202508181023

• Source: [[ @giubertoni2023a Multidimensional infrared diffusion-ordered spectroscopy in depletion mode distinguishes protein amyloids and monomers ]]
• Tags: #DOSY #IR-DOSY #spectroscopy

The idea behind Optical DOSY is relatively simple: build a microfluidic channel with two inlets an done outlet.

1. Send a sample from one inlet, and just pure solution on the other.
2. Stop the flow
3. Continuously measure spectra in the region where there was no sample
4. Analyse the results, considering that the time-axis now has information on the diffusivity of the species present.

Some of the challenges are from the system, and are associated to the integration of the flow cells into standard spectrometers. For example, a crucial limitation is the time it takes for the analytes to diffuse to the observation window. Big particles (or proteins) may take hours to get there.

On the other hand, since the spatial resolution is achieved by creating a rectangular window in front of the channel, the signal-to-noise goes down.

Initial tests with Dialanine and Acetone, show that the 2D spectrum generated by IR-DOSY nicely distinguishes the two species thanks to their different diffusion rates.

Interestingly, the approach works the other way around as well:

If instead of looking at the end of the channel, you look at the original one. If you know what particles are diffusing away, you can subtract that from the signal and know what remains (albeit, without being able to measure its diffusion coefficient.)

Outlook

The same approach can be used with different measurement methods, not just FTIR . Pulsed pump/probe systems, UV-VIS , etc. can all be leveraged to add another dimension to the diffusion measurements (or the other way around).

The biggest challenge is the diffusion-limit. Big particles will diffuse too slowly to make the method experimentally feasible. One way of overcoming the challenge would be to perform the measurement closer to the interface (diffusion time scales with $L^2$, which means going from $4mm$ channel to $400\mu m$ one has a factor 100 in the time it takes to perform a measurement).

However, to be close to the interface it also means increasing the spatial resolution of the readout. Just a slit approach may not be sufficient to ensure a $<100\mu m$ positional accuracy.