Discrete FET driver turns off slowly.

[CaptainCarp]

Run the simulation, looks like about 100mW dissipation in the fet.
Ripple current looks to be about 2mA pk-pk (why so tight?).

 

[FvM]

I would target to 20 or 30% ripple resulting in a few kHz switching frequency. I guess, you already planned steady-state current reduction. The only reason to go for 16 kHz or above would be a solenoid valve producing too much noise in a sensitive application. I remember a previous discussion about peak current rating of µC IO pins. It has been said, that a sufficient series resistor, or even setting the IO pin to low drive strength, is the most simple solution.

It's always instructive to study the home aplliances electronic boards from major manufacturer. You rarely won't find parts, that could be saved somehow.

Run the simulation, looks like about 100mW dissipation in the fet.

In this regard, please notice my previous comment about inappropriate transistor size.