Parameter assessment for scale-up of co- and counter-current photochemical reactors using non-collimated LEDs
Introduction
The scale-up of photochemical reactors has been studied extensively in the past years, more specifically in the field of cross-current illumination. With this illumination technique, the incident photons travel perpendicular to the direction of flow, and usually result in short absorption path lengths. Various applications of this commonly used illumination technique as lab setups are shown by Cambié et al. (2016); Knowles et al. (2012), Hook et al. (2005), Plutschack et al. (2017), Gilmore and Seeberger (2014). These reactors all contain a reactor channel illuminated sideways via e.g. a tubular reactor coiled around a mercury lamp. Even in the industrial field, this cross-current illumination technique is commonly applied as an alternative to batch reactors as shown by Oelgemöler and Shvydkiv (2011) and Noël (2017) with similar reactor setups to lab scale. This is contrary to co- and counter-current illumination, which emphasises the use of longer path length by illuminating the reactor in the direction of flow. Scale-up studies using this novel illumination technique are absent and usually non-transferrable from cross-current illumination as different challenges are posed.
Recently, strategies were developed capable of predicting operating conditions for a scaled up cross-current flow reactor. These scale-up strategies are diverse and include experimental variation of operating conditions until decent conversion is obtained (Sambiagio and Noël, 2019; Su et al., 2016; Zhao et al., 2018) as well as the use of several (non-dimensional) parameters for scale up. Some of these parameters do not include absorption of the reagents or photocatalysts, the efficiency of the light source or the illuminated surface area (Elliott et al., 2018) and would therefore be only applicable in similar reactor setups. Other parameters which are more robust take the absorbed light flux into account, which has shown to be much more robust when using a batch reactor and various continuous reactors of different scales (Corcoran et al., 2020). This latter parameter, based on the absorbed light flux, uses a similar approach as the work by Meir et al. (2020a) in which the ‘quantum photon balance’ as a reactor parameter was introduced. In this work, an approach similar to Corcoran et al. (2020) was used to determine the relative absorbed light flux as a reactor parameter for the design of an efficient photochemical reactor.
Meir et al. have shown that the use of co- and counter-current illumination shows significant improvements compared to the state-of-the-art cross-current photoreactor illumination (Meir et al., 2020a,b). Conversion, space-time yield (STY) and photochemical space-time yield (PSTY) increased for equal incident light flux entering the same reactor geometry. The results for co- and counter-current illumination were obtained using a high intensity collimated LED as light source. However, scaling-up such a reactor in an industrial context proves to be difficult as these collimation optics are expensive and the LED efficiency decreases with the presence of these collimation optics. This efficiency decrease is shown in the lamp used by Meir et al., which only had a measured efficiency of 8% in the presence of these collimation optics, whereas a typical LED in this wavelength range (365 nm) has an efficiency of 30–35%. Hence, by omitting the collimation optics, a higher efficiency is achievable. As the derived model explicitly takes collimated light into account, a difference in performance is expected when switching to standard UV-LEDs. However, the use of standard UV-LEDs would significantly improve scalability of the co- and counter-current illumination scheme for homogeneously non-catalysed photochemical reactions.
In this work, we investigated the possibility of using non-collimated light sources to drive homogeneous photochemical reactions to enable easier adoption in industry by decreasing initial cost of the reactor and potentially also operation cost. It is expected that a discrepancy shows up between the derived model by Meir et al. (2020b) assuming ideal conditions regarding light influx and flow regime. In this work, non-collimated light sources are used to irradiate the reactor, which are not considered by the ideal model and are therefore assessed via the use of ray tracing. In this paper, we studied the influence of reactor geometry, reactor material, LED type and absorbance which are important parameters w.r.t. scaling-up co- and countercurrently illuminated photochemical reactors. After optimisation of this set of parameters, the initial cost and operational cost of the co- and counter-currently illuminated photochemical reactor can decrease as no collimation optics are required and the associated energy losses are avoided. This study is performed via the use of experimental and simulation data, with the latter being performed via the ray tracing toolbox in COMSOL.
Note that this study is aimed towards the use of homogeneous photochemical reactions rather than homogeneous photosensitized or photocatalyzed reactions. Single phase photochemistry without a homogeneous catalyst is preferred when the reaction is selective and has a decent quantum yield. Homogeneous photocatalysis is often more selective and has decent quantum yield with commonly used photocatalysts and sensitizers. The disadvantage of homogeneous photocatalysis is the increased need for adequate mixing in the reactor combined with the mismatch between absorption and conversion. At higher conversions, not every absorbed photon (corrected for the quantum yield) yields a converted reagent molecule. This mismatch is absent in case of homogeneous photochemistry as in most cases only the reagent is capable of absorbing light and subsequently converting to its product. Furthermore, when using a catalyst or sensitizer, an extra separation step is often required, thereby increasing operation cost.
Section snippets
Model equations for co- and counter-current photochemical reactors
Meir et al. (2020a,b) derived a model to predict conversion and reactor efficiency for co- and counter-current photoreactors under ideal conditions. These conditions imply the use of homogeneous and collimated light and a known residence time distribution.
In this paper, we investigated the applicability of the derived model by Meir et al. for standard non-collimated LEDs for co- and counter-current illumination by studying the influence of the viewing angle, illumination configuration with
Materials and methods
A combination of experimental and modelling data is used in this work. Experimentally, the flow rate, the light intensity, the LED type and, to a smaller extent, the absorbance were varied to assess the effect of each parameter. To acquire this experimental data, a reactor, a light source, and a photochemical reaction are used as described in Sections 3.1–3.4. Using these experimental data, a ray tracing model in COMSOL is validated. With this model, the effect of the non-collimated light is
Results and discussion
The reactor, as mentioned in Section 3.2, and light source, as mentioned in Section 3.3, were used in this analysis. The aim was to check the reactor's performance and compare it to the previously derived MATLAB model, as described in Section 4. These results are shown for LED60. Prior to these results, the illumination efficiency of the reactor is evaluated.
Conclusion
The results of this paper show that when using non-collimated light, reflection losses occur resulting in a deviation compared to the model and validation as described by Meir et al. (2020a,b). This model can still be applied in conjunction with a ray tracing model implemented in COMSOL. This model is capable of determining absorbance, which can be implemented in the MATLAB model for accurate prediction of conversion and reactor performance.
Based on a parameter study, guidelines were derived
Declaration of interests
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
Acknowledgements
Glen Meir thankfully acknowledges FWO-Flanders for a SB PhD fellowship (1SA6219N). Mumin Enis Leblebici thankfully acknowledges FWO-Flanders for a postdoctoral fellowship (39715). The authors would like to extend their gratitude to Senne Fransen for his help regarding model derivations and to Kenneth Simoens for his help regarding HPLC measurements and analysis.
References (22)
- et al.
Experimental study of flow regime and mixing in T-jets mixers
Chem. Eng. Sci.
(2012) - et al.
Media controlled photo-Favorskii type rearrangement of a-chloro acetophenones: synthesis of phenylacetic acids
Tetrahedron
(1997) - et al.
Flow imbalance and Reynolds number impact on mixing in confined impinging jets
Chem. Eng. J.
(2015) Characterization of mixing in T-jets mixers
Chem. Eng. J.
(2012)- et al.
A convenient numbering-up strategy for the scale-up of gas–liquid photoredox catalysis in flow
React. Chem. Eng.
(2016) - et al.
Continuous photochemistry: the flow synthesis of ibuprofen via a photo-Favorskii rearrangement
React. Chem. Eng.
(2016) - et al.
Applications of continuous-flow photochemistry in organic synthesis, material science, and water treatment
Chem. Rev.
(2016) - et al.
Photon equivalents as a parameter for scaling photoredox reactions in flow: translation of photocatalytic C−N cross-coupling from lab scale to multikilogram scale
Angew. Chem. Int. Ed.
(2020) - et al.
Using batch reactor results to calculate optimal flow rates for the scale-up of UV photochemical reactions
React. Chem. Eng.
(2018) Properties of metals
Nature
(1973)
Continuous flow photochemistry
Chem. Rec.
Cited by (4)
Energy transition technology comes with new process safety challenges and risks
2023, Process Safety and Environmental ProtectionDesign and evaluation of co-currently illuminated two-phase bubbly flow photochemical reactors
2023, Chemical Engineering JournalCombining computational fluid dynamics, photon fate simulation and machine learning to optimize continuous-flow photocatalytic systems
2023, Reaction Chemistry and Engineering