Characterization of structural relaxation in inorganic glasses using length dilatometry

The processes that govern how a glass relaxes towards its thermodynamic quasi-equilibrium state are major factors in understanding glass behavior near the glass transition region, as characterized by the glass transition temperature (Tg). Intrinsic glass properties such as specific volume, enthalpy, entropy, density, etc. are used to map the behavior of the glass network below in and near the transition region. The question of whether a true thermodynamic second order phase transition takes place in the glass transition region is another pending question. Linking viscosity behavior to entropy, or viewing the glass configuration as an energy landscape are just a couple of the most prevalent methods used for attempting to understand the glass transition.;The structural relaxation behavior of inorganic glasses is important for more than scientific reasons, many commercial glass processing operations including glass melting and certain forms of optical fabrication include significant time spent in the glass transition region. For this reason knowledge of structural relaxation processes can, at a minimum, provide information for annealing duration of melt-quenched glasses. The development of a predictive model for annealing time prescription has the potential to save glass manufacturers significant time and money as well as increasing volume throughput. In optical hot forming processes such as precision glass molding, molded optical components can significantly change in shape upon cooling through the glass transition. This change in shape is not scientifically predictable as of yet though manufacturers typically use empirical rules developed in house. The classification of glass behavior in the glass transition region would allow molds to be accurately designed and save money for the producers.;The work discussed in this dissertation is comprised of the development of a dilatometric measurement and characterization method of structural relaxation. The measurement and characterization technique is comprised of three main components: experimental measurements, fitting of configurational length change, and description of glass behavior by analysis of fitting parameters. N-BK7 optical glass from Schott was used as the proof of concept glass but the main scientific interest was in three chalcogenide glasses: As40Se 60, As20Se80, and Ge17.9As19.7 Se62.4.;The dilatometric experiments were carried out using a thermomechanical analyzer (TMA) on glass sample that were synthesized by the author, in all cases except N-BK7. Isothermal structural relaxation measurements were done on (12 mm tall x 3 mm x 3 mm) beams placed vertically in the TMA. The samples were equilibrated at a starting temperature (T 0) until structural equilibrium was reached then a temperature down step was initiated to the final temperature (T 1) and held isothermally until relaxation concluded. The configurational aspect of length relaxation, and therefore volume relaxation was extracted and fit with a Prony series. The Prony series parameters indicated a number of relaxation events occurring within the glass on timescales typically an order of magnitude apart in time. The data analysis showed as many as 4 discrete relaxation times at lower temperatures. The number of discrete relaxation decreased as the temperature increased until just one single relaxation was left in the temperature range just at or above Tg. In the case of N-BK7 these trends were utilized to construct a simple model that could be applied to glass manufacturing in the areas of annealing or PGM. A future development of a rather simple finite element model (FEM) would easily be able to use this model to predict the exponential-like, temperature and time dependent relaxation behaviors of the glass. The predictive model was not extended to the chalcogenide glass studied here, but could easily be applied to them in the future.;The relaxation time trends versus temperature showed a definite region of transition between a low temperature state with many relaxations to a high temperature state with only a single relaxation. Evidence was found for the existence of a definitive transition of some kind in the range of Tg possibly relating the idea of a percolation temperature (T*) as defined by Carmi. The results of the measurements showed substantial support for both the Adam-Gibbs interpretation of decreasing entropy towards the Kauzmann temperature, while also displaying trends compatible with energy landscape theory and the idea of broken ergodicity of glass configuration below Tg. In addition effective relaxation energies were calculated and the energy needed for relaxation showed a definite upward trend with decreasing temperature also supporting the idea of reduced entropy and configurational freedom at lower temperatures. The effective relaxation energies are not purely thermodynamic in nature because they also characterize the effects of viscosity and the kinetics of the material that was relaxing. (Abstract shortened by UMI.).