Hi Vivek,

Measuring the leakage current at high temperature would ease and speed up your measurmentes without requiring significant area costs.
Although the leakage current levels will depend on the particular parameters of the process you are working with, I estimate you could easily get a few nanoamps just considering either a bipolar emitter-base junction or a drain-body junction for  - say - W=10u measured at  T=150C.
I use to design for such high temperatures and the leakage currents grow exponentianlly into the nanoamp range above about 100C.

So definitely if reaching those tempeartures for characterizing your leakage current is possible then just by means of a picoampmeter you should be all set.

If measuring at those temps is not an option and you have to do it at room temp I would try your third option with a small cap (a very large cap - say 50pf -  might have an oxide leakage in the order of the few picoamps which might introduce some error in your measurement). You could control the diode anode voltage in such a way that you can charge the cap and bring it to a high reference voltage. Then take such anode down to ground such that it is reverse biased and leaks current from the cap. You could either measure the ramp at the cap node (with a buffer as you suggested)  or put a very simple comparator with some controlled hysteresis in such a way that when that hysteresis is overcome then the comparator trips and you only have to compute the time it took to trip in order to get the acutal slope (and therefore the leakage current).

Hope this helsp
Tosei