(103e) Using Microreactors for Liquid-Liquid Extraction of Vanillin

Within the development to ?lab on a chip?, all unit operations should be performed in a microreactor to get a powerful tool with all important processes available. In chemical engineering, extraction is an important process for cleaning carrier fluids or to separate the products from educts. As an example, we investigated the extraction of vanillin from the aqueous phase with toluene, which is part of an industrial process for the production of vanillin from lignosulfonates [1].

In order to extract vanillin out of the aqueous phase in a microreactor, three aspects had to be optimized in this project:

1) creating a large interfacial area and a stable flow pattern

2) achieving a sufficient long residence time

3) separating both phases effectively.

The used microreactor was designed in a way to fulfill all these requirements. The width was 300 µm, the height 200 µm. The two-phase channel length was 3.4 cm. Based on our experiences in gas-liquid flow, we designed the inlet in a way to achieve a stable flow without oscillations and a large interfacial area. An important aspect in extraction is the phase separation on the chip. Guenther et al. [2] demonstrated gas-liquid phase separation based on the capillary pressure difference: a rake-like shaped separator consisting of 20µm-wide capillaries sucks the continuous liquid phase out of the main channel. Based on this phenomenon, we designed a liquid-liquid separator. As PDMS is a very hydrophobic material, the continuous organic phase preferably wets the enhanced surface at the separator.

The flow has been characterized via Laser Induced Fluorescence (LIF), micro-Particle Image Velocimetry (µ-PIV) and standard microscopy. For the determination of the separation grade we used volumetric measurements. The concentrations of vanillin were determined by UV/Vis spectroscopy of the aqueous phase and by gas chromatography of the organic phase.

The separation of the organic and aqueous phase with the rake-like separator depended on the velocity: At higher velocities (u>0.04 m/s) separation degrees of more than 0.95 were observed. Whereas the separation grade was defined as the ratio of the organic phase volume to the total volume collected at the separator outlet.

As we used microreactors made of PDMS, the swelling behaviour of different solvents had to be investigated. The sorption of toluene into PDMS led to a mass increase. Saturation was achieved after 3 hours with a mass uptake of 100%. As the volume of the PDMS is large compared to the microchannel, geometrical changes of the structure can be neglected.

Using the microreactor described above gave for very small velocities (u<0.03 m/s) different flow patterns depending on the initial wetting of the reactor. For Taylor flow the extraction was more effective than for stratified flow. At higher velocities we obtained only Taylor flow. For a certain flow pattern the amount of vanillin left in the aqueous phase depended strongly on the residence time in the reactor. For residence times >2s in Taylor flow with optimized mass transfer the distribution ratio (concentration of vanillin in the organic phase versus the concentration of vanillin in the aqueous phase after the extraction) correlated with the thermodynamic equilibrium.

Based on these experiments we propose a microreactor to perform a multi-step continuous extraction. Volumetric flow rates are in the range of 30-500 µl/min. Possible applications are the continuous operation mode for the extraction in ?lab on a chip? application and, as we are not mass transfer limited in the microscale [3], an analysis tool for kinetic measurements of extraction.

References

1. Hocking, M.B., Vanillin: Synthetic Flavoring from Spent Sulfite Liquor Journal of Chemical Education, 1997. 74(9): p. 1055-1059.

2. Guenther, A., Khan, S.A., Thalmann, M., Trachsel, F., Jensen, K.F., Transport and reaction in microscale segmented gas-liquid flow. Lab Chip, 2004(4): p. 278-286.

3. Fries, D.M., S. Waelchli, and P. Rudolf von Rohr, Gas-liquid two-phase flow in meandering microchannels. Chemical Engineering Journal. In Press, Corrected Proof.