SOME ELECTROPHYSICAL PROPERTIES OF A HOMOLOGOUS SERIES OF NOVOLAC-TYPE PHENOL-FORMALDEHYDE RESINS

R. M. KESSENIKH, V. G. SOTNIKOV, V. G. TRITSHEL,
A. V. PETROV, Yu. P. POKHOLKOV, Yu. N. SHUMILOV

(Presented by the scientific seminar of the organic departments)

Introduction

At present, one of the principal tasks of polymer physics [1] is to establish the relationship between the properties and the structure of polymers. The study of the behavior of a homologous series of novolac-type phenol-formaldehyde resin approaches this task. The present work contains the results of an investigation of certain dielectric properties and of the thermal expansion of a series of phenol-formaldehyde resins.

I. STRUCTURE AND CHEMICAL COMPOSITION OF THE RESINS INVESTIGATED

A homologous series of phenol-formaldehyde resin was investigated, prepared in the polymer problem laboratory of the Chemical-Technological Faculty under the direction of L. I. Aristov.

The structural formulas and principal properties are given in Table 1. All the properties listed were studied in the polymer laboratory.

As is evident from the structural formulas, the reactivity of the substances we investigated is different. In this case, the nature of the course of the reactions (rate, activity, etc.) that determine the production of the resins can directly influence a number of physical properties (molecular weight, heat resistance). An investigation of the structure of the resins (Table 1) by means of X-ray structural analysis on a URS-50I showed that all four resins are characterized, to a greater or lesser degree, by ordering in the short-range order (Fig. 1). On the X-ray diffraction patterns two maxima are observed; the first, near \(2\Theta \simeq 14^\circ\), is equally distinct for all the resins investigated. The second maximum is observed near \(2\Theta \simeq 24^\circ\); in this case the decrease of the maximum proceeds in the following sequence: phenol-formaldehyde resin → ortho-cresol-formaldehyde → meta-cresol-formaldehyde → para-cresol-formaldehyde. In the same sequence the regularity of the structure also decreases.

Table 1

Fig. 1. X-ray diffraction pattern of phenol-formaldehyde resins: 1 — para-cresol-formaldehyde resin; 2 — ortho-cresol-formaldehyde resin; 3 — phenol-formaldehyde resin; 4 — meta-cresol-formaldehyde resin.

II. TECHNOLOGY OF SAMPLE PREPARATION

The samples were prepared by pressing at a pressure of 150 kg/cm² and a temperature of 85–90°C. After pressing, the samples were cooled in the mold at a rate of 1.5–2° per minute. Before pressing, the powdered resin was subjected to preliminary drying. For the investigation of dielectric properties, the prepared samples had a diameter of about 40–50 mm and a thickness of 0.7–0.8 mm.

Electrodes made of thin aluminum foil were placed on the powdered resin and, together with the resin, were heated under pressure. For the investigation of thermal expansion, samples were prepared by pressing and also by casting.

III. RESULTS OF THE INVESTIGATION

§ 1. Dielectric constant and dielectric-loss tangent \((\varepsilon\) and \(\mathrm{tg}\,\delta)\)

The dielectric constant and dielectric-loss tangent of all four resins (Table 1) were investigated in the frequency range from 20 Hz to 17,000 Hz and in the temperature interval from +80 to −100°C. To obtain the indicated frequency range, an unbalanced bridge was used, constructed by diploma student A. V. Petrov after the type of B. I. Vorozhtsov’s bridge [2]. The frequency dependence studied for the dielectric constant (Fig. 2) and the dielectric-loss tangent

Fig. 2. Dependence of dielectric constant on frequency: 1 — phenol-formaldehyde resin; 2 — ortho-cresol; 3 — meta-cresol; 4 — para-cresol.

of dielectric losses (Fig. 3), as well as the temperature dependence of \(\varepsilon\) (Fig. 4) and \(\mathrm{tg}\,\delta\) (Fig. 5), indicate that the losses are of a relaxation character. In the frequency dependence of the dielectric-loss tangent a maximum is observed, the most distinct one being for the phenol-formaldehyde and ortho-cresol-formaldehyde resins (curves 1 and 2, Fig. 3), and a blurred one for the meta-cresol-formaldehyde and para-cresol-formaldehyde resins (curves 3 and 4). The influence of the frequency of the alternating field on the magnitude of the dielectric constant and the dielectric-loss tangent of the dielectri-

dielectric losses regularly decreases in the sequence: phenol-formaldehyde resin → ortho-cresol-formaldehyde → meta-cresol-formaldehyde → para-cresol-formaldehyde. We note that ortho-cresol-formaldehyde resin (curve 2, Fig. 3) seems to fall out of this regularity. For this resin, in the range

Fig. 3. Dependence of \(\operatorname{tg}\delta\) on frequency:
1 — phenol-formaldehyde; 2 — ortho-cresol;
3 — meta-cresol; 4 — para-cresol.

of frequencies from 20 to 500 Hz, the tangent of the dielectric loss angle decreases, which indicates the presence of conductivity losses. At frequencies exceeding 500 Hz, relaxation losses already become predominant. In the temperature course of the tangent of the dielectric-loss angle (Fig. 5), the maximum is located in the region of low temperatures, which agrees well with the temperature dependence of the dielectric permittivity (Fig. 4), in the region of the maximum of the tan-

Fig. 4. Dependence of \(\varepsilon\) on temperature:
1 — phenol-formaldehyde resin;
2 — ortho-cresol-formaldehyde resin;
3 — meta-cresol-formaldehyde resin;
4 — para-cresol-formaldehyde resin.

...the tangent of the dielectric-loss angle, and the dielectric constant has a reduced value. In the region of elevated temperatures, a sharp increase in \(\tg \delta\) is observed for all resins, which is due to conduction losses, especially in the region of the softening temperatures. From these same curves (Figs. 4 and 5) it follows that the effect of temperature on \(\varepsilon\) and \(\tg \delta\) of the resins decreases in the same sequence as the effect

Fig. 5. Dependence of \(\tg \delta\) on temperature:
1—phenol-formaldehyde resin;
2—ortho-cresol-formaldehyde;
3—meta-cresol-formaldehyde;
4—para-cresol-formaldehyde.

of frequency. To illustrate the reliability of the experimental results, the temperature dependence of \(\tg \delta\) is given for phenol-formaldehyde (Fig. 6) resin at four frequencies: 20, 200, 2000, and 10000 Hz. From the presented

Fig. 6. Dependence of \(\tg \delta\) of phenol-formaldehyde resin on temperature for various frequencies:
1—20 Hz; 2—200 Hz; 4—10000 Hz.

dependence it is seen that, as the frequency increases, the maximum of the tangent of the dielectric-loss angle shifts toward higher

high temperatures, and the higher the frequency, the less pronounced are the conductivity losses. On the basis of the temperature–frequency dependence of the tangent of the dielectric-loss angle, the activation energy \((U)\) was calculated; its absolute value lies within the range from 4.10 to

\[ 6.10 \ \frac{\text{kcal}}{\text{mol deg}} \]

The smallest value of the activation energy is found for para-cresol-formaldehyde resin \((4.10 \ \frac{\text{kcal}}{\text{mol deg}})\), and the largest value for phenol-formaldehyde resin \((6.10 \ \frac{\text{kcal}}{\text{mol deg}})\).

Analyzing the experimental material obtained in the study of dielectric losses in the homologous series of phenol-formaldehyde resins, the following regularity may be noted. Polarization and dielectric losses depend to a great extent on the mutual arrangement of the methyl \((\mathrm{CH}_3)\) and hydroxyl \((\mathrm{OH})\) groups that are part of the resins studied. As the methyl group approaches the para position of the hydroxyl group, the polarity of the resin decreases. Thus, for example, the dielectric constant and the tangent of the dielectric-loss angle for phenol-formaldehyde resin at a frequency of 20 c/s are respectively equal to 7.2 and \(23 \times 10^{-4}\), whereas for para-cresol-formaldehyde resin these same characteristics are equal to 4.95 and \(13 \cdot 10^{-4}\). This regularity is observed most sharply in the region of higher frequencies, where relaxation processes predominate. Thus, the resins we investigated decrease their polarity in the same sequence in which the regularity of the structure changes. Apparently, the greater the intermolecular interaction of amorphous resins, the greater should be the degree of short-range ordering.

§ 2. Study of Specific Volume Resistivity

The specific volume resistivity \(\rho_V\) of the resins was studied under direct current by means of a “Kaktus” type vacuum-tube electrometer under atmospheric conditions and in the temperature interval from 20 to \(90^\circ\mathrm{C}\). The resistance was measured at a voltage of 200 volts and was calculated according to Ohm’s law. The through current was determined one and two minutes after the voltage was applied. Therefore, in Fig. 7 two values of \(\rho_V\) are plotted. For all four resins studied, typical dependences were obtained; at elevated temperatures the current-decay process proceeds more rapidly. From the given dependence of the logarithm of the specific volume resistivity of the resins on the reciprocal absolute temperature (Fig. 7), it follows that the curves pass through two regions. In the first region, at low temperatures, when the resins are in the glassy state, a slow decrease of \(\rho_V\) is observed. In the second region, at elevated temperatures, a sharp decrease in specific resistivity is characteristic. The onset of the sharp drop in resistivity corresponds to the softening interval of the resins. Since the softening temperatures of the resins differ from one another, the “break” points are different for the different resins.

From this same dependence (Fig. 7) it is seen that curves 1, 2, 3 are grouped close to one another, while curve 4 is separated from them.

All the resins we investigated are polar (Table 1), and their absolute values of specific volume resistivity differ little from one another. The exception is ortho-cresol-formaldehyde resin, which has the lowest value of specific resistivity (Table 2).

The calculated activation energy of the resins is summarized in Table 3. It should be noted that, in the region of higher temperatures, the absolute values of the activation energy lie within the range from 99.0 to 114.0

\[ \frac{\text{kcal}}{\text{mol}\cdot\text{deg}} \]

. Such elevated values of the activation energy in the temperature region of softening cannot be explained. In [3] it is indicated that such activation-energy values are obtained as a result of failing to take into account changes in the degree of molecular interaction.

Comparing the experimental material on the study of the specific volume resistance of novolac-type phenol-formaldehyde resins with their structure, we are unable to establish any connection between them. In this case, the influence of the softening temperature of the resins is manifested most strongly.

Fig. 7. Dependence \(\lg \rho_v = f\!\left(\frac{1}{T_0}\right)\)

1 — phenol-formaldehyde resin; 2 — meta-cresol-formaldehyde; 3 — para-cresol-formaldehyde; 4 — ortho-cresol-formaldehyde.

Thus, for example, meta-cresol-formaldehyde and phenol-formaldehyde resins have practically the same softening interval, equal to 54–58°C and 53–57°C, respectively, and their specific resistance lies within the range \((1.73 \div 1.82)\cdot 10^{16}\ \Omega\cdot\text{cm}\) at a temperature of 30°C; para-cresol-formaldehyde resin has a softening interval of 38–42°C, and \(\rho_v = 1.22\cdot 10^{16}\ \Omega\cdot\text{cm}\); ortho-cresol-formaldehyde has a softening interval of 28–33°C and a specific resistance equal to \(6.7\cdot 10^{14}\ \Omega\cdot\text{cm}\). Thus, the pres—

...the presence of polar groups alone in the resin is not sufficient for the appearance of high conductivity. Dissociation of impurities or of the principal molecules of the substance is necessary. In addition, the mobility of the current carriers also becomes of great importance [4].

Table 2

Table 3

§ 3. Thermal Expansion of the Resins

The thermal expansion of the resins (Table 1) was studied with the aid of a linear quartz dilatometer with an IZK-50 microinterferometer (5, 6) over the temperature interval from −50 to +60°C. The study was carried out in a cooling–heating cycle at a rate of

\[ 0.75\ \frac{\text{degree}}{\text{minute}}. \]

In the measured temperature interval, the thermal elongation (Fig. 8) changes linearly with the change in temperature, which indicates that the temperature coefficient of linear expansion (in the measured temperature interval) is independent of temperature. In the region of elevated temperatures a smooth bend is observed (Fig. 8), followed by a further sharp decrease in elongation. The temperature at which this decrease is observed for each resin corresponds to the beginning of the temperature interval of softening. The resins investigated pass directly from the glassy state into a viscous-flow state, which is characteristic of low-molecular-weight compounds. In the present case

the flow temperature \((T_k)\) coincides with the softening temperature \((T_p)\). The observed deformation of the resin is due to the pressure of the quartz rod and the dilatometer. Table 4 summarizes the experimental data for determining the softening temperatures and the temperature coefficients of linear expansion \((TK\alpha_l)\).

Fig. 8. Dependence of thermal elongation on temperature: 1 — phenol-formaldehyde resin; 2 — meta-cresol-formaldehyde; 3 — para-cresol-formaldehyde; 4 — ortho-cresol-formaldehyde.

The linear dependence of thermal elongation on temperature that we obtained would seem, above all, to indicate that thermal expansion is due to intramolecular displacements. However, the obtained values of \(TK\alpha_l\) for phenol-formaldehyde and meta-cresol-formaldehyde resins practically coincided, \((40 \div 41)\cdot 10^{-6}\), while for para-cresol-formaldehyde and ortho-cresol-formaldehyde resins the temperature coefficient of linear expansion lies in the interval \((75 \div 68)\cdot 10^{-6}\ 1^\circ\mathrm{C}\). Thus, in the latter two resins the thermal expansion is greater than in the first two. The indicated

Table 4

This circumstance, apparently, can be explained by the different values of the softening temperatures of the resins.

Conclusions

1. The clearest relationship is that between the structure and chemical constitution of the studied low-molecular-weight resins of the homologous series of novolac-type phenol-formaldehyde resin and the dielectric permittivity and dielectric losses. As the methyl group approaches the para position relative to the hydroxyl group, the magnitude of the dielectric permittivity and of the dielectric loss tangent decreases. The observed decrease in polarity is unambiguously associated with a decrease in the degree of short-range order, as revealed by X-ray structural analysis.

2. No dependence of the volume resistivity of the studied resins on the presence of a methyl group and its position relative to the hydroxyl group was found; however, it was established that the absolute values of \(\rho_v\) are lower for resins with a lower softening temperature.

3. The thermal expansion of the studied resins in the measured temperature interval depends linearly on temperature. The influence of the position of the methyl group relative to the hydroxyl group on thermal expansion is not manifested in a regular way.

Literature

1. G. P. Mikhailov. Journal of the All-Union Chemical Society named after D. I. Mendeleev, vol. VI, 4, 1961.
2. B. I. Vorozhtsov. Izvestiya of the Academy of Sciences of the USSR, No. 1, 1959.
3. N. I. Shishkin. ZhTF, vol. 26, issue 7, 1956.
4. P. P. Kobeko. Amorphous Dielectrics. Gosenergoizdat, 1952.
5. V. A. Presnov. Proceedings of the Siberian Physico-Technical Institute, issue 36, 1958.
6. V. M. Anikeenko, V. M. Kessenikh. Bulletin of the Electrical Industry, No. 1, 1961.