Comment on “Following Molecular Mobility during Chemical Reactions: No Evidence for Active Propulsion” and “Molecular Diffusivity of Click Reaction Components: The Diffusion Enhancement Question”

Paramagnetic Relaxation Agents for Enhancing Temporal Resolution and Sensitivity in Multinuclear FlowNMR Spectroscopy

Sensitivity in FlowNMR spectroscopy for reaction monitoring often suffers from low levels of pre-magnetisation due to limited residence times of the sample in the magnetic field. While this in-flow effect is tolerable for high sensitivity nuclei such as 1 H and 19 F, it significantly reduces the signal-to-noise ratio in 31 P and 13 C spectra, making FlowNMR impractical for low sensititvity nuclei at low concentrations. Paramagnetic relaxation agents (PRAs), which enhance polarisation and spin-lattice relaxation, could eliminate the adverse in-flow effect and improve the signal-to-noise ratio. Herein, [Co(acac)3 ], [Mn(acac)3 ], [Fe(acac)3 ], [Cr(acac)3 ], [Ni(acac)2 ]3, [Gd(tmhd)3 ] and [Cr(tmhd)3 ] are investigated for their effectiveness in improving signal intensity per unit time in FlowNMR applications under the additional constraint of chemical inertness towards catalytically active transition metal complexes. High-spin Cr(III) acetylacetonates emerged as the most effective compounds, successfully reducing 31 P T1 values four- to five-fold at PRA concentrations as low as 10 mM without causing adverse line broadening. Whereas [Cr(acac)3 ] showed signs of chemical reactivity with a mixture of triphenylphosphine, triphenylphosphine oxide and triphenylphosphate over the course of several hours at 80° C, the bulkier [Cr(tmhd)3 ] was stable and equally effective as a PRA under these conditions. Compatibility with a range of representative transition metal complexes often used in homogeneous catalysis has been investigated, and application of [Cr(tmhd)3 ] in significantly improving 1 H and 31 P{1 H} FlowNMR data quality in a Rh-catalysed hydroformylation reaction has been demonstrated. With the PRA added, 13 C relaxation times were reduced more than six-fold, allowing quantitative reaction monitoring of substrate consumption and product formation by 13 C{1 H} FlowNMR spectroscopy at natural abundance.

Response to Comment on "Following Molecular Mobility during Chemical Reactions: No Evidence for Active Propulsion" and "Molecular Diffusivity of Click Reaction Components: The Diffusion Enhancement Question"

In their Comment (DOI: 10.1021/jacs.2c02965) on two related publications by our group (J. Am. Chem. Soc.2022, 144, 1380–1388; DOI: 10.1021/jacs.1c11754) and another (J. Am. Chem. Soc.2021, 143, 20884–20890; DOI: 10.1021/jacs.1c09455), Huang and Granick refer to the diffusion NMR measurements of molecules during a copper-catalyzed azide–alkyne cycloaddition (CuAAC) “click” reaction. Here we respond to their comments and maintain that no measurable diffusion enhancement was observed during the reaction. We expand on the physical arguments presented in our original JACS Article regarding the appropriate reference state for the diffusion coefficient and present new data showing that the use of other reference states, as suggested by Huang and Granick, will still support our conclusion that the two reactants and one product of the CuAAC reaction do not exhibit boosted mobility during the reaction.

Response to Comment on "Following Molecular Mobility during Chemical Reactions: No Evidence for Active Propulsion" and "Molecular Diffusivity of Click Reaction Components: The Diffusion Enhancement Question"

In their Comment (DOI: 10.1021/jacs.2c02965) on two related publications by our groups (J. Am. Chem. Soc. 2021, 143, 20884-20890; DOI: 10.1021/jacs.1c09455) and another (J. Am. Chem. Soc. 2022, 144, 1380-1388; DOI: 10.1021/jacs.1c11754), Huang and Granick discuss the diffusion NMR measurements of molecules during a copper-catalyzed azide-alkyne cycloaddition (CuAAC) "click" reaction. Here we respond to these comments and maintain that no diffusion enhancement was observed for any species during the reaction. We show that the relaxation agent does not interfere with the CuAAC reaction kinetics nor the diffusion of the molecules involved. Similarly, the gradient pulse length and diffusion time do not affect the diffusion coefficients. Peak overlap was completely removed in our study with the use of hydrazine as the reducing agent. The steady-state assumption does not hold for these diffusion measurements that take several minutes, which is the reason monotonic gradient orders are not suitable. Finally, we discuss the other reactions where similar changes in diffusion have been claimed. Our conclusions are fully supported by the results represented in our original JACS Article and the corresponding Supporting Information.