SUBNANOMETER STABILITY OF NANOSTAGE SUPPORTS

Citation
H. Vanderwulp et al., SUBNANOMETER STABILITY OF NANOSTAGE SUPPORTS, Journal of vacuum science & technology. B, Microelectronics and nanometer structures processing, measurement and phenomena, 15(3), 1997, pp. 566-573
Citations number
15
Categorie Soggetti
Physics, Applied
ISSN journal
10711023
Volume
15
Issue
3
Year of publication
1997
Pages
566 - 573
Database
ISI
SICI code
1071-1023(1997)15:3<566:SSONS>2.0.ZU;2-U
Abstract
In order to obtain atomic resolution in a transmission electron micros cope, a piezo-driven nanostage with subnanometer stability is under de velopment. Inertial sliding motion is used to move the nanostage table in the plane perpendicular to gravity. One of the most important para meters on the way to such a nanostage is the design of the nanostage t able supports. Through several experiments, the stability of two diffe rent nanostage table supports is studied: a kinematic and a nonkinemat ic support. In order to explain the submicrometer and nanometer stage table drift measured in the direction parallel to gravity, a subnanome ter contact theory is presented. This theory explains the stage table drift by the following parameters: the size of the apparent contact ar ea of the support, the gravity forces working on the support, the mult imolecular layer of adsorbed water molecules on all contact surfaces, creep in the contact points and settling of the contact through micros liding at the contact points. For the initial placement of a nonkinema tic support in ambient air with an apparent contact area size of 15x15 mm(2), a stage table drift of 100 nm over 30 min was measured, which almost exactly followed a logarithmic curve. This drift reduced to abo ut 45 nm when the support was placed in a low vacuum, where the number of layers of adsorbed water molecules on the support surfaces is redu ced to one or two. Filling the vacuum chamber with nitrogen gas result ed in an even lower stage table drift (25 nm). Stage table drift after single inertial sliding steps is about 25% of the initial amount of d rift. In case of a kinematic support, the apparent contact area reduce s significantly and stage table drifts after initial placement of 1 up to 3 nm were found. The drift after single inertial sliding steps is on the same order of magnitude. These drifts are attributed to contact creep, which was minimized by optimization of the material selection of both contact surfaces. A combination of two hard surfaces showed al most no creep at the subnanometer level. Therefore, this highly stable kinematic support suits the nanostage application very well. (C) 1997 American Vacuum Society.