Microchip-based Electroosmotic Pump Protocol

The electroosmotic pump incorporated in the microchip design is similar to earlier reported designs.^{32 ,44 (link)} The microfluidic components of the pump consist of a tee intersection where the end of the separation channel, the side channel, and short transport channel leading to the electro-spray tip meet (Figure 1). As described above, all of the channels except the side channel were coated with a polyamine to reverse the surface charge on those channels. When all of the channels were filled with the CE background electrolyte (50% methanol, 0.2% acetic acid), the electroosmotic mobility (μ_eof) in the polyamine-coated separation channel was ~5 × 10⁻⁸ m² V⁻¹ s⁻¹ (anodic EOF) based on the measured migration time of injected analyte bands. The electroosmotic mobility in the uncoated side channel was significantly lower in magnitude and in the opposite direction (cathodic EOF). The actual value of the EOF in the side channel was not determined; however, optical imaging of a neutral fluorescent marker (introduced via the side channel reservoir) revealed only a small amount of flow (significantly lower than the flow in the separation channel) under typical operating conditions. When the voltage at the side-channel reservoir was more positive than the voltage at the CE injection cross, the EOF in both the separation channel and the side channel flowed toward the intersection near the ESI corner. The short length of channel between the intersection and the ESI orifice was essentially field-free (due to the relatively large electrical resistance of the air between the microchip and the MS inlet), so there was no EOF in this channel segment. The electroosmotically driven flow entering the tee intersection from the separation and side channels generated a pressure that could easily drive flow through the field-free segment. It was estimated that a pressure of less than 1 mbar would be required to drive liquid through the 150-μm-long field-free channel at a flow rate of 40 nL/min. The separation and side channels had a hydraulic resistance more than 100 times greater than the electrospray channel; thus, the pressure-driven flow in these channels was negligible.