Microfluidics is a traffic control system for sampling, sorting, and mixing mesocopic objects. The objects are often biological---cells, proteins, chromosomes in a solvent---and the platform is often a lithographically patterned chip on which fluids are urged through microchannels using volts, heat, or even peristaltic pressure. Microfluidics was a large topic at this week's March Meeting of the American Physical Society (APS) in Austin, Texas (http://www.aps.org/meet/MAR03/baps/index.html ). Here are some highlights.

Carl Hansen (Caltech) described a device with the largest degree of integration yet achieved: a chip with 1000 250-picoliter chambers with attendant valves for controlling flow and mixing (see also Science, 18 October 2002). Another device in the Caltech lab of Stephen Quake allows the careful metering of reagents in order to facilitate protein crystallization under a variety of conditions (pH, viscosity, surface tension, 48 different solvents, etc.) on a huge scale (144 parallel reactions can take place) and with a minimum of means---only 10 nl of precious protein samples are needed, 100 times less than with usual methods (see also Proc. Natl. Acad. Sci., 24 Dec 2002). In this way, many proteins have been turned into crystals, often in the space of hours rather than days. Indeed some protein species were crystallized for the first time. The crystals can then be bombarded with x rays in order to determine molecular structure.

David Grier (Univ. Chicago) reported on a method called holographic optical tweezers, in which a beam of laser light, sent into a hologram, is divided into a myriad of sub-beams which can independently suspend and manipulate numerous tiny objects for possible transportation, mixing, or reacting. Grier showed movies of ensembles of micro-spheres moved into patterns and even set to spinning by the holographically sculpted light fields. Applied to fluid samples of biomolecules, the holographic multiplexing produces what Grier calls "optical fractionation," an optical equivalent of gel electrophoresis, in which electric fields are used differentially to drive and separate macromolecules. In the flexible Chicago approach, there is no viscous gel, and a deft change in the computer-generated hologram or the laser wavelength can quickly bring about sorting of objects ranging from the 100-nm size (viruses) up to the 100-micron size scale.

Meanwhile, Jochen Guck (Univ. Leipzig) subjects fluid-borne cells to a pair of laser beams which stretch the cells and probe their elasticity. In general sick cells are softer (by a factor of 2 to 10) than healthy cells. In this way, Guck's "optical stretcher" can "feel" the difference between normal and abnormal at a rate of hundreds of cells per hour, compared to typical rates of 10 cells per day using other elasticity-measuring methods, thus reducing the need for biopsies requiring larger tissue samples. The Leipzig device might even be able to tell the difference between ordinary cancerous cells from the even softer metastasizing-capable cells.