LE RAYONNEMENT 𝛾 COMME METHODE DE MODIFICATION DES PROPRIETES

Après avoir étudié la radiolyse des matrices à travers l’étude des espèces radicalaires formées et leur impact sur la formation d’un réseau, cette partie se consacre à apporter une réponse ‘‘Matériaux’’ à la problématique initialement posée qui consiste à limiter le gonflement des matrices PVDF par un électrolyte tout en conservant de bonnes affinités physico-chimiques avec celui-ci. Une première section vise ainsi à déterminer l’impact du rayonnement γ sur les propriétés macroscopiques du PVDF et des deux copolymères p(VDF-co-HFP) et p(VDF-co-CTFE). Une stratégie de réticulation ayant été retenue dans un premier temps pour limiter le gonflement, la caractérisation du réseau formé lors de l’irradiation constitue une part importante du travail effectué. La tenue mécanique, critère important pour le procédé d’élaboration des supercapacités, est aussi étudiée. L’influence de la dose d’irradiation et de la nature de la matrice polymère est présentée. Enfin, une attention particulière est accordée à l’amélioration de la densité de réticulation. Nous verrons comment, à travers l’incorporation d’un agent réticulant, le triallyl isocyanurate (TAIC) nous avons pu atteindre une partie des objectifs initiaux (Partie I). Dans la seconde section, une approche différente relative à la modulation des propriétés de surfaces du PVDF sera présentée. Les polymères fluorés étant connus pour leur résistance chimique élevée, le radiogreffage d’un monomère permettant une modification ultérieure sans toutefois diminuer la stabilité électrochimique du PVDF est envisagée. Ainsi, cette stratégie vise à allier les bénéfices liés à la réticulation sous rayonnement de la matrice et la possibilité de moduler les interactions de surface polymère/électrolyte. Nous décrirons alors les procédés de radiogreffage d’un monomère fluoré, le 2,3,4,5,6-pentafluorostyrene (PFS) et la post-modification mis en œuvre à travers une réaction chimiosélective avec un mercaptoalcool. En particulier, les moyens de caractérisation associés aux deux étapes de modification seront présentés. Enfin, les propriétés de surface seront discutées à travers l’étude de la mouillabilité avec différents liquides sondes, tels que l’eau et l’acétonitrile, ce dernier constituant le milieu de référence en tant que solvant utilisé dans les applications supercapacités (Partie II).

. 𝛾-IRRADIATED VINYLIDENEFLUORIDE BASED POLYMERS IN SITU REINFORCED BY A CROSSLINKING AGENT

Polyvinylidene fluoride and its copolymers containing either hexafluoropropene or chlorotrifluoroethylene were subjected to γ-radiation in argon up to doses of 500 kGy. The macroscopic behavior of the irradiated materials was investigated and correlated to the ability of the polymers to form crosslinks. The network formation was revealed by swelling tests while crosslinking density was evaluated from rheological measurements. The influence of polymer matrix was compared and a part of the study was then focused on the increase of crosslinking efficiency. For this purpose, a chemical crosslinker, namely triallyl isocyanurate (TAIC) which is sensitive to free radical reactions was incorporated into the polymer matrix. The effect of crosslinker concentration for a given dose was studied with the 3 different fluorinated matrix. Significant increases of both mechanical properties and gel content, as well as a relevant decrease of solvent uptake as determined with using two solvents, dimethylformamide and acetonitrile, have been observed until the rate of TAIC reaches 10 wt%. This improvement is especially noticeable in the case of copolymers. Studies of the effect of radiation dose for a given TAIC rate have shown that the properties reach a threshold value from 300 kGy and that the major influent parameter is the crosslinker concentration. INTRODUCTION Owing to their exceptional chemical and thermal resistance combined with their low surface energy poly(vinylidenefluoride) (PVDF) based polymers were applied in a wide range of applications, from the medical field to the energy one.1 Thus the use of fluorinated polymers was considered as microporous membranes for conventional batteries or supercapacitor according to their excellent mechanical properties and electrical stabilities. In such application affinity with the liquid medium is also required as well as a conservation of the dimensional integrity. Indeed, the compromise between a good wetting of the polymer membrane and a limited swelling without any dissolution is one of the most important parameters needed in electrical separators.2 These antagonist properties could be partially reached with PVDF or PVDF copolymers since crystalline regions can provide sufficient mechanical integrity to the membrane.3 Another strategy consist in using polymers blend in which one of the component is not soluble in the electrolyte medium as polystyrene (PS) in the case of p(VDF-co-HFP) /PS blends.4 Crosslinking the membranes was also reported as a relevant strategy to fulfill the aforementioned requirements, by using either a crosslinking agent such as polyethylene glycol dimethacrylate (PEGDMA)5 or by submitting a p(VDF-co-HFP) copolymer to the irradiation of Li ion.6 The latter approach can be applied on materials already processed and does not need a pre-modification of the matrix or addition of further initiator. Despite such advantages, and to the best of our knowledge, no other work upon the crosslinking of PVDF-based materials induced by irradiation has been mentioned in the literature, while the use of ionizing radiation applied on polymers has been intensively studied since the 1960s and reviewed several times.7,8 Radiation involves the generation and reaction of radical species and can be summarized as a competition of mechanisms of crosslinking and chain scission. Based on numerous reported works related to many fluoro-based polymers irradiated in different conditions, Lyons et al.7 have ordered fluoropolymers with increasing crosslink efficiency, and PVDF homopolymer turns out to be comprised between poly(vinylidene fluoride – co – chlorotrifluoroethylene) p(VDF-co-CTFE), the presence of a chlorine atom reducing the crosslinkability and the poly(vinylidene fluoride – co – hexafluoropropene) p(VDF-co-HFP). However with radiation doses commonly envisioned in industrial field (< 100 kGy), the solvent uptake factor, i.e. the weight fraction of solvent contained in polymer gel, remains higher than 1000 % whatever the VDF-based polymers.9 Obviously, if the crosslink density were increased, the solvent uptake would decrease. One way to proceed is by thermal control: as an example, it has been reported that when temperature rises from 0 to 120 °C, PVDF samples lead to higher gel content, from 20% to 60 % when irradiated at a given dose of 150 kGy. 10, 11 However this method is limited because the temperature has to be controlled and tunable during the irradiation step. By the way, the most common method to promote the radiation crosslinking is the use of additional polyfunctional monomers incorporated in the matrix prior to irradiation. Crosslinking efficiency depends on both their miscibility with the polymer and their functionality as shown by Makuuchi et al. 10 for a series of acrylate monomers. The latter are generally di- and tri-functional vinyl compounds but allyl functions are in some cases more suitable due to their lower reactivity with temperature.7 Surprisingly the triallylisocyanurate or TAIC, commonly used to crosslink VDF-based polymers with the use of free radical initiators such as peroxides, 12 has scarcely been employed in combination with γ-radiation. Only few patents13-16 claim the crosslinking of VDF-based polymers containing CTFE and HFP units whereas the works of Forsythe et al.17 proposed a mechanism of crosslinking fluoroelastomers. Herein we present a detailed comparison of the γ-radiation crosslinking of three different fluoropolymers: PVDF, p(VDF-co-HFP) and p(VDF-co-CTFE) and the impact of the introduction of TAIC at various concentrations. In particular, the mechanical properties as well as the uptake of solvent were investigated in depth in relation with the crosslinking density. 

METHODS

 The extrusion process was performed using a twin screw DSM microcompounder. Pristine polymers were sheared with at 100 rpm speed, at 200 °C for the homopolymer and 190 °C for the copolymers, and injected in a 1.5 cm3 mold at 25 °C to obtain dumbbell-shaped specimens. TAIC was pre-incorporated in the polymer matrix using a dissolution process to avoid excessive loss of TAIC during the extrusion process. Granulated polymers and TAIC with percentage of 3, 7, 10, and 13 wt% were extruded at 190 °C and 180 °C for the homopolymer and the copolymers respectively. The mixtures were sheared during 3 min and the rate of TAIC was verified using a thermogravimetric analysis. Samples were stocked at -18 °C to prevent any diffusion and exudation of TAIC. Specimens were placed into Schlenk tubes, vacuumed at ambient temperature, purged with argon several times and finally kept under argon atmosphere. The tubes were then placed into a vacuum bell jar under argon to avoid any oxygen contamination. Irradiations were carried out using an industrial 60Co gamma source at room temperature with dose rate of 0.7 kGy.h-1. An annealing at 100 °C under argon was performed after the irradiation in order to favor the reaction of remaining radicals. 18-20 DSC measurements were carried out by using Q20 (TA instruments). Samples were heated from -80 °C to 200 °C under nitrogen flow (50 mL.min-1) with a rate of 10 °C.min−1, held at 200 °C for one minute and cooled to -80 °C at the same rate before a second heating run. Cristallinity yields were calculated with using the enthalpy change of melting as measured during the second heating run, corrected to the weight fraction of polymer contained in the sample, and compared to the enthalpy change of melting of a 100% crystalline PVDF homopolymer ∆𝐻𝑓 0 (104.7 J.g-1).   γ-irradiated VDF-based polymers in situ reinforced by a crosslinking agent Stress-strain curves in tension were obtained using an MTS 2/M electromechanical testing system at 22 ± 1 °C and 50 ± 5% relative humidity at cross-head speed of 10 mm.min-1 corresponding to a 𝜀̇ of 8.33 10-3 s −1. Sol-gel analyses were performed with two different solvents, i.e. DMF and acetonitrile. About 1 g was taken from dumbbell specimen then introduced in a closeable flask. After determining with accuracy the weight (wi) a large excess (60 mL) of solvent was introduced. Solutions were heated at 80 °C (with DMF) or 60 °C (with acetonitrile) for 48 h to allow complete extraction of the soluble part. Swollen samples were carefully wiped with a tissue and weighted (wg). The solvent was then evaporated under vacuum 24 h at 100 °C to determine the weight of dried gel (wdg) i.e. the weight of insoluble. The gel content (% gel) and the solvent uptake in the gel (% sol) were obtained by following equations: % 𝑔𝑒𝑙 = 𝑤𝑑𝑔 𝑤𝑖 × 100 and % 𝑠𝑜𝑙 = 𝑤𝑔 − 𝑤𝑑𝑔 𝑤𝑑𝑔 × 100 Dynamic mechanical properties were evaluated using a Rheometric Scientific Rheometer ARES. Sample with cross section of about 24 mm² and a length comprised between 10 and 15 mm were cut from the central part of dumbbell-shaped specimens. Measurements were carried out using a 2000 g.cm transducer and a low torque transducer (200 g.cm with a lower resolution limit of 0.2 g.cm) over a frequency of 6.28 rad.s-1 (1Hz) from -80 °C to 200 °C. The heating rate was set at 5 °C.min-1 and the 1% applied strain was taken in the linear domain.