Separation of bacteria from complex samples constitutes a difficult engineering problem with important ramifications to food safety and health care. Not only pathogenic and non-pathogenic bacteria are morphological very similar and small (around 2 microns in diameter), but also, very low concentration of bacteria (e.g. 10 Colony Forming Units, CFU) can pose real threats to human health. Current technologies for bacterial detection rely on DNA analysis and can be fast (e.g. PCR) if the bacterial count is large enough and if bacteria are suspended in "clean samples". But real samples are complex (e.g. feces, food and blood) and they often contain solid particulates, biopolymers, fibers, foreign DNA, eukaryotic cells and a myriad of non-pathogenic bacteria that make the direct use of PCR impractical. Therefore, sample preprocessing (1-2 days) is almost always required, including mechanical separation (e.g. separation by centrifugation) and bacterial enrichment in selective media to increase bacterial count and raise the signal to noise ratio. Transformative technologies for the rapid separation of low count of bacteria from complex samples would allow to address the current bottleneck in bacterial detection, and could have a profound impact in health care, food safety water treatment etc.
We are developing a microfluidic platform to separate bacteria using their ability to swim towards some chemical compounds (chemotaxis), thereby obviating the need for mechanical separation (bacteria will self-separate from the complex sample), and selective enrichment (we will use selective chemo-attractants to pull out specific strains). Engineered microfluidic devices are designed and used to systematically quantify the difference of chemotactic behavior between bacterial species and strains, and, in turn, these studies will inform and guide the design of new separation technologies for the selective fractionation of specific bacteria.