The separation of immiscible liquids is a critical operation in process chemistry since it is the final step in sample purification by liquid/liquid extraction (LLE). In a typical purification procedure, a solvent containing a target molecule and various impurities is brought into contact with a second solvent that preferentially dissolves some of the impurities, causing them to leech out of the original solvent. When the second solvent is physically removed, a partially purified solution of the target molecule is left behind in the first solvent. By carrying out repeated LLEs (with multiple extracting solvents if necessary), virtually all impurities may be removed, resulting in a highly purified solution of the target molecule in the original solvent. LLE is especially important in flow chemistry since it is one of the few flow-compatible purification procedures.
By combining flow synthesis with flow purification, it is in principle possible to prepare highly purified materials entirely in flow without the need for further work-up. In practice, however, LLE is difficult to carry out in flow, especially when sample volumes are low, since gravitational forces are typically too weak to induce the necessary separation of the two liquids. Liquid-liquid separation in flow is therefore a topic of continuing research interest.
Liquid-liquid separation in flow reactors is most effectively achieved using wetting-based methods that exploit differences in the tendencies of the two liquids to wet a surface or membrane. Recently, a new wetting-based method for separating immiscible liquids has been reported, that involves the use of commercially sourced porous polytetrafluoroethylene (PTFE) capillary. Using this approach, continuous separated streams of the two phases may be readily obtained over a broad range of flow conditions using a wide variety of liquid–liquid combinations. However, in order to achieve perfect separation, an appropriate pressure difference must be established between the two outlets. The need to find and maintain just the right pressure to achieve complete separation has inspired the development of an optically monitored motorized needle valve.
The use of the automated needle valve to dynamically control the pressure difference between the two outlets has yielded promising results. In a preliminary paper , researchers had already reported a fully automated separator that could induce complete liquid/liquid separation within minutes without manual intervention. However, the original device was built from a large number of discrete components, making both its construction and its use difficult. In recent work, Dr. Andrew Harvie (Postdoctoral fellow) and Professor John deMello from the Norwegian University of Science and Technology, Trondheim, Norway, together with Jack Herrington from Imperial College London, proposed several design modifications to reduce construction costs, simplify assembly, and improve the performance of the automated separator, thereby elevating it from a proof-of-principle prototype to a reliable and practical tool for flow chemistry. Their work is currently published in the research journal, Reaction Chemistry & Engineering. 
Technically, the separator relies on the two incoming liquids having different wettability to the wall of the porous capillary, with the more strongly wetting liquid seeping through the porous wall, while the other liquid passes unhindered through the core of the capillary. To achieve complete separation of the two phases, the fluid streams at the separator outlets are monitored optically, and the back-pressure at one outlet is iteratively adjusted using a motorized needle valve until smooth time-invariant signals are recorded at both outlets, signifying complete separation.
The research trio highlighted that the most significant change to their previous system was the integration of the porous capillary into a single block of machined aluminum, greatly reducing the component count and significantly simplifying the assembly process.
In summary, the study reported on a high-performance liquid/liquid separator constructed from a drilled-out block of aluminum, a short length of porous PTFE tubing, and a small number of off-the-shelf components and 3D printed parts. The new separator had a substantially reduced dead-volume compared to their previous design and was consequently less prone to unwanted trapping of air-bubbles – an occasional cause of misbehavior. In a statement to Advances in Engineering, Professor John deMello, the lead author pointed out that their improved separator may be applied to many liquid/ liquid combinations over a broad range of flow rates, can achieve complete separation within minutes of activation, and can maintain separation even after large changes of flow rate. He further noted that the simplified construction coupled with ease of use, low cost, and very competitive performance characteristics made the improved separator a promising tool for flow chemistry.
 James H. Bannock, Tsz Yin (Martin) Lui, Simon T. Turner and John C. deMello. Automated separation of immiscible liquids using an optically monitored porous capillary. Reaction Chemistry & Engineering, 2018, 3, 467.
 Andrew J. Harvie, Jack O. Herrington, John C. deMello. An improved liquid–liquid separator based on an optically monitored porous capillary. Reaction Chemistry & Engineering, 2019, 4, 1579.