Junghoon Yeom

Maximum achievable aspect ratio in deep reactive ion etching of silicon due to aspect ratio dependent transport and the microloading effect

When etching high-aspect-ratio silicon features using deep reactive ion etching (DRIE), researchers find that there is a maximum achievable aspect ratio, which we define as the critical aspect ratio, of an etched silicon trench using a DRIE process. At this critical aspect ratio, the apparent etch rate (defined as the total depth etched divided by the total elapsed time) no longer monotonically decreases as the aspect ratio increases, but abruptly drops to zero. In this paper, we propose a theoretical model to predict the critical aspect ratio and reveal its causal mechanism. The model considers aspect ratio dependent transport mechanisms specific to each of the reactant species in the three subprocesses of a time-multiplexed etch cycle: deposition of a fluorocarbon passivation layer, etching of the fluorocarbon polymer at the bottom of the trench, and the subsequent etching of the underlying silicon. The model predicts that the critical aspect ratio is defined by the aspect ratio at which the polymer etc...

Critical aspect ratio dependence in deep reactive ion etching of silicon

Aspect ratio dependent etching and microloading effects are two mechanisms leading to non-uniformities in the etching of silicon using deep reactive ion etching technology. This paper focuses on the apparent presence of a critical aspect radio when etching three-dimensional MEMS structures. A mask designed to separate the effects of microloading from aspect radio dependent etching was made, and various sizes of features (3 to 1000 /spl mu/m) are etched simultaneously for different etching times (10 to 180 min) using a Bosch etching process. Experimental data exhibiting three distinct regimes for the etch rate, the corresponding critical aspect ratios, and the dependence on microloading effects are presented. Possible mechanisms governing these regimes are also postulated.

The design, fabrication and characterization of a silicon microheater for an integrated MEMS gas preconcentrator

We report the design and fabrication of a microheater unit as a key component of an integrated micro gas preconcentrator that has an ultra-small preconcentrator volume (<0.25 µL) and microvalves for fast injection speeds (<1 ms). Monolithic integration of the microvalves into the microheater of the preconcentrator gives rise to challenges in designing the microheater and implementing thermal isolation for low power and energy consumption. A preconcentrator chamber, 3.5 × 1.5 mm2 in planform area and 40 µm deep, was built in the device layer of a silicon-on-insulator wafer and filled with an array of microposts with a preconcentrator volume of 0.2–0.25 µL. Different generations of the microheaters and their mating dies were fabricated to show the effects of thermal isolation and thermal mass of the system on the performance of the heater. The microheater assembly with the least thermal mass and most thermal isolation can reach 300 °C in 100 ms with 12.3 W of power and is expected to consume less than 2 J during the operation of each preconcentration cycle.

Low Reynolds number flow across an array of cylindrical microposts in a microchannel and figure-of-merit analysis of micropost-filled microreactors

Micropost-filled reactors are commonly found in many micro-total analysis system applications because of their large surface area for the surrounding volume. Design rules for micropost-filled reactors are presented here to optimize the performance of a micro-preconcentrator, which is a component of a micro-gas chromatography system. A key figure of merit for the performance of the micropost-filled preconcentrator is to minimize the pressure drop while maximizing the surface-area-to-volume ratio for a given overall channel geometry. Several independent models from the literature are used to predict the flow resistance across the micropost-filled channels for low Reynolds number flows. The pressure drop can be expressed solely as a function of a couple of design parameters: β = a/s, the ratio of the radius of each post to the half-spacing between two adjacent posts, and N, the number of microposts in a row. Pressure drop measurements are performed to experimentally corroborate the flow resistance models and the optimization scheme using the figure of merit. As the number of microposts for a given β increases in a given channel size, a greater surface-area-to-volume ratio will result for a fixed pressure drop. Therefore, increasing the arrays of posts with smaller diameters and spacing will optimize the microreactor for larger surface area for a given flow resistance, at least until Knudsen flow begins to dominate.

Sensitivity of nanotube chemical sensors at the onset of Poole-Frenkel conduction

We find that the applied electric field has an unexpectedly large effect on the sensitivity of a sensor consisting of a network array of carbon nanotubes. The sensors are insensitive to gas adsorption at low applied voltages and remains low until a critical potential is reached. The sensor response then rapidly increases over a small range of voltage. The critical voltage roughly corresponds to the barrier for electron hopping between defect sites. These results show that there is a correlation between the conduction mechanism in the nanotube and the sensitivity of the nanotube. Such a correlation has not been observed previously.

A regenerable oxide-based H2S adsorbent with nanofibrous morphology

Hydrogen sulphide is found in raw fuels such as natural gas and coal/biomass-derived syngas. It is poisonous to catalysts and corrosive to metals and therefore needs to be removed. This is often achieved using metal oxides as reactive adsorbents, but metal oxides perform poorly when subjected to repeated cycles of sulphidation and re-oxidation as a result of complex structural and chemical changes. Here, we show that Zn-Ti-O-based adsorbents with nanofibrous morphology can sustain their initial reactivity and sulphur removal capacity over multiple regeneration cycles. These nanostructured sorbents offer rapid reaction rates that overcome the gas-transport limitations of conventional pellet-based sorbents and allow all of the material to be used efficiently. Regeneration can be carried out at the same temperature as the sulphidation step because of the higher reactivity, which prevents sorbent deterioration and reduces energy use. The efficient regeneration of the adsorbent is also aided by structural features such as the growth of hierarchical nanostructures and preferential stabilization of a wurtzite phase in the sulphidation product.

Surface energy approach and AFM verification of the (CF)n treated surface effect and its correlation with adhesion reduction in microvalves

The purpose of this paper is to outline an approach that will determine the optimal surface pairs to use in a MEMS device with demonstrated stiction problems. The approach uses the contact angle and atomic force microscopy (AFM) pull-off measurements to predict adhesion at the solid?solid interface. The results are compared to microvalves that have been fabricated with different surfaces at the seat/membrane interface. For microfabricated mechanical devices with surfaces that touch or have a small gap distance, stiction can occur during fabrication or during use. Fabricating multiple devices with different surfaces to determine the lowest stiction can be costly and time consuming. Identifying the surface pair with the least amount of surface adhesion before fabrication can lead to a reduction in device failure due to stiction, and/or time it takes to find the lowest energy surfaces via trial and error. Surface energies are obtained using the van Oss equation based on the contact angle measurements, and surface energy can be used to show the relative adhesions between two surface pairs. An AFM pull-off test is performed using nano- and micro-sized tips to quantify the effect of the different surfaces on adhesion between the solid?solid surface pairs, including SiO2/PI, CFn/PI, CFn/SiO2 and CFn/CFn. The work of adhesion of the surface pairs is obtained using the Derjaguin?Muller?Toporotov (DMT) and Johnson?Kendall?Roberts (JKR) theories. The surfaces treated with a C4F8/Ar plasma to form a CFn coating showed the least amount of surface adhesion. The CFn surface treatment effects on adhesion are quantified and correlated with the reduction in the opening pressure of the microvalve whose interfaces are coated with a CFn film. The AFM pull-off test was more closely related to adhesive forces between the surfaces of the valves as seen in the opening pressure data. The adhesion calculation based on the contact angle measurements predicts the adhesion force with a similar trend but does not take into account the meniscus forces.

Air-Stable Humidity Sensor Using Few-Layer Black Phosphorus

As a new family member of two-dimensional layered materials, black phosphorus (BP) has attracted significant attention for chemical sensing applications due to its exceptional electrical, mechanical, and surface properties. However, producing air-stable BP sensors is extremely challenging because BP atomic layers degrade rapidly in ambient conditions. In this study, we explored the humidity sensing properties of BP field-effect transistors fully encapsulated by a 6 nm-thick Al2O3 encapsulation layer deposited by atomic layer deposition. The encapsulated BP sensors exhibited superior ambient stability with no noticeable degradation in sensing response after being stored in air for more than a week. Compared with the bare BP devices, the encapsulated ones offered long-term stability with a trade-off in slightly reduced sensitivity. Capacitance-voltage measurement results further reveal that instead of direct charge transfer, the electrostatic gating effect on BP flakes arising from the dipole moment of adsorbed water molecules is the basic mechanism governing the humidity sensing behavior of both bare and encapsulated BP sensors. This work demonstrates a viable approach for achieving air-stable BP-based humidity or chemical sensors for practical applications.

A Fully-Integrated MEMS Preconcentrator for Rapid Gas Sampling

In this paper, we present a new type of MEMS gaseous species preconcentrator (muPC) that has been fabricated and tested as a front end for a flame ionization detector (FID). A one microliter muPC filled with microposts is integrated with microvalves (response time < 50 mus) and a resistive microheater (ramping to 200degC in 0.5 seconds). The integrated muPC can sample a cubic centimeter of gas in 0.2 second at 49 kPa, adsorb targeted species, heat and desorb, and inject concentrated gaseous species with 50 microsecond pulses into separation columns and/or detectors. The unprecedented speed of this muPC is enabled by MEMS sizing and fabrication, allowing sniffing of phosphonates, toxic industrial chemicals (TICs), and other volatile compounds in seconds, rather than tens of minutes with conventional systems.

Nano-fabrication with a flexible array of nano-apertures.

We report fabrication and use of a flexible array of nano-apertures for photolithography on curved surfaces. The batch-fabricated apertures are formed of metal-coated silicone tips. The apertures are formed at the end of the silicone tips by either electrochemical etching of the metal or plasma etching of a protective mask followed by wet chemical etching. The apertures are as small as 250 nm on substrates larger than several millimeters. We demonstrate how the nano-aperture array can be used for nano-fabrication on flat and curved substrates, and show the subsequent fabrication steps to form large arrays of sub-micron aluminum dots or vertical silicon wires.