We have grown out-of-plane wires with a tunable PbSnTe composition in an ultrahigh vacuum system. This is the first time that uniform arrays of wires, with a good aspect ratio can be grown. The wires are single crystalline and have well defined {100} facets, important for studying the topological properties. We have revealed the growth mechanism of wires and flakes, by developing a detailed growth model with a special focus on the anisotropic shape of the structures. This work has resulted in 3 papers [Mater. Quantum Technol. 2 (2022) 015001, Nanotechnology 35 (2024) 325602, paper accepted for Nanotechnology (arXiv.2411.19627)] and invited talks at the ICMAT conference in Singapore in 2023, and CEMS/RIKEN in Japan in 2024. These wires are used to find the trivial to topological transition as a function of the PbSnTe composition.
These wires have been used to fabricate short channel devices. We have found ‘quantum dot’ behaviour, but with very small charging energies. Normally the charging energy is significant and plays and important role. leading to so-called Coulomb blockade. In the PbTe system the dielectric constant is very high (>1000, about two order of magnitude higher than in III-V semiconductors) and therefore the capacitance is large and consequently the charging energy is very small. The conductance spectra are now dominated by the orbital energies. From the Zeeman splitting the anisotropy of the g-factor has been determined. This worked has been published in 2 papers (SciPost Physics 13, 089 (2022), Nano Lett. 2022, 22, 17, 7049–7056).
When a magnetic field is applied parallel to the long SnTe nanowire axis, oscillations in the conductance are observed as a function of the field. These Aharonov Bohm oscillations indicate the presence of surface states. The absence of such signals for Sn-poor compositions, for which the system is expected to be trivial, indicates that these surface states arise from topology. In addition, in these and in other experiments we have observed Fabry-Perot interference fringes, which indicate ballistic transport in pure SnTe wires over length scales comparable to the total wire length, much longer than the mean-free path length in the bulk of the semiconductor. This is another manifestation of scattering-free transport in surface states. These two observations are strong indications of the presence of topological surface states. One paper has been published (Adv. Electron. Mater. 2025, 2500027), and another manuscript is ready to be submitted.
In parallel, we have grown PbTe, PbSnTe and SnTe in-plane by selective area growth on InP susbstrates. We have revealed a new growth mechanism, which entails the reorientation of nuclei to form a single crystalline layer. We have followed this process in time using different analysis techniques, and understand the crystal growth mechanism. In addition, we have developed a route to in-situ deposit a superconductor using so-called shadow-walls; the focus of the future work will be on studying the deposition details and the characterization of these devices. This approach has resulted in high quality films with promising transport properties, and has been published in 4 papers [Adv. Funct. Mater. 2021, 31, 2103062, Adv. Funct. Mater. 2022, 2208974, Phys. Rev. Mat. 7, 023401 (2023), Adv. Funct. Mater. 2023, 2305542]. This work has been presented on invitation at these important conferences: IOP condensed matter Physics in Manchester 2022 (CMD29), and the Compound Semiconductor week (CSW) in Lund 2024.
The in-plane PbSnTe nanostructures have been used to fabricate more complex devices with longer channel length and multiple contacts. These are used to study spin/momentum locking and induced superconductivity. This locking has been studied by a multicontact device, which includes 3 ohmic contacts and 1 ferromagnetic tunnel contact. A current is sent through the wire between the ohmic contacts, and the potential of the ferromagnetic contact is measured. We have observed that the current direction and the spin orientation are coupled in SnTe by a (large) voltage step between the ferromagnetic contact and ground as a function of the applied magnetic field. In the trivial PbTe system we have not seen this effect in the same device geometry and experiment, excluding artifacts. This strongly indicates spin/momentum locking in the topological SnTe. In addition, we have observed a spin memory effect. The spin orientation can be programmed with a large write current and read out with a smaller current. The lifetime is extremely long, and we are currently investigating the origin of this effect. It could be due to the nuclear spins in the system, a ferroelectric effect in SnTe, or due to the induced (p-wave) superconductivity. All these phenomena could be (very) interesting but must be further investigated/falsified.
