February Q&A with Dr Ulrich Hintermair

Post date: Mar 12, 2021 4:6:36 PM

• Is it realistic to assume that a reaction may continue in the jacketed tubing between the reactor and the spectrometer? How do you avoid this reaction progression?

We have done careful controls to show that the continuous action of flow does not disturb the reaction system, and that the reaction proceeds uniformly across the setup when used in recirculating batch mode. In continuous flow mode this is less relevant as one analyses a steady-state, but here too one can maintain reaction conditions up until the tip of the NMR flow tube if desired.

Q: How can one calculate the molecular weight of a copolymer from a DOSY spectrum?

The determination of its diffusion coefficient from DOSY can be performed in same way as in case of a homopolymer. In order to determine the MW one would need information on the composition of the copolymer, either from qNMR integration or confirmation from mass spec. As with all MW determinations by DOSY there will also be a shape factor to consider, so some knowledge of macrostructure of the copolymer will help the derivation of accurate MW from the diffusion constant and hydrodynamic radius.

Q: Why do we have higher dispersion for the DOSY results in your work compared to the Neve et al work in this diagram?

The slightly larger errors in our DOSY results compared to that of De Neve originate from the fact that we measured online from a flow reactor, whereas the literature data was collected on equilibrated static samples. These were our first attempts at online DOSY and we can now do better in flow (i.e. as good as static)!

Q: Have you tried or there any examples of doing this diffusion based work on a benchtop NMR system that has pulsed field gradients?

Although there are some low field instruments that offer gradients we have not tried any of them yet, partly because they are much less powerful and less accurate due to the higher sensitivity to eddy currents and convection than a shielded, cryogenic high-field instrument. I agree that it is worth trying though, so if anyone has data to compare with ours please let me know!

Q: Can we identify a new chemical compound by this technique?

Indeed, we (and others) have been able to see new reaction intermediates by FlowNMR in many instances - the closer you look the more you find!

Q: When having a reaction vessel under reflux, surely when the pump is taking the solution to go to the NMR tube to analyze and recirculate back to the vessel there is a considerable amount of ptfe tubing that is not insulated at the same temperature as the reaction vessel (lets say from 90 °C to rt). How do you deal with reactions that products/starting materials can potentially crash out in the short amount of period that are in the ptfe tubing that is not thermostatised? The tube can get block while the pump keeps putting pressure until fails.

This is purely an engineering challenge and depends on how important temperature control is for a given application. It is possible to effectively heat-regulate the entire flow path with a combination of active heating and passive insulation so that no significant cold spots exist throughout the system (say within +/- 1 degC) considering the short residence times of the sample in the small ID tubing. Although one would chose to avoid this situation if possible, we have successfully investigated heated reaction systems that were on the verge of their solubility limit. Blockages due to accumulation of solids is a concern in any type of flow chemistry of course, but there are safety features such as pressure relief valves one would use to avoid fatal failures due excessive pressure build-up. These and other engineering aspects of FlowNMR will be described in a forthcoming publication from our group.