Instrumental aspects of low-energy electron microscopy are reviewed with a view toward the future evolution of this reemergent technology. Both elastically scattered and inelastically excited electrons in the 0–1000-eV range may be used to form direct rather than scanned images of surfaces in the 3–10-nm resolution range. Different instrumental setups may be used to form images that selectively contain information about the topography, crystalline structure, chemical composition, or magnetic orientation of the first few monolayers. Frequently, parallel imaging allows observation of dynamic processes occurring during observation. Key electron optical elements and their systemic relationships are described in the context of a still hypothetical generalized instrument that would allow complementary exploitation of many contrast modes. Image quality issues such as resolution, sensitivity, statistics, and contrast selectivity are considered with a view toward their optimization, in many cases by drawing ideas and technologies from other fields of microscopy.

No abstract

No abstract

A set of multiple electron-beam ͑e-beam͒ aperture/blanker chips have been fabricated using silicon microelectro-mechanical systems ͑MEMS͒ techniques. The aperture sizes range from 8 to 4 m ͑nominal͒ squares, and the chip configurations feature either eight individually controlled monopolar blanker electrodes or four bipolar electrode pairs. The chips replace the shapers of a 20 kV AEBLE™ shaped e-beam lithography column. The apertures in the chips convert an incident 150 m diameter e-beam into multiple beamlets. Each beamlet can be independently blanked off of a 100 m aperture placed at the following beam crossover. Data are presented that demonstrates the ability to independently blank each beamlet by applying 10 V. Magnified images of the beamlets show square or rectangular shapes with sharp corners, indicating that the apertures were properly fabricated. The degree of electrostatic blanker crosstalk was measured and found to be up to 15% at the crossover plane for different pairs of beamlets, but no observable beam displacement occurred at the image plane. We compared the experimental results to a rough model that estimates the effect of the electrostatic field distribution of one excited blanker electrode on the unblanked beams. The results matched to within 20%.

No abstract