Session L32: Dynamics of Glassy Polymers Under Nanoscale Confinement II

Measuring dynamic mechanical properties of thin polymer films

Polymer dynamics under confinement on nanoscale exhibits substantial deviation from the bulk and has received a great deal of research interests in the past two decades. To date, some representative physical properties of confined polymers such as glass transition temperature and structural relaxation, have been studied extensively. Despite milestone progress in measuring Young’s modulus, creep, and stress-strain curves of thin polymers, mechanical properties of confined polymers are still far from well understood, and remain to be elucidated in many other aspects. Here we demonstrate an experimental apparatus for direct measurement of dynamic mechanical response of thin polymer films. We utilize a sinusoidal pressure wave on a polymer film to induce vibration, and the response of the material is record by a high-speed digital camera. By analyzing the data on stress, stain, and the phase difference in between, we obtained dynamic mechanical properties of polymer films with various thickness.
  

An explanation of how nanoconfinement affects the control of local dynamic relaxation

The dynamics of local segmental relaxation in glass forming systems is linked to fundamental material properties and control variables that reflect the influence of thermodynamics on the relaxation time, tau. In the Cooperative Free Volume (CFV) model the key variables are temperature (T) and free volume (Vfree), which is related to density. The general CFV result is that log tau ~ (1/Vfree)x f(T). We find that our result for f(T) works equally well for the bulk under highly varying pressure conditions, and for all types of confinement. The presence of Vfree in the CFV expression for log tau results in a sensitivity to density changes that becomes strong at low T, and also reflective of the presence of interfaces, whether in film or nanocomposite. In this talk these CFV features will be illustrated through extensive connection with experimental data on a variety of systems.
  

Is there a general compensation rule governing the relaxation dynamics of polymeric surface patterns?

Understanding the mechanism and decay kinetics of patterned polymeric surfaces is a pertinent issue in nanotechnology. Herein, we present the relaxation behaviour of nanoparticle-brush filled imprinted and wrinkled polymer films showing similar compensation effect. Entropy-enthalpy compensation (EEC) effect signifies a linear correlation between the activation parameters of a relaxation process and is routinely observed in the relaxation dynamics of many condensed materials such as molecular additives and glass-forming materials. For the first time, we experimentally observed a full mapped out transition of pattern decay kinetics as a function of temperature (below and above the glass transition of the matrix) and additive concentration by utilizing decay of polymeric surface wrinkles. We observe EEC effect for an athermal and a favorable interacting composite system, thereby ensuring the robustness of the observed phenomenon. As a consequence of this compensation effect, relaxation kinetics for composite wrinkled films is faster than the neat polymer film below the characteristic compensation temperature, T_comp and faster above the T_comp. EEC proves itself to be the underlying mechanism for patterned polymer decay, governing the kinetics of any polymeric surface with patterns.   

A simulation study on nonlinear mechanical responses of glassy polymer nanofibers

The confined polymer glasses, such as glassy polymer fibers, exhibit unique glassy behaviors that differ from bulk polymer glasses. We perform molecular dynamics simulations and study nonlinear mechanical responses of glassy polymer nanofibers under uniaxial deformation. We investigate not only nonlinear mechanical responses but also the dependence of mechanical properties on the strain rates of typical polymer glasses, which were also observed in previous experiments. We find from our simulations that the local stress in the surface regions of fibers is greater than that in the core region of fibers, for which the stress of glassy polymer fibers is greater than that of bulk polymer glasses in our simulations. The distance between monomers in glassy polymer fibers are more stretched than that in bulk polymer glasses. Also, the non-affine displacements in the surface regions of glassy polymer fibers are greater than those in bulk polymer glasses. These results indicate that the microscopic events during deformation relate closely to the mechanical responses of polymer glasses.   

Gradient overlap effects in the thin films

The dynamics of polymer and other glass-forming liquids can exhibit massive gradients in the nanoscale vicinity of interfaces – an effect that is not locally correlated with microscopic changes in structure. A major outstanding question is how these gradients behave in extremely thin films, where gradients emanating from distinct interfaces can interact. Here we report on the results of ultra-thin film simulations probing dynamics and glass formation locally and globally in this gradient-overlap regime. Results point to three general regimes of thin film behavior: one when the film thickness is greater than twice the gradient range; one in which the gradients overlap but do not individually span the film; and an ultra-thin-film limit in which each gradient span fully to the other interface. We report on distinct behaviors in these regimes in terms of the form of the gradients, the presence or absence of a bulk-like domain, and the breadth of the overall film glass transition. These findings have implications for the interpretation of dynamical data in ultra-thin films and for the underlying origin of alterations in dynamics in the nanoscale vicinity of interfaces.   

Dynamical gradients, barrier factorization and interface coupling in thick and thin films of glass-forming liquids

We have developed a microscopic theory for the spatially heterogeneous dynamics of glassy polymer liquids near a vapor interface. The key activated event involves cage scale hopping facilitated by a collective elastic distortion of the surrounding medium. Three coupled physical effects enter for thick films: reduction of neighbors and weakened caging constraints nucleated very near the surface, dynamical transfer of weakened constraints in a layer-by-layer manner into the bulk, and modification of the collective elastic barrier both near and far from the interface. Predictions include an exponential spatial variation of caging constraints and the local glass transition temperature, the near factorization of the temperature and spatial location dependences of the total activation barrier, a double exponential form of the alpha time gradient characterized by a nearly constant correlation length, and position-dependent power law decoupling of the relaxation time from its bulk analog. Generalization of the ideas to thin films predicts nonadditive dynamical gradient interference effects resulting in a further enhancement of relaxation and reduction of the film averaged effective barrier with decreasing film thickness. Comparisons to simulation and experiment will be presented.
  

Thickness dependence of surface glass transition temperature of polymer supported films

The surface glass transition temperature (T_g^surf) of polystyrene (PS) films supported by silica was studied for film thickness, h, from 7 nm to 100 nm by time-of-flight secondary ion mass spectrometry (ToF-SIMS). The width of the surface glass transition, ΔT_g^surf=(T_g^surf+-T_g^surf-) and T_g^surf=(T_g^surf++T_g^surf-)/2 were extracted from the end group intensity as a function of temperature. We found that T_g^surf decreases with decreases h and is ~20 K lower than the glass transition temperature (T_g) of the film for any given h. We attribute this observation to effects of the free surface. On the other hand, ΔT_g^surf increases with decreasing h starting from large h (> ~60nm). We explore possible origins for the noted broadening in surface glass transition of thin films.   

Dynamical phase transitions in amorphous thin films

Despite more than two decades of study, there remain many fundamental unanswered questions about the dynamics of glass-forming materials confined to thin films. In particularly, several experiments show evidence of a qualitative change in behavior upon confinement to sufficiently thin films. For example, the viscosity of amorphous thin films has been shown to exhibit a sharp transition from glassy to liquid-like behavior when film thickness is reduced below 30 nm [Y. Zhang et al., J. Chem. Phys. 145, 114502 (2016)]. Here, we provide evidence that this transition is due to the films inability to support an inactive, low-mobility, dynamic phase near a free surface. Active to inactive dynamical phase transitions have been found for a number of bulk glassy systems by biasing trajectories to low-mobility states using a field s. For a model polymer system, we find that thin films require a dramatically larger field strength than the bulk to reach the inactive phase suggesting that it may be inaccessible for thin enough films. This sheds light on why the dynamics on the surface of amorphous materials is so different from bulk behavior.   

Modeling the Glass Transition in Polymers using a Mean-Field “TS2” Model: Bulk and Thin Films

A phenomenological model is proposed to describe the equilibrium dynamic behavior of amorphous, glass-forming polymers. We postulate that a material can be represented by a lattice of cooperatively re-arranging regions (CRR), with each CRR having two states, the low-temperature “Solid” and the high-temperature “Liquid”. At low temperatures, the material exhibits two characteristic relaxation times, corresponding to the slow large-scale motion involving multiple “solid” CRRs (α-relaxation) and the faster local motion within individual CRRs (β-relaxation). At high temperatures, the α- and β-relaxation times merge, as observed experimentally and suggested by the “Coupling Model” framework. This approach is labeled “Two-state, two (time)scale model” or TS2. We show that the TS2 treatment can successfully describe the low-temperature Arrhenius a-relaxation time behavior described in several recent experiments. We also apply TS2 to describe the molecular-weight dependence of the glass transition temperature in bulk polymers, as well as its dependence on film thickness in thin films.   

Tuning the Effective Viscosity of Random Copolymer films of Styrene and 4-Methoxystyrene by Varying the Copolymer Composition

We show that thickness dependence of effective viscosity, η_eff(h₀), of random copolymer films of styrene (St) and 4-methoxystyrene (MeOS) supported by silica (SiOx) can be easily tuned by using different MeOS concentrations, X_MeOS. When X_MeOS is increased from 0% to 100%, η_eff of nanometer films changes steadily from suppressed to enhanced. At X_MeOS = 10%, η_eff(h₀) displays a non-monotonic variation. We explain our results by suggesting that MeOS interacts more strongly with SiOx than St does and the behavior of η_eff(h₀) is due to competitions between substrate effect and free surface effect. If substrate effect is stronger, η_eff of nanometer films is enhanced. If free surface effect is stronger, η_eff reduction occurs. When the two effects are comparable, our X_MeOS = 10% data suggests that substrate effect dominates in nanometer films but free surface effect dominates in thicker films, resulting in a non-monotonic η_eff(h₀) dependence.   

The Glass Transition Behavior and Structural Recovery of 2D Stacked Polystyrene Nanorods

The behavior of glass-forming materials confined at the nanoscale has been of considerable interest over the past two decades with conflicting results sparking debate. Here the focus is on the glass transition and associated structural relaxation kinetics of 20 and 350 nm stacked polystyrene nanorods using the Mettler Toledo Flash differential scanning calorimeter (DSC). The T_g of 20 nm stacked polystyrene nanorods is depressed by 20 K and 10 K at cooling rates of 0.1 and 1000 K/s, respectively, whereas bulk-like behavior is observed for 350 nm stacked polystyrene nanorods. Structural recovery is also performed on 20 and 350 nm stacked polystyrene rods as a function of aging time and temperature, and the evolution of the enthalpy is followed. The structural recovery rate is found to be enhanced in the case of 20 nm stacked polystyrene rods when compared to the 350 nm stacked rods. The effect of spatial dimensionality on T_g and structural recovery is also evaluated; in addition, a relaxation time map is constructed to better understand the glass transition and structural recovery kinetics. The results will also be discussed in the context of current controversies in the field.
  

The Importance of Density in Segmental Dynamics: Applications of the Cooperative Free Volume Rate Model and Connections with the Density Scaling Approach

A focus in our work is to make predictive connections with real experimental systems. For example, while segmental relaxation data is sometimes collected only at atmospheric pressure, deeper insight is only possible by accounting for pressure-dependent dynamics. This enables the analysis of the contributions due to independent changes in temperature (T) and volume (V), which gives a much deeper representation of the experimental system. It also leads to natural connections with that system's dynamics under confinement. In this talk we discuss our recent work in modeling and predicting alpha relaxation times, τ(T,V), using the cooperative free volume rate model (CFV), in which the system's (well-defined and thermodynamically quantified) free volume controls the molecular cooperativity, and thus the activation energy. In addition to presenting our analysis of experimental systems, we will also feature connections and comparisons with the widely applied density scaling approach. We will show how the key parameters of the two approaches are connected and contrast their predictive power. A strength of the CFV model is its more efficient use of the same thermodynamic information to characterize the form of the corresponding dynamics.