(304d) Microwave Assisted Flow Processing: Coupling of Electromagnetic and Hydrodynamic Phenomena

Microwave
assisted flow processing: coupling of electromagnetic and hydrodynamic
phenomena

Narendra
G. Patil,¹ Erik Esveld,² Faysal Benaskar,¹ Evgeny
V. Rebrov,³ Lumbertus A. Hulshof,¹ Jan Meuldijk,¹
Volker Hessel¹ and Jaap C. Schouten¹

¹Laboratory of Chemical
Reactor Engineering, Eindhoven University of Technology,
P.O. Box 513, 5600 MB Eindhoven, the Netherlands

²Wageningen UR Food and
Biobased Research, P.O. Box 17, 6700 AA, Wageningen, The Netherlands

³School of Chemistry and
Chemical Engineering, Queen's University Belfast, Stranmillis Road, Belfast,
BT9 5AG, United Kingdom

Introduction

Fast
and volumetric heating behavior of microwaves is attracting attention for their
application in continuous flow synthesis of specialty chemicals. However, the
use of tuned microwave field even in small defined reactors does not always
provide the expected temperature patterns.^1,2 This is majorly due to
a lack of position specific information with respect to velocity profile of
liquid. Additionally, the accuracy of these predictions in microwave heating is
limited when no space distribution of volumetric heating source is considered.¹
Understanding the space distribution of microwave heating process with detailed
modeling is a prerequisite for process development and control. In this work,
we take a closer look at the influence of liquid velocity profiles on the axial
and radial temperature profiles in a tubular microwave integrated millireactor
combined with a heat exchanger (Figure 1a).

Results
and discussion

Horizontal
co-current flow of a reactant (ethanol) and a microwave transparent coolant
(toluene) was established in a teflon supported quartz tube (i.d.: 3 mm, o.d.:
4 mm) and shell (i.d.: 7 mm, o.d.: 9 mm), respectively (Figure 1a). The
co-current flow of coolant was assumed to avoid overheating of the reactant by
inherently fast microwave heating. The reactor was inserted in a microwave
cavity with a length of 47 mm. The physical model included electromagnetic
interactions, fluid dynamics and heat transport (Figure 1b). The temperature
and flow profiles were obtained in a 3-D domain using the COMSOL software. As the dielectric properties of the reactant were
changing with temperature, it was expected that they should influence the
heating rate. However, this effect was of minor importance (Figure 1c) and the
remaining simulations were made in a 2-D domain. Effect of gravitation was
observed due to horizontal arrangement of the reactor-heat exchanger. The
effect of gravity was clearly pronounced in the coolant flow where a
recirculation pattern was observed due to difference in density (Figure 2). The
coolant flow rate had a minor effect on
temperature profiles and energy dissipation at a high heating rate  of 180 °C/s
(Figure 3, a and b). Higher temperatures were observed at the wall (stagnant liquid
films), leading to drastic drop in the energy dissipation (Figure 3, c and d). The
laminar flow of ethanol was disturbed by a temperature probe (Figure 4). The
probe was inserted either from the outlet (Figure 4a) or inlet (Figure 4b) side
of the assembly. In the former case, measured temperature was lower as compared
to the latter case (Figure 5). This
effect could be explained by local hydrodynamics near the tip of the probe. When
the probe was inserted from the outlet, the highest velocities were found at
the tip. Conversely, when the probe was inserted from the inlet, a stagnant
film assured lowest velocity at the probe tip. These observations were then
validated by experiments (data points, Figure 5). The temperature profiles
obtained by the modeling study agreed with the experimental observations.

Conclusions

The
stagnancy of the flow of the microwave absorbing fluid influenced the
temperature distribution in a microwave integrated flow reactor. The buoyancy
influence (as a result of gravitational forces) was already visible for
horizontal arrangement of reactor-heat exchanger assembly at millimeter sizes.
The stagnant layer formation caused by any insertion or at reactor walls yielded
higher temperatures and lower energy dissipation regions. The coolant flow,
unless used as a heated jacket to minimize heat losses, was found to be
ineffective. Lastly, complex 3-D geometry modeling efforts can be eliminated at
very early stages unless there are specific disturbances in electric field
pattern.

                                                                (a)

                                (b)                                                                                           (c)

Figure
1:
a) Schematic view and process details of the microwave integrated
reactor-heat exchanger assembly used for experimental validations, b) 3
D computational domain of the microwave cavity showing individual components of
the assembly, c) Electric field intensity in and around the microwave
integrated reactor-heat exchanger. Ethanol flow rate:  40 ml/min, toluene flow
rate: 100 ml/min. Input microwave power: 55 W.

Figure
2:
Effect of buoyancy in a horizontal arrangement of the microwave integrated
reactor-heat exchanger assembly. Gray lines (with thickness indicating the
strength) show velocity profiles of coolant. Axially sectioned colored pattern
shows temperature in the reactant as well as the coolant section.

                                                (a)
                                                                          (b)

                                                (c)
                                                                          (d)

Figure
3:
Temperature profiles (a,c) as a function of axial co-ordinate, and (b,d)
energy dissipation of microwaves (Q_MW, W/ml), temperature rise (T-T_in,
°C) and velocity profile (U, cm/s) as a function of radial co-ordinate. (a)
near stagnant coolant flow, F_EtOH = 100 ml/min, F_TOL =
0.5 ml/min. (b) high coolant flow, conditions are the same as those in
Fig 1c.

Figure
4:
Influence of the direction of probe insertion on the temperature profiles
obtained by modeling, a) probe inserted from the outlet, b) probe
inserted from the inlet. Conditions are the same as those in Fig 1c.

Figure
5:
Influence of the direction of probe insertion on the temperature profiles
obtained by modeling (lines) and experiments (data points). Solid line and squares:
probe inserted from inlet, dotted line and triangles: probe inserted from
outlet. Conditions are the same as those in Fig 1c.

Reference:

1. Patil, N. G.; Hermans, A. I. G.; Benaskar, F.;
Rebrov, E. V.; Meuldijk, J.; Hulshof, L.
A.; Hessel, V.; Schouten, J. C. Energy efficient and controlled flow
processing under microwave heating by using a milli reactor-heat exchanger. AIChE
J. 2011, Accepted. DOI
10.1002/aic.13713.

2.
Patil, N. G.; Benaskar, F.; Rebrov, E. V.; Meuldijk,
J.; Hulshof, L. A.; Hessel, V.;
Schouten, J. C. Continuous multi-tubular milli-reactor
with a Cu thin film for microwave assisted fine-chemical synthesis. Ind.
Eng. Chem. Res. 2012, submitted.