Nearly pure tritium filled capsule implosions to measure the time dependence of mix

We have designed, built, and fielded a unique experiment to measure the time dependence of atomic mix in directly driven plastic capsules. The time history of the burn is measured in two capsules, a reference capsule of 20 $\mu $m thick plastic (CH) and an experimental capsule of the same thickness but with a 1$\mu $m thick deuterated plastic (CD) layer on the inner surface. Both capsules are filled with 10 atm of nearly pure tritium gas containing 0.2 atm {\%} deuterium. Without atomic mix the DT yield of the experimental capsule should be comparable to the reference capsule ($\sim $1 10$^{11}$ with mix). Using the Scannapieco and Cheng model of fully atomic mix we calculated a reference capsule yield of 8 10$^{12}$. First results gave yields of 3 10$^{12}$, confirming the dominant role of atomic mix, but comparison with other plastic capsule implosions suggests that some of the mix is ``chunk,'' or non-atomic in character. In a simple model the ratio of yield rates is proportional to the time dependent amount of atomic mix. Fielding these capsules was a major accomplishment. Because the purest possible tritium was required, a special fill station was built at Los Alamos. Tritium diffusion through the plastic shell required storage and shipment to the University of Rochester at near liquid nitrogen temperatures.