There is a bag-type device we used in the lab to make small amounts of dry ice attaching it to a commercial tank of CO2. It was available from Aldrich Chemical in Milwaukee, and probably other scientific supply houses. Terribly inefficient, plus it made a fine flaky material which because of the high surface area sublimed way too fast. We had the stuff delivered in pellets the last 15 years or so and it was way better for acetone baths. This could be an interesting thread, though, cuz the stuff's $13 a brick here locally, and I can't waltz into a place I retired from and grab a couple handfuls for a "government job" anymore. I wonder if bubbling CO2 thru an appropriate liquid shouldn't have the same effect, though this would be a touchy process at best.