Hi, a peltier cooling device could help. But you have to think about the holder and the set up. You may have to build it yourself. I am not sure you will get a commercial set up for this.

Are you doing ultrafast or nanosecond transient absorption?

For nanosecond TA I've used a Unisoku cryostat in place of the standard sample holder

http://www.unisoku.com/products/CryostatForSpectrophotometer.html

I suppose in principle you could do the same for ultrafast, but the extra windows and 1 cm pathlength sample would probably make alignment difficult and add quite a bit of chirp.

These are very good, but expensive

http://www.linkam.co.uk/temperature-controlled-stages/