Solvent extraction (SX) is used throughout a range of different chemical industries: purification of pharmaceutical compounds, wastewater treatment, sample preparation, etc. In hydrometallurgy, SX even became one of the most important separation process . Traditionally SX is performed in packed columns, centrifugal extractors or mixer-settler tanks. Although, these systems have been the workhorses for decades now, important problems like foaming, flooding, unloading, the need for a density difference and stable emulsion formation still present important disadvantages . Non-dispersive contactors avoid such problems by contacting the fluids through a parallel flow profile. In order to maintain a stable interface and yet have a large operating range the interface needs to be stabilized either by placing coated guiding structures or equipping the contactor with a membrane –. This membrane can either be used in a flat sheet configuration or a hollow fiber configuration. Disadvantageous of the hollow fiber approach is that a bundle of fibers can lead to shell side by-passing and the potting adhesive is prone to attacks by the organic solvents. However, hollow fibers are self-supporting, resulting in inner diameters of typically 200-300 µm and walls of tens of micrometers thick . The flat sheet configuration however does require a support, quickly leading to channels of hundreds of micrometer to millimeters deep. As flows are most of the time laminar inside such membrane contactors the mass transfer is purely diffusion based, leading then to slow extraction kinetics. With current microfabrication technology it is now possible to fabricate structured supports, resulting in depths of only hundred micrometer or less reaching equilibrium in matter of minutes .
In this work the extraction and stripping efficiency of cobalt from a cobalt/nickel mixture inside such a flat sheet membrane microcontactor was studied. The impact of different parameters on the mass transfer kinetics such as channel depth, concentration, wetting properties of the membrane, pH, saponification of the extractant (Cyanex 272) and upscaling effects was investigated. With channel depths of 100 µm complete extraction was reached after 3 min. Lowering the channel depth further to 50 µm, only 1 min was needed to extract all of the cobalt. It was also seen that buffering the aqueous feed instead of saponification of the organic extractant, increased the extraction kinetics by a factor of 2.5. This was in conflict with what we expected as extraction rate is reported to increase with saponified extractant , . Due to the high partition coefficients it was also observed that the wetting properties of the membrane had a significant impact on the mass transfer kinetics. Using a hydrophilic membrane the stripping proceeded with a factor of 1.5 faster than with a hydrophobic membrane. By stripping cobalt from the loaded organic solution, the extractant was regenerated, allowing to operate continuously and recycle the organic phase between an extraction and stripping unit. Cobalt was successfully extracted for 6 hours leaving nickel behind in the raffinate. During this period no significant changes in extraction and stripping performance was observed and the interface remained stable. Only minimal breakthrough (1.87% of the total processed volume) at the stripping unit was observed, which could easily be removed from the buffering tank of the organic extractant, which can alternatively also be avoided by active pressure regulation.
J. Hereijgers greatly acknowledges the Agency for Innovation by Science and Technology (IWT) in the Flanders region for financial support. W. De Malsche is grateful to the Flemish Fund for Scientific Research (FWO) for financial support.
 D. S. Flett, “Solvent extraction in hydrometallurgy: The role of organophosphorus extractants,” in Journal of Organometallic Chemistry, 2005, vol. 690, no. 10, pp. 2426–2438.
 A. Gabelman and S. Hwang, “Hollow fiber membrane contactors,” Journal of Membrane Science, vol. 159, no. 1–2. pp. 61–106, 1999.
 D. Ciceri, J. M. Perera, and G. W. Stevens, “The use of microfluidic devices in solvent extraction,” Journal of Chemical Technology and Biotechnology, vol. 89, no. 6. pp. 771–786, 2014.
 Z.-X. Cai, Q. Fang, H.-W. Chen, and Z.-L. Fang, “A microfluidic chip based liquid-liquid extraction system with microporous membrane.,” Anal. Chim. Acta, vol. 556, no. 1, pp. 151–6, Jan. 2006.
 J. Hereijgers, T. Breugelmans, and W. De Malsche, “Breakthrough in a flat channel membrane microcontactor,” Chem. Eng. Res. Des., vol. 94, pp. 98–104, 2015.
 J. Hereijgers, M. Callewaert, X. Lin, H. Verelst, T. Breugelmans, H. Ottevaere, G. Desmet, and W. De Malsche, “A high aspect ratio membrane reactor for liquid-liquid extraction,” J. Memb. Sci., vol. 436, pp. 154–162, Feb. 2013.
 P. XING, C. WANG, S. XU, and Z. JU, “Kinetics of cobalt(II) extraction from sulfate aqueous solution by sodium salt of di-decylphosphinic acid (DDPA),” Trans. Nonferrous Met. Soc. China, vol. 23, no. 2, pp. 517–523, 2013.
 X. Fu and J. A. Golding, “EQUILIBRIUM AND MASS TRANSFER FOR THE EXTRACTION OF COBALT AND NICKEL FROM SULFATE SOLUTIONS INTO BIS (2,4,4-TRIMETHYLPENTYL) PHOSPHINIC ACID - ‘CYANEX 272’.,” Solvent Extraction and Ion Exchange, vol. 6, no. 5. Taylor & Francis Group, pp. 889–917, 12-Mar-1988.