APPLICATION OF COMPLEX DIELECTROMETRY TO COLUMN CHROMATOGRAPHY

V. P. Lopatinskii, V. T. Novikov, P. I. Gos’kov, V. A. Salikhov, P. P. Shadrin, Yu. K. Shelkovnikov, V. K. Katayukhin, V. L. Ivasenko, N. V. Zamyatin

(Presented by the scientific seminar of the chemical departments and laboratories)

Dielectric methods, distinguished by their simplicity, accuracy, and reliability, are widely used in qualitative and quantitative analysis. However, in determining the content in water of a number of high-molecular-weight substances such as flocculants and surfactants, the indicated methods have not hitherto been applied; only methods of chemical analysis have been used, which are either complex and very labor-intensive [1], or, with a comparatively simple procedure, are highly inaccurate and have poor reproducibility [2].

In the present work, the dielectric properties (permittivity and loss tangent) of two organic substances, very promising as flocculants and surfactants for intensifying the separation processes of liquid heterogeneous systems—suspensions and emulsions—were investigated. The studies were carried out with the aim of examining the possibility of recording changes in the indicated physical properties of aqueous solutions of the substances used, so that the results obtained could subsequently be used in the quantitative analysis of these substances.

The studies were carried out by comparing the dielectric properties of solutions containing a variable amount of organic substances with the analogous properties of the pure solvent, in

as which distilled water was used.

The properties of solutions of the following organic substances were studied:

1. A high-molecular flocculant—polyacrylamide (PAA).
2. Dissolvan—demulsifier FRT, a high-molecular (m.w. 2500–3000) polyalkylene glycol. Dissolvan belongs to compounds of the type of block copolymers of propylene and ethylene oxides [4].

The measurement of $\tg\delta$ and $\varepsilon$ of the solutions was carried out at a frequency of 100 kHz by the resonance method using contactless electrodes. The absence of contact between the electrodes of the measuring capacitance and the aqueous solution under study increased the accuracy and reliability of the results, since thereby the influence of various polarization phenomena at the metal electrode–liquid interface was reduced and shunting of the measuring circuit was sharply decreased. Changes in $\tg\delta$ and $\varepsilon$ were judged from the corresponding changes in the capacitance and quality factor of the measuring capacitor, which were recorded with a Q-meter. After thorough washing, the condenser was filled in turn with the reference liquid (solvent) and the liquid under study.

The results of measurements in the region of low concentrations are shown in Fig. 1. Here the coefficients $m_\varepsilon$ and $m_\delta$ are understood as follows:

\[ m_\delta=\frac{\Delta(\tg\delta)_i}{\Delta(\tg\delta)_{\min}} \qquad m_\varepsilon=\frac{\Delta\varepsilon_i}{\Delta\varepsilon_{\min}} \]

where the index $\min$ denotes the minimally distinguishable changes, in the course of the given experiment, of $\tg\delta$ and $\varepsilon$ of the solvent upon addition of one of the substances studied, while the index $i$ denotes the current changes of $\tg\delta$ and $\varepsilon$ of the solution.

Comparison of the dependences $m_\delta(N)$ and $m_\varepsilon(N)$ shows that, for aqueous solutions of dissolvan in the region of low concentrations ($N<100$ mg/l), the changes in $\varepsilon$ and $\tg\delta$ are of the same order; therefore it is immaterial by which of these characteristics the quantitative determination of dissolvan in water is carried out.

In the case of polyacrylamide, from the curves presented it follows that

that the addition of these substances to water changes $\varepsilon$ only very slightly and changes the $\operatorname{tg}\delta$ of water very strongly. Therefore, quantitative determination of PAA can be carried out reliably and with high accuracy by monitoring the change in $\operatorname{tg}\delta$ of solutions of this substance.

In Fig. 1, at $m_{\varepsilon\delta}=1$, one can find the minimum concentrations of the listed substances that could be determined using the method and apparatus employed. These data are summarized in Table 1.

Moreover, the numerator gives the minimum concentrations of the substance under study, in mg/l, determined from the corresponding changes in the properties of the solutions in this experiment; the denominator gives those limiting minimum concentrations of substances, in mg/l, that can be distinguished if a comparatively simple and quite accessible modernization of the measuring apparatus is carried out, i.e., if its sensitivity is increased appropriately. In our case,

\[ \Delta(\operatorname{tg}\delta)_{\min}=0.015,\qquad \Delta\varepsilon_{\min}=1 \]

With careful performance of the measurements, reproducibility was practically 100% when determining concentrations both from $\varepsilon$ and from $\operatorname{tg}\delta$. However, a comparison of the methods for measuring both characteristics, both in terms of actual resolving power and in terms of simplicity and convenience of measurements, as well as the possibility of automating the measurement process, especially in flow, shows the clear advantage of the monitoring method based on the use of the dependences $\operatorname{tg}\delta=\varphi(n)$. At the same time, as follows from Table 1, such a method will provide greater accuracy and much greater resolving power than the known chemical methods, with a very simple measurement procedure and uncomplicated measuring apparatus.

Table 1

The results obtained for the example of Dissolvan and PAA show that quantitative or qualitative analysis of organic

... substances by the dielectric method is advisable, and in some cases absolutely necessary, using a complex, i.e., combined, measurement of $\varepsilon$ and $\tg\delta$ of the solution under investigation. This substantially simplifies the measurement procedure and increases the accuracy and reliability of recording the phenomena being determined, especially in those cases where it is not known in advance which of the dielectric characteristics, and to what extent, will change during the investigation. What has been said may apply fully to chromatographic analysis, since, in this case, by using the complex determination of $\varepsilon$ and $\tg\delta$, it will be possible to obtain experimentally, with very high accuracy, chromatograms of the analyzed mixtures of substances. The dielectrometric apparatus required for this is comparatively simple; however, it should be noted at once that similar apparatus for the complex determination of $\varepsilon$ and $\tg\delta$ of substances is not known in the literature. Therefore, the authors [5] developed a differential instrument of this type, intended specifically for the complex determination of $\tg\delta$ and $\varepsilon$ in the high-frequency range, and it can be used in chromatographic analysis of organic substances. In addition, the authors carried out a detailed theoretical and experimental study of the complex combined determination of $\tg\delta$ and $\varepsilon$ in the microwave range, as well as of broadband measuring resonators for this purpose based on pin-type slow-wave systems. The results obtained in this work indicate the great promise of complex microwave dielectrometry in column chromatographic analysis.

The block diagram of the developed and constructed high-frequency instrument is shown in Fig. 2. Here 1 is a quartz generator with antiphase output; 2, decoupling stages; 3, measuring stages consisting of a resonant amplifier that includes a capacitive sensor with a dielectric; 4, adder; 5, high-frequency amplifier; 6, detector; 7, DC voltmeter.

A reference substance (comparison substance) is placed in one of the capacitive sensors, and the substance under investigation in the other sensor.

As can be seen from the block diagram, the instrument is a phase-sensitive voltmeter with an autonomous generator. We shall explain the operating principle of the instrument with the aid of diagrams.

Fig. 2. Block diagram of a differential-type RF instrument for the complex determination of \(\varepsilon\) and \(\tg \delta\) of substances.

Fig. 3. Vector diagrams.

After balancing, the voltages from the generator outputs, applied to the output of the adder, can be represented in the form of two vectors, \(OA\) and \(OB\), equal in magnitude and oppositely directed. The difference vector \(OC\) in this case is equal to 0 (Fig. 3a).

When the substance under investigation changes as a result of a change in the active losses in the sensor, the voltage vectors \((OA)\) and \((OB)\) are oppositely directed. The difference voltage vector \(OC\) in this case is not equal to zero (Fig. 3b).

If, however, only the dielectric permittivity of the substance under investigation changes, then the vectors \(OA\) and \(OB\) will be equal in magnitude, but not oppositely directed, and therefore the difference vector \(OC\) is not equal to zero (Fig. 3c).

If, however, both a change in \(\varepsilon\) and a change in \(\tg \delta\) are observed simultaneously, then the magnitude of the difference vector \(OC\) in this case (Fig. 3a) will be greater than in the cases in Figs. 3b or 3c, since it will now be determined simultaneously both by the difference in the magnitudes and by the phase difference of the vectors \(OA\) and \(OB\).

In those cases when the feedback is not equal to zero, the voltage at the output of summator 4 is different from zero. It is amplified in high-frequency amplifier 5, detected by detector 6, and measured by direct-current voltmeter 7.

The version of an instrument of this type that has been constructed to date provides measurement both of $\operatorname{tg}\delta$ and of $\varepsilon$ with an error not exceeding 0.5%; moreover, with further improvement of the instrument, an accuracy two orders of magnitude higher is entirely attainable.

References

1. Yu. Yu. Lur’e, A. I. Rybnikova. Chemical Analysis of Industrial Waste Waters. Goskhimizdat, 1963.

2. V. N. Utts, E. Reiter. A method for determining the amount of PAA in process waters after coagulation of coal slimes. Proceedings of the Chemical-Metallurgical Institute of the Academy of Sciences of the Kazakh SSR, No. I, 1963.

3. S. F. Kuz’min, V. P. Nebera, I. I. Mal’tseva. Application of polyacrylamide for thickening scheelite concentrates. Tsvetnye metally, 1960, No. 2.

4. D. N. Levchenko, et al. Emulsion of Oil with Water and Methods for Their Destruction. Moscow, “Khimiya,” 1967.

5. P. I. Gos’kov, V. A. Salikhov. A high-frequency instrument for integrated monitoring of the electrical properties of liquid dielectrics. Twentieth Scientific-Technical Conference on Radio Electronics, dedicated to the 50th anniversary of the formation of the USSR. Tomsk, NTORES named after A. S. Popov, 1972.