In situ dynamic investigation of rapid structural transformations in supercooled liquids

When a liquid is cooled below its normal melting point it enters a metastable region, in the sense that the crystal has a lower free energy than the liquid. A liquid that exists (in equilibrium) in a certain range of temperatures below melting represents a fascinating system to study and offer unique possibilities to investigate non-equilibrium phase transformation phenomena. Understanding structural transformations in supercooled liquids is a central goal of modern condensed matter physics and a fun-damental problem of non-equilibrium statistical physics. The study of liquid-solid phase transformations in supercooled liquids is in addition fundamentally relevant to a broad range of interdisciplinary research fields, ranging from atmospheric physics to material science. For example, homogeneous nucleation is one of the basic mechanisms for ice cloud formation in supercooled aqueous droplets that determine the microphysics of the earths atmosphere. Rapid solidification processes from a quenched melt provide the route for the realization of a wide range of materials with different structural, chemical, and physical properties. The ultimate ultimate fate of a supercooled liquid depends on the interplay of cooling rate and observation time scale. Usually, a first-order phase transition occurs, i.e., the formation of the crystal, the kinetic of which is governed by homogeneous nucleation and crystal growth. If the liquid is cooled "sufficiently" rapidly in order to avoid crystallization, then a glass may be formed. The glass formation is accompanied by a dramatic slowing down of the dynamics during supercooling, and a quantitative microscopic understanding of this extraordinary behavior represents a major scientific challenge.

In our group we employ microscopic liquid jets expanding in vacuum as a completely novel mean to investigate issues related to the above topics, allowing studies of structural transformations in supercooled liquids on previously inaccessible time scales (Kühnel et al., 2011). A laminar liquid jet that expands in vacuum rapidly cools below melting by surface evaporation at among the highest cooling rates that can be achieved in the laboratory, before it eventually undergoes a first-order phase transition. Owing to the univocal correspondence between the distance along the jet propagation direc-tion (say, z) and time, t = z/vjet, the temporal evolution of the phase transformation process can be efficiently probed in situ by, e.g., light scattering techniques (Fig. 1). With a typical jet velocity of vjet ≈ 100 m s-1 and a sub -10 μm light beam focus size, the distance-time relationship thus implies a direct access to the phase change dynamics at nanosecond resolution. A further unique feature of microscopic liquid jets is the suppression of heterogeneous nucleation sites as a result of the clean, continuously replenishing vacuum-exposed jet surface.


Figure 1
Figure 1. Artist's view of our novel approach for the time-resolved studies of phase transformations in supercooled liquids. A microscopic liquid jet issuing from a capillary orifice into vacuum rapidly cools below melting (TM), until it undergoes, e.g., a first-order phase transition. Eventually, freezing may not start until the spontaneous break up of the continuous cylindrical filament into droplets has occurred. The phase change dynamics is probed in-situ along the jet propagation axis by collecting the light scat-tered off the liquid jet. The time scale is univocally defined by t = z/vjet, where vjet is the jet velocity. The temporal resolution thus ultimately depends on the spot size of the focused probe photon beam and the jet velocity, providing typically nanosecond resolution (Kühnel et al., 2011).

We have recently demonstrated the experimental feasibility of the concept outlined above by investigating the rapid crystallization in the supercooled quantum liquid para-hydrogen (para-H2), the molecular species in the zero rotational ground state (Kühnel et al., 2011). These experiments have been performed at the Instituto de Estructura de la Materia at CSIC in Madrid, Spain, in the group of Prof. Montero, which is one of the leading groups worldwide on high-resolution Raman spectroscopy. Raman scattering offers in fact a particularly valuable probe for investigating the phase change dynamics in liquid hydrogen because quantum effects and the weak anisotropic interaction result in comparatively sharp lines in the Raman spectra (Kühnel et al., 2011).