«CURE KINETICS OF WOOD PHENOL-FORMALDEHYDE SYSTEMS By JINWU WANG A dissertation submitted in partial fulfillment of the requirements for the degree of ...»
predicting cure development across various temperature programs with only two vector parameters. C(β) and Eβ obtained from the isothermal and ramp data were utilized for assessing the MKF predictions on the mechanical cure development during ramp and isothermal cure of PF resins. Using the KAS parameters, Ck(β) and Eβ were extracted from ramp data to predict cure development both under ramp and isothermal temperature, and using Vyazovkin parameters from ramp data to predict cure development under isothermal conditions have been detailed elsewhere with DSC data (Wang et al. 2005). These algorithms also worked for DMA data. Figure 8.9 showed the experiment data at 1, 2, and 3 °C /min and KAS predictions; Figure 8.10 showed the experimental data at 120 °C and predictions by KAS, Friedman and Vyazovkin methods from ramp data. The parameters by Vyazovkin obtained from isothermal data can be used to predict cure behavior under isothermal temperature in the same manner as the Vyazovkin method from ramp data (Figure 8.11). With the MFK time-event method, substituting the parameters Eβ and Ct(β) into Eq. (69), the needed time to reach specific β can be obtained at specific isothermal temperature (Figure 8.11). Visually, all predictions were in agreement with the experiments.
Figure 8.9 Comparison of experimental mechanical degree of cure at 1, 2, and 3 °C /min for aluminum foil-wrapped PF bonded wood joints and KAS predictions from ramp data.
Figure 8.10 Comparison of experimental mechanical degree of cure at 120 °C for aluminum foil-wrapped PF bonded wood joints and predictions with parameters Eβ and C(β) from ramp data.
Figure 8.11 Comparison of experimental mechanical degree of cure at 120 °C for aluminum foil-wrapped PF bonded wood joints and predictions with the parameters from isothermal data.
CONCLUSION Previous research has shown that foil-wrapped PF joints facilitated the detection of gelation peaks on tan δ curves under linear heating regime and cured slowly as compared with un-wrapped counterparts (Wang et al. 2007). It was then hypothesized that cure conditions in a foil-wrapped specimen was similar with those in hot-pressing where moisture loss is inhibited by a bulk of mass volume and platens as well as those in sealed high pressure pans by DSC. Hence, mechanical degree of cure from foil-wrapped PF bonded wood joints was matched with chemical degree of cure determined by DSC. The mechanical cure was found to follow a sigmoidal relation with the chemical degree of cure and was fitted by a typical sigmoid function analogous to a two-parameter Weibull cumulative distribution function. The chemical degree cure at the gelation point as defined by DMA was relatively constant while the chemical degree of cure at vitrification points increased with isothermal temperature or linear heating rates. The maximum in dβ/dα was coincident with the vitrification points.
The magnitude of activation energy represents an important established method of reporting and comparing kinetic data. The mechanical cure development was modeled by model-free models. The activation energies dependence of mechanical degree of cure has been obtained by model-free kinetic methods for foil-wrapped PF bonded joints. The average activation energy by each method was well in agreement with each other and with those by model-fitting methods (Wang et al. 2007). Although physical interpretation of activation energy dependence on mechanical cure advancement was illusive, two vector parameters from model-free kinetics provided a powerful predictive algorithm. The parameters extracted from isothermal data or ramp data all gave a good prediction for cure evolution across various isothermal temperatures. Therefore, after either mechanical or chemical cure development is characterized, the other can be estimated through connection of correlation function between mechanical and chemical degree of cure.
REFERENCES Christiansen, A. W.; Follensbee, R. A.; Geimer, R. L.; Koutsky, J. A.; Myers, G. E.
Phenol-formaldehyde resin curing and bonding in steam-injection pressing. Part II.
Differences between rates of chemical and mechanical responses to resin cure.
Holzforschung (1993), 47(1), 76-82.
Han, C. D.; Lem, K. W. Chemorheology of thermosetting resins I. The chemorheology and curing kinetics of unsaturated polyester resin. Journal of Applied Polymer Science (1983), 28(10), 3155-83.
He, G.; Yan, N. Effect of wood species and molecular weight of phenolic resins on curing behavior and bonding development. Holzforschung (2005), 59(6), 635-640.
Laborie, M.-P. G. Investigation of the morphology of the wood/phenol-formaldehyde adhesive interphase. Doctoral dissertation, Virginia Polytechnic Institute and State University, (2002), Blacksburg, VA.
Malkin, A. Ya.; Gorbunova, I. Yu.; Kerber, M. L. Comparison of four methods for monitoring the kinetics of curing of a phenolic resin. Polymer Engineering and Science (2004), Volume Date 2005, 45(1), 95-102.
Prime, R. B. Thermosets. In: Turi E.A, editor. Thermal characterization of polymeric materials. New York: Academic Press, (1997), p435–569.
Steiner P.R.; Warren, S. R. Rheology of Wood-Adhesive Cure by Torsional Braid Analysis, Holzforschung (1981), 35 (6) 273-278.
Toffey, A; Glasser, W. G. Cure characterization of polyurethanes with lignin and cellulose derivatives. Holzforschung (1997), 51(1), 71-78.
Vazquez, G.; Lopez-Suevos, F.; Gonzalez-Alvarez, J.; Antorrena, G. Curing process of phenol - urea - formaldehyde -tannin (PUFT) adhesives. Kinetic studies by DSC and DMA. Journal of Thermal Analysis and Calorimetry (2005), 82(1), 143-149.
Wang, J.; Laborie, M.-P. G.; Wolcott, M. P. Comparison of model-free kinetic methods for modeling the cure kinetics of commercial phenol-formaldehyde resins.
Thermochimica Acta (2005), 439(1-2), 68-73.
Wang, J.; Laborie, M.-P. G.; Wolcott, M. P. Comparison of model-fitting kinetic methods for modeling the cure kinetics of commercial phenol-formaldehyde resins.
Journal of Applied Polymer Science (2006), accepted.
Wang, J.; Wolcott, M. P.; Laborie, M.-P. G. Model-fitting kinetic analysis of phenol-formaldehyde bonded wood joints. Chapter 7, 2007.
Yu, H.; Mhaisalkar, S. G.; Wong, E. H.; Teh, L. K.; Wong, C. C. Investigation of cure kinetics and its effect on adhesion strength of nonconductive adhesives used in flip chip assembly. IEEE Transactions on Components and Packaging Technologies, (2006), 29(1), 71-79
Two complementary methods of differential scanning calorimetry (DSC) and dynamic analytical analysis (DMA) were used to characterize the cure development of the core and face phenol-formaldehyde (PF) resol resins. Quantitative analyses with H-, 13C-NMR, and gel permeation chromatography demonstrated that the core PF resin has a higher molecular weight distribution than the face PF resin, and thereafter the core PF was labeled as PF-high and the face as PF-low. First, the resins were characterized in a neat state with DSC, in the blends of PF wood at various wood content levels with DSC, and in the thin film between two wood substrates, using DMA and the three-point bending test. The synergy of DSC and DMA techniques could pick up the intrinsic events and features of the PF curing processes, and characterization was comparable between the two techniques. Although the conclusions were based on PF resins, the developed kinetics and methodology should also be applicable to other thermosets.
In the neat state, both PF-high and PF-low resins exhibited two distinct exotherms that shifted to higher temperatures with an increasing heating rate. The PF-high resin reached similar degrees of cure, with ca. 10°C occurring earlier than that required for the PF-low resin. The chemical degree of cure was based on the reaction heat evolution recorded using DSC, and was assumed to be proportional to the crosslinking network formations. To model resin cure kinetics, model-fitting and model-free kinetics were used. Model-fitting kinetics of the nth-order Borchardt-Daniels (nth-BD), ASTM E698 (E698), autocatalytic Borchardt-Daniels (Auto-BD), and modified autocatalytic methods (M-Auto) were evaluated on PF-high and PF-low resins. The nth-BD, E698 and M-Auto methods all produced comparable values of activation energies (around 97 kJ/mol), while the Auto-BD method yielded aberrant values. For dynamic cure prediction, all model-fitting models failed to predict reaction rate, while degree of cure was reasonably well-predicted with all three methods. As a whole, the nth-BD method best predicted the degree of cure for both resins. Due to the limitations of model-fitting kinetics, the Friedman, Vyazovkin and Kissinger-Akahira-Sunose (KAS) model-free-kinetics algorithms were applied to the same DSC data to model and predict the cure kinetics of commercial resoles. Results demonstrated that the model-free kinetics of the Friedman and Vyazovkin methods can provide insight into the cure mechanisms of PF resoles, and that the KAS method can predict cure development under isothermal and linear heating regimes. However, these methods were not as effective for nth-BD predictions.
Although the cure development of neat PF resins can provide useful information for resin formulators, cure development in the presence of wood can elucidate wood-PF interactions, as well as the effects of wood presence on cure kinetics. DSC analysis showed that the curing behavior of the PF resin did not change significantly when wood content was below 20%. When wood content was over 35%, the overall DSC curve shapes and kinetic parameters changed. The wood addition accelerated one reaction and made it occur at a lower temperature; however, the main reactions did not change. Moreover, there was no significant difference in the effects of wood on PF curing behavior among the two species and their extracted counterparts. Additionally, the paper cellulose and xylan hemicelluloses did not change the cure behavior of PF resin at 35% wood content, while lignin and southern yellow pine extractives delayed the cure development. Both southern yellow pine and aspen extractives released similar heat reactions, while all other fillers reduced the reaction heat significantly. This suggests that the resins did not reach the same cure extent in the presence of wood and wood constituents, as compared with PF alone.
The in situ shear moduli of the PF resins were estimated from 0.01 to around 16 MPa during the curing process, and were in a typical rubber range. The storage
loss tan δ after cure were recommended for direct evaluating the performance of wood-adhesive systems using DMA analysis. Theoretically, the ratio R should approach a value of 4. DMA curves. Results showed that the PF-low bonded wood
well as a lower tan δ after curing than the PF-high bonded wood joints. With a similar resin load, the PF-low formed a very thin bond layer as compared with PF-high, suggesting a good interphase response, which is related to good stiffness and low viscoelasticity for cured PF-low wood joints. The transition temperatures of the curing process and cure development could be clearly assigned to tan δ. Vitrification was probed in all samples, while gelation point was only detected for foil-wrapped wood joints under linear heating regime. It was assumed that moisture loss in unwrapped joints muffled the gelation points. The activation energy for gelation and vitrification were approximately 40 and 48 kJ/mol respectively.
The mechanical degree of cure is based on the storage modulus development recorded by DMA. The activation energies using model-fitting kinetics of the autocatalytic, Prout-Tompkins, and Avrami-Erofeev models were similar, and these results were in agreement with those using time events of vitrification with isothermal data. The activation energies obtained from linear heating data by the Kissinger equation were a bit larger than those from isothermal data. The activation energy obtained from these methods under both linear heating and isothermal regime are around 50-70 kJ/mol, which are less than those obtained from the neat resins by DSC (85-100 kJ/mol) and in agreement with those of PF/wood mixtures obtained by DSC.
These results imply that the activation energy of cure processes decrease in the presence of wood.
The activation energy dependence of the mechanical degree of cure was obtained by model-free kinetics of Vyazovkin, Friedaman, KAS and isothermal time methods for aluminum foil-wrapped PF-high bonded wood joints. The average activation energy using each method were in good agreement with each other and with those using model-fitting methods. Two vector parameters from model-free kinetics provided a powerful predictive algorithm. The parameters extracted from isothermal data or ramp data all gave a good prediction for cure evolution across various isothermal temperatures.
The relationship between chemical cure and mechanical cure was correlated with an equation analog to the Weibull cumulative function. Therefore, when either mechanical or chemical cure development was characterized, the other can be estimated. From this relationship, it was found that the chemical degree of cure at gelation was independent of cure regime, while the chemical degree of cure at vitrification increased with cure temperature. At vitrification points, the maximum change rate for the mechanical degree of cure with respect to the chemical degree of