Current limits of actively cooled A140 (and variant) diodes? (below freezing)

Has anyone done testing of extremely well cooled diodes? I'm talking on the order of 15 degrees F. What are the limits of this diode's output with this type of cooling?

I realize that the output is not linear as driving current increases. At what point do gains become negligible? Thanks!

I don't know of any testing data, but I do seem to remember some banter about similar testing not having too much affect on being able to overdrive these diodes. And when you think about it, it's not that the chip heats up too much in the process of injection, but more that the "OC" facet is absorbing too much light and burning itself up. This is because all optics aren't 100% efficient, and when you overdrive the diode, it produces more light, more light means more light absorbed as an inefficiency.