Study Of The Confinement Effect On Glass Transition And Diffusion Behavior Of Polymers

The effects of nanoconfinement on the glass transition phenomena of glass forming liquids have been widely investigated since Jackson and McKenna first showed a depression for glass formers confined in nanopores in1991. But there are still a lot of controversies about how and whether the glass transition temperature changes from bulk value for polymers under nanoconfinement, and researchers still can’t reach a consensus view. Most of the existing studies have focused on polymer thin films, however, more and more attentions have been paid to investigate the glass transition and its dynamic behavior for polymer under different confining geometries in recent years. Here we introduced different polymer melt into the AAO nanopores by capillary force under high temperature, and studied their glass transition behaviors under confinement by traditional DSC method.Firstly, we investigated the glass transition behavior for polystyrene (PS) confined in AAO nanopores. Two distinct calorimetric Tgs were detected for PS with low moelcular wight confined in cylinderical nanopores. One is lower than the bulk Tg and the other is higher. The values of two Tgs show the nanopore size dependence, the higher one increases as the pore size decreases, while the lower one behaves on the contrary. By Monte Carlo simulations, we suggest that these behaviors should be related to the radial density distribution of the polymer chains in AAO nanopores. Interestingly, we found that Tg for PS recovered to the bulk value by simply removing the AAO walls, even the polymer was still in glassy state. This behavior unambiguously reveals that the nanopore wall is the dominant factor for the Tg variations. Besides, we modified the nanopore surface by chemical method, but we found it almost has no impact on the glass transition behavior. For PS with molecular weight higher than Mc, they show a complicated glass transition behavior, which has certain cooling rate dependence, this phenomenon maybe due to the entanglement between polymer chains. In addition, by the isothermal annealing experiments for liquid nitrogen quenched sample and sample that cooled from melt with10K/min, we found that PS confined in AAO nanopores is under a non-equilibrium state with higher energy compared with its bulk state.Secondly, we also investigated the glass transition behavior for polymethyl methacrylate (PMMA) confined in AAO nanopores. We found that oligmer PMMA shows three distinct calorimetric Tgs when it’s confined in native AAO nanopores, and this behavior could be explained by a three-layer model. The microstructure change of PMMA shows no impact on this glass transition behavior. Similar to the situation of PS under confinement mentioned above, once the AAO pore walls were removed, the Tg of PMMA nanorods recovered to the bulk state. Besides, we investigated the effects of cooling rate and chemical properties of AAO pore surface on the Tg behaviors for PMMA under confinement. For PMMA confined in native AAO nanopores, the Tg behavior of PMMA changes from the three-layer model to a two-layer model as the cooling rate decreases. While, for PMMA confined in modified AAO nanopores, PMMA only shows a single Tg under normal cooling rate; but it recovers to the two-layer model as the cooling rate continually decreases. These experiments reveal that, for PMMA confined in AAO nanopores, the two-layer model is a more stable state and closer to the equilibrium state compared to the single Tg behavior and the three-layer model.At last, we prepared a polystyrene/poly(phenylene oxide)(PS/PPO) demixed nanoparticle blend by the solution-extracting method, and investigated the diffusion process of liquid PS into glassy PPO by traditional DSC method. The diffusion experiments were promoted by annealing the l-PS/g-PPO system at several temperatures below the Tg of PPO matrix, and by tracing the Tgs of the PS-rich domain behind the diffusion front, we investigated the realtionships of chemical composition of PS-rich domain and the advancing diffusion front with the the elapsed diffusion time. We found that PS weight fractions in the PS rich domain vs. the logarithmic diffusion time obey the Boltzmann equation, and the curves of different diffusion temperature can be converted a master curve by time-temperature superposition (TTS), the shift factor and the diffusion temperature obey the Arrhenius equation. In addition, we assumed a core-shell two layer model and found that the calculated advancing diffusion fronts show a linear relationship with the square root of elapsed diffusion times, which is a characteristic of a Fickean diffusion mechanism. The Fickean linearity is departed as the diffusion time increases due to the limited liquid PS supply. With the results of rheology experiments, we conclude that the Fickean linearity offset maybe correspond to the variation in dynamics of PS chains during the diffusion process, which change from reptation model to constrained Rouse model. As the samples we parpared are disentangled, and we think this diffusion behavior, in a certain sense, can be expressed as a kind of diffusion process under nanoconfinement.