Yield, Surface Prep for Nano Devices

Contamination and defects are responsible for as much as 75% or more of yield loss in IC fabrication.² Defect removal processes constitute 80–100 out of 400 processing steps. When we start manufacturing nanoscale emerging research devices (i.e., nanomechanical memory, molecular memories, ionic memory), which may require the assembly of nanoelements, surface preparation and cleaning methods will be very different from what we are familiar with.

Surfaces prepared for nanoscale applications need to be free of particulate contamination on the order of 1 nm or less. To this extent, many of the techniques we currently use must be optimized to meet this challenge. However, once directed assembly of nanotubes or other nanoelements is utilized, then current techniques will remove contaminants as well as assembled device nanostructures. This is because current surface preparation is focused on the non-selective removal of contaminants. So what is needed to meet the challenges of nanomanufactured devices? There is a need to:

• Develop selective removal of nanoelements and impurities.
• Clean assembled nanostructures without destruction.

As a result, chemistry will play a much larger role than it presently does, and better understanding of the adhesion of particles and nanoelements will be needed.

Chemistry may be used to selectively etch or dissolve contaminants, or use other approaches. These techniques do not work, however, if the contaminant is the same material as the assembled nanostructure. In this case, physical chemistry and physical cleaning will have to be used. This will involve increasing the adhesion force of the assembled nanoelements by electrostatic charge (through zeta potential or through the substrate if it is conductive), changing the surface energy before or after assembly, or by using a substrate with a reversible surface energy that could be easily activated. The adhesion force could also be controlled by the selection of the substrate materials where the assembly occurs. The other half of the equation is the selectivity of the removal force applied through a cleaning technique. This can be performed by tuning the removal force to remove the contaminants, keeping it below the removal threshold of the assembled nanoelements. This has been demonstrated for nanoparticles as well as nanotube assembly.

Cleaning control during directed assembly — Polystyrene latex (PSL) nanoparticles are assembled using electrophoresis on conductive gold wires. The gold has a higher adhesion force to PSL particles than silicon. Particles deposited anywhere else (such as between the wires) can be removed by cleaning control (using high frequency flow rinse in this case). The Figure shows that fluorescent nanoparticles assembly without cleaning control leaves many particles between the wires (left). The right side of the Figure shows that when cleaning control is applied, most of the particles are gone. This can be accomplished only by understanding the adhesion and removal forces of these nanoparticles on different surfaces and precise control of the applied removal force.

Assembly of fluorescent nanoparticles assembly without cleaning control (left) and with cleaning control (right).

Selective removal of carbon particles from carbon nanotube films — Carbon nanotube (CNT) synthesis produces CNTs along with many carbon particles and other contaminants. When a standard semiconductor cleaning procedure is employed to remove carbon particles from a CNT film on a 6 in. wafer, the CNT's film and particles are completely removed. However, through understanding of the adhesion forces, pH and ionic strength control, the amorphous carbon particles can be removed without damaging or removing the CNT film.