Page 290 - Optofluidics Fundamentals, Devices, and Applications

264 Cha pte r Ele v e n

microfluidic chip containing a channel (width = 50 μm, height = 15 μm)
on top of the sensor chip. The system was illuminated with a halogen
2
lamp (~20 mW/cm —approximately equal to the intensity of sunlight).
The microfluidic channel was designed with a smooth funnel at
both ends. The channel was oriented at a small angle θ= 0.05 rad
with respect to the aperture arrays, which ensured that approxi-
mately 100 apertures in each row spanned the channel. Oxygen
plasma was used to make the inner surface of the PDMS microfluidic
channel hydrophilic. Prior to use, we additionally flushed the chan-
nel with a PEG solution (10% concentration) to reduce potential sam-
ple adhesion to the channel walls. We chose to operate the completed
system in the upright mode (Fig. 11-2a), so that gravity can drive the
flow and eliminate the need for bulky external pressure pumps.
When the specimen solution (newly hatched C. elegans at a number
concentration of ~20 per μL) was injected into the top funnel, the
solution would wet the microfluidic channel and the specimens
would be pulled continuously into the microfluidic channel by grav-
ity. To prevent excessive nematode wiggle motions, we immobilized
them by subjecting them to a 70°C heat bath for 3 min. The maximum
observed throughput was approximately five worms per minute.
However, the nematode flow speed v in the channel was fairly uni-
form and was approximately 500-μm/s. Imaging of each nematode
required approximately 2.5 s.
The OFM sampling scheme effectively establishes a virtual sens-
ing grid. Unlike the physical sensing grid in CCD and CMOS image
sensors, the pixel density of the OFM virtual sensing grid can be
adjusted by changing the number of apertures spanning the channel,
the flow speed of the target objects, and the pixel readout rate. For
our prototype, the grid spacing along the Y direction equals δY =
Lsinθ= 0.5 μm, and the grid spacing along the X direction equals δX =
v/f = 0.5 μm. We note that pixel density is distinct from system’s reso-
lution. In the case of the OFM, the pixel density is not limited by the
aperture size. Higher pixel density is helpful as it allows us to over-
sample the object and prevent undesirable aliasing artifacts from
appearing in the images.
Figure 11-3a shows a pair of OFM images acquired by the two OFM
arrays from the same wild-type C. elegans L1 larva. The image correla-
tion between them is 56%. Consistent internal structures are found in
both OFM images. For comparison, Fig. 11-3b shows an image collected
from a similar nematode that was placed directly onto an unprocessed
CMOS sensor (note that the pixel size is 9.9 μm × 9.9 μm); the nematode
was barely distinguishable in this poor-resolution direct projection
image. Figure 11-3c shows a conventional microscope image of a similar
larva acquired through a 20× Olympus objective lens (650-nm resolution
for 555-nm wavelength under Sparrow’s criterion) [4]. Similar internal
structures of C. elegans appear in both the microscope and the OFM