I'd just leave it mono. Most 'simulated' stereo mixes from a mono source sound off to me.
20-40 ms creates the most realistic effects.
I disagree re: delaying one channel behind the reference channel...I think 20ms (or more) is really excessive and would not try more than a fraction of that (like < 5-6ms max). My reasoning: Since sound generally travels at ~ 1ms/ft, at least as far as calculating the delay from the distance between disparate sources, the creation of an artificial time difference by delaying one source that much behind the other often reduces the clarity of the mono source by potentially introducing phase issues, and it does not really create or do a good job simulating a stereo soundfield. If you were to do an intentional A-B stereo recording with a 20-40' split, you would still have a relative similarity between the time sound from the stage/PA reaches each of the two mics. By delaying a copy of one track 20ms or more behind the reference track, it is more like placing one mic way up front and another 20 to 40' further back from the sound source...not the best way to make a 'stereo' recording.
Apart from that, bit of stereo reverb with a very short decay.
Also tone. The Beatles fake stereo records were famous for that, e.g., more trebly on one side, more bassy on the other.
^ these two are more 'realistic' options IMO if you're really intent on trying to simulate stereo from a mono source (without also introducing phase anomolies, etc.). You could also try a stereo/width adjustment plugin, as suggested by bgreen.
But, I'll ask again...why not just leave it mono?
Or, maybe the best (and determining) question is: which track did you capture - the mid or the fig. 8?