ABSTRACT
The small signals flanking the CH3CN resonance are traces of its 13C satellites. DI-
NMR, which was also first developed by our group [114,115], is significantly different. Rather than starting off with an NMR flow cell full of solvent and suffering the sensitivity losses caused by the ensuing dilution of the injected sample, the flow cell starts off empty in DI-NMR. A sample solution is injected directly into the NMR flow
cell. The spectrometer then acquires the SCOUT scan, analyzes it, and acquires the signal-averaged data (with WET). When finished, the
syringe pump pulls the sample back out of the NMR flow cell, in the reverse direction, and returns it to its original (or an alternate) sample container. Once the flow cell is emptied, clean solvent is injected and removed from the flow cell to rinse it. The solvents used for both the sample and the rinse must at least be miscible, and ideally should be identical, to avoid lineshape degradations caused by magnetic susceptibility discontinuities (due to liquid-liquid emulsions or precipitation of solutes in the flow cell)
or any plugging of the transfer lines. DI-NMR uses very different hardware than FIA-NMR. It uses the syringe pump in a
standard Gilson 215 liquids handler to deliver samples to the NMR flow probe. Samples go directly into the NMR flow cell, through the bottom, via an unswitched Rheodyne injector port. The Gilson liquids handler is capable of accepting a wide variety of samplecontainer formats, including both vials and microtiter plates.
