Laser beam and sample fluorescence in microfluidic channel. With the laser focus at depth , the collected light cone will have diameter at the top of the channel. , where is the index of refraction in the channel, and is the numerical aperture of the optical system used to collect the light. For and , . Optimal measurements at the full depth therefore require a channel width . In the present mixer design, .
Microfluidic device mixing region. Etched reference marks facilitate precise positioning of the confocal measurement volume. wide areas are used for measurements in the inlet channels at top, left, and bottom. The first position where a measurement can be made using the entire collection cone of an objective with numerical aperture is indicated by a small triangle below the observation channel.
Numerical calculation of flow pattern and photograph of mixing region. Inlet channels narrow from observation areas (see Fig. 2) to an intersection followed by a wide mixing channel. The mixing channel widens into the observation channel. (a) Calculation of ink concentration in the mixing pattern. Black lines are contours at 90% of the initial concentration and 110% of the final concentration. (b) Photograph of ink (Quink, Parker) mixing with water in the microfluidic device.
(a) Four-terminal lumped impedance model used for device design (Ref. 2). Volume flow rates are proportional to pressure drops and inversely proportional to impedances , which for uniform rectangular channels are completely determined by fluid viscosity and channel dimensions. (b) Photograph of central inlet region. The fluid inlet, extending through the thickness of the device, is formed by casting to prevent rough edges and damage to the channel pattern. The PTFE post used for this purpose is centered within a large waffle-patterned pad (see text). The wide central inlet channel is made to follow a serpentine path, increasing impedance and enabling precise control over flow rates and mixing.
Interface cross section. (a) Removable pressure cap with optical port. (b) Aluminum pressure manifold. (c) PTFE chip carrier with sample wells. (d) PDMS microfluidic device bonded to coverglass. (e) Spring-loaded aluminum clamping plate. Screws holding the chip carrier in the manifold are not visible in this section.
Assembled microfluidic mixing system (compare to Fig. 5). A microfabricated PDMS device, bonded to a fused-silica coverslip, is held against the PTFE chip carrier by a spring-loaded clamping plate. A large window in the clamping plate allows access for a microscope objective with high numerical aperture. Raised bosses around the chip carrier sample well outlets form seals with the inlets on the back side of the mixing device. Face-type static O-ring seals (Fig. 5) allow chip carriers with mounted devices to be rapidly swapped into and out of the aluminum pressure manifold.
Mixing system mounted on single-molecule instrument. The removable pressure cap enables sample exchange using gel-loading pipette tips. Tubing connected to the aluminum manifold delivers precision-regulated compressed air for driving flow in the microfluidic mixer. The microscope objective (not visible) screws into the stainless steel beam below and to the left of the mixing system.
Composite casting dish. (a) PTFE posts embedded in the lid are used to cast smooth fluid inlets through the PDMS devices. (b) Complete casting assembly. When the lid is screwed down, inlet posts make firm contact with the silicon wafer. PDMS flows into the mold through holes near lid corners. The current design fits seven devices on a 100 mm diameter silicon wafer.
Fluorescence measurement in the microfluidic mixer. Single-molecule bursts are seen as a 15 pM solution of Alexa Fluor 488 hydrazide dye in distilled water flows through the confocal detection volume at the center of the observation channel. The solution is fed into all three inlets at a pressure of 6.9 kPa with respect to the outlet. Under these conditions, the flow velocity at the confocal volume, located downstream from the mixing intersection, is 1.0 mm/s . The measured background level is .
Results of numerical mixing calculations (see text). Pressure (relative to the outlet) is applied to the mixer inlets, resulting in mixing delay , dead time , and central streamline velocities at the first observation point and in the asymptotic fully developed flow. and are defined as the transit times from the 90% contour [see Fig. 3(a)] to the 110% contour and the first observation point (Fig. 2), respectively.
Channel dimensions and flow impedances. Values listed are for major segments of the mixer channels. Unlisted segments (tapers and premixing observation regions) add on the order of 1% to the tabulated inlet channel impedances. Impedances vary linearly with viscosity. Those shown here are calculated using (water). Dimensions and are shown in Fig. 1.
Article metrics loading...
Full text loading...