Design, Microfabrication and Characterization of a Split-Flow Thin Fractionation (Splitt) Electrical Device for Continuous Extraction or Binary Separation of Proteins

Abstract

Protein fractionation and isolation are fundamental sample preparation methods, used in fields ranging from medicine to food analysis. The isolation of one or a few proteins from a mixture allows for their characterization (e.g. function, chemical and biological behavior, structure). In the field of lab-on-chip (LOC), point of care devices and analytical instrumentation, proteins can severely interfere with sensors because of their high fouling power on surfaces. This happens, for instance, in LOC using immunoaffinity biosensors. Classical protein fractionation and sample preparation methods such as electrophoresis and high-performance liquid chromatography (HPLC) do not operate in a continuous mode. The volume of sample is consequently predetermined and it is not possible to modify operating parameters while using the systems. Filters can suffer from clogging or fouling problems, and may be very expensive. Centrifugation has low resolution and can cause damage to proteins, while chemical separation techniques may cause protein denaturation. The broad goal of this work was to evaluate continuous-flow microfluidic separation methods that can overcome the limitations of classical laboratory techniques. They can be used with a wide range of particles, they have a good resolution, high throughput, with a reduced consumption of reagents and analytes, and they are low cost. Due to their specific characteristics, these techniques may find application in future in Lab-On-Chip (LOC) for chemical or biological analysis for “point-of-need” or “point-of-care” portable devices, since they can be used in sample preparation modules for downstream miniaturized systems. For instance, the integration of a microfluidic device for particle fractionation/extraction in a LOC system can provide an answer to the challenge of “chip-to-the-world interface”. In fact, generally, the functionality of the sensing part of detection systems of LOC is compromised if there are interferents in the processed sample. This usually happens with raw samples, such as milk, blood, or saliva. This problem is also known as “matrix effect”. Microfluidic continuous-flow separators can provide a sample where only a single species or a group of species of special interest is present. In particular, this work of thesis describes the design, microfabrication and characterization of a miniaturized Split-flow Thin (SPLITT) fractionation electrical system for continuous flow protein extraction or fractionation. Split-flow thin fractionation is a highly promising particle separation technique, since it allows for fast continuous sample processing with good resolution and higher throughput (e.g. hundreds of microliters or milliliters) with respect to other continuous-flow microfluidic separation methods such as Free Flow Electrophoresis or Free Flow Isoelectric Focusing. Particle separation in SPLITT can be electrical, centrifugal, magnetic, or based on gravity or on diffusion. Professor J. Calvin Giddings (University of Utah) invented SPLITT in 1985 and demonstrated the applicability of the method to separate binary protein mixtures, colloidal particles, pharmaceutical emulsions, and human blood cells, platelets and plasma proteins. Over the years, other research groups implemented SPLITT systems for the fractionation of marine sediments, small peptides, and bacteria. Since proteins carry a net charge if the pH of the surrounding medium is different from their isoelectric point, electrical based SPLITT is suitable for protein separation. This work of thesis presents several innovative aspects: (1) The miniaturisation of a split-flow cell presents significant advantages and challenges. In fact, the microfluidic channel has a high aspect ratio rectangular cross section, with submillimeter thickness. Moreover, the functionality of a SPLITT device relies on the presence of two suspended microstructures at both ends of the microfluidic channel, named “splitters”, which are difficult to realize with microfabrication methods. In this work of thesis, we demonstrate for the first time the complete miniaturization through microfabrication of the ideal split-flow cell. With respect to non-miniaturized systems, the implemented microfluidic cell is ten times shorter and four times thinner. Thanks to the reduced dimensions, the device is easy portable; it is suitable for integration in LOC, and for parallel processes on multiple streamlines. Moreover, since it is completely microfabricated, theoretically, it is possible to automate all manufacturing steps. In electrical-based SPLITT, miniaturization is also important because it allows for the application of high electric fields for protein fractionation with low voltages.

(2) The microfluidic network included in the implemented device is realized through multiple laminations and subsequent exposure and development of a photopatternable dry film, named Ordyl SY 355. The microfabrication process we describe represents a novel use of this technology, since it allows for the heretofore unrealized manufacturing of buried microfluidic channels and suspended microfluidic components by using laminated dry films. The fabrication process is easily transferrable and suitable for the realization of other types of lowcost microfluidic modules for LOC. Moreover, a suitable method to include nanopatterned electrodes in the microfluidic device is reported; (3) The efficacy of pulsed electric split-fractionation is demonstrated. Literature concerning non-miniaturized SPLITT mainly describes DC-based electric systems. Instead, in this work we demonstrate the advantages of pulsed fields, which allows for a better transfer of electric fields inside the microfluidic channel with respect to DC fields. We also describe the optimization of the applied electrical waveform, in order to obtain an efficient protein separation; (4) We demonstrate the functionality of the cell both for continuous-flow protein extraction (Full-Feed Depletion Mode) and for continuous binary protein separation (classical SPLITT mode) with high processing rate and high throughput (in the order of ml/min). With respect to non-miniaturized systems, we managed to achieve comparable fractionation rates; (5) We demonstrate the effect of different operating parameters on SPLITT efficacy, by performing specific measurements and providing a discussion about the results. This provides new knowledge about split-flow thin fractionation, giving the optimal experimental conditions for electrical separation.