This time I had a harder time keeping the temperatures together, with the top at 91K and the bottom at 84K. This 7 degree difference is still better than the 9 degree difference we had been seeing, but not nearly as good as yesterday. I conjecture that heating things up a bit slower this time was to blame, along with the lack of position offset that was used yesterday.
Since it seems pretty clear that the bottom is performing poorly due to lack of centers, it is probably for the best that the bottom was colder this time. The bottom has received much less dose than the top; it is a feedback loop: the worse a material performs, the less we want to use it, the less dose it gets, the worse it performs. We should be sure to put some good cold dose on this stuff before the next anneal. I expect it will continue to perform poorly until we do.