[Urbanium]

I have a very general question, out of pure curiosity:

is it possible to set up an experiment in the NMR sample tube (if the conditions allow us that, by that I mean no gas bubbling, no extreme temperatures, no overly corrosive or reactive substrates or reagents), and follow the reaction in situ from minute to minute by taking proton spectra (the idea is that the sample tube is in the NMR machine all the time)? Of course the quantities of compounds would be such that good spectra can be generated within a single minute or two (i.e. 5-6 scans, no more)?

Would it be possible to plot a 3D spectrum in which first two dimensions would belong to the ordinary 1D spectra, and 3rd dimension would be time? Do you know any software capable to do that?

I'm asking because I'm tired of doing a larger scale procedure in the lab and then running every 10 minutes downstairs to the NMR lab to take new spectra.

[discodermolide]

Yes, such experiments are easily done in an NMR tube, at any temperature depending upon your solvent.
The software from ACD labs can overlay spectra. It's freeware, all you need are the FIDs.

[TheUnassuming]

This is an excellent way to watch a reaction, especially if you are looking at kinetics or similar aspects of the reaction.  Just remember to use d-solvent and that everything is extra clean and dry so you get the best spectra possible.
I've only used a hand full of NMR's in my time, but in every case the stock software used to take the spectra was able to overlay like you want.  The key is usually poking around the software a bit to learn everything it can do for you, which is quite a lot in most cases. 

[Urbanium]

Ok, thank you all for the answers. I realized it's possible to do something like that, I wasn't sure if it's a common practice.

[Elan]

That is such a cool idea... someone make a GIF animation of this if you can