Multifunctional polymer composites for thermal energy storage and thermal management

Thermal energy storage (TES) consists in storing heat for a later use, thereby reducing the gap between energy availability and demand. The most diffused materials for TES are the organic solid-liquid phase change materials (PCMs), such as paraffin waxes, which accumulate and release a high amount of latent heat through a solid-liquid phase change, at a nearly constant temperature. To avoid leakage and loss of material, PCMs are either encapsulated in inert shells or shape-stabilized with porous materials or a nanofiller network. Generally, TES systems are only a supplementary component added to the main structure of a device, but this could unacceptably rise weight and volume of the device itself. In the applications where weight saving and thermal management are both important (e.g. automotive, portable electronics), it would be beneficial to embed the heat storage/management in the structural components. The aim of this thesis is to develop polymer composites that combine a polymer matrix, a PCM and a reinforcing agent, to reach a good balance of mechanical and TES properties. Since this research topic lacks a systematic investigation in the scientific literature, a wide range of polymer/PCM/reinforcement combinations were studied in this thesis, to highlight the effect of PCM introduction in a broad range of matrix/reinforcement combinations and to identify the best candidates and the key properties and parameters, in order to set guidelines for the design of these materials. The thesis in divided in eight Chapters. Chapter I and II provide the introduction and the theoretical background, while Chapter III details the experimental techniques applied on the prepared composites. The results and discussion are then described in Chapters IV-VII. Chapter IV presents the results of PCM-containing composites having a thermoplastic matrix. First, polyamide 12 (PA12) was melt-compounded with either a microencapsulated paraffin (MC) or a paraffin powder shape-stabilized with carbon nanotubes (ParCNT), and these mixtures were used as matrices to produce thermoplastic laminates with a glass fiber fabric via hot-pressing. MC was proven more suitable to be combined with PA12 than ParCNT, due to the higher thermal resistance. However, also the MC were considerably damaged by melt compounding and the two hot-pressing steps, which caused paraffin leakage and degradation, as demonstrated by the relative enthalpy lower than 100 %. Additionally, the PCM introduction decreased the mechanical properties of PA12 and the tensile strength of the laminates, but for the laminates containing MC the elastic modulus and the strain at break were not negatively affected by the PCM. Higher TES properties were achieved with the production of a semi-structural composite that combined PA12, MC and discontinuous carbon fibers. For example, the composite with 50 wt% of MC and 20 wt% of milled carbon fibers exhibited a total melting enthalpy of 60.4 J/g and an increase in elastic modulus of 42 % compared to the neat PA. However, the high melt viscosity and shear stresses developed during processing were still responsible for a not negligible PCM degradation, as also evidenced by dynamic rheological tests. Further increases in the mechanical and TES properties were achieved by using a reactive thermoplastic matrix, which could be processed as a thermosetting polymer and required considerably milder processing conditions that did not cause PCM degradation. MC was combined with an acrylic thermoplastic resin and the mixtures were used as matrices to produce laminates with a bidirectional carbon fabric, and for these laminates the melting enthalpy increased with the PCM weight fraction and reached 66.8 J/g. On the other hand, the increased PCM fraction caused a rise in the matrix viscosity and so a decrease in the fiber volume fraction in the final composite, thereby reducing the elastic modulus and flexural strength. Dynamic-mechanical investigation evidenced the PCM melting as a decreasing step in ’; its amplitude showed a linear trend with the melting enthalpy, and it was almost completely recovered during cooling, as evidenced by cyclic DMA tests. Chapter V presents the results of PCM-containing thermosetting composites. A further comparison between MC and ParCNT was performed in a thermosetting epoxy matrix. First, ParCNT was mixed with epoxy and the mixtures were used as matrices to produce laminates with a bidirectional carbon fiber fabric. ParCNT kept its thermal properties also in the laminates, and the melting enthalpy was 80-90 % of the expected enthalpy. Therefore, ParCNT performed better in thermosetting than in thermoplastic matrices due to the milder processing conditions, but the surrounding matrix still partially hindered the melting-crystallization process. Therefore, epoxy was combined with MC, but the not optimal adhesion between the matrix and the MC shell caused a considerable decrease in mechanical strength, as also demonstrated by the fitting with the Nicolais-Narkis and Pukanszky models, both of which evidenced scarce adhesion and considerable interphase weakness. However, the Halpin-Tsai and Lewis-Nielsen models of the elastic modulus evidenced that at low deformations the interfacial interaction is good, and this also agrees with the data of thermal conductivity, which resulted in excellent agreement with the Pal model calculated considering no gaps at the interface. These epoxy/MC mixtures were then reinforced with either continuous or discontinuous carbon fibers, and their characterization confirmed that the processing conditions of an epoxy composite are mild enough to preserve the integrity of the microcapsules and their TES capability. For continuous fiber composites, the increase in the MC fraction impaired the mechanical properties mostly because of the decrease in the final fiber volume fraction and because the MC phase tends to concentrate in the interlaminar region, thereby lowering the interlaminar shear strength. On the other hand, a small amount of MC enhanced the mode I interlaminar fracture toughness (Gic increases of up to 48 % compared to the neat epoxy/carbon laminate), as the MC introduced other energy dissipation mechanisms such as the debonding, crack deflection, crack pinning and micro-cracking, which added up to the fiber bridging. Chapter VI introduces a fully biodegradable TES composite with a thermoplastic starch matrix, reinforced with thin wood laminae and containing poly(ethylene glycol) as the PCM. The wood laminae successfully acted as a multifunctional reinforcement as they also stabilized PEG in their inner pores (up to 11 wt% of the whole laminate) and prevent its leakage. Moreover PEG was proven to increase the stiffness and strength of the laminate, thereby making the mechanical and TES properties synergistic and not parasitic. Finally, Chapter VII focused on PCM microcapsules. The synthesis of micro- and nano-capsules with an organosilica shell via a sol-gel approach clarified that the confinement in small domains and the interaction with the shell wall modified the crystallization behavior of the encapsulated PCM, as also evidenced by NMR and XRD studies and confirmed by DSC results. In the second part of Chapter VII, a coating of polydpamine (PDA) deposited onto the commercial microcapsules MC. The resulting PDA coating was proven effective to enhance the interfacial adhesion with an epoxy matrix, as evidenced by SEM micrographs. XPS demonstrated that the PDA layer was able to react with oxirane groups, thereby evidencing the possibility of forming covalent bond with the epoxy matrix during the curing step.