Silica gel should be heated to about 105C, give or take. A toaster oven set to 220F or even 250F is fine. Leave the gel in the toaster oven for a half hour. The color changes much faster than that so you need to leave the gel in the oven for the full time to get all the moisture out. A low temperature bake of 60C is not effective.
As you found out it's possible to ruin silica gel in a microwave. Basically you can't adequately control the temperature and a lot of it gets overheated and destroys the structure of the gel which is responsible for adsorption.
I just put mine in a foil lined tin, spread it out thin, set the timer for about 30 minutes and walk away. Come back after the ding of the bell and leave it there until it cools off. Then pour into a glass mason jar with a rubber seal. It stays fresh indefinitely, until you need it. This has always worked well for me.
For a standad reel size bag, the RH goes to about 10% within 5 minutes, with a spool of filament in it at atmospheric pressure.
I've been looking for some data on drying Silica Gel. I did find a chart and did some calculations based on that. Higher temperatures are definitely better, but even 60C does dry silica gel so it has about 2/3 of the water capacity compared to 50% RH at room temperature capacity, if I read the graphs correctly.
https://www.sorbentsystems.com/desiccants_charts.html various charts and data
Calculations below are based on water capacity of Silica Gel at 50% RH at room temperature of about 25g/100g. At higher RH the higher moisture capacity would make these percentages even larger, but in my use case the RH is generally in the 50% range so this is a more realistic calculation.
60C 140F would dry to about 8g/100g or 32% removing 68%
65C 150F would dry to about 7g/100g or 28% removing 72%
75C 170F would dry to about 5g/100g or 20% removing 80%
100C 210F would dry to about 2g/100g or 8% removing 92%
120C 250F would dry to about 1g/100g or 4% removing 96%
The only reason I'm interested in drying at such low temperatures is the filament dryer I have will do this so I don't have to mess with the kitchen appliances. But clearly drying at 100C or 120C produces a more effective result and shortens the time required. The time required is variable depending on many factors with 0.5-2 hours recommended at the higher temperatures and 6-7 hours at the lower end (but it also depends on Silica packaging and airflow). The filament dryer recommends 65C in the instructions, based on this research I will probably use the maximum 75C setting next time. The filament dryer has good airflow across the drying area so it may be more effective than the temperature alone indicates. Convection ovens are recommended for similar reasons, and occasionally opening the oven door to allow moist air to escape.
I have definitely dried a no longer effective set of silica gel packets at 65C in this filament dryer for 8 hours or so and they immediately dropped to 10% RH on a meter once they cooled off. These little meters are very handy to tell what is going on, whether desiccant is actually still active or not.
Regarding the silica gel packs that I cooked in the microwave and then dried at 65C overnight, they did eventually drop to 10% but they seem to be much slower than the other packs that were not abused in the microwave. So they still work, but not as well as their peers. I don't plan to use the microwave in the future, at least not for Silica Gel. I might consider trying it for Zeolite molecular sieve since oven drying is at fairly extreme temperatures, though we can see on this chart at 350F one can get down to 5g/100g which is still an 80% effective result. So that is probably safer than the microwave as well. Molecular Sieve is useful for drying High Pressure Air (4500 psi). I wouldn't use it for filament.
In doing this research I looked at many sites recommendations, and they do vary a lot from 60C to 120C. But many of them do include microwaving at medium power for some minutes with some rest periods. I still think it isn't controlled enough for good desiccant life. But it is quick and these folks are happy to sell you more desiccant.
There is one advantage of drying the desiccant less than fully. In museum applications they talk about conditioning the desiccant to produce a stable RH at some level that is most suitable for protecting the items in display cases. Minimum humidity is not the goal, constant humidity at a certain range is. In our case filaments like PLA are actually damaged by cycling humidity. Moisture breaks the polymer bonds and drying the filament does not repair them, but prepares the polymer for more moisture damage on the next cycle of increasing RH. So we want low enough humidity but not a lot of drying. So going for the driest desiccant is not really a priority as long as we have enough desiccant for the drying capacity needed.
The Repkord Repbox folks have a series of videos on utub about filament and drying where they have a guest polymer scientist who goes into some of the details in an understandable fashion.
