Drift-insensitive distributed calibration of probe microscope scanner in nanometer range: Approach description

Abstract

A method is described intended for distributed calibration of a probe microscope scanner consisting in a search for a net of local calibration coefficients (LCCs) in the process of automatic measurement of a standard surface, whereby each point of the movement space of the scanner can be defined by a unique set of scale factors. Featureoriented scanning (FOS) methodology is used to implement the distributed calibration, which permits to exclude in situ the negative influence of thermal drift, creep and hysteresis on the obtained results. The sensitivity of LCCs to errors in determination of position coordinates of surface features forming the local calibration structure (LCS) is eliminated by performing multiple repeated measurements followed by building regression surfaces. There are no principle restrictions on the number of repeated LCS measurements. Possessing the calibration database enables correcting in one procedure all the spatial distortions caused by nonlinearity, nonorthogonality and spurious crosstalk couplings of the microscope scanner piezomanipulators. To provide high precision of spatial measurements in nanometer range, the calibration is carried out using natural standards -constants of crystal lattice. The method allows for automatic characterization of crystal surfaces. The method may be used with any scanning probe instrument.

Gwyscan: a library to support non-equidistant scanning probe microscope measurements

“…Raster scan drifts are also more difficult to evaluate algorithmically and advanced analysis is necessary, as e.g. in [33]. In contrast to this, measurement along more complex scan paths enables seeing even relatively small drift by eye immediately because it results in discontinuities in the image.…”

Computer vision distortion correction of scanning probe microscopy images

“…Continued improvements of the mechanical stability, as well as the introduction of closed loop scanners 8 - allowing the precise position of the tip with respect to the sample to be determined - have greatly simplified the problem of nonlinearities and drift. In systems where such hardware improvements were not possible, software algorithms have been introduced for an online correction of the tip positioning, taking into account the nature of the piezoelectric tube scanner - with the requirement of complex calibration and limited success in real-world applications 9 , 10 . Finally, offline software correction algorithms applied directly to the acquired images were developed, mapping topographic features detected during sequential scans onto each other to correct for thermal drift induced position shifts during the actual scanning process 11 , 12 .…”