Capillary zone electrophoresis (CZE) devices fabricated in poly(dimethylsiloxane) (PDMS) require continuous voltage control of all intersecting channels in the fluidic network in order to avoid catastrophic leakage at the intersections. This contrasts with the behavior of similar flow channel designs fabricated in glass substrates. When the injection plugs are shaped by voltage control and leakage from side channels is controlled by the application of pushback voltages during separation, fluorescein samples give 64 200 theoretical plates (7000 V separation voltage, E = 1340 V/cm). Native PDMS devices exhibit stable retention times (± 8.6% RSD) over a period of five days when filled with water. Contact angles were unchanged (± 1.9% RSD) over a period of 16 weeks of dry storage, in contrast to the known behavior of plasma‐oxidized PDMS surfaces. Electroosmotic flow (EOF) was observed in the direction of the cathode for the buffer systems studied (phosphate, pH 3—10.5), in the presence or absence of hydrophobic ions such as tetrabutylammonium or dodecyl sulfate. Electroosmotic mobilities of 1.49 × 10—5 and 5.84 × 10—4 cm2/Vs were observed on average at pH 3 and 10.5, respectively, the variation strongly suggesting that silica fillers in the polymer dominate the zeta potential of the material. Hydrophobic compounds such as dodecyl sulfate and BODIPY® 493/503 adsorbed strongly to the PDMS, indicating the hydrophobicity of the channel walls is clearly problematic for CZE analysis of hydrophobic analytes. A method to stack multiple channel layers in PDMS is also described.

We have fabricated fully released nano-electro-mechanical system (NEMS) cantilevers of various geometries from metallic alloy nanocomposite films. At thicknesses of 4.3 and 20.0 nm, these are the thinnest released metal cantilevers reported in the literature to date. Such device dimensions are very difficult to achieve using conventional metal films. We were able to overcome this limitation by using room-temperature co-sputtering to synthesize nanocomposite alloy films of Al–Mo. A systematic investigation of microstructure and properties as a function of Mo content resulted in an optimum film composition of Al–32 at.%Mo with a unique microstructure comprising a dense distribution of nano-scale Mo crystallites dispersed in an amorphous Al-rich matrix. These films were found to exhibit unusually high nanoindentation hardness and a very significant reduction in roughness compared with pure Al, while maintaining resistivity in the metallic range. A single-anchored cantilever 5 µm long, 800 nm wide and 20 nm thick showed a resonance frequency of 608 kHz, yielding a Young’s modulus of 112 GPa, in good agreement with a reduced modulus of 138 GPa measured by nanoindentation.

The distortion of images by the finite radius of the tip is a significant artifact in atomic force microscopy ͑AFM͒ imaging of thin films. This article examines the effect of the finite radius of the AFM tip on the measurement of five surface roughness parameters: R a and R rms roughness, horizontal roughness, bearing ratio, and power spectral density. The effect of the finite size of the AFM tip on the measurement of these roughness parameters was studied using simulations. We found that the vertical roughness measures ͑R a and R rms ͒ were the least affected by the tip induced distortion. For a spherical tip, the error in R a and R rms roughness was less than 15% as long as the radius of the major features in the AFM image ͑R AFM ͒ was greater than twice the radius of the tip ͑R tip ͒. The horizontal roughness of the surface was sensitive to the distortion due to the tip. For distorted AFM images ͑R AFM /R tip ϭ5͒ the bearing ratio curve was similar in shape to the curve for the original surface. The bearing ratio curve of the severely distorted profile ͑R AFM /R tip ϭ1.5͒ was significantly different from the curve for the original trial surface. The general shape of the power spectral density ͑PSD͒ curves for the original and both distorted profiles were similar. However, the lower frequency peaks in the PSD curves for the distorted profiles differed from those in the PSD curve for the original surface.

We have fabricated 3D photonic crystals based on the woodpile structure by means of the two-photon polymerization technique. The crystals have a photonic bandgap at a wavelength of 1.2 lm and show superprism effects at wavelengths slightly below their PBG. We have demonstrated these wavelength-dependent dispersion properties experimentally, and they are in good agreement with our theoretical calculations. We were able to observe a change in the propagating angle of more than 10°for a wavelength tuning range of 20 nm. This effect is two orders of magnitude larger than in conventional prisms. The results show the great potential of low-refractive-index 3D PCs, fabricated in a very fast and single-step process, to serve directly as functional micro-optical devices in the NIR wavelength regime. ExperimentalFor the fabrication of the woodpile structures, organic/inorganic hybrid polymers of high optical quality, called Ormocers, were used. After laser irradiation, the samples were developed in a mixture of 4-methyl-2-pentanone and isopropyl alcohol (volume ratio of 1:2) for approx. 1 min and subsequently rinsed in isopropyl alcohol. To fabricate the structures by means of 2PP, we used an optical parametric oscillator (Coherent Mira OPO) at a wavelength of 580 nm, a repetition rate of 90 MHz, and pulse duration of 100 fs. The pulses were focused into a thin Ormocer film with an immersion oil objective lens (Olympus; numerical aperture, NA= 1.4, 100×). In order to move the laser focus three-dimensionally through the resin, the sample was mounted on a computer-controlled piezo-driven nanoscanner (PI). The transmission spectra were measured with a FTIR spectrometer (Thermo Nicolet) in combination with an IR microscope. The IR light was focused onto the sample using a 32× reflective microscope objective. The superprism characterization was performed with a tunable Ti:sapphire laser operated in continuous-wave mode (Spectra Tsunami), allowing for a maximum wavelength of approx. 1020 nm. The light was focused onto the crystals with a 6.3× objective lens (NA= 0.2), leading to a focal size of approx. 10 lm. The angles of propagation were derived quantitatively by computational analysis of the CCD images of the propagating light (a Gaussian function was fitted to the recorded light distribution at different positions inside the crystal. The accuracy of the fitting determined the size of the error bars). All of the band structures shown were calculated using the freely available MIT Photonic Bands software package [23]. Calculations of the transmittance and reflectance of the woodpile structures were not performed.

We report a study of tip artifacts in atomic force microscope (AFM) images of thin film surfaces. Specifically, a troublesome artifact that occurs when an AFM tip images surface features with a radius of curvature less than the apex of the tip. These artifacts are not easily detected, since the affected AFM images are similar to those expected for a thin film with a columnar microstructure. In a study of 23 thin films, we found that for a significant fraction, the AFM image was affected by this type of tip artifact. In the worst cases, the AFM images consisted almost entirely of images of the AFM tip. We discuss a simple technique to determine the extent of these tip artifacts.