Bulk Melting of Ice at the Limit of Superheating - The Journal of

For energy depositions beyond the limit of superheating (330 ± 10 K) partial melting in two steps is observed and assigned to (i) catastrophic meltin...
0 downloads 0 Views 109KB Size
J. Phys. Chem. B 2007, 111, 11271-11275

11271

Bulk Melting of Ice at the Limit of Superheating Marcus Schmeisser, Hristo Iglev,* and Alfred Laubereau Physik-Department E 11, Technische UniVersita¨t Mu¨nchen, D-85748 Garching, Germany ReceiVed: May 14, 2007; In Final Form: July 13, 2007

The ice-water phase transition after an ultrafast temperature jump is studied in HDO:D2O (15 M) ice with use of 2-color IR spectroscopy. The OH-stretching vibration is applied for rapid heating of the sample and for fast and sensitive probing of local temperature and structure. For energy depositions beyond the limit of superheating (330 ( 10 K) partial melting in two steps is observed and assigned to (i) catastrophic melting within the thermalization time of the excited ice lattice of 5 ( 2 ps and (ii) secondary melting with a time constant of 33 ( 5 ps that is assigned to interfacial melting at the generated phase boundaries. The latter process is found to consume energy amounts in agreement with the latent heat of melting and is accompanied by an accelerated temperature and pressure decrease of the residual ice component.

I. Introduction The ice-water phase transition is the most common structural transition, and considerable experimental and theoretical effort has been invested to study this phenomenon.1-5 The melting usually observed is the so-called heterogeneous melting. The process starts at the surface at temperatures above the melting point.6-9 On the other hand, we recently demonstrated10-12 that shock laser heating of bulk ice can avoid the common surface melting, leading to substantial superheating of the ice lattice. Maximum superheating of bulk ice to 330 ( 10 K was observed.12 The superheated ice state is found to be amazingly stable and persists for more than 1.3 ns. For excitation levels beyond the limit of superheating strong evidence for homogeneous melting was observed.11,12 It is of fundamental interest to understand the different melting processes, at the surface and in the bulk above the superheating limit.13-18 The structural changes by melting are governed by the properties of the hydrogen-bonded network of the molecules.19,20 The relaxation dynamics of the latter include many elementary steps in the femtosecond and picosecond time domain.21-24 Obviously, an improved microscopic understanding of the icewater phase transition requires investigations on ultrashort time scales. Here, we present an experimental study on bulk HDO:D2O ice using the recently developed ultrafast temperature jump technique.10,11 The OH-stretching vibration is applied for rapid heating by intensive subpicosecond infrared (IR) pulses. The same mode is known as a fast and sensitive probe of local temperature, pressure, and structure.25-29 The amplitude and spectral shape of steady-state differential spectra for a fixed temperature rise ∆T vary slightly with initial sample temperature Ti and pressure, whereas the spectral position shows a distinct, nearly linear dependence.10 The features allow the extrapolation of the steady-state spectra to temperatures above the melting point and the measurement of a transient superheating of the sample. The spectroscopic determination of the temperature is also supported by a comparison with the deposited IR energy. It is important to recall the isochoric character of the ultrafast * Address correspondence to this author. E-mail: ph.tum.de. Fax: +(49) 89 289 12842.

hristo.iglev@

temperature jump because of the slower volume expansion of the sample;10,11 in other words, a pressure increase is involved that can also be measured with our technique. II. Experimental Section The subpicosecond infrared spectrometer used in this investigation was described recently.10 The laser system provides IR pulses tunable in the range 1700 to 3700 cm-1 (2300 to 3700 cm-1) with a duration of 0.7 ps (0.9 ps), spectral width of 24 cm-1 (19 cm-1), and typical energy of 10 nJ (5.5 µJ). Numbers in parentheses refer to the pump pulses. The probe beam diameter in the sample of approximately 75 µm is a factor of 2 smaller than that of the pump, so that only the central part of the interaction volume with maximum excitation is monitored. The energy transmission T(ν) of the probing pulse through the excited sample is measured for perpendicular polarization with respect to the linear polarization of the pump beam and compared with the probe transmittance T0(ν) for the blocked excitation beam. The absorption change, defined as ∆OD(ν) ) -log (T(ν)/T0(ν)), is plotted in the figures. We have verified experimentally that accumulative effects related to the repetition rate of the laser system of 43 Hz are negligible. The investigated samples are produced by slow cooling of thin de-gassed layers of isotopically mixed water (15 M HDO in D2O) to 240 K between two CaF2 windows with a spacer and adjusting the desired temperature later on. The HDO:D2O liquid is prepared by isotopic exchange in a mixture of appropriate amounts of D2O (>99.9 atom % D) and H2O (Ultrapur, impurity