OK, you know that Darlington pairs are generally rather slow, right? The propagation delay from one transistor to another is cumulative, but there is also the saturation aspect - and it's difficult to predict how long it will take for a transistor to come out of saturation.
Why don't you tell us what you're really trying to accomplish? Simply driving a coil at 10s of kHz really doesn't say a lot.
Zener diodes might accomplish the function you're looking for. They'll have to be pretty high Wattage though, and have good heatsinks.
