Joint research by our group at IBM Research Europe in Zurich and collaborators at Technion, Israel, has led to a new method for separation of particles and molecules from small volume samples. The technique exploits differences in diffusivity – a molecular property which correlates well with size. We are currently adapting the method for rapid and direct detection of coronavirus from throat swabs.
Sometimes, slowness can come in handy in terms of moving faster. As counterintuitive as that may sound, it is exactly what we used in creating a new method to separate particles and molecules in a microfluidic device. The details are discussed in our paper Tunable bidirectional electroosmotic flow for diffusion‐based separation published recently in Angewandte Chemie and designated by the journal as “Very Important Paper”. Our new method and device take advantage of differences in diffusivity of particles and molecules immersed in liquids in a way that the larger, less diffusive entities basically get “carried away” in the direction of flow of the liquid while their smaller and more diffusive counterparts lag behind.
Video caption: Experimental visualization of the bidirectional flow.
The device makes use of virtual channels, a concept that we presented a year ago in a paper in Proceedings of the National Academy of Sciences, wherein unique flow fields can be generated in a microfluidic chamber using electric field actuation. Here, we have used the technology to create bidirectional flows – alternating stripes carrying fluid in opposite directions. Such a flow field is impossible to create using traditional pumps and valves, and when particles and molecules are introduced into this flow they behave in a well explained, yet initially non-intuitive manner: small particles remain stationary, while large particle flow away quickly.
All particles in a fluid move in random directions in a process called Brownian motion. This is the same mechanism that allows us to smell a small drop of perfume from across the room – the molecules simply make their way randomly in a process also known as diffusion. However, small particles diffuse much faster than large ones, and they move across the opposing streams very quickly. This makes them move very slightly back and forth but overall – stay in place. Larger molecules or particles diffuse much slower and end up being carried away by the flow. Our team calls this method BFF, meaning ‘bidirectional flow filter’. According to one of the paper’s reviewers, this separation mechanism represents “a fundamentally significant contribution to the field that only comes along every 10-20 years”.
Microfabricated device for creating opposing streams of flow of liquid to perform diffusion-based separations.
The underlying principle here is pretty simple. But to our surprise, it hadn’t been done before, most likely because of technological limitations. While developing the concept certainly took time and a number of iterations, with today’s microfabrication capabilities the final device is a rather easy to fabricate solid-state device, which should facilitate production on a large scale.
In the paper, we demonstrated the separation of antibodies and particles from small molecules, and provided the theory and engineering guidelines for separation of a wide variety of biomolecules. The reason this might be very useful is because the majority of biological assays rely on a reaction between a probe and the target molecule in the sample, followed by removal of the excess probe molecules that did not find their target. This last step is often very involved and it is extremely challenging when the volume of the sample is small. Our method accomplishes the separation step very well, provided that the two reacting elements are of sufficiently different size.
We are currently working hard towards adapting the method for rapid detection of coronavirus. Fortunately, the coronavirus is fairly large – about 100 nm in diameter. This is much larger than antibodies or other probes that can be used to bind to it. Using our method, we hope to be able to put a patient’s sample into our microfluidics chip where it will mix with visible probes, and then see only the viruses flowing out while the unbound probes stay behind.
This work was funded by the European Research Council (MetamorphChip) and by the BRIDGE program (project 40B1-0_191549), funded by Innosuisse and the Swiss National Science Foundation.
Tunable bidirectional electroosmotic flow for diffusion‐based separation, Vesna Bacheva Federico Paratore Shimon Rubin Govind V. Kaigala Moran Bercovici, https://doi.org/10.1002/anie.201916699
In a new paper published in Frontiers in Digital Health journal, we present the first empirical evidence of tablet-based automatic assessments of patients using speech analysis — successfully detecting mild cognitive impairment (MCI), the transitional stage between normal aging and dementia.
Our study "Comparison of methods to reduce bias from clinical prediction models of postpartum depression” examines healthcare data and machine learning models routinely used in both research and application to address bias in healthcare AI.