig. 2 (a), an arrangement is shown in which one electrode is embedded in the support and the other gently applied over the thin layer. We found that this system produced cells the apparent volume of which extended too far within the layer, thus giving rise to enlarged peaks. In Fig. 2 ( b ) , a comb-like type of electrode is shown, which gave rise to a slight perturbation of the flow and to tailed peaks.r 1 5 4 2 Fig. 1. Detector system: 1, support; 1 a~itl 3, electrodes; 4, leads; and 5, thin layer Fig. 2. Different arrangements of elec- trodes In Fig. 3 is shown the design of a special column, which was developed in order to allow the conductivity detector to be applied to column systems. A large conventional column could easily accommodate at least one pair of electrodes, but is not suitable for the separation and detection of micro-amounts of substances. A capillary column does not afford enough room for the electrodes, so that the annular space enclosed between two concentric columns, which may be regarded as a thin-layer column, was used. This space is currently 1 mm wide but could be further reduced, possibly to 0.3 to 0.4 mm, thus obtaining a cylindrical layer suitable for micro-analytical purposes.Filling such a column does not present particular difficulties and can be achieved by carefully pouring the substance chosen as the chromatographic support, in the form of a slurry, from the top of the column while the latter is maintained in a vertical position and in contact with a plate vibrating at about 50 Hz and with moderate amplitude. The two conductivity cells, applied to the separating layer, constitute two adjacent arms of a Wheatstone bridge to which an oscillator feeds a 2-V, 2000-Hz signal. The output of the bridge is fed into a differential amplifier, which could be replaced by a lock-in amplifier for greater accuracy.This amplified signal is fed both into an integrator and into a chart recorder. If the cells are arranged in such a way that one electrode of one of the cells is short-circuited with one of the electrodes of the other cell, while the signal from the oscillator is applied to the other two electrodes, the zero line is almost independent of variations in the flow of the eluting agent and changes in the environmental conditions. Electrodes can be set side by side in pairs located symmetrically with respect to the cylinder axis. Samples are applied with a microsyringe through neoprene diaphragms located upstream at convenient distances from the electrodes.April, 19731 FOR PAPER, THIN-LAYER AND COLUMN CHROMATOGRAPHY 253 I Fig. 3. Chromatographic column THEORY OF THE RESPONSE- across its terminals when a chromatographic spot is flowing through the working cell is-- It can easily be demonstrated that the response of the bridge in terms of the potentid kX (C--CZ+ .. . ) Valkh (0 + aJ2 + 0 1 P = where p is the instantaneous root mean square of the potential across the bridge; I/ the root mean square of the supply potential; a1 the conductivity of the arm adjacent to the working cell; a the conductivity of the working cell before and after the spot has passed, whose value is usually very close to a,; k the constant of the working cell, a function of the geometry of the electrodes and of the electrical path between them; h the equivalent con- ductivity of the spot material x 103; and C the instantaneous equivalent concentration of the spot material.As a first approximation the response is proportional to concentration, which has been confirmed experimentally as will be shown later. The higher the supply potential the higher the response; it was found that, as with most conductimeters, 2 V are sufficient to obtain a good and clean response. The response is higher if the conductivity of the eluting agent is low, but the conductivity of the latter cannot be too low or the stability of the zero line and linearity of response with varying concentration will be affected. In a series of experiments with a diethyl ether based eluting agent, which gave rise to a resistance across the electrodes of 250 kQ, it was found that trace amounts of moisture in the sampling device are capable of rendering the entire system unstable for a long period of time.Thus, if the conductivity of the eluting agent is reduced to very low values, the electronics of the device must be suitably designed and great care should be used in handling the apparatus. If we set (0 + aJ2 * = Vqkh it can be seen that t,b is the constant of the detector and as a first approximation c = #P The complete response with time, that is, the chromatogram, is a peak, and the amount of the substance under examination is proportional to the area of the peak.264 DI STEFAN0 AND MARINI AN ELECTRICAL CONDUCTIVITY DETECTOR [Anat!ySt, VOl. 98 RESULTS AND DISCUSSION The logical first application of such a detector is to separation systems for inorganic ions and the work carried out on some of these systems is reported below.In the course of experiments performed with the various types of electrodes described in this paper it was found that better results were obtained with cell configurations of the type shown in Fig. 1 and interest was accordingly concentrated on these cells. In Fig. 4 (a) a separation of Fe3+ from Zn2f ions in a mixture of these ions on a thin layer of silica gel G is reported; the eluting agent used was acetone - 12 N hydrochloric acid - 30 20 10 0 - I I I I I I I I I I 0 10 20 30 40 50 60 70 80 30 - (4 (4 Fe3+ - - I I IApril, 19731 FOR PAPER, THIN-LAYER AND COLUMN CHROMATOGRAPHY 265 hexane-2,5-dione (100 + 1 + 0.5). The sample analysed consisted of 10 pl of a solution containing 0.2 pg p1-1 of both iron and zinc ions.On the y-axis of the graph the output of the amplifier is given in millivolts. The true resistance of each cell was about 400 SZ. In Fig. 4 (b) the chromatogram for the mixture of Fe3+ and Mn2+ ions, which was obtained under the same conditions as for the previous chromatogram, shows that the Fe3+ peak appears after almost exactly the same time as the Fe3+ peak in Fig. 4 (a). The same is true of Fig. 4 (c), which relates to the separation of a mixture of Fe3+ and Co2+ ions carried out under the same conditions as before. With the mixture of Mn2+, Co2+ and Zn2+ ions, separa- tions were not attempted as efforts were devoted more to testing the properties of the detector than to studying separation techniques. The sensitivity of the entire system depends largely on the characteristics of the elec- tronics used and the level obtained with the prototype is shown in Fig.4 (c), for which the sample taken amounted to only 0.75 pg of Co2+. The chromatogram shown in Fig. 4 (a) is the result of an attempted separation of a mixture of Fe3+, Cr2f and Co2+ ions, by using as eluting agent acetone - pentyl acetate - pentane-2,4-dione - 6 N hydrochloric acid (150 + 10 + 1.5 + 1.5). For every substance it is necessary to plot a calibration graph and consequently care should be taken to maintain constant experimental conditions. No difficulties arose and thermostatic control of the system was found to be unnecessary. The only precautions necessary were the maintenance of a sufficient supply of eluting agent to the layer and prevention of evaporation of the standard solutions prepared with the same eluting agent.The calibration graph for iron is plotted in Fig. 5. On the abscissa is the area originally obtained in millivolts per minute, but it has been plotted in square centimetres, with an approximation of 10 per cent., in order to demonstrate the extension of the peak. Standard solutions with different concentrations were used in order that samples should be at constant volume, but it was noticed that the volume of the eluting agent had no efkect and all of the straight lines obtained could be extrapolated to zero, which was further confirmed with a sample of the pure eluting agent. Nevertheless, we found that a higher dispersion was obtained with samples at constant concentration and with variable volume, which was caused by the graduation and calibration of the microsyringe, and we therefore finally resorted to the use of samples of constant volume.Peak area/cm2 Fig. 5 . Calibration graphs for different ions: A, Co*+; 13, I:@+; and C, Zn2+ The calibration graphs for Zn2+ and for Co2+ ions are also shown in Fig. 5. From day to day, or after any interruption of the flow of eluting agent through the chromatograph, we256 DI STEFAN0 AND MARINI noticed some variations in the angular coefficients of the calibration graphs, so that for analytical purposes it is advisable to introduce a standard sample at the beginning of each analytical cycle. In order to observe the effect of time upon the angular coefficient of the calibration line, we examined, once every hour for 7 hours, a standard volume of 4 pl of a solution containing about 1 pg p1-1 of iron.The result of this experiment is shown in Table I. The mean value was 4.11 pg, the standard deviation 0.25 pg and the coefficient of variation 6.2 per cent. TABLE I EFFECT OF TIME ON ANGULAR COEFFICIENT OF CALIBRATION LINE FOR IRON Tirne/hours . , .. 0 1 2 3 4 5 6 7 Amount of ironlpg . . 4.2 4-0 4.2 3.9 4.1 4.3 4.0 4.2 CONCLUSION A conductivity detector consisting of one or two electrode pairs applied to a chromato- graphic layer appears to be a simple device that is capable of minimising difficulties and manipulations associated with quantitative liquid chromatography. For paper and thin-layer chromatography, the best results were obtained by applying, in contact with the layer, narrow strip electrodes embedded in the supporting material. With column chromatography, good results were obtained by filling the annular space between concentric columns with the solid phase, while the electrodes were supported by the glass of the column. Samples from a solution in the eluting agent of the substances to be analysed are applied directly to the layer with a precision microsyringe. In order to obtain accurate results, analyses should be made and calibration graphs obtained by using sample solutions with a constant volume of eluting agent. 1 . 2 . 3. 4 . 5 . 6. 7 . 8 . 9 . REFERENCES Petrowitz, H. J., Mitt. dt. Ges. Holzforsch., 1962, 48, 57. Fisher, R. B., Parson, D. S., and Morrison, G. A., Nature, Lond., 1948, 161, 764. Seher, A., Nahrung, 1960, 4, 466. -, Mikrochim. Acta, 1961, 308. Zollner, N., Wolfram, G., and Amin, A., KZin. Wschr., 1962, 40, 273. Neubauer, D., and Mothes, K., Planta Med., 1961, 9, 466. Hefendehz, F. W., Ibid., 1960, 8, 65. Ganshirt, H., and Morianz, K., Arch. Farmaz., 1906, 293, 1066. Schaulze, P. E., and Wenzel, M., Angew. Chem., 1962, 74, 777. Received May 31st, 1972 Accepted November 7th, 1972