(404g) Resorcinol-Added Phenolic Resins for Carbon/Carbon Composite Fabrication: Kinetics for Curing Prediction and Application

Phenolic resins are thermosetting polymers that have a wide range of applications. Because of their high carbon yield and aromatic nature, they are commonly used as carbon precursors in the fabrication of Carbon/Carbon (C/C) composites [1,2]. In this DoE-funded research project, a novel C/C composite with an embedded channel network is proposed for use in concentrated solar power (CSP) gas receivers. CSP systems work by focusing sunlight into a single receiver, where a heat transfer fluid is heated and later used to produce electricity. Recent developments show that the use of micro-channel nickel-alloy based receivers has the potential to achieve up to 90% global efficiency, based in simulated solar testing [3]. These traditional metal alloy receivers do however suffer from thermal fatigue due to daily startup operations, which currently limits their operating temperatures to around 530-750 °C. On the other hand, C/C composites have a much lower coefficient of thermal expansion, which can virtually eliminate thermal fatigue problems. In addition, these composites have high strength, maintain a comparable thermal conductivity and are over four times lighter, which can contribute to the decrease in installation costs. In this work, for the first time, the curing behavior due to the addition of resorcinol (10-30 wt%) in a commercially available phenolic resin (GP^® 5236H – GP is a trademark licensed to Georgia-Pacific Chemicals LLC) is investigated. This specific resin with additions of resorcinol has never been investigated before and so far, no kinetic study has been done on resorcinol addition.

The fabrication of C/C composite modules with micro-channels starts with the production of a prepreg [2], which in this work is done by combining a phenolic resin with carbon fibers. The quality of the prepreg dictates the properties of the final C/C composite and thus this initial fabrication step is very crucial. The micro-channels (which will be used to transport the heat transfer fluid) are implemented during this early stage by using a VaSC (vaporization of sacrificial component) technique [4], where 3D printed PLA (poly lactic acid) micro-channels (diameter between 0.25 to 0.75 mm) are placed between layers of carbon fiber and act as a template for the finalized channels. The 3D printing approach allows for a very versatile channel network to be used with an architecture that can be optimized for heat transfer. Following the curing of the resin, the sacrificial PLA micro-channels are depolymerized at 295 ËšC under vacuum, leaving a cleared channel network in their place.

Curing of phenolic resins comes from reactions between phenol and formaldehyde, forming monomers and dimers with different configurations [1]. These curing reactions generally require heating, starting around 60 °C and finalizing around 150 °C. In this work, the resin is cured at temperatures up to 150 ˚C in a vacuum bag and in-house designed autoclave setup. Because the T_g (glass transition temperature) of the 3D printed PLA channels is around 54 ˚C, the use of high pressure from the autoclave (160 psi) during curing can cause the channels to warp or even be completely compressed during heating. It was then theorized that B-stage curing (gelling) the resin below the T_g of PLA could provide enough resistance to the compaction forces while allowing the resin to set and be free of voids. However, pure phenolic resin starts curing only at 60 ˚C, and gelling (B-stage cure) without excessively long times can only be achieved above 80 ˚C, well above PLA’s glass transition temperature. To increase curing rates and be able to achieve gelling at lower temperatures, the use of catalyst has commonly been adopted. The addition of these components can however be detrimental to the performance of the finalized C/C composite, particularly when it comes to carbon yield and the addition of species that can act as active sites for oxidation. An alternative to catalysts is the addition of a modifier, such as the resorcinol evaluated in this work. Resorcinol was chosen because it has a very similar structure to phenol, with an additional hydroxyl group that makes it more reactive and consequently can decrease curing times [5]. Its similar molecular structure also means that its addition to phenolic resins has essentially no impact on carbon yield, being completely incorporated into the resin matrix with no contamination. Thus, we hypothesize that resorcinol-added phenolic resins can B-stage cure below the glass transition temperature of PLA, preventing channel warping during the autoclave curing of prepregs.

In this work, curing kinetics of pure, 10, 20 and 30 wt% resorcinol-added phenolic resins are investigated with DSC (differential scanning calorimetry) analysis with an aim of determining optimal concentrations and heat treatment times to produce void-free prepregs that maintain the channel’s geometry. Constant heating rates of 0.5, 1, 2.5 and 5 ◦C min^−1 were used to obtain thermograms to fit different curing models. To evaluate the accuracy of curing models, isothermal DSC was done to obtain experimental curing times under constant temperature conditions. The Kissinger method, which depends only on peak temperature with a constant activation energy (E_a) assumption, provides the least accurate results. Isoconversional mathematical approaches, such as the Kissinger-Akahira-Sunose, the Flynn-Wall-Ozawa and the Vyazovkin methods provided the best predictions, attributed to their adoption of an activation energy that depends on the extent of cure α (E_a(α)). Finally, Thermogravimetric Analysis (TGA) revealed that resorcinol decreases the curing times by up to 71% without lowering carbon yield.

Based on the kinetic study performed, prepregs with resorcinol-added phenolic resins were prepared using different concentrations. Microscopy analysis was done to evaluate channel fidelity and void content. For prepregs with 30 wt% resorcinol addition, curing took place too fast, which essentially did not leave time for the gases from the curing reaction to evolve out of the composite layup. On the other hand, the addition of 10 wt% resorcinol produces the best results with void-free prepregs at lower gelling and curing times when compared to pure phenolic resin prepregs. Microscopy revealed that in this configuration, PLA channels can be implemented in the prepreg with exceptional fidelity to the original print.

Subsequent processing of these resorcinol-added phenolic resin prepregs include several steps. Carbonization at 1000 ËšC under nitrogen converts the resin matrix into a continuous carbonaceous matrix, followed by a chemical vapor deposition densification with methane at 1000 ËšC to increase matrix carbon content. Graphitization is promoted at 2200 ËšC under argon to produce the finalized C/C composite with an embedded channel network to be used as a modular gas receiver in CSP systems.

References

[1] Pilato, Louis. “Phenolic Resins: 100 Years and Still Going Strong.” Reactive and Functional Polymers, vol. 73, no. 2, Elsevier, Feb. 2013, pp. 270–77.

[2] Chung, D.D.. “Carbon composites: composites with carbon fibers, nanofibers, and nanotubes”. Butterworth-Heinemann. 2016.

[3] Zada, K. R., Hyder, M. B., Drost, M. K. & Fronk, B. M. J. Sol. Energy Eng. 138, 61007 (2016).

[4] Esser-Kahn, A. P., Thakre, P. R., Dong, H., Patrick, J. F., Vlasko-Vlasov, V. K., Sottos, N. R., Moore, J. S., & White, S. R. (2011). Three-Dimensional Microvascular Fiber-Reinforced Composites. Advanced Materials, 23(32), 3654–3658.

[5] Chow, S. “A Curing Study of Phenol-Resorcinol-Formaldehyde Resins Using Infrared Spectrometer and Thermal Analysis.” Holzforschung, vol. 31, no. 6, 1977, pp. 200–05.