Chemical and Biomolecular Engineering at Illinois

Paul Kenis Research

Microchemical Systems: Microreactors, Microfuel Cells, and Microfluidic Tools

Kenis Research Group

Dr. Paul J.A. Kenis

In the Kenis research group, we exploit the capability of exquisite control over transport phenomena at the microscale to study fundamental phenomena (including protein chemistry, cell biology) and to develop novel technologies for a range of applications, including energy conversion, chemical synthesis, and fundamental biological studies. To conduct research in these interdisciplinary areas, we have developed core expertise in characterization of electrochemical systems, microfabrication, microfluidic technologies, as well as analytical and computational modeling of transport phenomena, and analytical and material characterization techniques such as various types of spectroscopy and microscopy.

Currently the group pursues research projects in the following areas:

1. Electrochemical systems for carbon dioxide conversion and fuel cells
2. Microfluidic platforms for crystallization of proteins and pharmaceuticals
3. Microfluidic platforms for studying inter and intra-cellular processes
4. Microreactors for chemical synthesis
5. Manufacturing technologies for microfluidics
6. Emerging microfluidic ‘bio’ projects

1. Electrochemical systems for carbon dioxide conversion and fuel cells

1a. Electrochemical reduction of CO2:

The CO2 levels in the atmosphere have been rising steadily, which has led to a negative impact on global climate. Multiple strategies, such as carbon capture and sequestration, switching to cleaner fuels, expanding utilization of renewable energy sources, and increasing the energy efficiency of buildings, need to be employed simultaneously to curb this rise. Electrochemical reduction of CO2 into value added chemicals or their intermediates is another approach to address this challenge. This process can be driven by the excess power from intermittent renewable sources, thereby providing a means to store excess intermittent renewable energy while simultaneously recycling CO2 as an energy carrier. Furthermore, by utilizing CO2 as the starting material for chemical production, society’s dependency on fossil fuels is reduced.


For the electrochemical reduction of CO2, our group aims at improving the product selectivity, energetic efficiency and conversion rate through the development of novel catalysts, application of suitable electrolytes, and optimization of the electrode structure and the reactor design. For example, we decreased the cell overpotential to less than 0.2 V by using an aqueous solution containing 1-ethyl-3-methylimidazolium tetrafluoroborate (EMIM BF4), which presumably stabilizes a reaction intermediate (Rosen et al. Science, 2011). We also developed silver-based organometallic catalysts that exhibit high catalytic activity at low Ag loading (Thorson et al., J. Am. Chem. Soc., 2012). As a support material, TiO2 is used to minimize Ag particle size and increase catalyst activity, resulting in a drastically lower Ag loading without sacrificing the performance towards reducing CO2 to CO (Ma et al., ChemSusChem, 2014). Also, engineering the catalyst layer structure provides an approach to maximize catalyst utilization and overall performance. An automated airbrushed catalyst deposition method led to high performance on CO2 reduction with reduced catalyst loading while unwanted H2 evolution was suppressed (Jhong et al., Adv. Energy Mater., 2013).

Currently we continue to pursue research towards better catalysts, electrodes, and operation conditions for the electrochemical conversion of CO2 into chemicals of interest. Some of this work is in collaboration with others: Nakashima, Lyth in Kyushu, Japan; and Rich Masel at Dioxide Materials.

1b. Fuel cells:

(2) Microfluidic platforms for crystallization of proteins or pharmaceuticals

Crystallization of proteins and pharmaceuticals can quickly become very expensive because of the large amounts of material needed for screening for optimal crystallization conditions. Despite the availability of automated robotic crystallization screening instruments that can utilize nanoliter-sized drops, the large investment in capital required makes such instruments practical only to a few well-funded laboratories or crystallization centers. Our microfluidic platforms for protein and pharmaceutical crystallization (i) enable high-throughput screening and optimization of crystallization conditions while using a few nanoliters per trial; (ii) are simple to use, cost-effective alternative to crystallization robots for the average laboratory; and (iii) are compatible with analytical techniques by appropriate selection of materials (e.g., high transmission of X-ray, UV, and IR). Being X-ray transparent, our chips can be mounted directly in an X-ray beam for data collection bypassing the step of manually harvesting the crystals. Our microfluidic platforms enable studies of fundamental science of crystallization (crystal seeding, nucleation and growth rates) as well as applied science (structural analysis, solid form screening) for both protein and pharmaceutical crystallization.

2a. Membrane protein crystallization:

Membrane proteins (MPs) reside within the cellular membrane and act as mediators for signal, energy, and material transduction into and out of the cell. Not surprisingly, the malfunction of membrane proteins has been linked to numerous diseases (Quick and Javitch, PNAS, 2007). MPs are thus common drug targets. Various analyses have indicated that MPs constitute nearly 30% of the proteins encoded in the genomes of Escherichia coli, Saccharomyces cerevisae, and Homo sapiens (Seddon et al., Bba-Biomembranes, 2004). kenis5_0Despite their overwhelming preponderance in the cell, MPs account for less than 1% of protein structures deposited in the Protein Databank. Structure determination of membrane proteins has been hampered by difficulties in obtaining sufficient quantities of the proteins due to low abundance and their inherent amphiphilicity, and subsequent difficulties in crystallization. In our group, we have developed X-ray transparent microfluidic platforms for in surfo and in meso MP crystallization. In addition, our research includes X-ray transparent platforms that enable the study of lipidic cubic phase diagrams and microseed matrix screening, two powerful yet typically inaccessible crystallization techniques for membrane proteins. The overall goal of our research is to crystallize large, well-ordered (“diffraction-quality”) crystals for X-ray analysis and structure elucidation. We have crystallized several targets and solved their structures using data collected solely on chip Current efforts are focused on crystallizing respiratory membrane proteins in collaboration with Prof. Robert Gennis, Department of Biochemistry.


2b. Solid form screening of candidate pharmaceuticals:

During the early stages of pharmaceutical drug discovery, scientists search for solid forms of active pharmaceutical ingredients (APIs) that have appropriate physical and chemical properties (i.e. solubility, bioavailability, stability) that can later move through the drug development pipeline. Unfortunately, success in finding a crystalline solid form of an API with optimized properties using conventional screening procedures (well plates) is limited by small quantity of API available during the early stages of drug discovery. To address this issue we have developed microfluidic platforms for pharmaceutical solid form screening with the goals of (i) reducing the quantity of active pharmaceutical ingredients (APIs) needed for solid form screening, (ii) increasing compatibility between solid form screening platform and analytical instruments, and (iii) determining if a microfluidic approach to solid form screening allows for elucidation of novel solid forms. We have validated microfluidic platforms based on free interface diffusion (Thorson et al., LOC, 2011) and controlled evaporation (Goyal et al., LOC, 2013) that reduce the amount of API needed per solid form screening condition by an order of magnitude (from 5 mg to 5 μg for each conditions), with comparable results to traditional evaporation based solid form screening experiments. Reduction in sample quantity allows scientists to perform solid form screens earlier in the drug discovery process when minimal amounts of API are available, and allows for more extensive screen enabling the discovery of novel solid forms. We designed the microfluidic platforms to be optically transparent allowing for easy identification of crystalline solids, and to show minimal signal in Raman spectroscopy and X-ray diffraction allowing for on-chip identification of solid forms (Goyal et al., Crys. Growth & Des., 2012). Currently, we are pursuing research towards solving crystal structures of unknown cocrystals by utilizing our microfluidic platform to grow diffraction quality crystals. This work is in collaboration with AbbVie.

(3) Microfluidic platforms for cell studies

Microfluidic platforms provide several characteristics that better facilitate studying cellular and inter-cellular processes compared to traditional petri dish- or well plate-based techniques. Examples include the ability to study single cells in highly controlled environments, superior control over the cellular microenvironment in space and time, and convenient integration with different types of microscopy. In our group, we develop microfluidic platforms for the following applications:

3a. Antibiotic susceptibility testing:

Effective treatment of clinical infections is critically dependent on the ability to rapidly screen patient samples to identify the susceptibility of the infecting pathogens to antibiotics. Existing methods for antibiotic susceptibility testing (AST) suffer from several issues, including long turnaround times (days), excess sample and reagent consumption, poor detection sensitivity, and limited combinatorial capabilities. These factors preclude the timely administration of appropriate antibiotics, complicating management of infections and exacerbating the development of antibiotic resistance.


To address these issues, we develop microfluidic platforms for AST that provides several advantages compared to conventional methods, including higher detection sensitivity, rapid results (<6 hours), reduced consumption of reagents, and more quantitative results. For example, in collaboration with Prof. Schroeder we have used our microfluidic platforms for studying the susceptibility of various pathogenic bacteria, such as E. coli, P. aeruginosa, and K. pneumoniae, against different antibiotics (Mohan et al. Biosens. & Bioelect., 2013). We have also used the platform to study the interaction amongst different species of bacteria (polymicrobial cultures) and the effect of these interactions on antibiotic susceptibility. Currently, we are applying the microfluidic platform in conjunction with using the resulting experimental data for pharmacokinetic- pharmacodynamic (PK/PD) modeling to provide better information towards the best way to treat a given infection.

3b. Studying cells under controlled oxygen conditions:


As tumors grow outward away from the local vascular architecture formation of variable hypoxic (sub-physiologic tissue oxygenation) regions occur throughout the solid mass. These hypoxic regions have been associated with therapeutic resistance, metabolic reprogramming, and the epithelial-mesenchymal transition. Many questions remain regarding the effects of hypoxia on these outcomes, yet only few methods enable both precise control over oxygen concentration and real-time imaging of cell behavior. Microfluidic platforms are particularly well suited to control oxygen concentration while enabling real-time imaging due to their control over temporal and spatial chemical conditions. In addition to control over the local microenvironment, the reduced length scale in microfluidic platforms compared to conventional methods provides shorter equilibration times. Utilizing the advantages of microfluidic platforms, we have developed an arrayed device capable of controlling oxygen concentration from 0.5% to 21%. In collaboration with Professor Rex Gaskins (Department of Animal Sciences), we utilize these platforms to study real-time changes of organellar redox potential in cancer cells under hypoxia.

(4) Chemical synthesis in microreactors

Microreactors provide several advantages for the study and actual execution of chemical synthesis compared to traditional ‘wet-lab’ approaches. For example, smaller, precisely engineered platforms provide enhanced heat and mass transfer, reduced consumption of reagents, and are more amenable for automation. In our group, we develop microreactors for the following applications:

4a. Synthesis of radiopharmaceuticals:

kenis7_0Radiopharmaceuticals are a class of drugs used in the diagnosis and treatment of several diseases and disorders, including certain types of cancer and heart diseases. The amounts of the precursors for the synthesis of these drugs are typically small (a few microliters) due to limited availability, high costs, and upper limits on the amount of radioactivity that can be handled safely. The inability of the conventional ‘wet-lab’ methods to efficiently manipulate low reagent volumes not only leads to synthesis of low-quality drugs for clinical applications, but also hampers the development of new drugs. We try to address these issues by developing microfluidic technologies, or better microreactors, for the synthesis of these radiopharmaceuticals. By integrating different microfluidic modules, we envision that these compounds can be made much more reliably and at higher yield.

We have shown that the microfluidic technologies provide several advantages for each step compared to conventional methods, including improved reaction yields, reduced consumption of reagents, and amenability for automation (Goyal et al., Sens. & Act. B, 2014; Hairong et al., LOC, 2014; Hairong et al., Bioconj. Chem., 2014; Zeng et al., Nuc. Med. & Bio., 2013; Wheeler et al., LOC, 2010). Currently, we are further optimizing the microreactors and developing an integrated system for clinical and research use. This project is in collaboration with Prof. David Reichert’s research group in the department of Radiological Chemistry at Washington University, St. Louis.

4b. Microreactors for quantum dot synthesis:

kenis9Fluorescent semiconductor nanoparticles show promise in solid-state lighting and display technology due to significantly higher photoluminescence and better spectral behavior than conventional phosphor technology. These nanoparticles also have potential uses in medical imaging and quantum computing. High production costs due in part to a lack of reliable methods for the production of high quality, monodisperse nanoparticles currently greatly inhibit their widespread usage. Conventional batch synthesis methods suffer especially from batch-to-batch variation of nanomaterial quality. Batch syntheses, owing to slow heat and mass transfer lack the ability to precisely control on size, morphology and composition of nanoparticles. Continuous flow reactors provide potential solution to these problems. The efforts in Kenis group are focused on development of high throughput continuous reactors affording fast mixing and heating times at high temperatures to synthesize high quality semiconductor nanoparticles of varying composition and morphology. For example, we successfully synthesized nanorods using one of our continuous flow reactors (see figure). We are studying both Cd containing and Cd-free systems, reaching quantum yields as high as 60%, which is comparable to commercial products.

(5) Manufacturing technologies for microfluidics

In our research group, we explore various manufacturing technologies to advance the development of microfluidic devices. The focus in this area is to facilitate the integration of microfluidics with end-applications. Currently, we are pursuing research in two directions:

5a. Microfluidic components to enhance portability and scale-out of devices:

The advent of very large scale integration (VLSI) microfluidics has enabled multi-step and high-throughput applications with massively parallel operations to be performed on a single chip. Key to these advances was the development of pneumatic microvalves, which are fabricated with soft-lithographic techniques. Despite successful integration of such pneumatic microvalves in microfluidic chips for diverse applications, these microvalves require bulky ancillaries, which limit the scalability and portability of these microfluidic chips. We address these issues in two ways:


Use of a normally-close (NC) valve architecture valve architecture: Devices employing conventional normally-open (NO) valves have limited portability in applications that require continuous closed state for operation, as these valves need bulky ancillaries (pumps, nitrogen gas cylinders, pneumatic peripherals) for actuation. NC valves not only address the above limitation of restricted portability, but also retain the ease of fabrication and integration into microfluidic devices. To enable integration of NC valves, we used a combination of analytical and computational modeling, and systematic experiments to formulate design rules for developing optimal NC valves with the objective of minimizing actuation pressures and facilitating fabrication of these valves (Mohan et al., Sens. & Act. B, 2011). The figure shows the actuation pressure needed as a function of the width of the fluid channel for different microvalve shapes (straight, v-shaped, anddiagonal). We have used these valves for a variety of applications, such as protein–antibody interactions virus detection, protein crystallization, solid form screening, and exploring other applications (Schudel et al., LOC, 2011; Thorson et al., CrystEngComm, 2012; Guha et al., Sens, & Act. B, 2012; Mohan et al., Biosens. & Bioelect., 2013; Tice et al., JMEMS, 2013).


Use of electrostatic microvalves to replace or supplement pneumatic microvalves: Our microvalves based on electrostatic actuation retain the small footprint ( 1), for membrane thicknesses ™ of 5 μm. The design parameter space is estimated for the presence of air (darker), oil (hatched), or water (lighter) in the fluidic channel. Another interesting application that we are exploring is the use of electrostatic microvalves to control pneumatic microvalves. This combination of pneumatic and electrostatic microvalves will greatly simplify the ancillaries, and aid in realizing the goal of ‘lab-in-a-chip’ rather than the ‘chip-in-a-lab’.

5b. New materials and fabrication processes:


Poly(dimethylsiloxane) or PDMS has been the preferred material for fabrication of microfluidics devices, mainly because the use of PDMS allows for simple, rapid, and inexpensive fabrication of devices with varying degrees of complexity. However, PDMS suffers from several limitations, a key one being incompatibility with a wide range of organic solvents and analytical techniques. In our research group, we are exploring a variety of polymeric materials as an alternative to PDMS to manufacture microfluidics devices; some of these materials are thiolene, cyclic-olefin copolymer and Teflon. We used these materials to develop microfluidic devices that are compatible with a range of organic solvents and analytical techniques, such as X-ray and Raman. We also show that hybrid devices, which combine the advantages of different materials, are superior alternatives to devices comprising one or two materials.

(6) Emerging microfluidic ‘bio’ projects

6a. Microfluidic platforms for time-resolved FTIR spectroscopy:

Our overall goal is to develop an innovative microfluidic technology for time-resolved Fourier-transform infrared (FT-IR) spectroscopy of biomolecular reactions or interactions. Protein folding, enzyme catalysis, and protein-ligand interactions are critical to maintaining healthy cells and tissues. The root of many chronic or genetic diseases can be traced back to the malfunction of such reactions in proteins – e.g., plaque formation by misfolded beta-amyloid peptide in Alzheimer’s disease.


Investigations to reveal reaction mechanisms at the molecular and intermolecular level are essential for developing novel therapeutics from rational drug design as well as to their testing – e.g., beta-amyloid folding pathways can reveal targets on which candidate drugs against plaque formation can be tested and optimized. Fourier transform infrared (FTIR) spectroscopy provides several advantages compared to other spectroscopy techniques, including non-requirement of extrinsic labeling, simple sample preparation, and easy acquisition of a range of information (high-resolution molecular details to low-resolution protein-protein interactions).

However, several limitations with current FTIR flow cells, including low time-resolution, cost, and requirement of large sample volumes, have prevented the wide-spread use of FTIR. We address these issues by developing microfluidic FITR flow cells out of low-cost, IR-transparent materials. Preliminary results with ubiquitin have validated our approach and we are optimizing the flow cell for conducting experiments with clinically relevant proteins. This project is in collaboration with Prof. Rohit Bhargava in the Department of Bioengineering.

6b. Microfluidic technologies for improving the islet transplantation process:


Diabetes is a devastating disease that affects 25.8 million Americans (8% of population). Human islet transplantation is a promising therapy for Type I diabetes mellitus (TIDM). This procedure, however, is not very reproducible and consistent. To improve the outcomes of islet transplantation, several clinical, biological, and engineering issues need to be addressed. In our research group, we are developing microfluidic technologies to address some of these issues, including maintenance of optimal conditions during isolation of islets from donor pancreas, automation of the islet isolation and separation process, and preservation of the islet viability and functionality during the transplant process. This project is in collaboration with Prof. Jose Oberholzer’s research group in the Division of Transplant Surgery at the University of Illinois at Chicago.

6c. Microfluidic platform for freeze-quench EPR studies:


Most of the interesting phenomena in many biochemical reactions occur during the first few milliseconds of the reactions, e.g., ATP synthesis mediated by the cytochrome bc1 complex. Structural and functional studies of these early-stage intermediate products will not only elucidate the mechanism of these reactions, but will also enable rational design of drugs to treat diseases and disorders associated with the malfunctioning of these reactions. Freeze-quench electron paramagnetic resonance (EPR) is a powerful technique to study these reactions, where the intermediate products of these reactions are rapidly frozen to prevent further reactions and later analyzed using EPR. However, the limitations of the current apparatus for freeze-quench EPR, mainly the slow mixing of reagents, has prevented the application of this technique to study ultra-fast biochemical reactions. In our research group, we are developing a microfluidic device for rapid mixing of reagents (~20 µs) and subsequent ejection of the mixed reagents in the form of an ultra-thin jet onto a frozen copper wheel set-up. We have validated this approach with a model biochemical reaction and are exploring the application of clinically relevant biochemical reactions. This project is in collaboration with Prof. Tony Crofts from the Department of Biochemistry.

6d. Determining pharmaceutical-target interactions:


All of biology, and by extension all of pharmacology, depends on the interaction of proteins with other molecules. Electron Paramagnetic Resonance (EPR) combined with Spin Labeling (SLEPR) can be used to detect such interactions in real time, in vitro or in vivo, and to track the ratio of bound to unbound proteins, with minimal perturbation of the biology. This makes it an ideal tool to directly study the effects of pharmaceutical agents on their biological target and on related biochemical systems, improving the accuracy of early stage development predictions of efficacy and toxicity of drug candidates. However, current wet-lab methods for preparation of the small samples required by EPR spectrometers tend to be wasteful, imprecise, and slow (taking 24 hours or more). In our group we are developing devices for rapid and precise labeling of proteins, taking full advantage of the combinatorial nature of microfluidic chips to create a series of samples at multiple concentrations or with a variety of partners, and incorporating on-chip cell culture when necessary. This project is in collaboration with New Liberty Proteomics.