Dr. Eric B. Cummings and Dr. Yolanda Fintschenko LabSmith, Inc. Livermore, CA
A PDF version is available in the July 2011 AES Newsletter.
Introduction: Particle-image velocimetry (PIV) is a flow measurement technique having many uses in microfluidics1-10. PIV works via a comparison of images taken at different times of a flowing stream of densely seeded particles. Usually, cross-correlation math1 is used to estimate a two-dimensional histogram of particle displacement2 between images for one or more locations within an image. The signal-to-noise ratio of this statistically derived 'histogram' scales ideally with the square root of the number of measurements, N (# pixels x # frame-pairs) used to construct the histogram. The histogram is broadened by the finite image size of particles, Brownian motion, Zeta-potential variations, spatially irresolvable flow gradients (out-of-plane parabolic flow) spatially resolvable flow gradients (if spatial averaging), and flow unsteadiness (if time-averaging). Originally developed in the 1980s for macroscale flows, in 1998 PIV was adapted into a technique by Santiago et al. for measuring steady fluid flow in microfluidic devices3. Called micro-PIV, this adaptation typically flood-illuminates a microchannel to excite fluorescently labeled ~100-1000-nm particles, captures particle-fluorescence images through microscope optics, and records or processes arrays of image pairs on a computer to produce measurements of velocity. Some applications process histograms further to extract parameters such as unresolved parabolic flow and temperature from Brownian motion4. Micro-PIV applications in microfluidics2-10 include high-resolution velocity-field imaging2,4 and real-time velocity sensing9,10.
|Figure 1. A. High resolution micro-PIV experiment obtained by processing a 30-second video of fluorescence from 200 nm green-yellow FluospheresTM in an electric field of 2 V/mm applied top to bottom, from a standard NTSC camera. B. Real-time micro-PIV experiment. Pressure driven flow (5 ul/min) was used with a 500 nm diameter Flu- osphereTM seeded solution at pH 7 in a 1 mm wide by 100 um deep microchannel in a Topas chip. Taken with a LabSmith SVM340TM synchronized video microscope black and white camera. Both A. & B. used a 10X DIN objective and blue LED illuminator.|
High Resolution Velocity Field Imaging: If the flow is steady, PIV can produce accurate, single-pixel-resolution2, diffraction-limited flow field images, which are useful for characterizing devices and comparing theory and experiment4. For example, Fig. 1a, shows theoretical and PIV-measured electrokinetic flow fields in a microchannel containing an array of 93-μm circular posts on 200-μm centers. Fringes are contours of constant speed (24.5 μm/ s) and measurements differ from the ideal flow by ~2 μm/s out of a peak of 320 μm/s, mostly due to surface flaws. These PIV measurements were processed from a 30-second video from a standard NTSC camera of 200-nm particle fluorescence through a 10x DIN microscope with a blue-LED ring illuminator, a system similar to LabSmith's SVM340TM. Each histogram was constructed with N = 3600 (2 x 2-pixel x 900 frames). Reference-quality PIV does not require expensive cameras or imaging systems, but it does require 1) large N, 2) sophisticated processing software to optimize dynamic range and ignore flaws such as particle aggregates, and 3) tenacity.
Real Time Micro-PIV: On the other end of the spectrum, PIV is useful as a real-time velocity sensing technique. PIV-based software probes can be "rapidly prototyped" to automate a control loop that would otherwise be hand-tweaked. PIV probes and intensity probes, which track intensity within user-defined regions, can trigger real-time actions, such as changing voltages or flow rates, moving to the next step in a sequence, etc. Configurations can be edited, saved, and opened like any document, so control-system conceptualization and testing is fast and inexpensive. For example, Figure 1B shows real-time velocity-vector and intensity probes of a 5 μl/min pressure-driven microchannel flow processed using LabSmith's uScopeTM software. Every 16 ms, probe measurements are captured, processed, displayed, and recorded.
Conclusions: Micro-PIV can eliminate uncertainty and save development time for researchers and engineers. Using off-line processing, high-resolution flow-field imaging can detail flows and reveal device issues. Real-time software probes provide a low-cost "rapid-prototyping" alternative for developing microsystem sensors and closed-loop controls. Micro-PIV can accelerate microfluidic system prototyping and refinement, key steps for capitalizing a startup and commercializing a technology.