Unlocking the Power of Flow Chemistry

The Process Diagram

A process diagram shows the interconnections between different elements of your flow setup, together with online and atline measurements. However, the purpose of such a diagram is more than showing connections, but to begin to describe the conditions under which the reaction takes place, with a particular focus on the reactor itself. A top level description will describe the physical environment of the reaction - for example if the reaction is taking place in a tubular reactor then describe the characterisitcs of the reactor (pipe diameter and pipe length) and the characteristics of the fluid (composition, viscosity, density and flow rate). This, alongside the temperature will allow replication of your process, but also if required the ability to describe the conditions more fundamentally - that is the local environment under which the reaction runs, in addition to the global characteristics of the reactor. As well as information to support kinetic understanding, this also provides a framework on which to scaleup processes.

A list of common symbols used in flow chemistry are shown under the flow chemistry equipment page.

Reaction and reactor parameters

As much as possible, you should describe both your reaction parameters and your reactor parameters. Your feed compositions, temperature and pressure are essential descriptions of your setup. Ideally with temperature, you measure a temperature most representative of the reaction, rather than the temperature of the reactor exterior. Since this is a flow reaction, you should also describe your flow rates of the feed materials.For multiphasic systems, the startup procedure can affect which phase is continuous and which is disperse. In regards to your reactor there may be additional parameters such as stirring speed, or conditions associated with for example light (photochemistry) or electricity (electrochemistry) that you need to report. Finally give some thought to describing the reactor - the volume, geometry, mixing elements. Of course you may have additional measurements e.g. within the reaction (e.g. pH, composition) that you take to help you determine the reaction itself.

For single phase reactions, the flow rate of material is enough to specify the hydrodynamics of the reaction.

For liquid-liquid reactions, the way you start the reactor, alongside the flow of material, defines the state within the reactor. For a liquid-liquid system, you can have an organic in water phase or a water in organic phase - what you end up with is due to the relative rates of flow, the startup of the reactor (you generally fill with your continuous phase first) and mixing conditions. Additionally, surfactants can change this behaviour.

For gas-liquid reactions, the gas will ideally be small bubbles within the liquid, giving good mass transport. The gas will occupy a certain volume of the reactor and the liquid the remainder. Knowing the volume fraction helps you establish the residence times of the two phases.

For solid-liquid reactions, the quantity of particles in the reactor is important. Ideally for a well mixed reactor, the particles remain well suspended, but if there is potential for particles to collect in dead-regions in the flow, there maybe pockets of solids. Mix well and this problem goes away!

Of course, it isn't always possible to collect all this data, but at a minimum reporting carefully what you have set up and how you have run the reaction together with visual observations will allow further study if, for example, you wanted to scale-up.

Mixing and micro-mixing in reactors

Mixing is the process of creating a reaction environment where the conditions, for example temperature, concentration or phase fractions, are uniform. Macro-mixing takes place on the length-scale of the vessel; for example the blend time, where the time to reach 95% homogeneity is a macro-mixing property. In flow chemistry, where the kinetics of the reaction can be fast, due to the rate enhancements brought about through high temperatures, the micro-mixing which is at a scale towards the molecular, can also be very significant, as it controls the rate molecules are brought together for a (now fast) reaction step.

For some reactions, poor micro-mixing merely results in a slower reaction (for example a neutralisation reaction), whilst for others the mixing rate can significantly affect the conversion and selectivity of the reaction – the reaction is not only affected by the molecules themselves, but the local environment in which it takes place.

In multiphasic systems, micromixing controls the rate of heat and mass transfer across the phase boundaries and this can also have a significant effect on the reaction. Without good micromixing, this transfer can be slow.

To study the rate of micromixing, competitive multi-step reactions are used – for example the Villermaux/Dushman reaction, where a fast neutralisation reaction which dominates under good mixing competes with a parallel but slower redox reaction. The latter gives iodine as a product allowing detection through spectrophotometry. In multiphasic systems, examining the transfer between two phases can give an indication of the micromixing.

Reactors with active mixing where energy is added externally (for example fReactor-Classic which has a physical stirrer bar) or those with passive mixing where energy is transferred through the flow (for example a static mixer) will create a better mixed environment than flows dominated by diffusion alone, such as in a laminar pipe flow.

Notes

Within the residence time distribution section there are a number of assumptions about the fluid and the flows.

• The laminar pipe flow reactor assumes the fluid is Newtonian, which leads to a parabolic flow profile. The effect of diffusion is neglected in this analysis - it is a purely convective model.
• For the continuous cstr models, the reactor is assumed to be perfectly mixed - ie if you added a drop of fluid into the reactor, it would immediately be dispersed giving a uniform concentration throughout the reactor. Inherent in this assumption is the mixing is instantaneous and there are no dead-zones or bypass zones (regions of fluid the main flow doesn't access) within these models.

The concept of residence time distribution is well covered in Chemical Reaction Engineering by Levenspiel and from which the models for RTD for the pipe flow reactor and CSTRs are taken. Another classical reference is Danckwerts, P. V. Continuous-flow systems. Distribution of residence times. Chem. Eng. Sci. 1953, 2, 1.