Thermal properties and stability of boron-containing phenol-formaldehyde resin formed from paraformaldehyde

Abstract

A boron-containing phenol-formaldehyde resin (BPFR) was synthesized from boric acid, phenol and paraformaldehyde. The structure change of BPFR during curing was studied by Fourier-transform infrared (FTIR) spectroscopy. The glass transition temperatures (T_g) were determined by differential scanning calorimetry (DSC) and torsional braid analysis (TBA). The thermal degradation of BPFR was studied by FTIR and thermogravimetry analysis (TGA). The results indicate that the degradation of BPFR begins with the breaking of B–O bonds. The TGA results show that the dynamic thermal degradation of BPFR appears as four reaction stages in nitrogen and three in static air. The kinetic analysis results show that the reaction follows a first order mechanism to the second and third stages in air, and first order to the second stage and third order to the third and fourth stages in nitrogen.

Introduction

The boron-containing phenol-formaldehyde resin (BPFR) is a modified phenolic resin. It is obtained by introducing boron into the main chain of a phenolic resin. BPFR has high performance properties, such as thermostability, mechanical strength and electric properties. Some reports have been appeared on the synthesis and application of BPFR [1], [2], [3], [4], [5], [6]. The thermal stability of BPFR formed from phenol, boric acid and formalin has been investigated [5], while thermal properties and stability of BPFR formed from paraformaldehyde are not investigated until now.

In this work, BPFR was synthesized by the paraformaldehyde method [3], [4], the structure change of BPFR during curing and thermal degradation was monitored by Fourier-transform infrared (FTIR) spectrometry, the variation of glass transition temperatures (T_gs) for BPFR during isothermal curing conditions was studied by differential scanning calorimetry (DSC) and torsional braid analysis (TBA), and the weight changes were monitored by thermogravimetry analysis (TGA).

Boric acid, phenol and paraformaldehyde were all analytical reagents, were supplied by Beijing chemical reagent company.

The BPFR was synthesized according to the literature [3], [4], its structure is mainly the two forms shown in Scheme 1.

A Shimadzu DSC-41 differential scanning calorimeter operating in a nitrogen atmosphere was used to determine the glass transition temperatures (T_g) of BPFR. The DSC instrument was calibrated by indium standard. α-Al₂O₃ was used as the reference material.

About 10 mg of the BPFR sample was placed in a sample cell, then a series of samples were cured at different temperatures for various periods of time. Thereafter, the cured samples were cooled rapidly to 20 °C, and then subjected to a dynamic scan at 20 °C/min to determine the T_g. The glass transition process appears as baseline shift, and the T_g is taken as the midpoint of the heat capacity change (ΔC_p).

A cleaned glass fibre braid was first dipped in the BPFR-trichloromethane solution for about 30 min, and then taken out. When the solvent in the braid was completely evaporated in vacuum, the braid was placed in a heated oven with a fixed temperature of 180 °C. After curing at this temperature for some time, the braid was withdrawn and cooled to room temperature, and the T_g was determined using a GDP-3 torsional braid analyzer with a heating rate of 2 °C/min.

A Bio-Rad FTS-40 FTIR spectrometer was used to investigate the structure changes of the BPFR during the curing and thermal degradation. The BPFR sample was dissolved in trichloromethane and then coated as a thin film on a potassium bromide plate. When the solvent in the film was completely evaporated in vacuum, the potassium bromide plate was scanned by the FTIR instrument. Thereafter, the plate was placed in a heated oven with a fixed temperature of 210 °C. During the curing reaction at this temperature, the plate was repeatedly withdrawn at regular time intervals for analysis.

To determine the structure change during the thermal degradation, the BPFR was first scanned by the FTIR instrument after curing at 210 °C for 6 h, then it was scanned at different temperatures during the dynamic degradation reaction at a heating rate of 10 °C/min.

The principal absorption bands of BPFR are: the hydroxyl group –OH is at 3300–3500 cm⁻¹, the benzene ring is at 1600 and 750 cm⁻¹, the phenol borate B–O is at 1350 cm⁻¹, phenol hydroxyl group is at 1230 cm⁻¹, the –CH₂– group appears at 2920 and 1450 cm⁻¹, benzyl hydroxyl group is at 1020 cm⁻¹, ether linkage C–O is at 1050 cm⁻¹, carbonyl group is at 1650 cm⁻¹.

A Shimadzu TGA-40 thermogravimeter was used to determine the weight loss behaviour of BPFR during degradation. About 8 mg of the cured BPFR sample was introduced into the thermobalance, then heated to 900 °C at different heating rates: 2.5, 5, 7.5 and 10 °C/min. The experiments were carried out under nitrogen and static air atmosphere, respectively.

Fig. 1 shows the IR absorption variations of BPFR during the curing reaction. In the initial stage of curing, the absorption of phenol borate B–O at 1350 cm⁻¹ increases. With the reaction proceeding, the absorptions of phenol hydroxyl group at 1230 cm⁻¹ and benzyl hydroxyl group at 1020 and 3300 cm⁻¹ decrease, due to the formation of borate from the reaction of phenol and benzyl hydroxyl groups in the BPFR with unreacted –OH group in boric acid, and the formation of –CH₂– from the reaction of paraformaldehyde with the benzene ring. In addition, different from the BPFR formed from formalin method [5], this BPFR shows no ether bonds formed between benzyl hydroxyl groups and no carbonyl group formed from oxidation of –CH₂– and –CH₂–O–CH₂– groups during the curing reaction. The final molecular structure may be described as in Scheme 2.

Fig. 2 shows the IR absorption variation of BPFR during thermal degradation. With rising temperature, the absorption of B–O at 1350 cm⁻¹ decreases first, then the absorptions of phenol hydroxyl and –CH₂– groups at 1430, 1390, 1230 and 1100 cm⁻¹ decrease gradually. Other absorptions related with the benzene ring, such as 750 and 650 cm⁻¹, decrease rapidly after beyond 400 °C. After 550 °C, these absorptions almost disappear. These IR absorption variations indicate that the BPFR degradation begins with the breaking of B–O bonds, and then the –CH₂–, and finally the benzene rings are oxidised or broken.

Generally, the T_g of a crosslinked system is related to the conversion of reacting groups [7], [8], [9], which depends upon the curing conditions, such as temperature, time and heating rate. With these curing conditions varying, the T_g of the system will be changed, correspondingly. Now, it is accepted that the variation in T_g is attributed to various parameters such as the molecular weight, the stiffness and the free volume entrapped in the network. By monitoring the T_g variation during curing, the thermal properties of the material will be better understood [10]. In this work, a series of DSC experiments was carried out for BPFR with the curing temperature and time varying. Some DSC curves are shown in Fig. 3, Fig. 4, respectively, and all the results are listed in Table 1, in addition, some TBA results are also listed in Table 1.

As shown in Fig. 3, the glass transition of the uncured BPFR appears a steep slope in the DSC curve, this may due to the lower molecular weight, more flexible chain, and the large free volume. In addition, the DSC trace of uncured BPFR shows that the glass transition region appears over a relatively narrow temperature range of 25.5–40.8 °C, and the heat capacity variation is 0.45 J/(g °C).

For the cured BPFR, Fig. 4 shows that different transition behaviours appeared with different degrees of cure. For the BPFR cured at low temperature and in a short period of time, the dynamic DSC scan gives a exothermic peak following the glass transition, this is attributed to the continuing of the residual curing reaction between the uncured groups at high temperature. With the increasing isothermal curing temperature and time, the degree of cure is increased, so the residual exothermic effect is gradually decreased. In addition, as shown in Table 1, T_g increases with the curing temperature and time. The T_g can reach the highest value of 225.5 °C.

Generally, the temperature range of the glass transition region in a homopolymer is about 20–30 °C. Comparing the transition behaviour of cured BPFR with that of the uncured BPFR, the distinction can be obviously seen in the transition temperature range, strength and associated heat capacity. When the BPFR was lightly crosslinked, that is to say, the BPFR was cured at low temperature for a short period of time, the influence of the crosslinking on the T_g can be seen apparently. With increasing curing temperature and time, the influence gets significant. With the reduction of free volume, the motion of the chain segment with large groups gets difficult, so, as a result, the glass transition temperature region becomes broad.

The TGA, derivative thermogravimetry (DTG) and DTA curves of BPFR are shown in Fig. 5, Fig. 6, Fig. 7, respectively. It can be seen that the thermal degradation of BPFR appears as a multi-stage reaction and they are different in nitrogen and static air. According to these curves, the degradation process can be divided into four stages in nitrogen and three stages in static air, which indicates that the influence of the atmosphere on the degradation is different for various stages.

In the initial stage, the total weight loss in nitrogen and air are almost the same, about 2%, and may correspond to the loss of some small end groups, such as –CH₂OH.

However, in the following stages, the effect of the atmosphere on the degradation can be seen from Fig. 5 and Table 2, especially at low heating rate, such as 2.5 °C/min. In nitrogen, the weight loss in the second stage between 428 and 610 °C is about 16% and in the third stage between 610 and 719 °C it is about 6%, and the total weight loss at 900 °C is about 41%. In air, the weight loss in the second stage between 410 and 607 °C is about 24%, which is close to the total weight loss in both the second and third stages in nitrogen. In the third stage in air, the influence of the atmosphere on the reaction is significant, and the total weight loss at 900 °C is about 88%. This indicates that the degradation of BPFR in nitrogen is a thermal decomposition process and in air is a thermal oxidation reaction.

Considering the structure changes shown in Fig. 2, the weight loss in the second and third stages in nitrogen or the second stage in air may be corresponding to the loss of the small groups and weaker bonds in the chains of BPFR, such as –OH, –CH₂–, and the weight loss in the fourth stage in nitrogen or the third stage in air may be corresponding to the loss of the benzene rings.

To determine the kinetic parameters of the decomposition from the thermogravimetric data, the first step is to evaluate the conversion of the reaction. In dynamic TGA experiments, the weight change of the sample is regarded as a function of temperature, and the conversion can be expressed as

$$\text{α=}\text{w}{}_{\mathrm{i}}\text{−w}{}_{\mathrm{T}}\text{w}{}_{\mathrm{i}}$$

where w_i is the sample weight in i stage, w_T is the residual weight of the w_i at temperature T. Therefore, with the TGA curves and using the Eq. (1), the conversions are calculated for different degradation stages.

In this work, considering the multi-stage degradation process of BPFR, the TGA data were analyzed on the basis of the Madhusudanan–Krishnan–Ninan method [11], which can be expressed by the following equation:

$$\text{ln}\text{g}\text{α}\text{T}{}^{1.92}\text{=}\text{ln}\text{AE}\text{βR}\text{+3.77−1.92}\text{ln}\text{E−}\text{E}\text{RT}$$

where A is the pre-exponential factor in the Arrhenius equation, E is the apparent activation energy, R is the universal gas constant, β is the heating rate, T is absolute temperature, and g(α) is the integral form of the conversion dependence function. The correct form of g(α) depends on the proper mechanism of the decomposition reaction [11]. Different expressions of g(α) for some solid state reaction mechanisms can be described as following: first order: F₁, −ln(1−α); second order: F₂, $\text{1}\text{(1−α)}$; third order: F₃, $\text{1}\text{(1−α)}{}^2$.

According to the Eq. (2), the activation energy for every g(α) function can be obtained at different heating rates from fitting the $\text{ln}\text{g(α)}\text{T}{}^{1.92}$ versus $\text{1}\text{T}$ plots. In this work, the apparent activation energies and pre-exponential factors are calculated for the conversions in the range of 5–20%. For different degradation stages in both nitrogen and air, the conversions werer all tested for various mechanism functions. Some results are listed in Table 3, Table 4, respectively.

As shown in Table 3, Table 4, for the same degradation stage at a given heating rate, the correlation values for different mechanisms are different. According to the principle that the probable mechanism has high correlation coefficient value and low standard deviation value, the probable mechanism functions of the thermal degradation reaction are deduced from the calculated results: F₁ to the second and third stages in air, F₁ to the second stage and F₃ to the third and fourth stages in nitrogen, respectively. Correspondingly, the kinetic parameters are listed in Table 5, Table 6 for these mechanism functions in various stages at different heating rates in nitrogen and air, respectively.

As shown in the Table 5, Table 6, for the same degradation stage at different heating rate, the values of E and A for the same mechanism are relatively close and within a certain range. For different degradation reaction stages, the ranges of E and A values are different, and they are affected by the atmosphere.

• The cured structure of BPFR formed from paraformaldehyde method is different from BPFR formed from formalin method. The structure in this curing BPFR does not contain ether bonds and carbonyl groups. So the thermal stability of this BPFR is better than BPFR formed from formalin.

• The degradation of BPFR begins with the breaking of B–O, corresponding to IR absorption at 1350 cm⁻¹. The TGA results show that the dynamic thermal degradation of BPFR appears as four reaction stages in nitrogen and three in static air. The kinetic results calculated with Madhusudanan–Krishnan–Ninan method show that the reaction follows F₁ mechanism to the second and third stages in air, F₁ to the second stage and F₃ to the third and fourth stages in nitrogen.

• The glass transition behaviour of the cured BPFR is significantly affected by crosslink density, molecular motion and free volume in the system. With varying curing conditions, the glass transition behaviour of BPFR varies correspondingly. The highest T_g value of BPFR is 225.5 °C.