«CURE KINETICS OF WOOD PHENOL-FORMALDEHYDE SYSTEMS By JINWU WANG A dissertation submitted in partial fulfillment of the requirements for the degree of ...»
Figure 5.1 Baselines of southern yellow pine (SYP) at two moisture contents scanned at 10 oC/min in a high pressure crucible.
Comparison among different SYP contents
with the first and third peaks depressed and the intermediate peak 2 increased (Figure 5.2). With southern yellow pine contents at or above 35%, the first peak disappeared and the third peak gradually vanished into a shoulder. Regardless of wood content, all three peaks, when present, occur at the same temperature and have a similar activation energy (Table 5.1). Apparently, time to cure completion is also unaffected by the wood content (Figure 5.2) which has practical implications in terms of utilizing the cure characteristics of neat resin for panel hot-pressing cycle determination. This suggests that the overall kinetics of the main cure reactions is marginally affected by the presence of wood.
70% 1.0 0.8 0.6 0.4
lower temperature, between 60 and 100°C, and the heat of reaction associated with this peak increases with increasing wood content, from 5 % of the total heat of cure at 35% wood content to 15% for a 70% wood content mixture. The activation energy for the new peak is around 50 kJ/mol, approximately half that of the other peaks (Table 5.1). While the origin and reaction underlying the new peak, labeled N peak, cannot be defined with certainty based on these data, a range of reactions can be proposed.
First, as the appearance of the new peak is concommittant with the disappearance of peak 1, similar reactions may underlie both peaks and be simply catalyzed by SYP. In fact, He and Riedl (2004) and he and Yan (2005) observed this new peak as a shoulder in his thermogram of a 50/50 weight % mixture of spruce and PF, which they assigned to a catalyzed substitution reaction in the PF resin.
Table 5.1 Summary of DSC features at different southern yellow pine contents.
0 118 99 130 85 149 98 363 1.00 20 118 97 130 80 147 95 344 0.91 35 74 42 123 82 145 93 316 0.86 50 81 51 131 85 283 0.78
Similarly, this new peak may also represent PF condensation reactions. If this is the case, the activation energy for PF cure reactions is extensively depressed, from 99 to 50 kJ/mol by the presence of wood reaction, supporting previous reports of wood catalytic effects (Pizzi et al. 1994; He and Riedl 2004). The new exotherm may also originate from a new reaction, such as wood-resin covalent bonds as hypothesized early on (Chow 1969). Indeed, Chow (1969) proposed that the activation energy for wood-resin covalent bond formation is one half that required to form a resin-resin bond, which would agree with our measurements of activation energies. With increasing wood content, the greater availability of hydroxyl groups from wood would allow greater proportions of wood-resin bonds versus resin self-condensation which is consistent with the thermograms. Finely ground wood flour provides a large surface area, upon which the PF species may adsorb facilitating their polymerization.
Interestingly, the resin heat of cure also decreases with increasing wood content (see Table 5.1), indicating in accordance with previous findings (He and Riedl
2004) that wood limits the state of full cure for PF resins. A lower state of full cure for PF resins in the presence of wood has previously been ascribed to wood imparting physical separation and lower mobility to PF species by wood absorbing water in resin, hence reducing their ability for self-condensation (He and Riedl 2004).
Wood-resin reactions have also been proposed to release one fourth the heat released by PF self-condensation (Jones 1946; Chow 1969).
The catalytic effect of wood is also evident when evaluating conversion as a function of temperature during a heating scan (Figure 5.3). With increasing wood content, cure develops faster at the low degree of cure (Figure 5.3), whereas at α0.7 there is no difference in PF conversion with temperature between neat resin and wood/PF mixtures.
Figure 5.3 Experimental cure development of PF/southern yellow pine at various wood contents.
Table 5.2 Kinetic parameters by the nth order Borchadt-Daniels method for southern yellow pine mixtures.
0 24 99 1.13 0.93 20 24 100 1.11 0.93 35 13 66 0.85 0.88 50 11 56 0.83 0.87 70 9 49 0.83 0.83 C: wood content level; A: preexponential factor; E: activation energy; n: reaction order; R2: coefficient of determination.
Next, we turn to evaluating the information provided by the various cure kinetic models for wood/PF mixtures. With the Kissinger equation, the individual peaks are evaluated, and the presence of wood does not significantly change the activation energy of the main cure exotherms of PF resin. The Kissinger approach also allows measurement of the activation energy of the new peak, at approximately half that of the other PF reactions (Table 5.1). The ability to monitor individual exotherms is valuable when a mechanistic understanding of the cure chemistry is desired. On the other hand, the nth-BD method gives a single activation energy for the overall cure process (Wang et al. 2006). With nth BD, a decrease in activation energy with increasing wood content from 99kJ/mol to 49 kJ/mol (Figure 5.2) is measured, clearly showing wood’s catalytic effect, while providing no insight on the mechanistic origins for this effect. It is noted that the preexponential factor A (Table 5.2) declines, and if this is interpreted in terms of the probability of collision of reactive species, this is consistent with the lower mobility and lower accessibility of reactive species with increasing wood content.
If a mechanistic understanding of the reactions is desired, then model free kinetics and the Vyazovkin method (Vyazovkin 1997) in particular are likely best suited for understanding the in situ PF resins cure (Wang et al. 2005). Figure 5.4 compares the activation energy changes as a function of conversion for the neat PF 0% E (kJ/mol) 20%
Figure 5.4 Activation energy changing patterns of PF/southern yellow pine mixtures at various wood contents by Vyazovkin method.
and the SYP/PF mixtures. The activation energy of the PF/wood mixtures follows similar patterns to that of pure PF, albeit in the mixture, the activation energy peaks reach lower values and they are delayed at high wood content. At a conversion between 0 to 0.2, the activation energies are significantly depressed with increasing wood content. This confirms that wood catalytic effect occurs mainly at low conversions. At a higher degree of cure, however, the activation energy for all the wood/PF mixtures follow a similar trend to that of the neat PF, and the activation energy measured in all PF/ wood mixtures merge to a same value. Using the KAS algorithm, He & Riedl (2004) also reported a slower increase in activation energy for wood/PF mixture below 10% degree of cure and similar Ea at higher degrees of cure.
As previously observed for neat PF resins (Wang et al.2005), the Vyazovkin method is more sensitive to changes in mechanisms than the KAS method, and in the case of wood/PF mixtures, differences in Ea are more marked with the Vyazovkin method. In any case, wood influence at low conversions from the Ea curves is consistent with the results from Kissinger equation, which established the influence of wood on the low temperature/ low conversion exotherm only (Table 5.1).
Figure 5.5 Experimental cure development of pure PF at 120 oC and model-free predictions at 120 oC and 80 oC by Vyazovkin method for pure PF and 70% PF/southern yellow pine.
Using the kinetic parameters obtained from the Vyazovkin method, the cure development of neat PF resins and wood/PF mixtures with 70% wood content have been predicted under isothermal cure conditions at 80°C and at 120 oC (Figure 5.5).
Furthermore, experimental cure data for the neat resin at 120°C are plotted to demonstrate the validity of the Vyazovkin method (Wang et al. 2005). It is clear that at the lower cure temperature (80°C), which is close to that of the new peak exotherm, wood influences the cure of PF resins; whereas at 120°C the wood has little impact of the cure of PF resin (Figure 5.5). Again, this is consistent with wood’s influence on PF cure, and is more prominent at low conversion and low temperature.
1.0 0.8 0.6 0.4 0.2 surface, and thus does not have as much contact with wood as it would in the form of powder. Similar wood content at 70% level on a small disc surface shows that wood effect on curing development is small, at a level of only about 20% wood content (Figure 5.6). To further investigate the impact of wood on PF cure and cure kinetics, the influence of wood constituents on PF cure was examined at a mixture level of 35
0.00 0.6 40 60 80 100 0.4
Figure 5.7 DSC thermograms of southern yellow pine (SYP) and extracted SYP/PF mixtures at 35% wood content.
Comparison among wood constituents and species
and prompted a small peak (N peak in Figure 5.7) in a low temperature, with around 5% of total area under the new peaks (Figure 5.7). The DSC curves for SYP and extracted SYP were similar, with peak temperatures differing slightly. The one-way ANOVA results and Tukey multiple comparison tests indicated no significant
Figure 5.8 Comparison of DSC thermograms for cellulose, xylan, lignin, and SYP extractives mixtures at 35% substrate content at a linear heating rate of 5 oC/min.
difference for all three peak temperatures at 2, 5, 10, and 20˚C/min. There was also no significant different in the heat of reactions between SYP and extracted SYP. However, the area under the new small peak for the extracted SYP/PF mixture was larger than that for SYP mixture. This indicates that extraction enhanced catalytic effects. DSC thermograms of cellulose/PF and xylan/PF mixtures followed similar trends as PF alone (Figure 5.8), whereas lignin and SYP extractives changed the DSC traces (Figure 5.8). SYP extractives decreased the onset cure temperature and extended the end of the cure to a higher temperature. The cure development for the lignin mixture follows a distinct pathway from the pure resin. Lignin has a similar onset cure temperature compared with pure PF resin, but the cessation temperature extended to a higher temperature (Figure 5.8). Like SYP and extracted SYP, a small new peak appears at a low temperature. This suggests that lignin was one of the contributors to the early catalytic effect of wood.
There is a small but significant difference with ANOVA analysis at α = 0.05 for the heat of reaction between cellulose/PF, xylan/PF mixtures and PF alone (Table 5.3). The reduction of reaction heat may be due to a physical separation effect, which accounts for a 3% loss in heat of reaction as compared with pure PF. ANOVA analysis shows that the PF/lignin mixture released significantly less heat in reaction than any other mixtures. The extent of reduction of reaction heat cannot only be ascribed to physical separation; it may be due to interactions between the resin and lignin. The ANOVA analysis shows that there were no significant differences in the heat of reaction between SYP extractives and pure resin. The reduction in heat reaction due to physical separation by SYP extractives was compensated by reactions between PF and extractives so that the total heat of the reaction did not change. It was also envisioned that this compensation effect might come from decreased pH value due to addition of acidic extractives, which may increase the reactivity of functional groups of PF resins (He and Yan 2005). There is around 14% less heat of reaction for SYP and extracted SYP as compared with pure resin. It is assumed that around 11% reaction heat reduction was due to wood-resin interaction, since around 3% reduction can be ascribed to physical separation as indicated by the lack of catalytic paper cellulose and xylan.
E (kJ/mol) Figure 5.9 Comparison of activation energies with theVyazovkin method for southern yellow pine (SYP), extracted SYP, SYP extractives, and lignin/PF mixtures at 35% substrate content.
Table 5.3 DSC local peak temperatures (T) at 5 oC/min, activation energies (E) of the peaks by the Kissinger equation, heat of reaction (∆H) and ratio of reaction heat to that of pure PF resin at 35% wood contents.
Table 5.3 lists the peak temperatures and corresponding activation energies calculated with the Kissinger equation for various mixtures.
Cellulose and xylan do not change the peak temperature and activation energy as compared with pure PF resin. SYP and extracted SYP prompted a new peak with activation energy half of that disappeared peak 1, whereas the activation energy of main exotherms does not change significantly as compared with pure PF. Lignin decreases the activation energy, while SYP extractives increases the activation energy. The activation energies by nth-BD are summarized in Table 5.4, supporting the thesis that wood, lignin, and extractives decreased the activation energy and catalyzed the PF curing. The model-free activation energy changing patterns are depicted as in Figure 5.9. The activation energy of cellulose and xylan mixtures follows a similar changing pattern with that of pure PF, while the model-free activation energy of SYP, extracted SYP, lignin and SYP extractives follows different patterns (Figure 5.9). These observations corroborate the thesis that cellulose and xylan contribute little to the catalytic effects of wood on PF curing, but that lignin and extractives is responsible for wood catalytic effects.
Table 5.4 Comparison of kinetic parameters by the nth order Borchadt-Daniels method for various wood/PF mixtures at 35% wood contents.
SYP Extractives 14 71 0.82 0.91 SYP: southern yellow pine; A: preexponential factor;