How to make a 3D micromixer

Microfluidics is a rapidly emerging technology that enables miniaturization and integration for biological, chemical, and medical applications. By integrating fluidic functions such as valving, metering, mixing, transport, and separation on a single substrate, microfluidic systems can be used to control and manipulate tiny volumes of liquids with high precision, enabling downsizing of both chemical and biological analyses.¹ Fluid mixing is an essential function required by most microfluidic applications; however, the laminar nature of microflows makes fast and efficient fluid mixing inside microchannels difficult to achieve. Viscous forces are dominant, which produces a smooth, constant flow rather than the turbulence needed for mixing. To overcome this difficulty, Carrière showed that a passive micromixer consisting of symmetrical 3D units could in theory provide efficient mixing by repeatedly dividing a liquid stream into two substreams, rearranging the substreams in 3D microchannels, and rejoining the substreams into a single stream (see Figure 1).² Unfortunately, such 3D microstructures are difficult to make with today's mainstream microfluidic fabrication techniques, which still heavily rely on 2D planar lithography methods, and remains difficult even with the assistance of additional stacking and bonding procedures.

Figure 1. Close-up schematic diagram of the 3D passive microfluidic mixer

As a direct fabrication technique requiring no masks, femtosecond laser direct writing provides a straightforward approach for high-precision, spatially-selective modification inside transparent materials through nonlinear optical absorption. Previously, hollow microfluidic structures embedded in glass have been fabricated by femtosecond laser irradiation followed by chemical wet etching.³ Unfortunately, with this technique the length of the microfluidic channel is usually less than 1cm, due to the limited etch ratio between the areas with and without the femtosecond irradiation. Recently, we have developed a technique to fabricate microchannels in glass with nearly unlimited lengths and with arbitrary geometries.⁴ The main fabrication process includes two steps: direct formation of hollow microchannels in a porous glass substrate immersed in water by femtosecond laser ablation (see Figure 2:a); and postannealing the glass substrate at ∼1150^°C (see Figure 2:b). This consolidates the nanoscale pores by causing them to collapse, while the fabricated microchannels survive due to their larger size. The glass substrate, which is opaque due to scattering by the nanopores, becomes highly transparent after annealing (see Figure 2:c).

Figure 2. The fabrication process. Femtosecond laser direct writing in porous glass immersed in water (a). Postannealing of the porous glass for collapsing the nanopores (b). Microchannels in consolidated glass contain liquid samples without leaking (c).

Figure 3. Schematic diagrams of the 3D passive microfluidic mixer: overview (a) and close-up view images (b). Optical micrographs of the fabricated 3D microfluidic mixer: overview (c) and close-up view micrographs (d). 1D (e) and 3D (f) microfluidic mixing experiments.

We fabricated a 3D micromixer using this method (see Figure 4:a–d for schematics and micrographs).⁵ The channel cross-section was elliptical with a width of ∼50μm and a depth of ∼75μm.⁴ We controlled microchannel dimensions by adjusting machining parameters, such as the numerical aperture of objective, the laser pulse energy and the translation speed. To test the mixer, we combined two fluorescent dye solutions (fluorescein sodium and Rhodamine B) in 1D and 3D micromixers fabricated under the same laser conditions. In the 1D microfluidic channel, efficient mixing cannot be achieved for a much greater distance than in the 3D mixer; after passing through three mixing units of the 3D mixer, the fluids are well mixed (see Figure 4:e and f, respectively).⁵

Figure 4. A microfluidic lantern embedded in glass composed of hollow microchannels (a). Filling the lantern with different fluorescent dye solutions produced green (b) and red (c) colors.

We also fabricated a 3D microfluidic lantern composed of ring-shaped microchannels using this technique (see Figure 4:a). When either fluorescein sodium solution or Rhodamine B solution was injected into the microchannel and excited by a laser operated at 490nm or 540nm wavelengths, the lantern produced either green or red colors (see Figure 4:b and c).

In summary, we have shown that 3D microfluidic components with arbitrary geometries can be directly formed in glass using femtosecond laser direct writing, without using extra stacking and bonding procedures. This makes rapid prototyping possible for 3D microfluidic systems with enhanced efficiency, flexibility, fabrication precision, and cost-effectiveness. Using this technique, we have fabricated a 3D micromixer that shows superior mixing performance in comparison with its 1D counterpart. This opens up promising prospects for a broad spectrum of applications based on compact and complex 3D microfluidic networks. We are now designing and fabricating 3D micro- and nanofluidic chips for DNA and cell analysis.