Speciality Chemicals Magazine SEP / OCT 2024 | Page 39

FLOW CHEMISTRY
a Pump 1

�t-

Pump 2
b Pump 1
Pump 2
Automated reaction injector 1

I •

Automated reaction injector 2
Figure 3 – Fluidic set-up with convergent ( a ) & non-convergent ( b ) systems
The number of compounds that can be synthesised depends on the system used . Automated reagent injectors that can combine numerous diverse reactants , an automated sample collector with the capacity for a large number of vials and suitable process control software are all essential .
Efficient , automated reagent injectors allow the introduction of reactant segments faster than the typical reaction ( residence ) time , enabling the movement of multiple reaction segments through the flow system at any time . This minimises the time taken to synthesise the compounds of interest , making the process more efficient .
Automating segmented flow & segment tracking
Manual reagent injectors require calculation of the exact time to switch the injection loop in and out of line to introduce the correct volume of aliquot , which is highly dependent on flow rates . In practice ,

I

Reactant slug

+

t Reactant slug

/

Inert fluid
Product slug

+

Automated reaction Reactant Inert fluid injector 1 slug
Reactant

I + / slugs

-- ----- -------
Automated reaction injector 2

I \

Automated collector

.

;,
.,
�� .....
-:., this means using a stopwatch and performing the injection process in person , which is open to error and not very efficient .
Figure 3 shows how , when multiple reactants are introduced into a flow system , the ‘ segments ’ are required to converge at the mixing junction in order to react . The sample convergence must be checked , and the dispersion within the different injected sample volumes appraised to take advantage of automated compound generation by flow chemistry .
In a simple , single-step , tworeactant experiment performed in a convergent system ( Figure 3a ), the injected samples ( shown in red and blue ) meet at the same time as the streams are mixed , resulting in a collection of reaction products ( purple ). In a non-convergent system , the misalignment of the reactant slugs could , in a worst-case scenario , lead to a lack of reaction and the collection of only starting material and carrier fluid ( Figure 3b ).
Automated collector

� ---- m.:i No reaction � t Reactant slug

Convergence becomes even more important with increasing experimental complexity . For example , where three or more reactant segments converge , or when multiple telescoped reaction steps with multiple reactant segments are involved , advanced automation is required to track every segment accurately , ensuring convergence at the right junctions .
Convergence within a flow system can be investigated by injecting known solutions of compounds or dyes , followed by quantitative analysis using analytical techniques such as UV-Vis absorbance measurements . Although a minimal amount of dilution can be expected , the system is convergent if the measured UV-Vis absorbance is close to the expected value . Detection of the carrier fluid is not expected – or desired – as this would indicate non-convergence .
Minimising effects of dispersion & diffusion
Convergence is not the only consideration when performing a flow chemistry experiment . It is also important to consider dispersion within the system , and how this can be minimised . In continuous flow reactions , the percentage of dispersion is negligible compared with the size of the slug , so the effect is generally quite small ( Figure 4a ).
In an ideal segmented flow regime , the reaction moves through the system with no interaction with the carrier solvent . Diffusion or dispersion of the sample is not expected and the concentration should not change .
However , in non-ideal – i . e . realworld – cases , some dispersion and diffusion phenomena can occur when using a miscible carrier solvent , mainly in the longitudinal direction . Often , the reaction and carrier slugs mix , forming a blurred region at the front and back of the slug reactions ( Figure 4b ). 1 , 2
This creates a solute concentration gradient that can , for concentration-
SEP / OCT 2024 SPECCHEMONLINE . COM
39