The primary objective of this study was to analyze the influence of condensate obtained during the direct steaming of beech timber on the curing kinetics of UF resin and to estimate the potential of its application for the preparation of UF adhesives for the application in the WBP industry.
Characterization of UF adhesive
Commercial UF resin was characterized as pure and with the addition of BSC and the technical specifications of UF and UFC resins are given in Table 1. Properties, such as solid content, density, pH, viscosity, free formaldehyde content, and gel time were determined. The addition of the condensate did not affect the technical properties of the UF resin in terms of density, solid content, and viscosity. The pH value slightly decreased as expected after the addition of BSC. Gel time was also reduced, indicating a faster condensation reaction of UFC resin. Previous studies have suggested that more branched structures with many methylene ether linkages exhibit shorter gel time Since previous studies pointed out that more branched structures with many methylene ether linkages showed shorter gel time (Kim and Park 2022). Therefore, the shorter gel time of the UFC resin implies that it formed a more networked structure compared to pure UF resin, which typically contains de-branched hydroxymethyl groups and linear structures (Wibowo et al. 2022).
Table 1
The technical properties of UF and UFC resins
Sample | Property |
Solid content (wt%) | Density (g/cm3) | Gel timea (s) | pH | Free-F (wt%) | Viscosity, F420 (s) |
UF | 66.0 | 1.290 | 111 | 8.0 | 0.14 | 84 |
UFC | 66.5 | 1.290 | 71 | 7.9 | 0.10 | 84 |
a Gel time determined for UF resin with solid content reduced to 55% |
Characterization of BSC
Condensate obtained in the process of direct steaming of beech wood appeared as a dark brown solution with a pH value of 5 (Table 2, Fig. 1a). After evaporation, brown powder was obtained and used for further investigations.
Table 2
The properties of condensate
Sample | Property |
Solid content (wt%) | pH | TPC (mg GAE/mL) |
BSC | 1.19 | 5 | 1.54 |
It is already known that water-soluble wood extractives are compounds of a phenolic nature that readily react with formaldehyde (Tanase et al. 2019). Expectedly, the diverse chemical compounds identified in this study involve polar molecules soluble in water (phenols, polyphenols, and flavonoids) and lignin fragments. Due to elevated temperatures, steam atmosphere, and the presence of microorganisms, the wood steaming process leads to the degradation of lignocellulosic structures, followed by hydrolysis and extraction of obtained water-soluble compounds into the wastewater solution (Rowell 2005). This process is also followed by the gradual increase in acidity that additionally facilitates the breakdown of covalent bonds.
The high TPC value (Table 2) and preliminary chemical analysis obtained by the ultra-high performance liquid chromatography-quadrupole time-of-flight mass spectrometry (UHPLC Q-ToF MS) confirmed the presence of numerous phenolic compounds in BSC. More than 28 phenolic compounds (phenolic acids, their derivatives, and flavonoids) were identified and among them predominant are: gallic acid, vanillic acid, syringic acid, abscisic acid, quinic acid, hydroxybenzoic acid, dihydroxybenzoic acid, coniferin, taxifolin, epicatechin, vanillin, and methyl-galloyl ester, present as aglycones or glycosides (mostly hexosides).
FTIR spectroscopy further elucidated the chemical composition of the crude BSC by revealing the presence of the functional aromatic and phenolic OH groups (Fig. 1b) as confirmation of chemical analysis results. Broad and strong absorption bands observed at 3250 cm− 1 can be attributed to hydroxyl group (OH) stretching vibrations. The existence of a sharp peak and a small shoulder at 2922 cm− 1 are associated with the symmetric and asymmetric C-H stretching vibrations of CH2 and CH, respectively. The absorption in the range between 1600 cm− 1 and 1500 cm− 1 is attributed to the C = C stretching vibrations of aromatic rings. The bands in the region of 1375 cm− 1 can be assigned to the O-H stretching vibration of the phenolic group and the C-H deformation of the aliphatic group. Peaks between 1350 cm− 1 and 1200 cm− 1 indicate the presence of phenolic and anthocyanic compounds. Absorption bands that correspond with symmetrical stretching vibrations of C-O-C and the in-plane bending vibrations of aromatic C-H and C-O stretching vibrations occurred at 1010 cm− 1 and 1270 cm− 1. Additionally, the pattern of the out-of-plane C-H bending bands in the region 900 − 675 cm − 1 characteristic of the aromatic substitution scaffold can be found at 772 cm− 1.
The reaction mechanism of the condensation reaction
The conventional synthesis of UF resins generally involves two types of reactions, namely addition and condensation reactions. In the first stage, addition reactions occur in which urea and formaldehyde react together to form mono-, bis-, or tris(hydroxymethyl) ureas, whereas, in condensation reactions, hydroxymethyl compounds combine to form an oligomer, which is then further reacted to form a three-dimensional cross-linked network (Fig. 2). Methylene, methylene ether, and oxymethylene linkages are generally formed during condensation reactions with their proportions being mostly related to the synthesis conditions (Wibowo et al. 2020; Nguon et al. 2018). The polymerization of urea and formaldehyde is a very complex reaction since many parameters or conditions can influence the reaction rate as well as the characteristics of the end products. The most important parameters that determine the properties of resins are: (1) the formaldehyde-to-urea (F/U) molar ratio, (2) the temperature of the reaction, and (3) the pH at which the addition and condensation reaction takes place.
Each methylolation step has its own rate constant, k, with different values for the forward and backward reactions. The reversibility of this reaction is one of the most important features of UF resins and is responsible for both the low resistance against hydrolysis caused by the attack of moisture or water and the subsequent formaldehyde emission. Free formaldehyde in the resin originates from the slight hydrolysis of weakly bonded formaldehyde. The formation of oxymethylene groups mostly depends on the F/U molar ratio, with higher molar ratios increasing the tendency to form highly methylolated species.
The curing behavior of UF and UFC resin studied by FTIR Spectroscopy
The chemical structure and changes during curing and cross-linking reactions were investigated and confirmed also via FTIR spectroscopy, as shown in Fig. 3. The FTIR spectra depicted include pure UF adhesive (Fig. 3a) and the UFC mixture (Fig. 3b) before curing, as well as after curing (cUF and cUFC, respectively).
Both resins (UF and UFC) behave similarly after curing. The wide bands at 3320 cm− 1 in Fig. 3a (or 3310 cm− 1 for UFC, Fig. 3b) originated from stretching vibration of O-H and N-H and the presence of byproducts such as water and excess formaldehyde that can hydrogen bond to the CH2OH, NH2, and NH functional groups present in the observed resins and their intensities expectedly decreased in cured samples. A sharp, but small peak, occurring at 2959 cm− 1 belongs to the stretching C-H vibrations of methylene groups as the result of the contribution of dimethylene ether bonds (CH2-O-CH2), hydroxymethyl (CH2OH) and methylene bonds (N-CH2). The increased intensity after curing indicated crosslinking of the resins. The characteristic amide bands that appear at 1628 cm− 1 (i.e. 1638 cm− 1), 1531 cm− 1 (i.e. 1539 cm− 1), and 1243 cm− 1 (i.e. 1248) cm− 1 represent amide I (C = O stretching vibration in -CONH), amide II, and amide III bands (stretching vibration of C-N and deformation of vibration of N-H in NH-CO). The intensity of all amide bands also increased during curing implying the formation of amide bonds. The weak band of UF at 1361 cm− 1 in Fig. 3a decreases, while a similar peak at 1377 cm− 1 for UFC emulsion (Fig. 3b) becomes stronger after curing. This area of UF polymer spectra is usually assigned to C-H vibrations in methylene structures. The methylene ether bridge (CH2-O-CH2) characteristic absorption peak at 1129 cm− 1 (i.e. 1135 cm− 1) belongs to the C-O stretching vibration of hydroxymethyl and for UFC becomes stronger after curing compared with UF resin, indicating the formation of new ether linkages in a greater extent. The peaks about 1011 cm− 1 (i.e. 1003 cm− 1) were assigned to ν(C-O-C) stretching vibrations of ether linkages and decreases with crosslinking, but to a lesser extent for UFC resin pointed out that cured UFC resin contained more ether linkages compared to UF. A small peak at 772 cm− 1 which appears in cured samples is the characteristic absorption peak of the uronic structure that is formed.
In conclusion, while the spectra did not exhibit significant differences, the main distinction was observed in the region attributed to vibrations of ether linkages. Since they were more pronounced after curing in UFC spectra, the resulting adhesive contained a higher degree of crosslinking. These differences between spectra originated from the presence of numerous phenolic compounds in the BSC that contain reaction centers available for the reaction with formaldehyde (Bekhta et al. 2021), especially in the first stage of the reaction, thus subsequently influencing the latter condensation step. Numerous previous studies have also shown that catechin, epicatechin, and quercetin react with formaldehyde 5–6 times faster than gallic acid (Takagaki et al. 2000). Furthermore, they react with formaldehyde, similar to phenolic acids, even at 25°C in a slightly acidic environment (pH 6), conditions closely resembling those encountered during the preparation and utilization of the UFC adhesive system. The flavonoid molecules also possess reaction centers that can attach formaldehyde (Arbenz and Averous 2015), so flavonoids proved to be capable of binding approximately 10% of formaldehyde relative to their weight. Likewise, the syringic acid and vanillic acid molecules can bind 15% and 17% of formaldehyde relative to their weight, respectively (Garro Galvez et al. 1996).
The curing behavior of UF resin determined by DSC
The curing behavior of the UF resin was assessed using the DSC method, a well-established technique for studying curing reactions, i.e. the condensation and crosslinking between urea and formaldehyde. The effect of BSC addition to UF adhesive on its hardening characteristics was also investigated using this methodology.
The curing reaction started at Tonset temperature, the so-called initiation temperature. The conversion rate reached a maximum during a dynamic scan of the reaction at Tcuring temperature. Both values were obtained from the DSC curves. The curing reactions of UF resins are typically exothermic, so it was hypothesized that the reaction enthalpy (ΔH) would simply be proportional to the degree of conversion (α) during the curing process. The curing reactions of UF resins were represented by an exothermic region with a single peak, and the total area under the respective thermographs was assumed to be the total heat of the reaction for the respective reaction systems. An exothermic peak corresponded to a continuing exothermic reaction with an increase in temperature and it was also visible in the DSC scan. It is generally accepted that, in the presence of an acid catalyst, the exothermic peak arises from the polycondensation reaction of primary amino groups of unreacted urea with hydroxymethyl groups of resin (Xing et al. 2005; Kim et al. 2006; Gao et al. 2007). The endothermic reaction in the DSC curves after resin curing is related to vapor loss (Réh et al. 2021), which can be attributed to the hydrolysis of the cured UF resin.
Figure 4 presents the DSC thermograms of the curing reaction of UF adhesive and the UFC adhesive system. The values of peak temperature and the relevant heating rate were used to plot the dependence of ln(β/Tp2) vs. 1/Tp and to form the corresponding regression line, as presented in Fig. 5. This enabled the calculation of activation energy (Ea) for each UF adhesive system, following the Kissinger method (Eq. 2).
When UF curing is evaluated by DSC, the peak temperature (corresponds to the curing rate of the UF resin and it is directly proportional to the rate constant (Gao et al. 2007). With increasing heating rate, the curing peak temperature shifts to a higher value in both adhesive systems (Fig. 4) and the same is observed for the peak area (the area under the exothermic DSC curve) as well. Since the peak area corresponds to the amount of energy required to complete the cure of the resin (ΔH, Table 3), it should be also proportional to the amounts of methylene linkages in the UF resins (Lee and Kim, 2013). The introduction of BSC to the UF resin formulation significantly affected the curing behavior and curing kinetics of the resulting resin. The observed shift in curing peak temperatures to lower values in UFC compared to UF indicates a faster reaction, which is in agreement with the lower gel time obtained for UFC (Table 1). This acceleration can be attributed to the reaction of phenolic compounds present in the BSC with free formaldehyde in the UF adhesive system which occurred at lower temperatures and facilitated crosslinking reactions at a later stage of the reaction.
Activation energy (Ea) determined by DSC analysis
The Ea of the curing reaction for both UF and UFC resins was calculated using the Kissinger equation, based on the variation of the peak reaction temperature (Tmax) as a function of the heating rate in the DSC scan (Fig. 5, Table 3). The activation energy describes the effect of temperature change on the curing reaction and can serve as a reference in the evaluation of UF resin curing behavior. A lower activation energy indicates that the curing rate is more sensitive to temperature change compared to that of a resin with a higher activation energy.
The addition of condensate into the UF adhesive resulted in slightly lower peak temperatures at heating rates of 5°C/min and 20°C/min, while this parameter did not change at the heating rate of 10°C/min. If assumed that the peak temperatures represent the temperatures at the maximum rate of reaction, these results suggest that the addition of condensate has slightly improved the curing reaction of the UF adhesive at slow and fast heating rates.
Table 3
Peak temperatures of curing reaction, activation energy (Ea), enthalpy (ΔH), and pre-exponential factor (A) for the UF and UFC adhesives
Adhesive | Tmax (°C) | Ea kJ/mol | r2 | ΔH J/g | A |
β = 5°C/min | β = 10°C/min | β = 20°C/min | 1011 s− 1 |
UF | 82.12 | 90.20 | 99.96 | 79.44 | 0.9981 | 74.32 | 1.85 |
UFC | 81.39 | 90.21 | 98.51 | 82.59 | 0.9989 | 63.90 | 5.78 |
However, the addition of condensate has increased the energy of activation from 79.44 kJ/mol (pure UF adhesive) to 82.59 kJ/mol (UFC). Although, at first in contrast to lower peak temperature, this increase in Ea suggests a more complex curing process of the UFC adhesive system. The presence of phenolic compounds in the condensate enhances reactions that may not occur in the pure UF resin, potentially altering the balance between different curing reactions. Moreover, the reaction between phenolic compounds and free formaldehyde in the UF resin reduces the availability of free formaldehyde, leading to a lower ΔH (Table 3) of UFC adhesive system. This affects the degree of crosslinking of the cured resins and the reactivity of the hardening reaction, as both depend on the availability of free formaldehyde in the system (Reh et al. 2021). Additionally, the increased activation energy for the UFC resin at the same time resulted in a more stable crosslinking structure, as its degradation requires more energy (the hydrolysis as a backward reaction also has a higher Ea). The rate constant (i.e. rate of chemical reaction) of UF curing is determined by both the activation energy and the pre-exponential factor (A) which is obtained from the intercept of fitted linear, and it can be used to evaluate the number of effective collisions of activated molecules. Collision theory also indicates that the number of successful collisions is defined by both temperature and the characteristics of the reactant particles. The higher the pre-exponential factor for UFC - the more effective the collision (Table 3), thus indicating that the addition of condensate increases the production of active sites and the number of collisions during the crosslinking reactions.
The enthalpy of the curing reaction (ΔH) was also calculated (Table 3), showing a decrease from 74.32 J/g for pure UF adhesive to 63.90 J/g for UFC. As the ΔH value indicates the amount of heat energy emitted during the curing reaction of the UF adhesive, its decrease indicates that the condensate has affected the intermediate crosslinking reactions of the UF adhesive. Given that ΔH represents the energy created by the collision of molecules in the reactions of adhesive curing, its decrease indicates that it was spent for the activation of possible reactions of extracted compounds, both with the catalyst and with the reactive groups of the adhesive. BSC decreases the curing enthalpy through diffusion control, as decreased mobility of the molecules and their reactive groups resulted in crosslinking between molecules at the latter stages of the curing process. The reduction in reaction enthalpy observed may also be explained by the lower energy needed for UFC curing due to the presence of phenolic compounds in BSC that can react with formaldehyde. The activity of UF curing increases with higher amounts of formaldehyde scavenger introduced into the samples (Reh et al. 2021).
Iso-conversional methods
Since the Kissinger method is based on Eq. 2 which is the rate equation for a single-step reaction, the problem arises for the reactions in the condensed state which typically involve multiple steps, and, therefore, face more than a single energy barrier that has important implications for the temperature dependence of the reaction rate. A more sensitive way of detecting the reaction complexity is to use iso-conversional method that allows to determination of the activation energy as a function of conversion. In that way, the Ea change can be followed during the course of the reaction since during various reaction steps processes Ea vary. Hence, a detailed analysis of the reaction kinetics was made using the iso-conversional methods of Kissinger-Akahira-Sunose (KAS) and Friedman (Eqs. 3 and 4). These methods were applied to evaluate the reaction kinetics at various degrees of reaction, as well as to make predictions of UF adhesive curing at iso-thermal conditions.
The thermograms obtained from the DSC scans, as presented in Fig. 4, were integrated using the TA Universal analysis software in order to obtain the values of reaction degree as a function of temperature at different heating rates, for both UF and UFC adhesives, as presented in corresponding diagrams on the Fig. 6.
The obtained kinetic data (Fig. 6) were used to calculate the activation energy values (Ea) at a certain degree of reaction (α). These values for different iso-conversional models are presented in Fig. 7, and Tables 4 and 5.
Table 4
Activation energies (Ea) and pre-exponentional factors (A) of the curing reaction of UF and UFC adhesive systems obtained by the KAS iso-conversional model
α % | UF | UFC |
Ea kJ/mol | r2 | A 1010 s− 1 | Ea kJ/mol | r2 | A 1010 s− 1 |
5 | 77.87 | 0.9976 | 22.0 | 78.54 | 0.9993 | 31.4 |
10 | 76.86 | 0.9979 | 12.6 | 78.89 | 0.9997 | 28.8 |
20 | 76.55 | 0.9982 | 9.04 | 78.62 | 0.9997 | 20.5 |
30 | 76.61 | 0.9984 | 8.11 | 78.54 | 0.9997 | 17.2 |
40 | 76.53 | 0.9985 | 7.18 | 78.49 | 0.9998 | 15.3 |
50 | 76.10 | 0.9984 | 5.73 | 78.28 | 0.9999 | 13.0 |
60 | 75.29 | 0.9982 | 4.04 | 77.73 | 1.0000 | 9.85 |
70 | 74.44 | 0.9977 | 2.76 | 77.33 | 0.9999 | 7.71 |
80 | 74.52 | 0.9967 | 2.47 | 77.79 | 0.9988 | 7.68 |
90 | 76.05 | 0.995 | 3.30 | 78.53 | 0.9957 | 7.67 |
95 | 77.93 | 0.993 | 5.05 | 78.22 | 0.9925 | 5.55 |
Table 5
Activation energies (Ea) and pre-exponentional factors (A) of the curing reaction UF and UFC adhesive systems obtained by the Friedman iso-conversional model
α % | UF | UFC |
Ea kJ/mol | r2 | A 1011 s− 1 | Ea kJ/mol | r2 | A 1011 s− 1 |
10 | 71.92 | 0.9988 | 6.63 | 80.57 | 0.9996 | 128 |
20 | 75.22 | 0.9994 | 28.5 | 77.72 | 0.9999 | 68.8 |
30 | 77.18 | 0.9996 | 81.6 | 78.23 | 0.9998 | 117 |
40 | 75.87 | 0.9992 | 67.4 | 78.24 | 1.0000 | 143 |
50 | 70.87 | 0.9968 | 14.5 | 76.08 | 0.9986 | 75.3 |
60 | 65.33 | 0.9916 | 2.18 | 72.07 | 0.9921 | 18.4 |
70 | 66.10 | 0.9871 | 2.12 | 74.04 | 0.9785 | 26.1 |
80 | 75.80 | 0.9845 | 3.03 | 80.92 | 0.9680 | 147 |
90 | 83.20 | 0.9862 | 156 | 81.98 | 0.9679 | 99.6 |
95 | 87.72 | 0.9838 | 305 | 77.85 | 0.9663 | 12.8 |
The application of the KAS iso-conversional method also showed that the addition of beech condensate solution into UF adhesive increased the activation energy values (Fig. 4a). Hence, the Ea values for the UFC curing reaction ranged from 77.33 to 78.89 kJ/mol, throughout the conversion, while the UF adhesive exhibited Ea values in the range of 74.44–77.93 kJ/mol. The lower Ea for the pure UF adhesive complies with the results of the average Ea calculated by the Kissinger model (Table 3). In addition, the Ea values remain relatively uniform throughout the conversion. However, in the range from 30–60% of conversion, the Ea values slightly decrease for both adhesive systems. It is followed by an increase in Ea towards the end of conversion, which is especially pronounced for the UF system. According to Fig. 7, it is obvious that curing activation energy was different at different conversions, mainly because the condensation reaction forms various chemical bonds at different decomposition stages.
Since the Friedman iso-conversional method does not involve any mathematical approximation, the obtained activation energy (Ea) values are more sensitive to the changes in the curing mechanism (Yang et al., 2013). Hence, in comparison to the results of the KAS model, the Friedman iso-conversional method shows more complex curing behaviour for both the UF and UFC adhesive systems, i.e. more pronounced changes in Ea values (Fig. 7b). It is interesting that both the KAS and Friedman methods show an increase in Ea values at the end of conversion for UFC adhesive system. This suggests that the final stage of the UF adhesive curing reaction is less dependent on kinetic factors and that the diffusion-controlled reaction becomes dominant (Gao et al., 2006). However, this behaviour is more pronounced for the pure UF adhesive with the Friedman model depicting this effect in more detail.
In the range from 30–60% of conversion, the curing reactions UF and UFC adhesives show a decrease in Ea values from 77.18 to 65.33 kJ/mol, and from 78.23 to 72.07 kJ/mol, respectively (Friedman model). Afterward, starting from the 70% conversion, the Ea of the pure UF adhesive consistently increases towards the end of the conversion, finally reaching the value of 87.72 kJ/mol. At the same stage of reaction, the UFC system also shows an increase of Ea values, but with an abrupt drop at the very end of the curing reaction. The maximum Ea value for the UFC system was 81.98 kJ/mol. According to the Friedman iso-conversional method, the most interesting aspect is that in the final stage of reaction, above 90% of conversion, the curing reactions in the UFC adhesive showed a decrease in Ea values, with the Ea value for UFC becoming lower than for pure UF (Fig. 7b). It might be due to the fact that in the deficiency of formaldehyde in the final stage, phenolic compounds in BSC can react and promote additional crosslinking, leading to more stable resin structure.
The application of the Friedman iso-conversional method confirmed the results of the KAS model in terms that the addition of beech condensate resulting in higher values of activation energy (Ea) throughout most of the conversion range, with pure UF adhesive surpassing the Ea of the UFC system except above the conversion level of 90%. Such activation energy dependence on conversion, found in both iso-conversional models, favours the pure UF adhesive, which requires less activation energy for the curing reaction. This is, however, in contrast with the results of peak temperatures at various constant heating rates (Table 3), showing slightly lower values for the UFC adhesive system, thus suggesting a possible catalytic effect of the condensate addition. This discrepancy, primarily on the side of activation energy, may be explained by the change in the UF adhesive reaction mechanism caused by the addition of condensate.
The KAS and Friedman iso-conversional methods were also used to calculate the curing time for both the UF and UFC adhesive system under isothermal conditions (Fig. 8). Figure 8a, b, and c present the time predictions obtained by the KAS model at the arbitrary temperatures of 60, 90, and 120°C, while Fig. 8d, e, and f illustrate the time prediction diagrams obtained by the Friedman model.
According to both KAS and Friedman prediction models, at low curing temperatures (60°C), the addition of beech condensate prolongs the curing time of the UF adhesive. However, both models indicate that increasing the curing temperature offsets this retarding effect. Moreover, the Friedman prediction model for 120°C (Fig. 8f) clearly demonstrates a reduction in curing time within the conversion range of 30–80%, thus suggesting that the addition of beech condensate could have a catalytic effect on the curing reaction of UF adhesive at higher temperatures.
This behaviour could prove itself significantly valuable in the context of UF adhesive application in real processing conditions. When mixed with hardener, the UF adhesive is a highly temperature-dependent system, required to cure fast with the application of heat (e.g. in the hot press operation, at temperatures that could exceed 200°C). On the other side, the extended pot life of the prepared UF adhesive at lower temperatures may also prove beneficial, since it provides additional flexibility in the adhesive application.
In respect to the findings of this research, the liquid beech condensate did show positive effects on the UF adhesive cure, when used as a water substitute in the adhesive preparation step. Therefore, there is an opportunity for coupling the two different technological processes − the steaming of beech wood and the WBP production) − by using the by-product of one process as a raw material for another. In addition, the utilization of beech wood condensate to modify UF adhesive offers significant environmental and economic benefits. It addresses the problem of the substantial amount of condensate generated during common steaming processes in the wood industry. By repurposing this condensate to modify adhesives, it can effectively and economically transform what would otherwise be wastewater into a valuable resource. This approach eliminates the need for additional and costly post-treatments of liquid condensate since it will be completely utilized for adhesive preparation in the process of wood-based panel production, and finally sequestered in the panel products. This not only addresses environmental concerns but also contributes to the circularity of the wood industry as a whole, aligning with principles of sustainability and resource efficiency.