Unraveling surface enabled phenomena in low dimensional systems

Milos Baljozovic

Paul Scherrer Institute, Switzerland

Future quantum technologies rely on the detailed understanding of the interaction between different electronic states in atoms or molecules. Surface supported atomic and molecular systems provide a base for such investigations with the particular advantage of addressability. In our work we establish on-surface architectures which exhibit extraordinary magnetic and quantum properties originating from the reduced dimensionality of the self-assembled and atomically precise architectures.  Quantum well arrays, for example, can be produced by the interaction of porous on-surface networks with 2D Shockley-type surface states. Interestingly the periodicity of these (lossy) confinements causes band formation by the coupling between the individual quantum well. In our most recent work the quantum wells have been used modified by the adsorption / condensation of Xe atoms. Remarkably the quantum well state can be programmed by the number (1-12) of Xe in the quantum box. Also the delocalized electronic states across the 2D array can be modified by designing the Xe filling pattern. This provides a quantum breadboard in remembrance of a breadboards used for testing electronic circuitry.  2D ‘checkerboard’ architectures of magnetic molecules containing different e.g. Fe, Mn spins, on the other hand, exhibit surface enabled magnetic phenomena. On magnetic substrates their spin state is determined by the substrate up to elevated temperatures and can be selectively switched by ligation to e.g. NH3. On non-magnetic Au(111), we have observed the first example of 2D ferrimagnetic long-range order and remanence. Magnetic ordering here is due to RKKY interaction of Kondo underscreened moments mediated by the surface states of the support.  Self-assembled 2D architectures contribute to our understanding of fundamental interactions involved in low dimensional systems. In addition, we can put these in the framework of existing models towards their further development.