PCSI2027 Session PCSI-WeM2: Molecular, Organic and Hybrid Interfaces

To date, various scanning probe lithography (SPL) techniques already exist, in which scanning probe microscopes are used to directly create nanostructures on surfaces. Thereby, different physical and chemical effects are used to locally change the structure of the sample surface beneath the probe. Examples include thermal or thermochemical SPL, oxidation SPL, and dip-pen nanolithography [1].

Here, we present a method allowing the direct writing of nanostructures consisting of monolayers of N-heterocyclic carbene (NHC) molecules. NHCs constitute a well-established class of organic compounds with applications in homogeneous catalysis, main-group chemistry, and functional materials [2]. In recent years, they have also emerged as important ligands for surface modification and functionalization. While initial research focused mainly on noble metal surfaces, other substrates such as semiconductors have recently come into focus [3,4].

In the NHC-SPL method demonstrated here, the writing is induced by the electric field between the STM tip and the surface, thus it can be regarded as a sort of bias-induced SPL. During writing, molecular precursors, which are present on the surface as a highly mobile molecular gas are attracted towards the tip-sample gap and immobilized underneath the tip. This is accompanied by a covalent bonding to the surface resulting in the formation of highly stable NHC monolayers. The method allows for the creation of arbitrary structures with resolutions in the range of single nanometers. The writing furthermore works on a variety of (modified) semiconductor surfaces and with different NHC molecules, making a wide range of applications conceivable.

[1] R. Garcia. et al., Nat. Nanotechnol. 9, 577 (2014).

[2] M. N. Hopkinson et al., Nature 510, 485 (2014).

[3] M. Franz et al., Nat. Chem. 13, 828 (2021).

Chirality-induced spin selectivity (CISS) is the magnetically non-trivial electronic response of chiral molecules: the current through a chiral molecular junction depends on the magnetization of a ferromagnetic electrode and on handedness, and electrons photoemitted through a chiral layer emerge spin-polarized, with a sign set by handedness [1]. The effect is large and survives at room temperature, The origin of the CISS effect is a major open question, but experimental evidence point strongly to an interface effect [2].

The “Spinterface mechanism” [1,3] suggests that the CISS is an interface phenomena. It relies on an assumption (motivated by the appearance of magnetization in metal-molecule hybrid systems discussed above) that the interface between the metal and molecule gives rise to some localized magnetic moment, whose direction is undetermined. The adsorption of a chiral molecule drives a transient charge redistribution in a chiral environment, accompanied by a pulse of magnetic field along the chiral axis; with exchange coupling and dissipation into the metal, this stabilizes the fluctuating surface moments along that axis. This is the only theoretical approach to CISS that permits direct fitting of raw experimental data [3].

This is not the only magnetic effect arising in hybrid metal-molecule interfaces. In fact, evidence for magnetic effects in hybrid interfaces with achiral molecules is over two decades old [4]. And yet, to the best of our knowledge, no one in the CISS or other community has made the connection between the two.

In this talk we demonstrate that the “spinterface mechanism” provides a direct and natural link between these two families of phenomena. The magnetism arises from the same reasons, namely the presence of magnetic moments at the metal-molecule interface, and chirality is a director for the effect, supplying an out-of-plane advantage that prevents cancellation that typically washes out the effect in achiral molecules. Thus, CISS may thus be a legible case of emergent interfacial magnetization, and not a unique effect.

[1] S. Sarkar, A. Sharoni, O. L. A. Monti, and Y. Dubi, ACS Nano 19, 37484 (2025).

[2] Nguyen et al., Chem. Sci. 15, 14905 (2024); Aragònès et al., JACS 147, 36453 (2025).

[3] O. L. A. Monti and Y. Dubi, JACS 146, 32795 (2024); S. Alwan and Y. Dubi, JACS 143, 14235 (2021).

[4] I. Carmeli et al., J. Chem. Phys. 118, 10372 (2003); M. Baljozović et al., ACS Nano 20, 4143 (2026).

+ Author for correspondence: jdubi@bgu.ac.il

Interactions between two fully occupied atomic orbitals are conventionally expected to be repulsive because both the bonding and antibonding molecular orbitals are occupied, resulting in zero bond order[1]. Here, we demonstrate that proximity to a metallic substrate can qualitatively alter this picture and stabilize a covalent interaction between two atoms with closed-shell valence orbitals. Using scanning tunnelling microscopy, atom manipulation, inelastic electron tunnelling spectroscopy, and first-principles calculations, we investigate Ho dimers manually assembled on two monolayers of MgO on Ag(100). Individual Ho atoms possess fully occupied, atomic-like 6s orbitals and show no spin excitations within the investigated energy range[2][3]. In contrast, Ho dimers exhibit intense spin-flip excitations between approximately 60 and 130 meV. The spatial distribution and energy of these excitations, combined with electronic-structure calculations and a tight-binding model, identify an unpaired electron in an antibonding molecular orbital derived predominantly from the Ho 6s states. We show that formation of this state results from partial depopulation of the antibonding orbital through coupling to the metallic substrate. The resulting imbalance between bonding- and antibonding-state occupations produces a net attractive covalent interaction despite the closed-shell character of the isolated atoms. Furthermore, changing the adsorption-site coordination modifies atom–substrate hybridization and systematically tunes the spin-flip excitation energy. These results reveal a substrate-enabled mechanism for covalent bond stabilization between surface adsorbates and illustrate how interfaces can fundamentally modify chemical bonding at the atomic scale.

[1] Pauling, L. Atomic Radii and Interatomic Distances in Metals. Journal of the American Chemical Society 69, 542–553 (1947).

[2] Singha, A. et al. Spin Excitations in a 4 f − 3 d Heterodimer on MgO. Phys. Rev. Lett. 121, 257202 (2018).

[3] Natterer, F. D. et al. Reading and writing single-atom magnets. Nature 543, 226–228 (2017).

The self-assembly of organic molecules into ordered layers on solid surfaces is of considerable interest for potential applications in advanced optical and electronic devices. While ordered molecular growth on metal surfaces is well established, comparatively little attention has been paid to silicon, which remains the predominant material in modern semiconductor technology. On pristine silicon surfaces, the high density of dangling bonds typically impedes the formation of well-ordered organic films, making appropriate surface modification essential.

Here, we demonstrate that rare-earth-modified Si(111) surfaces provide suitable templates for ordered molecular growth [1]. Using scanning tunneling microscopy (STM) at room temperature, we investigated the adsorption behavior of organic molecules with widely differing degrees of chemical reactivity. These include highly reactive N-heterocyclic carbenes (NHC), moderately reactive PTCDA molecules, and weakly interacting transition metal phthalocyanines. At submonolayer coverages, most molecules remain mobile on the surface, whereas highly ordered films form at monolayer coverage for all systems investigated. Figure 1 shows representative STM images of the ordered structures formed by IPr-NHC, PTCDA, and cobalt phthalocyanine. In all three cases, the molecules form highly ordered commensurable monolayers, with the corresponding periodicities indicated in the STM images.

These findings establish rare-earth-modified silicon surfaces as promising templates for the growth of ordered organic thin films and demonstrate their suitability for molecular layers with markedly different chemical reactivities.

[1]M. Kubicki, M. Franz, and M. Dähne, J. Phys. Chem. C 128, 13347 (2024).

⁺Author for correspondence: milan.kubicki@physik.tu-berlin.de

Boron exhibits exceptional chemical stability, making it an attractive material for applications in catalysis and energy storage [1]. This stability arises from its strong covalent bonding, which makes elemental boron highly resistant to oxidation. When oxidized to boron oxide (B₂O₃), it is thermodynamically stable and difficult to reduce, even in reducing environments such as CO or hydrogen. 2D boron, which is characterized as borophene or a transition metal boride, depending on the substrate, is often grown on Cu(111) [2]. There is limited knowledge in the current literature regarding the chemistry of 2D boron. However, partially oxidized 2D copper oxide species are considerably more reactive and can be reduced more readily than fully oxidized 2D copper oxide [3]. This same concept can be applied to partially oxidized 2D boron. Despite the importance of these oxidation states, their formation and evolution have received little attention for 2D boron surfaces. Understanding the oxidation process is therefore critical for controlling the chemical stability, reversibility, and performance of 2D boron, motivating the investigation of early stage oxidation presented herein.

2D B/Cu(111) synthesis was first optimized by post annealing while recording in-situ low energy electron microscopy (LEEM) and low energy electron diffraction (LEED) data. The results were confirmed with scanning tunneling microscopy (STM) (Fig. 1) and auger electron spectroscopy (AES). Boron deposition on Cu(111) produced a clean 2D copper boride film through reversible diffusion between the surface and subsurface. Oxidation experiments revealed the reaction pathway in which copper boride first forms partially oxidized boron oxide (Fig. 2), reacts preferentially with water at elevated temperature to produce boric acid, and subsequently dehydrates into stable B₂O₃ upon reduction with CO.

[1] Z. Huang, Angew Chem Int Ed Engl. 6;59(23):8800–8816(2020)

[2] H. Liu, Sci Rep, 3;3238 (2013)

[3] F. Xu, J. Phys. Chem. C. 126 (44): 18645–18651 (2022)