Speaker
Description
Science and industry ask for faster microscopy techniques to resolve dynamic processes at the atomic scale. Prominent research examples include catalytic reactions at surfaces, corrosion processes and semiconductor devices. Conventional imaging with a scanning tunneling microscope (STM) takes several minutes per frame. We developed a new high-speed STM that can acquire a single frame in less than 25 ms. This means, we increased the frame rate from 0.01 to almost 100 Hz. For this, we changed to an unconventional scan pattern. The scan control, data acquisition, and image visualization are integrated into the EPICS framework.
A VMEbus - System running EPICS on RTEMS controls the high-speed scan. An arbitrary waveform generator provides the scan input signal, which the user can modify at will.
A digitizer writes the data into PVs, which are provided by a PV server on a Unix- System. A monitor subscribes to these PVs. This monitor acts as a producer and puts the data chunks to a queue on a local machine. A consumer thread saves the data from the queue to an HDF5 file. This file contains all relevant PVs, spatial coordinates, and measuring points.
A parallel thread reads single data chunks from the PV, processes the data and provides a separate PV for data visualization. The Phoebus interface displays the generated live video.
The acquired HDF5 file contains all necessary data. For the image and video analysis we developed a purely python based and customizable program that allows us to characterize the dynamics at the atomic scale.
