1. I'm pretty sure he means that the step for the separation COULD potentially last up to 45 mins (or if you are unlucky even more).
So the thing is: You throw your non polar solvent (NPS) on it and then you want it to separate obviously. But if the solution is not filtered beforehand and still contains small particles flowing in it or if you shook it wayyyy to hard, then it may take much longer to separate.
This number shall only tell: if you don't feel that it has fully separated, you should wait further. Because an unfinished separation means potential loss of product.
But my experience is that it would normally only take like 5 minutes.
2. So you mean a plastic syringe instead of a glass one? Or what other material? I dont know the protocol, but all I can say is that plastic will have no impact, so it will be fine.
3. It says to repeat the pulling with the NPS, because it is very likely that the first pull will not get 100 % of it out. My observation: Pull 3 will already have taken nearly everything out. But of course you want ot be safe.
My recommendation: Keep every pull on its own and check at what pull you did not get anything further. But of course this numbers that you will aquire will only be usable when you do the next time exactly the same. But if you got your most efficient method, then you wont change your method anyways. So it is nice to know if 3 or 4 are already enough, instead of doing 5 or 6 all the time.
And to the time question: He means that you should again wait to be safe that it fully separated.
If the 1. pull was not separated 100% but just 90%, its not that bad, because the second pull may get that last 10 %. Only the very last pull should be sure to get a 10000% separation.
Also I recommend to heat your NPS to its boiling temperature. Naphtha always has a broad range for boiling, which normally starts at 60 °C. So using your Naphtha AND your soup at 60 °C will make you needing only the least pulling steps. It may become more inpure, but better get everything out fast and then recrystallizing afterwards. This will be the most efficient process!
GL