For a dielectric solid surrounded by an electrolyte and positioned inside an externally biased parallel-plate capacitor, we study numerically how the resulting induced-charge electro-osmotic (ICEO) flow depends on the topology and shape of the dielectric solid. In particular, we extend existing conventional electrokinetic models with an artificial design field to describe the transition from the liquid electrolyte to the solid dielectric. Using this design field, we have succeeded in applying the method of topology optimization to find system geometries with non-trivial topologies that maximize the net induced electroosmotic flow rate through the electrolytic capacitor in the direction parallel to the capacitor plates. Once found, the performance of the topology optimized geometries has been validated by transferring them to conventional electrokinetic models not relying on the artificial design field. Our results show the importance of the topology and shape of the dielectric solid in ICEO systems and point to new designs of ICEO micropumps with significantly improved performance.

For a microchamber filled with a binary electrolyte and containing a flat un-biased center electrode at one wall, we employ three numerical models to study the strength of the resulting inducedcharge electro-osmotic (ICEO) flow rolls: (i ) a full nonlinear continuum model resolving the double layer, (ii ) a linear slip-velocity model not resolving the double layer and without tangential charge transport inside this layer, and (iii ) a nonlinear slip-velocity model extending the linear model by including the tangential charge transport inside the double layer. We show that compared to the full model, the slip-velocity models significantly overestimate the ICEO flow. This provides a partial explanation of the quantitative discrepancy between observed and calculated ICEO velocities reported in the literature. The discrepancy increases significantly for increasing Debye length relative to the electrode size, i.e. for nanofluidic systems. However, even for electrode dimensions in the micrometer range, the discrepancies in velocity due to the finite Debye length can be more than 10% for an electrode of zero height and more than 100% for electrode heights comparable to the Debye length.

Microfluidic chips have been fabricated in Pyrex glass to study electrokinetic pumping generated by a low-voltage ac bias applied to an in-channel asymmetric metallic electrode array. A measurement procedure has been established and followed carefully resulting in a high degree of reproducibility of the measurements over several days. A large coverage fraction of the electrode array in the microfluidic channels has led to an increased sensitivity allowing for pumping measurements at low bias voltages. Depending on the ionic concentration a hitherto unobserved reversal of the pumping direction has been measured in a regime, where both the applied voltage and the frequency are low, V rms Ͻ 1.5 V and f Ͻ 20 kHz, compared to previously investigated parameter ranges. The impedance spectrum has been thoroughly measured and analyzed in terms of an equivalent circuit diagram to rule out trivial circuit explanations of our findings. Our observations agree qualitatively, but not quantitatively, with theoretical electrokinetic models published in the literature.

We present a solid polymer dye laser based on a single-mode planar waveguide. The all-polymer device is self-contained in the photodefinable polymer SU-8 and may therefore easily be placed on any substrate and be integrated with polymer-based systems. We use as the active medium for the laser the commercially available laser dye Rhodamine 6G, which is incorporated into the SU-8 polymer matrix. The single-mode slab waveguide is formed by three-step spin-coating deposition: a buffer layer of undoped SU-8, a core layer of SU-8 doped with Rhodamine, and a cladding layer of undoped SU-8.