Many important micrometer-sized devices-from arrays of microlenses [1,2] through microsensors and actuators [3] to passive microfluidic mixers [4,5] -rely on curvilinear and/or multilevel surface topographies. Although fabrication of such structures from elastomers [6] or polymers [7] is relatively straightforward by casting/molding against appropriately shaped masters, [8][9][10] these materials have several unsatisfactory properties (mechanical wear, [11] ground collapse, [12] swelling by organic solvents, [13] or gas permeability [14] ), limiting their uses in many practical applications. [1,2,5,15] On the other hand, microstructuring of hard, durable solids with 3D surface reliefs requires complicated and expensive procedures such as reactive-ion etching, [16] laser ablation, [17] powder blasting, [18] or electrochemical micromachining. [19] Importantly, fabrication with these methods is serial, and each device has to be fabricated anew. Here, we describe a straightforward and inexpensive experimental method based on reaction diffusion [1,2,9,[20][21][22] that overcomes these limitations and allows for bench-top prototyping of curvilinear and/or multilevel reliefs in solid materials (especially glass and silicon) with lateral resolution down to several hundreds of nanometers. In our maskless process, hydrogel stamps are replicated-as in soft lithography [8] -from micropatterned masters, soaked in a chemical etchant, and then used to literally "print" entire microdevices into solid substrates with full retention of the stamp's topography and with excellent surface smoothness. We demonstrate the versatility of our method by fabricating microlens arrays, passive microfluidic mixers, [4] and gradient diffusers in glass, chirped and 3D diffraction gratings in silicon, and two-level line arrays in common semiconductors. Figure 1a illustrates the experimental strategy for etching glass and silicon (for semiconductors, the procedure is analogous, but uses different etchants; see Fig. 3). First, a 10 wt % aqueous solution of high-gel-strength agarose (EM Sciences, Darmstadt, Germany) is cast against a micropatterned master that has a desired surface topography and is prepared from a soft material (photoresist, epoxy, poly(dimethylsiloxane) (PDMS), etc.) by any of the conventional multilevel fabrication schemes; [2,9,23] importantly, after the agarose is solidified and peeled off, the master can be reused many times. Next, the gel layer is cut into ca. 3 cm× 3 cm rectangular blocks (stamps), which are soaked for 4 h in a C 0 = 0.6 M aqueous solution of hydrofluoric acid with 0.1 vol % Triton X-100 surfactant, dried on a filter paper for 10 min, and blown dry with nitrogen for 15 s. [24] The stamps are then placed (feature side up) in a Petri dish containing light mineral oil. The surfactant helps the oil to penetrate between the features of the stamp, thus preventing any spilling or lateral spreading of the etchant and limiting etching to the areas of contact between the stamp's microfeatures and the substrate. In addi...

A methodology is shown for predicting the time-dependent reliability (probability of survival) of ceramic components against catastrophic rupture when subjected to thermal and mechanical cyclic loads. This methodology is based on the Weibull distribution to model stochastic strength and a power law that models subcritical crack growth. Changes in material response that can occur with temperature or time (i.e. changing fatigue and Weibull parameters with temperature or time) are accommodated by segmenting a cycle into discrete time increments. Material properties are assumed to be constant within an increment, but can vary between increments. This capability has been added to the NASA CARES/Life (Ceramic Analysis and Reliability Evaluation of Structures/Life) code. The code has been modified to have the ability to interface with commercially available finite element analysis codes such as ANSYS executed for transient load histories. Examples are provided to demonstrate the features of the methodology as implemented in the CARES/Life program.

Arrays of planar, Fresnel-like microlenses are prepared by a spontaneous chemical process of periodic precipitation (PP) occurring in a thin layer of a dry gel, and initiated by wet stamping. The PP lenses focus white light more efficiently than the conventional Fresnel zone plates of similar dimensions. Nanoscale topographies of the micropatterned gels can be replicated into transparent elastomers, and used for focusing based on optical path differences. Experimental observations for both types of structures are in agreement with the Fresnel diffraction calculations.

This letter describes a wet-stamping technique for the fabrication of highly regular arrays of microlenses through reaction-diffusion-swelling processes occurring in a thin film of ionically doped gelatin. Geometrical parameters of the lenses depend on and can be controlled by the geometries of the stamped features and the concentrations of chemicals diffusing and reacting in the gelatin matrix. Surface topographies of the lenses are studied experimentally, and are reproduced by a lattice gas reaction-diffusion model.

No abstract