Bereiche | Tage | Auswahl | Suche | Aktualisierungen | Downloads | Hilfe

O: Fachverband Oberflächenphysik

O 99: Post-Deadline Session

O 99.5: Vortrag

Donnerstag, 21. März 2024, 20:18–20:30, H 0105

AC/DC - STM — •Marcel Rost¹, Kier Heeck¹, and Milan Allan^1,2 — ¹Huygens-Kamerlingh Onnes Laboratory, Leiden University, Leiden, The Netherlands — ²Munich Center for Quantum Science and Technology (MCQST), München, Germany

STM is a powerful tool for measuring atomic structures on metal and semiconducting surfaces as well as to perform spectroscopy. However, when it comes to insulating surfaces, STM quickly hits its limits: it only works if the insulating (oxide) barrier is thin enough (<1nm). In principle, as the impedance drops with frequency, not only the conductance through the oxide increases, but also the number of tunneling electrons decreases. How many electrons are needed to operate an STM in full AC mode, and would it be possible to tunnel on glass? At 1 nA and 1 GHz one would only need around 6 electrons doing the dance between tip and sample, which one probably could pick up from surfaces charges. The major problem, however, is the inherent tip-sample capacitance, which is parallel to the tunneling junction. Acting as a low-pass filter, this capacitance creates an undesired current, which completely overshadows the real tunneling current: hitting the uA regime already at 10 MHz. We have invented a broadband compensation circuit that cancels this capacitance, enabling us to tunnel in full AC mode (without any DC component). We demonstrate tunneling with only ~ 600 electrons by measuring on Au(111) as well as on top of 25 nm SiO2 at 10 MHz thereby receiving even atomic step resolution. Our findings pave the way for high-frequency time resolution without pump-probe, tunneling on insulators, and even faster STM imaging.