Reply to the comment on “Pressure-induced phase transition of ice in aqueous KOH solution”

Abstract

The conclusions that have been drawn in the previous papers [1–4] regarding the pressure-induced phase transition of ice in aqueous electrolyte salt solutions were drawn on the basis of the compressions of quick pre-cooled samples. I suppose that the main difference in the results between Yoshimura and Handle et al. lies in the (initial) cooling rate of the sample, i.e. path (1) in Figure 1 in the original paper [3]. Supporting evidence for this is that if I cooled the sample solution at a slower cooling rate (e.g. 5 K/min), I found that the samples show a “normal” transition behavior (i.e. Ih-HDA-VII) at the reported pressures, as stated in the previous papers [2–4]. This is probably because the cooling rate should influence the morphology, the dimension and the distribution of ice crystals. In actual fact, methods reported for the rapid quenching of aqueous solutions were developed for the purpose of achieving small ice crystals [5]. The transition of the ice crystal size from macroscopic to nanometer dimensions occurs discontinuously at a critical cooling rate (and also over a very limited temperature range) [6]. I propose that most of the salt segregates at grain boundaries, and such segregation would lead to very small particles of ice in the sample and reduce the size of the ice crystals. It is possible that, this effect could cause the unexpected transition. In the measurements, I used a specially designed cooling apparatus (flow-type liquid N2 cryostat) as shown in Figure 1. First, the whole of a small-size diamond anvil cell (DAC) (SRDAC-KYO-3-1, Kyowa Co. Ltd; screw-type diamond anvil cell, o.d. = 25 mm, height = 55 mm) with the sample solution was directly immersed into liquid nitrogen. In doing so, the overall cooling rate of ∼150 K/min was achieved. The samples were milky in appearance either because of the inclusions of salts or because of the concentrated salt solution. In this procedure, small transient clusters (seeds) may exist in the sample before the transition upon compression. On the other hand, generally if we cool the sample slowly, the solutes are largely rejected into the liquid and do not make the solid solution upon freezing of aqueous solutions. It should be, however, noted that Klotz et al. [7] recently reported that it is possible for salts to enter the ice lattice structure and form the salty ice under high pressure at lower temperatures. Then, the sample was compressed isothermally keeping liquid nitrogen temperature during the in situ Raman measurements