Microfluidics explained

How fluids behave in channels thinner than a hair, why that is useful, how the chips are made, and what still makes them hard to commercialize.

Reference guide written and published by biochip.com. Draft, not yet independently reviewed; last substantive update September 7, 2026.

Microfluidics is the science and engineering of fluids moving through channels that are tens to hundreds of micrometres across, roughly the width of a human hair down to the size of a cell. At that scale fluids stop behaving the way they do in a beaker. Flows are smooth and predictable, mixing happens only by diffusion, surface forces dominate gravity, and heat moves in and out almost instantly. Microfluidic engineers use those properties to move, split, mix, heat and sort picolitre-to-microlitre volumes with precision that is hard to achieve by hand. The technology underpins lab-on-a-chip diagnostics, droplet-based single-cell sequencing, inkjet printing and organ-on-a-chip culture systems.

Why small channels behave differently

The key quantity is the Reynolds number, a ratio of inertial forces to viscous forces. In kitchen plumbing it is in the thousands and flow is turbulent. In a 100-micrometre channel it is typically below 1, so flow is laminar: two streams introduced side by side run in parallel without swirling together, and only molecular diffusion blends them. This is a nuisance when you want to mix and a gift when you want to keep reagents separate, create precise chemical gradients, or focus a stream of cells into single file. Squires and Quake's review lays out this physics carefully, including the scaling of diffusion, surface tension and electrokinetic effects with channel size (Squires and Quake, Rev. Mod. Phys. 2005).

Two consequences matter for nearly every device. First, surface-to-volume ratio is enormous, so the chemistry of the channel wall (what sticks to it, whether it is wettable) can dominate the behaviour of the liquid inside. Second, capillary forces are strong enough to pull liquid through a channel with no pump at all, which is how paper-based and many plastic point-of-care tests work.

How microfluidic chips are made

The field grew out of the silicon micromachining of the 1980s, but silicon and glass are expensive to pattern and opaque or fragile in inconvenient ways. The turning point for academic research was soft lithography in PDMS (polydimethylsiloxane), a transparent, gas-permeable silicone rubber. A channel design is written into a photoresist master by photolithography, liquid PDMS is poured over it and cured, and the peeled-off slab is bonded to glass. Duffy and colleagues described the workflow in 1998 (Duffy et al., Anal. Chem. 1998), and it let a graduate student go from drawing to working device in a day. Stacking two PDMS layers with a thin membrane between them yields pneumatic valves and pumps, and thousands of them can be integrated on one chip (Unger et al., Science 2000).

PDMS is rarely used in commercial products because it absorbs small hydrophobic molecules, swells in solvents and is hard to mould in volume. Production devices are typically injection-moulded or hot-embossed thermoplastics such as COC, PMMA or polycarbonate, sometimes glass for optical or high-pressure work. At the other extreme, paper microfluidics patterns hydrophobic barriers onto paper so that capillary wicking routes a sample to test zones; Martinez, Whitesides and colleagues introduced the approach in 2007 as a route to very cheap diagnostics (Martinez et al., Angew. Chem. 2007).

Ways of moving liquid

  • Pressure-driven flow from syringe pumps or pressure controllers: simple and dominant in research, but external hardware is bulky.
  • Capillary flow: no power needed; the basis of lateral-flow tests and paper devices, at the cost of limited control.
  • Electrokinetic flow: electric fields move liquid and charged molecules; used in chip electrophoresis.
  • Centrifugal ("lab-on-a-disc"): a spinning disc drives liquid outward through channels; used in several commercial blood analyzers.
  • Droplet microfluidics: aqueous droplets in oil act as millions of tiny independent reactors; this is what makes high-throughput single-cell RNA sequencing possible.
  • Digital microfluidics: droplets are moved on an electrode grid by electrowetting, with no channels at all.

What microfluidics is used for

Whitesides' 2006 review traced the field's roots to analytical chemistry, genomics and defence-funded detector programmes, and predicted that its largest impact would be in biology (Whitesides, Nature 2006). That has largely proven true. Droplet devices enabled single-cell genomics at scale. Microfluidic cartridges sit inside many molecular diagnostic instruments. Inkjet print heads, the largest microfluidic product by volume, predate the field's name. Organ-on-a-chip systems use microfluidic perfusion to keep cultured tissue alive under realistic flow. Sackmann, Fulton and Beebe assessed, with some frankness, why adoption in ordinary biology labs has nonetheless been slower than the technology's advocates expected (Sackmann et al., Nature 2014).

Limits and open problems

The world-to-chip interface. Getting a real sample into a chip, connecting tubing without leaks or bubbles, and reading the result usually costs more effort than the microfluidic part. Bubbles and clogging are persistent failure modes in small channels. Surface effects mean that proteins and cells adsorb to walls and change device behaviour over time. Manufacturing transfer from PDMS prototypes to moulded plastics often changes performance and requires redesign. Standardization is weak: there is no equivalent of the electronics industry's standard packages and connectors, so each device is a custom system. These are engineering problems rather than physical barriers, but they explain why many impressive academic demonstrations never become products.

Related terms

  • Lab-on-a-chip: a device that uses microfluidics to integrate several laboratory steps; see the lab-on-a-chip guide.
  • Nanofluidics: channels below about 100 nanometres, where the electrical double layer and single-molecule effects dominate.
  • Soft lithography: patterning by moulding, stamping or embossing elastomers rather than etching silicon.
  • Laminar flow: smooth, layered flow with no turbulence, the normal regime inside microchannels.

References

  1. Squires TM, Quake SR. Microfluidics: fluid physics at the nanoliter scale. Reviews of Modern Physics, 2005.
  2. Duffy DC, McDonald JC, Schueller OJA, Whitesides GM. Rapid prototyping of microfluidic systems in poly(dimethylsiloxane). Analytical Chemistry, 1998.
  3. Unger MA, Chou HP, Thorsen T, Scherer A, Quake SR. Monolithic microfabricated valves and pumps by multilayer soft lithography. Science, 2000.
  4. Martinez AW, Phillips ST, Butte MJ, Whitesides GM. Patterned paper as a platform for inexpensive, low-volume, portable bioassays. Angewandte Chemie International Edition, 2007.
  5. Sackmann EK, Fulton AL, Beebe DJ. The present and future role of microfluidics in biomedical research. Nature, 2014.