I'm not sure what occurs in the solid state, but in solution spin diffusion cross-peaks are 180 ° out of phase compared to chemical exchange peaks.

Thank you David.

I know that in solution one can use ROESY to distinguish between spin-diffusion and chemical exchange but since in solid-state NMR the dipolar coupling is working in a different mechanism than the NOE mechanism, I don't think we will be able to get this effect for ssNMR.

There is no simple way to distinguish between spin diffusion and chemical exchange in solids.

One thing that you can do is change the temperature. Chemical exchange processes will change the rate constant while spin diffusion is independent of the temperature. Another option would be to change the field. Again chemical exchange processes will show a strong field dependence through the chemical shift but spin diffusion might also show some field dependence. How big it is is not always easy to predict and one has to be very careful.

Another option would be to replace the DARR by a normal PDSD experiment. If it is chemical exchange, you would expect no changes but spin diffusion might change. Again it depends on the details of the system and a very careful analysis is important.

An interesting question indeed. From the physicist's point of view, spin diffusion occurs via various kinds of spin-spin interactions, in practice prevalently via the direct dipole-dipole interaction, and is activated by random molecular motions, including the tiny ones like molecular and phononic vibrations in solids. The so called chemical exchange, instead, occurs when a spin system "jumps" between two or more Hamiltonians. Despite this difference, both processes cause similar coherence dephasing and magnetization transfer and may in some contexts appear equivalent among themselves as well as with other mechanisms (e.g., "normal" diffusion).

In the context of ssNMR it is probably impossible to distinguish between the two things, unless you bring in a dependence on temperature (either variations in the spectra and/or in relaxation times, including T1). Plain spin diffusion should lead to relatively limited and smoothly varying temperature effects, while chemical exchange effects may vary drastically and in a nonlinear way, provided you are lucky and get close to a proper temperature range. The temperature dependences of both spin diffusion and chemical exchange were well studied and you should be able to find ample references.

Btw, comparing spectra measured at different fields (e.g., 200 and 700 MHz) might also give some insight, since all the involved effects depend on the product * with the tau related to random molecular dynamics. You either change the tau by changing temperature, or you change the omega by changing magnetic field.

In any case, to do it properly, it will NOT be a snap-fingers task :-( Is distinguishing between the mechanisms really necessary for your final goal?

Spin diffusion may be dependent on the temperature, if the temperature will affect the molecular mobility, which could affect the dipolar couplings.

What you say is of course correct but typically, I would expect a stronger temperature dependence for chemical exchange. As I said before, there is no simple solution to this question and one has to be very careful when investigating such problems to make sure that the answer is unambiguous.

one possible way could be, if ones spins the sample fast enough the spin diffusion processes could be quenched or slowed down enough while chemical exchange should still be operative. Depending on the chemical shift difference and the dipolar coupling (in this case density of protons and the P-P couplings) it should be possible to make reasonable estimate of the required MAS frequency.

A series of MAS dependent experiment should lend support to the conclusion

We were doing some simulation for heteronuclear decoupling, where we found spin diffusion being gradually removed by increase in MAS frequency. Lines from abstract " Simulations performed to probe the role of the homonuclear 1H–1H bath show that the robustness of the decoupling schemes improves with the size of the bath. In addition, these simulations reveal that the homonuclear 1H–1H bath also leads to broad baselines at high MAS rates."

Article An analysis of phase-modulated heteronuclear dipolar decoupl...

Experimentally, one can see variation in spin diffusion with MAS. It might be useful. Some experiments may be in my thesis.

https://www.pascal-man.com/book/thesis-RSThakur_april2009.pdf

Thank you so much for all of your detailed answers.

We have decided to try low spinning rate with DARR vs. high spinning rate with PDSD. But we are also considering using different temperatures.