Photon Transport and Hydrodynamics in Gas‐Liquid Flows Part 1: Characterization of Taylor Flow in a Photo Microreactor

Abstract

Gas-liquid photoreactions are increasingly implemented in microreactors. Taylor flows containing an inert dispersed phase were previously used to increase the conversion of photochemical reactions in comparison to using a single liquid phase. However, identifying the optimal flow conditions requires an extensive experimental effort. This work aims to understand the photon transport and hydrodynamics in a Taylor flow photo microreactor so that the reactor behavior can be understood and predicted. Chemical actinometry, flow imaging and residence time distribution experiments were used to develop a multi-region photochemical reaction model. This model shows that the conversion is significantly affected by the liquid distribution, and not by the light scattering or liquid mixing. Moreover, an empirical relation was proposed to predict the optical pathlength in gas-liquid flows. The knowledge gained in this study helps to optimize the performance of Taylor flow photo microreactors, but also to design improved multiphase flow photochemical systems.

Photon Transport and Hydrodynamics in Gas‐Liquid Flow Part 2: Characterization of Bubbly Flow in an Advanced‐Flow Reactor

“…Figure 2 illustrates images at three different instances for each investigated case. In contrast to the Taylor flow investigated in our previous work, [7] the dispersed bubbles in the G1 AFR had a relatively large size distribution. Moreover, the bubble size changes continuously along the reactor due to bubbles breaking and merging in the heart-shaped cells.…”

“…However, the difference in axial dispersion between single and two-phase flow is much smaller than observed in Taylor flow. [7] However, this effect is not present at the highest gas fraction of β G = 0.52, as the RTD responses in single-phase and two-phase flow are identical. In Section S2.4 in ESI, the axial dispersion model (ADM) is used to quantify the axial dispersion in single and two-phase flow.…”

“…The quantum yield of this photochemical reaction varies with the irradiation wavelength and is comprised between 0.015 at 480 nm and 0.027 at 620 nm as reported by Sumi et al [13] In our previous work, the photon flux per liquid volume, I v 0 , and the optical pathlength, l reactor , were determined in single-and two-phase flow in a microreactor. [7,12] This was realized by performing a series of measurements at different starting concentrations of DAE CF and a fixed light intensity. In this study we used the same approach, therefore solutions with different starting concentrations of DAE CF were converted in the G1 AFR under illumination of green LEDs.…”

“…As the optical pathlength, l reactor , is unknown, it was fitted together with I v 0 using a nonlinear regression model in Matlab (fitnlm). [7] For each flow condition, the resulting I v 0 values are randomly distributed around the average value as illustrated in Figure 5, and the standard deviation did not exceed 5 % as shown in Table 2. This indicates that the methodology used to determine the photon flux and optical pathlength in the photo microreactor works well also for the G1 AFR with a more complex geometry.…”

“…This was proven in our previous study on Taylor flow in a photo microreactor. [7] The presence of a long liquid film around the gas bubbles coupled with short liquid slugs positively affected the conversion and reduced the distance traveled by the light in the liquid phase. While the gas-liquid photoreactions scaleup using capillary reactors involved Taylor flow, the reactors with integrated static mixers such as Corning® G1 Advanced-Flow™ Reactor (G1 AFR) promoted bubbly flow.…”