maximum intensity of the beam as a function of its position across the channel [Fig. 4]. Figure 4 confirms that the intensity from the shaped beams is drastically improved over the control device (no lens system) and the high intensity cross-channel plateaus are formed in the center of the channel. The plateau of uniform illumination-where the beam is within 5% of the maximum intensity in the center of the channel-is measured to be 10.4-and 6.2-lm long, for the 10-and 3.6-lm lens systems respectively, spanning the center of the channel. These plateau regions are difficult to form as the depth of focus is minimized considerably when focusing a beam to a spot in the channel. However, by using aberrations to warp the Gaussian shape, we have formed large regions where uniform excitation is possible across the entire sample flow. In fact, if the plateaus are measured using the standard depth of focus, it is found that these regions span 26.2 and 14.1 lm for the 10-and 3.6-lm devices, respectively-significant portions of the entire channel, and that intensity varies by 25% from the maximum intensity over this region.It should be noted that there is a noticeable trade-off on the size of the beam waist and the plateau of uniform intensity: the smaller the waist, the smaller the plateau that can be formed, as evidenced by the much smaller region defined by the 3.6-lm device. This is due to the inverse relationship between beam waist and divergence. This is of little consequence because when smaller cells are analyzed, our designs shrink the beam waist to eliminate the occurrence of double detections while the uniform region is kept large enough to accommodate the particle to deviate within the cell flow. A larger maximum intensity is also noted for larger beam waist due to the larger input NA of the lens system allowing greater collection of light from the waveguide.
CONCLUSIONSExact formation of ideal beam shapes for cytometric analysis has been demonstrated experimentally. These formed beams are highly desirable in a flow cytometry application because they exploit the ability to control aberrations in the lens design to specifically tailor beam shapes to form as near an ideal a shape for illumination as possible. These bowtie-shaped beams will significantly improve the sensitivity of a photonic-integrated microchip-based flow cytometer by offering an improved area of illumination over previously shown devices [16]. Beam shaping on a photonic-integrated microfluidic flow cytometer is a major step forward to achieving feasible devices for clinical, laboratory, and point-of-care medical applications.
ACKNOWLEDGMENTS