May_EDFA_Digital

edfas.org 13 ELECTRONIC DEV ICE FA I LURE ANALYSIS | VOLUME 24 NO . 2 introduced by the probe tips. In contrast, when switching to lower and lower electron beam acceleration voltages, the electrons are indeed affected by the probe tips’ pres- ence and the image quality rapidly deteriorates (Fig. 1). One way to reduce the susceptibility of the electron beamtoenvironmental influences is to reduce theworking distance. This, however, is limited by the geometry of the microscope’s pole piece (objective lens) and the probing system’s mechanical envelope. In addition, factors such as the sample size and probe tip bending angles play a significant role. Another way tomitigate these effects is to ensure that the probes are arranged in a symmetrical fashion—in 3D space—meaning they must be positioned at the same height above the sample surface as well as being placed on the perimeter of a (perfect) circle. Even when these conditions are not met, it may be possible to obtain a good image by realigning themicroscope’s apertures and correcting for astigmatism (as far as these corrections are possible). Now, it is important to realize that every time one of the probes is moved, the image quality will (more or less) instantly deteriorate so that it is impossible to pre- cisely determine the probes’ positions. This necessitates realignments/readjustments of apertures and astigma- tism. Without these adjustments the image is unusable, and the operator is “flying blind.” WORKFLOW The workflow described here applies to the Kleindiek Nanotechnik Prober Shuttle (Fig. 2) but will be similar for any other nanoprobing setup. In order to place the probes on the sample surface without employing high acceleration voltages, automa- tion is required. By equipping the micromanipulators with positional feedback and indexing the probe tips’ positions, these can be automatically positioned in a symmetrical pattern on the same plane above the surface. Once positioned the image can be optimized as described above. Using an automated workflow, the probes can be positioned as described without operator interaction so that constant tedious image adjustments are not neces- sary during this procedure. The result of this step is shown in Fig. 3. The automatic pre-positioning step is followed by a manual procedure where the probes are brought closer together using intuitive drag and dropmotion input com- mands. These inputs are applied using the same periph- eral hardware (keyboard/mouse) that is used to drive the microscope itself. Motion in the x and y axes is executed by dragging the mouse. Motion in the z direction is driven using a separate jog-shuttle (dial). Upon completing this step, all eight probe tips arepositioned ina circular pattern approx. 2 µm from the image center (Fig. 4). The next step is to move the sample into view using a suitable sample stage that allows positioning the sample relative to the probes. Typically, the stage is also equipped with positional feedback and the sample is referenced to a CAD design such that a specific area can be targeted without the need to visually inspect the sample with the electron beam. Once the sample is in place, the SEM image is used to address individual exposed contacts or metal lines on the sample surface. After executing the Fig. 2 KleindiekNanotechnikProber ShuttlePS8e equipped with eight nanomanipulators as well as a three-axis substage for a total of 27 axes. All axes have sub-nm resolution as well as absolute positioning capability. Fig. 3 Eight probe tips automatically arranged in a circle with a radius of roughly 40 µm at the same height above the sample (electron beam acceleration voltage: 100 V).