Continuous Photo-Oxidation in a Vortex Reactor: Efficient Operations Using Air Drawn from the Laboratory

We report the construction and use of a vortex reactor which uses a rapidly rotating cylinder to generate Taylor vortices for continuous flow thermal and photochemical reactions. The reactor is designed to operate under conditions required for vortex generation. The flow pattern of the vortices has been represented using computational fluid dynamics, and the presence of the vortices can be easily visualized by observing streams of bubbles within the reactor. This approach presents certain advantages for reactions with added gases. For reactions with oxygen, the reactor offers an alternative to traditional setups as it efficiently draws in air from the lab without the need specifically to pressurize with oxygen. The rapid mixing generated by the vortices enables rapid mass transfer between the gas and the liquid phases allowing for a high efficiency dissolution of gases. The reactor has been applied to several photochemical reactions involving singlet oxygen (1O2) including the photo-oxidations of α-terpinene and furfuryl alcohol and the photodeborylation of phenyl boronic acid. The rotation speed of the cylinder proved to be key for reaction efficiency, and in the operation we found that the uptake of air was highest at 4000 rpm. The reactor has also been successfully applied to the synthesis of artemisinin, a potent antimalarial compound; and this three-step synthesis involving a Schenk-ene reaction with 1O2, Hock cleavage with H+, and an oxidative cyclization cascade with triplet oxygen (3O2), from dihydroartemisinic acid was carried out as a single process in the vortex reactor.


Design of the Continuous Photo-Oxidation Vortex Reactor
Our continuous photo-oxidation reactor was designed to have strong mixing and shear to increase the exposure of the reactants to the light and to promote self -cleaning of the reactor walls. A vortex reactor was chosen in which the reactor is an annular space between two cylinders. When the inner cylinder rotates, a series of toroidal vortices are formed. These are known as Taylor vortices as first observed and studied in 1923 by Taylor. 1 Flow between two concentric cylindrical surfaces where one or both of the cylindrical surfaces is or are rotating is laminar flow when the rotating speed is low and is known as rotating Couette flow. When the rotating speed is increased, the flow becomes unstable if the angular velocity exceeds a critical value and a steady axisymmetric secondary flow in the form of regularly spaced toroidal vortices is generated. This is known as Taylor-Couette flow.
Alternate vortices rotate in opposite directions and generate a complex toroidal vortex pattern in the gap. The stability of flow depends on the rotating speed of cylinder, fluid properties and geometry of the annulus. For the case of a stationary outer cylinder and a rotating inner cylinder, the critical value of the dimensionless Taylor number 2 (Ta) where vortices are generated (for an annulus width that is small compared to the inner radius) in the flow is given by: The Taylor number is a dimensionless number based on these parameters, w here r i , r o , and Ω i are the radius of inner surface, outer surface, and rotational speed of inner surface respectively, as shown in Figure S1. While ν is the kinematic viscosity of fluid in the gap between two surfaces.
Further increasing the rotational speed, many regular flow configurations can exist beyond the first transition to Taylor vortex flow; these flow configurations are generally called wavy vortex flow. They are characterized by a pattern of traveling waves, which are periodic in the circumferential direction and appear just above the critical Taylor number. It is found that the presence of an axial flow will tend to suppress the formation of the Taylor vortices and would therefore reduce the mixing. However, at the flow rates used in this investigation the axial velocity is very low relative to the rotational speed and so any effect would not be noticed. .

Figure S1. A typical Taylor vortex flow
Taylor vortex flow theory can be applied to design the vortex reactor to generate a wavy vortex flow in the gap between the two cylindrical surfaces. If the outer cylinder is transparent the reactants in the annular gap can be irradiated with light and the vortices will provide a rapid mixing in the flow refreshing the reactants exposed to the light close to the outer cylinder. The strong shear over the outer cylinder surface also helps to keep the surface clean.
The concept of reactor is shown in Figure S2 and consists of a stationary outer tube and rotating inner tube. The outer tube is made of glass to allow light to be introduced. Fluid is continuously introduced from a central tube within the rotating tube and exits the reactor from top edge of the annulus gap. Air is sucked into the reactor from the free surface at the top of the annulus gap when the inner tube is rotating due a negative pressure that is generated near the inner rotating surface by the centrifugal force in the rotating fluid region.
S4 Figure S2. The vortex reactor concept A simple test rig was used to investigate the shaft diameter. A common chemical glass test tube, 20 mm diameter, was selected for the outer stationary tube whilst the inner rotating tube was made of stainless steel to give a high shaft stiffness. The gap between the surfaces was then controlled by changing the diameter of inner rotating shaft. Figure S3 shows how the Taylor number varied with the inner diameter of rotating shaft for a fixed outer tube of 20 mm. As the gap is increased, the flow becomes less stable at the same rotating speed and the critical Taylor value occurs at lower rotating speed. A shaft diameter of 18 mm was selected for the final reactor design. Figure S3. Critical Ta number of various rotating shaft diameters use water as fluid.

S5
The reactor was designed as described above and is shown in Figure S4.

Computation Fluid Dynamics (CFD)
Computational Fluid Dynamics (CFD) were used to visualise the flow pattern in the reactor.
Commercial CFD software Ansys Fluent 3 was used for flow simulation of the reactor in 3D.
In order to save the computational time, only 20 degree of test section was simulated with cylindrical periodic boundary conditions. Figure S5a shows

S7
The velocity plots show that the strength of the vortices gives radial and axial veloc ities in the vortices that are typically 10%-20% of the peripheral speed of the inner rotor. The vortices are approximately circular in the axial/radial plane and thus provide a significant convective mass transfer across the annulus that is much greater than would be achieved by diffusion alone in a purely annular flow. The vortices also provide greater shear over the outer walls of the annulus, which would help to keep them clean. Future research will include development of the CFD modelling to include the photochemical reactions and effect of the second phase. This will then allow scale up to be modelled with greater confidence.

Dissolved oxygen studies
It has been shown previously that pyrogallol dissolved in basic solution can act as a reporter as to the amount of dissolved oxygen. 4,5 This can be measured by UV-Vis absorption spectroscopy as the starting colourless solution changes colour on exposure to oxygen. We

Experimental Methods and Spectra General Information
All solvents and reagents were used as obtained without any further purification. Biphenyl was used an internal standard for determining the yields and/or conversion of 2 and 8. For the yield/conversion of 5 and 7, 1,3,5-trimethoxybenzene was used as an internal standard. The spectra were assigned comparing chemical shifts to existing literature values and commercial samples.

General Reactor Operation
The LED recirculating chiller was turned on and set to 5 °C whilst the reactor chiller was set to the appropriate temperature for the reaction. The chillers were left to equilibrate for ca. 20 min. The reaction mixture was prepared as describe d below and the inlet pipe for the HPLC pump was fed into the solution and the pump primed. The desired flow rate was set on the HPLC pump and the solution pumped. The peristaltic pump to remove the products from the reactor was set to 50 rpm. The rotation motor was turned on and slowly increased until the desired rotation speed was achieved. Then the LEDs were turned on at full brightness. Two full reactor volumes were allowed to pass before taking a sample for analysis, this ensured that the reactor had reached a steady state.
Once the operation was complete, the LEDs and the HPLC pump were turned off, and the rotation motor was slowly decreased to 0 rpm. The recirculating chillers were also turned off, providing the LED blocks were not excessively hot. The inlet pipe was switched to a container with a compatible solvent (usually the reaction solvent) and the reactor was flushed at 2-5 mL min -1 . The rotation speed was set to 200 rpm during this time and increased to 4000 rpm occasionally to ensure that all material was removed from the reactor.
Once the reactor was clean, the motor and pumps were turned off.