It has long been observed that the glass transition temperature of a large variety of polymer films decreases with decreasing film thickness. The most frequently cited model, the layer model, assumes that a highly mobile layer exists at the free surface of the films and engenders a reduction in the overall dynamics of the films. It is still not known how the surface mobile layer brings about the observed change in the glass transition temperature. In a recent experiment, the temperature dependence of the viscosity of short-chain polystyrene films supported by silicon was measured, and found to display a good consistency with the experimental glass transition temperature. More importantly, the data was found to be fully accountable by a two-layer model assuming the total flow mobility of the film to be the sum of the flow mobility of a surface mobile layer and a bottom, bulk-like layer. In that experiment, only polystyrene films on silicon with a single molecular weight was studied. In this program, a series of experiments will be carried out to explore the validity of the layer model for a wider range of polymer molecular weights and polymer films with different polymer-substrate interactions. Molecular weight is well-known to have a strong effect on the dynamics of polymers. Specifically, it can bring about qualitatively different dynamical behaviors depending on whether the polymer is entangled or not. At the same time, different polymer-substrate interactions have been attributed to be the cause of the increase in the glass transition temperature with decreasing film thickness observed in some polymer film systems. The goal of this program is to develop an encompassing layer model that would allow an effective way of predicting the viscosity and glass transition temperature of a broad range of thin film systems.
NON-TECHNICAL SUMMARY:
Glass transition temperature is an important physical property of polymers in a myriad of technological applications. It determines when the material softens and eventually loses the ability to hold its shape. Ample evidence accumulated over the past 15 years shows that this property can deviate substantially from that of the bulk when the polymer is made into nanometer thick films. In this program, a physical model will be developed to describe the phenomenon. If successful, the model will enable strategic design of the glass transition temperature of nanometer polymer films, and thereby significantly streamline the adaptation of polymers in nano-scale applications. The major impact of this program would be in the generation of new knowledge and training of personnel in the STEM field. Participants of this program will not only receive training on the state-of-the-art characterization techniques and processing of nanometer polymer films, they will also be exposed to an international environment due to the on-going collaborations between the PI and a number of international groups. Two graduate students will be trained. The PI will also actively recruit undergraduates and high-school students to the program. She has a strong record of engaging junior students in her research.
Recent experiments showed that the properties of polymer nanometer films can be different from the bulk, and cannot be inferred from conventional knowledge. This project aims to develop the new knowledge set governing the non-bulklike viscosity of these films. Previously, we found that the viscosity of short-chain polystyrene films coated on silica can be described by a two-layer model assuming hydrodynamic coupling between a nanometer mobile layer sitting on top of a bulklike layer. The goal of this project is to examine whether similar layer models can explain the viscosity of polymer nanometer films in general. To this end, we set out to verify the layer model in: (1) Polystyrene-on-silica containing sufficiently long polymer chains that they entangle among themselves. In bulk polymer, entanglement is known to cause a fundamental change in the polymer flow behavior. (2) Polystyrene-on-silica containing sufficiently long polymer chains that Rg, the unperturbed size of the polymer (which normally folds up into a yarn-ball-like coil) exceeds the film thickness. (3) A system sufficiently different from polystyrene-on-silica that a different layer model would be needed to model the data. To date, all the planned tasks were completed and two additional issues were addressed (please see the Appendix for details). The activities had led to 11 publications, 23 presentations and training of three Ph.D. students, one exchange Ph.D. student from China, one undergraduate and one high-school student. The postdoc is now a professor of Soochow University, China. All students have gained development in scientific research work. The three Ph.D. students completed their degree and had joined the STEM workforce in the US. The exchange student is transitioning in the PI’s laboratory, planning a career in the STEM field. Appendix: Details of the Key Scientific Findings 1. Model Validity for Entangled Polystyrene-on-Silica We observed that the viscosity of entangled polystyrene-on-silica films increases with time initially and eventually comes to a steady value. This steady value is fully explainable by the same two-layer model found applicable to the unentangled films. These findings imply that the entangled films evolve from a less entangled to a more entangled state, and the surface mobile layer exists at equilibrium and can dominate the dynamics of unentangled and entangled polymer films in a similar way. 2. Model Validity for Polystyrene-on-Silica over Wide Range of Chain-lengths We found that the two-layer model can also provide a good description to the steady-state viscosity of polystyrene-on-silica for a wide range of chain-lengths (or Rg, from 1.3 to 41 nm). While this result shows that the two-layer model is able to describe (mathematically) the thin film viscosity at all the Rgs investigated, we are skeptical about its viability at long chain-lengths. From fittings to the two-layer, we find that the surface mobile layer is about 3 nm thick, independent of the Rg. This implies that for the films with Rg >> ~ 3 nm, the surface chains must reside simultaneously in the surface mobile and bulklike regions. We estimate that such surface chains can be highly unphysically stretched. We are currently working to understand the physical origin of the non-bulklike viscosity of the long chain-length films. 3. Model Validity for Polymethylmethacrylate-on-Silica The above investigations had involved polystyrene-on-silica, which is a representative of the films dominated by the polymer-air interface. We have also studied polymethylmethacrylate (PMMA) coated on silica, a representative of the films dominated by the polymer-substrate interface and exhibit slower dynamics. To incorporate this property, we model the films to be three-layers containing a bulklike layer sandwiched between a mobile top layer and an immobile bottom layer. Such a layer model was found to describe the viscosity of these films well. Importantly, given a sub-set of the viscosity measurements, the model is able to reasonably predict the rest of the measurements, demonstrating its predictive power. 4. Development of a Formal Proof for the Model used to Analyze the Entangled Film Data We worked out a formal proof for the model we had used to analyze the steady-state viscosity of the entangled polymer films. This legitimizes the method we had used to measure the viscosity of these films and strengthens the conclusions arising from those results. 5. Development of a Theoretical Model to Analyze the Free-standing Film Data Free-standing films provide the simplest configuration for studying the influence of a free surface on the properties of polymer nanometer films, because they possess only one kind of interface - the polymer-air interface. All the analyses we had developed before apply only to films that are supported by a substrate surface. In collaboration with Prof. Chi-Hang Lam of HK Polytechnic University, we developed a theoretical description for the dynamics of free-standing films. By incorporating the experimental conditions, we found satisfactory agreement between the model prediction and experiment. With the model, more detailed investigations can be performed in the future.
