Silicon Chemical Vapor Infiltration

Selected Publications

K. Zufelt, R. Vanfleet, and R. C. Davis

An issue that often impacts x-ray and electron analysis of transmission electron microscopy (TEM) samples is the presence of high-Z atoms in the sample substrate. In many cases, it is also desirable that the chosen substrate be resistant to chemicals and various processing methods. We present an improved TEM grid made by carbon-infiltrated carbon nanotube templated microfabrication (CNT-M). These grids provide a significant advantage in analytical TEM applications due to the absence of high-Z atoms and the improved chemical resistivity which allows for a wider range of sample preparation and processing techniques. We have refined the CNT-M process by developing a method for preventing delamination of highly carbon-infiltrated CNT-M structures from the growth substrate. We further present a scalable method for suspending thin films (<30 nm) across large gaps (>100 um) between CNT-M defined features. Several membranes were deposited on the grids including amorphous carbon, boron carbide, silicon dioxide, and alumina. These results are of significance to CNT-M MEMS design and production.

Anthony D. Willey, Josh M. Holt, Richard R. Vanfleet, and Robert C. Davis (et al.)

The objective of this study was to test the mechanical durability of a polyurethane/organoclay nanocomposite modified with perfluoroalkyl methacrylic copolymer in conditions that replicate extended rain impact. Samples were impacted with 1.4 mm droplets at a velocity of 24 m/s and a flow rate of 0.78 gpm for a period of 5 hours by an axial full cone nozzle. The cases of the spray nozzle being placed vertically above a sample as well as at a 40° angle were examined. After the spray period, samples were heated at 100° C to allow saturated liquid to evaporate from the surface. Contact angle and sliding of the superhydrophobic surfaces were measured before and after the spray period. A decrease in performance for both samples was observed, with the vertically sprayed sample seeing greater degradation. SEM images of superhydrophobic samples before and after spray impact revealed large amounts of circular indentations on the surfaces caused by the impacting droplets which, along with leaching of the fluoroacrylic copolymer, was likely the cause of the decrease in performance.

Anthony C. Pearson, Bhupinder Singh, Matthew R. Linford, Barry M. Lunt, and Robert C. Davis

We have fabricated nanoscale tellurium fuses using electron beam lithography for long term data storage applications. The tellurium fuses contain a narrow resistive region(250 nm to 10 micron). Application of a sufficiently high voltage accross the tellurium fuse causes resistive heating of the narrow region. After a sufficient temperature is reached, the tellurium melts and begins to flow until a gap is formed in the material. After the gap is formed there is no longer a path for electric current flow, causing a large permanent change in the resistance of the device. We have measured teh resistance change following gap formation and have experimentally determined the effect of voltage on the resulting gap size. Finite element simulations have been used to estimate the effect of applied voltage on the temperature of the fuses.

Barry M. Lunt, Matthew R. Linford, and Robert Davis
This paper has provided details on the LE of the data on magnetic tape, hard-disc drives, flash memory, and optical discs. For archival purposes, only the M-DISC presently meets the requirement of long-term data retention (more than 100 years). This disc has been rigorously tested, and the results are very positive. The main weakness of the M-DISC is its relatively small capacity. This weakness is being addressed by our current research on a permanent solid-state solution, one whose density will enable a capacity of several terabytes in the format of a solid-state drive. Though it will be much more expensive than an M-DISC, the convenience of having several terabytes in a single, removable form of storage should be very attractive to the archival community.
Brian Davis, Hiram Conley, David Jones, John N. Harb, and Robert C. Davis
Dielectrophoresis has been used as a technique for the parallel localization and alignment of both semiconducting and metallic carbon nanotubes (CNTs) at junctions between electrodes. A variation of this technique known as floating potential dielectrophoresis (FPD) allows for a self-limiting number of CNTs to be localized at each junction, on a massively parallel scale. However, the smallest FPD geometries to date are restricted to conductive substrates and have a lower limit on floating electrode size. We present a geometry which eliminates this lower limit and enables FPD to be performed on non-conducting substrates. We also discuss experiments clarifying the self-limiting mechanism of CNT localization and how it can be used advantageously as devices are scaled downwards to smaller sizes.
Kellen Moulton, Nicholas B. Morrill, Adam M. Konneker, Brian D. Jensen, Richard R. Vanfleet, David D. Allred, and Robert C. Davis
This paper examines the effect of iron catalyst thickness on the straightness of growth of carbon nanotubes (CNTs) for microelectromechanical systems fabricated using the CNT-templated-microfabrication (CNT-M) process. SEM images of samples grown using various iron catalyst thicknesses show that both straight sidewalls and good edge definition are achieved using an iron thickness between 7 and 8 nm. Below this thickness, individual CNTs are well aligned, but the sidewalls of CNT forests formed into posts and long walls are not always straight. Above this thickness, the CNT forest sidewalls are relatively straight, but edge definition is poor, with significantly increased sidewall roughness. The proximity of a device or feature to other regions of iron catalyst also affects CNT growth. By using an iron catalyst thickness appropriate for straight growth, and by adding borders of iron around features or devices, a designer can greatly improve straightness of growth for CNT-MEMS.