A compact and affordable optical trap for cell sorting
Optical trapping in a microfluidic device uses an ordinary diode laser bar.
David L. Shenkenberg
Microfluidic devices can be used to perform cellular assays for biomedical research, for drug screening and for medical laboratory tests. The devices fit on a microscope stage and can be manufactured inexpensively through lithographic methods already being used in factories.
Fluid flow drives cells through channels in microfluidic devices; however, without a control system, fluid flows somewhat haphazardly. An optical trap can manipulate cells in microfluidic channels precisely and safely, and it can enable label-free characterization via interferometry or other methods. However, the method usually requires a bulky and expensive laser system, which defeats the purpose of having a small and low-cost microfluidic device. Now an ordinary diode laser bar has been integrated with a microfluidic device for the optical trapping of cells, thanks to a team from Colorado School of Mines and from a spin-off company, Metafluidics Inc., both in Golden, Colo.
According to professor David W.M. Marr, the Ti:sapphire, Nd:YAG or Nd:YLF laser systems typically used cost many thousands of dollars, whereas diode laser bars cost about $60. And whereas conventional solid-state laser systems occupy a tabletop, the optical trap and microfluidic device fit in the palm of a hand; the bar alone occupies an area of merely 200 × 1 μm.
The microfluidic device could count cells such as white or red blood cells, and it could sort cells based on physical appearance or on expression of certain genes. Marr said that the small size makes the device much more portable than a flow cytometer. As such, it could be used as a point-of-care device, providing doctors and patients with instant results. In contrast, flow cytometry remains mostly restricted to specialized laboratories, which is why medical laboratory results do not come in until after a couple of days.
Most optical trapping setups have used a single Gaussian beam from a continuous-wave laser focused through a high-numerical-aperture microscope objective. The beam creates an optical intensity gradient that must overcome gravity and scattering forces to capture a particle in three dimensions. In contrast, the system developed by Marr and the other researchers uses a nearly two-dimensional microfluidic channel that restricts the motion of the cells, greatly reducing the optical requirements and enabling them to use a common diode laser bar.
In their experiments, the researchers used optical trapping to manipulate polystyrene beads ranging from 2.0 to 10.7 μm in diameter because the beads are similar to cells in terms of size, shape and refractive index. Optical trapping of these beads within the microfluidic device was imaged using a homebuilt microscope with a CCD camera and a 10×, 0.25-NA objective. The microfluidic device contained several simple linear channels with one input and output.
Underneath and parallel to the microfluidic device, the researchers placed a diode laser bar with an integrated cylindrical microlens (Figure 1). The diode laser bar emitted 808-nm light with 2 W of average power, a wavelength that mostly is not absorbed by cells and that has an optical power conveniently high for optical trapping. They used an ordinary polymethyl methacrylate fiber to focus the emission from the bar onto the microfluidic device. A 1-mm-diameter fiber provided the best balance between light collection with minimal losses and the refractive index, which was 1.49. Marr said that such fibers cost a few cents.
Figure 1. Researchers integrated a microfluidic device with a diode laser bar for optical trapping. Using a microscope and CCD camera, they imaged the optical trapping of polymer beads with this setup. Images reprinted with permission of Applied Physics Letters.
In results detailed in the Jan. 7 issue of Applied Physics Letters, the researchers observed that particles were pushed near the microfluidic channel wall, whereas a traditional Gaussian optical trap would hold particles in the center of the microfluidic channel.
The required laser intensity for trapping is lower near the wall because the fluid velocities are lower. As such, they discovered that they could use the diode laser bar to trap polystyrene beads both large and small, whereas the high fluid velocities in the center of the channel make it difficult to trap larger particles with a Gaussian trap. The requirements for lower laser intensities and lower fluid velocities would place less stress on cells compared with traditional devices, the researchers said. Marr noted that these fluid velocity gradients also could be exploited for chemical separations.
The scientists developed a mathematical model based on Mie ray optics that enables prediction of the trapping force for a given particle size and laser intensity (Figure 2). They experimentally determined the stretching forces on the polystyrene beads at various optical trapping angles and found that the model agreed well with the experimental results.
Figure 2. Using classical Mie ray optics, the scientists predicted the forces required to trap particles with the diode laser bar (left) versus a traditional Gaussian beam optical trap (right) for a given particle radius (σ) and laser intensity (x). The insets show the maximum restoring force for a particle with a 5-μm radius.
“Our goal is to use cell physical property modeling to predict cell deformation as a function of applied optical and hydrodynamic forces,” Marr said. This information will enable them to optimize the design to ensure the safety of cells.
MORE FROM PHOTONICS MEDIA