In physics, the term 'glass' is used to denote materials which solidify without crystallizing. Besides common window glass, many materials such as certain polymers, organic compounds and metallic glasses fall into this category. Despite the long history of glass research, the molecular nature of glass transition and the associated relaxation processes is still not well understood. We use inelastic and quasi-elastic neutron scattering to study these dynamics on a microscopic length scale.
The most prominent glass-specific dynamics is the α relaxation. It cannot be described by a simple exponential function and does not follow the Arrhenius law. Nevertheless, it is universally observable by methods ranging from rheology to quasi-elastic neutron scattering. To date, there is no complete theory describing all features of the α relaxation over the whole temperature range.
The first quasi-elastic neutron scattering experiments were performed using single wave vectors q and proved that the behaviour of the α relaxation is the same for microscopic and macroscopic length scales (universality). Recent experiments concentrate on the q-dependence of the α relaxation which is expected to reveal its spatial structure, e.g. whether it is of an homogeneous or heterogeneous nature. In this respect a cross-over was found from the validity of the Gaussian approximation at low q to a range where it is invalid. This q-dependence can be consistently explained by a model involving discrete jumps.
Another relaxation phenomenon considered typical for glass formers is the β relaxation. Its temperature dependence is weaker, usually Arrhenius-like. This leads to the fact that at a certain temperature, it merges with the α relaxation. Quasi-elastic neutron scattering has contributed significantly to the understanding of this merging. In particular, it was found that it is not a simple linear superposition but rather a convolution of both processes. In addition, the spatial nature of the β relaxation could be determined from the q-dependence for some polymers.
A recent discovery is that of a fast process in the picosecond range by neutron time-of-flight spectroscopy. This process is postulated by some theoretical concepts of glass transition (mode coupling theory, coupling model, vibration-relaxation model). Nevertheless, it could not be found experimentally for a long time because commonly used relaxation experiments (e.g. dielectric spectroscopy, rheology) do not cover this high frequency range.
This process is immediately adjacent to the low-temperature vibrational anomaly which is another characteristic of glass: the so-called “boson peak”. This feature was originally inferred from anomalies of the specific heat and heat conductivity at low temperatures. Using inelastic neutron scattering, it could be shown to exist in the vibrational density of states. The origin of this excess over and above the Debye density of states is still unclear.
In current experiments, the influence of system size on the glass specific dynamics is being studied. For this purpose glass-forming liquids (salol) and polymers (PMMA, PMPA) are put into controlled pore glasses with pore sizes of 2.5 – 20 nm. Dielectric spectroscopy has shown that this confinement leads to a broadening of the α relaxation time distribution. This could be confirmed by quasi-elastic neutron scattering on the molecular level. A new effect discovered using inelastic neutron scattering was a cut-off in the boson peak density of states. This effect is very pronounced and extends to surprisingly large pore sizes for some systems.