A dual tunneling-unit scanning tunneling microscope (DTU STM) was developed for nm order length measurement with wide scan range. The crystalline lattice of highly oriented pyrolitic graphite (HOPG) was used as reference scale. A reference unit was set up on top of a test unit. The reference sample holder and the probe tip of test unit were attached to one single XYscanner on either surface, while the test sample holder was open. This enables simultaneous acquisition of wide images of HOPG and test sample. The length in test sample image was measured by counting the number of HOPG lattices. An inchworm actuator and an impact drive mechanism were introduced to roughly position probe tips. The XY scanner was designed to be elastic to eliminate image distortion. Some comparison experiments using two HOPG chips were carried out in air. The DTU STM is confirmed to be a stable and more powerful device for length measurement which has nanometer accuracy when covering a wide scan range up to several micrometers, and is capable of measuring comparatively large and heavy samples.

Since scanning probe microscopes scan the probes mechanically along the samples, accuracy of lateral scales of acquired images is mainly determined by the calibration of the movement of the end point of the probes. A dual tunneling unit scanning tunneling microscope (DTU-STM) with an xy stage for simultaneous lateral scanning of both the sample and the scale-reference crystal was developed. It enables calibration of the lateral scale of the sample image under the assumption that the lattice spacing is constant. Accuracy and problems of the proposed method were evaluated by comparing images of graphite simultaneously or consecutively acquired with the DTU-STM. For simultaneously acquired images in the 10 nm range, calibration of drift rates of the tips to the samples, and tilt of the samples to the xy plane were found to be effective in improving the accuracy of comparison measurement to 98.9±1.5%, regardless of orientation. The accuracy marked a higher value of 99.7±0.25% in the direction of the line scan, since the effect of thermal drift is less dominant in the direction. Consecutive single-line scanning of 150 nm with a rate of 380 ms/line, at 30 min intervals, gave an accuracy of over 99.98±0.036%.

No abstract

No abstract