Achieving ignition in inertial confinement fusion (ICF) requires the preparation of high-quality deuterium-tritium (D₂-DT-T₂) ice layers in targets under cryogenic conditions. The target production processes primarily focus on metrics such as ice layer uniformity, roughness and defects. Most ICF theoretical designs and simulations generally assume the ice layer is homogeneous, meaning the distribution of fuels within the ice layer is uniform. The differences in the thermodynamic properties of hydrogen isotopes can cause isotope fractionation during the crystallization process, leading to uneven ice density, which increases the hydrodynamic stability during implosion compression and, in severe cases, can reduce implosion performance. This article theoretically and experimentally studies the fractionation effect during the crystallization of the 1:1 H₂-D₂ system at low temperatures.
Based on Raoult's law, the changes in the composition of gas, liquid, and solid phases under infinitely slow cooling and quasi-thermodynamic equilibrium conditions were analyzed. Theoretical calculations provide the upper limit of fractionation under ideal conditions, with the initial and final deuterium content in the solid phase differing by more than double. In actual crystallization growth, due to non-ideal crystallization, the actual degree of fractionation is lower than the theoretical calculations.
By using a self-developed low-temperature device combined with a specially designed target chamber, high-quality hydrogen-deuterium crystals are obtained based on the melt growth method. First, opaque snowflake-like crystals are formed through rapid freezing (Fig 1-a), then the temperature is gradually raised until almost all crystals in the target chamber have melted, leaving only a few residual crystals. Fine control of residual crystals produces seed crystals (Fig 1-b), and a highly transparent crystal was obtained by slowly growing from seed crystals (Fig 1-c).
Low-temperature Raman spectroscopy tests reveal that H₂/D₂ mixtures exhibit fractionation behavior during crystallization(Fig 1-d). In the vertical direction, from bottom to top, the D₂ component content gradually decreases, indicating phase separation between H₂ and D₂ during cooling process. The isotopic fractionation leads to D₂ crystallizing first in the lower part of the target chamber, while H₂, with a triple point lower than D₂, crystallizes later during cooling process, resulting in higher D₂ content in the lower section of the target chamber.
However, experimental data show that phase separation in the existing mixture is not severe (approximately 1.85%/mm), aligning with theoretical expectations. The results also indicate that there was no phase separation between o-H₂(D₂) and p-H₂(D₂). Compared to previous reports, our results are slightly lower than the experimental results (2.6%/mm) reported by the U.S. LLE Laboratory using infrared absorption measurement methods.