Marcel Mibus

Dielectric breakdown and failure of anodic aluminum oxide films for electrowetting systems

We study electrical properties and breakdown phenomena in metal/aluminum oxide/metal and electrolyte/aluminum oxide/metal contacts, with the aim to achieve a better understanding of failure modes and improve the performance of model electrowetting systems. Electrical conduction in anodic aluminum oxide dielectrics is dominated by the presence of electrically active trapping sites, resulting in various conduction mechanisms being dominant within distinct voltage ranges until hard breakdown occurs. Breakdown voltage depends on its polarity, due to the formation of a p-i-n junction within the oxide; such asymmetric behavior tends to disappear at larger oxide thickness. Electrolyte/dielectric contacts present an even more pronounced asymmetry in breakdown characteristics: a cathodic bias results in breakdown at low voltage, while under anodic bias high field ionic conduction starts before breakdown occurs. These phenomena are interpreted in terms of electrochemical reactions occurring at the surface: cathodic...

Failure Modes during Low-Voltage Electrowetting

Low-voltage electrowetting devices allow significant contact angle changes below a 50 V bias; however, operation under prolonged cycling and failure modes have not yet been sufficiently elucidated. In this work, the failure modes and performance degradation of Cytop (23-210 nm)/aluminum oxide (15-44 nm) bilayers have been investigated. Contact angle and leakage current were measured during stepped voltage measurements up to failure, showing three electrowetting response regimes: ideal Young-Lippmann behavior, contact angle saturation, and dielectric breakdown. The onset of ionic conduction in aluminum oxide and the resulting breakdown control when the layer would ultimately fail, but the thickness of the Cytop layer determined the achievable contact angle versus voltage characteristics. Cyclic electrowetting measurements studied the repeatability of contact angle change using an applied voltage above or below the voltage drop needed for polymer breakdown (VT). Results show repeatable electrowetting below VT and a rapidly diminishing contact angle response above VT. The leakage current and injected charge cannot be used to comprehensively assess the stability of the system during operation. The contact potential difference measured with a Kelvin probe provides an alternative means of assessing the extent of the damage.

Impedance spectroscopy and electrical modeling of electrowetting on dielectric devices

Using impedance spectroscopy, we have determined models for the elements which determine the ac electrical behavior in electrowetting on dielectric (EWOD) systems. Three commonly used EWOD electrode configurations were analyzed. In each case, the impedance can be modeled by a combination of elements, including the solution resistance, the capacitance of the dielectric layer, and the constant phase impedance of the electrode double layers. The sensitivity of the systems impedance to variations in the electrowetted area is also analyzed for these common configurations. We also demonstrate that the impedance per unit area of typical EWOD systems is invariant to bias voltage.

Performance and Reliability of Electrowetting-on-Dielectric (EWOD) Systems Based on Tantalum Oxide

The electrowetting-on-dielectric behavior of Cytop/Tantalum oxide (TaOx) bilayers is studied by measuring their response vs applied voltage and under prolonged periodic cycling, below and above the threshold voltage VT corresponding to the breakdown field for the oxide. TaOx exhibits symmetric solid state I-V characteristics, with electronic conduction dominated by Schottky, Poole-Frenkel emission; conduction is attributed to oxygen vacancies (6 × 1016 cm-3), resulting in large currents at low bias. Electrolyte/Metal Oxide/Metal I-V characteristics show oxide degradation at (<5 V) cathodic bias; anodic bias in contrast results in stable characteristics until reaching the anodization voltage, where the oxide thickens, leading eventually to breakdown and oxygen production at the electrode. Electrowetting angle vs applied voltage undergoes three different stages: a parabolic variation of contact angle (CA) with applied voltage, CA saturation, and rebound of the CA to higher values due to degradation of the polymer layer. The contact angle remained stable for several hundred cycles if the applied voltage was less than VT; degradation in contrast is fast when the voltage is above VT. Degradation of the electrowetting response with time is linked to charge accumulation in the polymer, which inhibits the rebound of the CA when voltage is being applied.

Improving dielectric performance in anodic aluminum oxide via detection and passivation of defect states

The electronic and ionic transports in 32–56 nm thick anodic aluminum oxide films are investigated before and after a 1-h anneal at 200–400 °C in argon. Results are correlated to their defect density as measured by the Mott-Schottky technique. Solid state measurements show that electronic conduction upon annealing is hindered by an increase in the Schottky emission barrier, induced by a reduction in dopant density. Using an electrochemical contact, the films fail rapidly under cathodic polarization, unless defect density is decreased down to 1017 cm−3, resulting in a three order of magnitude reduction in current and no visible gas evolution. Under anodic polarization, the decrease in defect density delays the onset of ionic conduction as well as further oxide growth and failure.

Interrogation of Droplet Configuration During Electrowetting via Impedance Spectroscopy

Several parameters associated with the configuration of a microliter aqueous droplet are determined by the electrical impedance measurements conducted during electrowetting. In each case, the wetted area of a dielectric-coated electrode is first determined from the impedance data, and the known dielectric thickness and permittivity. The wetted area data is then employed in a series of identification problems. In the simplest of these, we demonstrate that the contact angle can be accurately found as a function of applied bias voltage. Thus, the Lippmann-Young (LY) curve of an electrowetting system can be determined from the impedance data alone without the use of any optical measurements. In another test, we show that the volume of a microliter droplet can be estimated from electrical impedance measurements with an error of <;1%. A similar algorithm was employed to determine the surface tension of the interface between the droplet and the surrounding oil to an accuracy of 4%. In still another identification experiment, the impedance data from multiple bias voltages were used as a batch to identify several unknown configuration parameters simultaneously (volume, interfacial tension, and zero-voltage contact angle). It is shown that identification accuracy is degraded in this case, because there exists a direction within the parameter space of low error sensitivity. Determining multiple parameters together, therefore, requires that the ideal LY relation fits the electrowetting behavior well over the range of voltages employed and that the dielectric parameters are accurately known.