edfas.org 17 ELECTRONIC DEVICE FAILURE ANALYSIS | VOLUME 27 NO. 4 such as creep and hysteresis, ensuring stable long-term positioning. Finally, the control software supports seamless correlation of SEM images with the device CAD layout, enabling automated alignment and direct probe contact to the targeted transistor structures. Collectively, these features allow for efficient, low-dose, and highly reproducible nanoprobing of advanced semiconductor devices. These features converge in what the authors refer to as semi-blind positioning: a novel operational paradigm in which probes are navigated based on high-resolution position data rather than continuous visual feedback. But why “semi-blind?” While motion operations are no longer reliant on real-time SEM feedback, verification of the final probe position is still reasonable. Thus, imaging is performed discretely—typically at key checkpoints. For example, an initial SEM frame (acquired at 200 eV with a frame time of 120 ms) is used for CAD-based targeting (see Fig. 3) or probe positioning via point-and-click (see Fig. 4). After the target is reached, a follow-up frame verifies accurate probe placement. With active position-holding and drift suppression, no intermediate images are needed. Using the semi-blind positioning, it is possible to contact a transistor with six discrete movements, resulting in six SEM images. While using probing with visual feedback, the contacting of a transistor takes a few minutes, resulting in hundreds of images. RESULTS FROM SEMI-BLIND NANOPROBING To systematically investigate the extent of electron beam-induced degradation in beam-sensitive technologies, a controlled experiment was designed using a fieldeffect transistor (FET) implemented in a 22 nm FD-SOI technology node. The experimental setup consists of the SmarAct SMARPROBE nanoprobe, a Carl Zeiss Sigma 300 SEM, and a Keithley Parameter Analyzer 4200A (Fig. 5). The FD-SOI architecture was chosen, as it is known to exhibit increased susceptibility to radiation-induced degradation. As a reference, a pristine pFET device was first characterized without any exposure to the electron beam, using an atomic force probe (AFP), which allows electrical characterization in ambient conditions with zero e-beam interaction. This measurement provided baseline parameters, most notably the threshold voltage (Vth), derived from I–V characteristics, against which all subsequent measurements could be compared. Following the baseline measurement, the same device was characterized using SmarAct’s SMARPROBE Fig. 4 The probes can be positioned manually using point and click. Because there are no parasitic movements due to piezo creep, one image per movement is sufficient as visual feedback. Fig. 5 SEM-based nanoprobing combines the advantage of accurate probe placement and e-beam analysis techniques like electron beam induced current (EBIC). This picture shows the experimental setup based on the SMARPROBE nanoprobe.
RkJQdWJsaXNoZXIy MTYyMzk3NQ==