Exchange interactions and intermolecular hybridization in a spin-1/2 nanographene dimer

In this study, researchers investigate exchange interactions and intermolecular hybridization in a spin-1/2 nanographene dimer to understand spin physics relevant for quantum technologies. The gist: Inelastic electron tunneling spectroscopy (IETS) reveals singlet–triplet excitation arising from interphenalenyl exchange, and theoretical analysis identifies third-neighbor hopping as essential for activating kinetic interphenalenyl exchange, which is further modulated by substrate coupling.

Synthesis and Characterization of the Dimer

The researchers synthesized the phenalenyl dimer, referred to as diphenalenyl, using a combined solution and on-surface synthesis approach. The preparation involved a sequence of nine steps starting from naphthalene in solution, followed by activation using atom manipulation with a scanning tunneling microscopy (STM) tip. Specifically, the precursor was deposited onto a clean Au(111) surface after preparing the substrate by sublimation of NaCl on Au(111). Sequential tip-induced cleaving of hydrogen atoms from the sp3 carbon atoms yielded the target diphenalenyl diradical, which was confirmed by constant-height nc-AFM measurements. This process demonstrated that "hydro precursors has the advantage that activation of the target open-shell compound can be achieved using atom manipulation with a scanning tunneling microscopy (STM) tip and does not necessarily require the catalytic action of metal substrates."

Spin Excitation Energies and Exchange Interactions

Inelastic electron tunneling spectroscopy (IETS) was used to probe the magnetic properties. The data revealed inelastic steps at ±41 meV on Au(111) and ±48 meV on a monolayer of NaCl on top of Au(111). These excitations are assigned to singlet–triplet excitation arising from interphenalenyl exchange. Theoretical analysis using a generalized Hubbard model, which includes long-range Coulomb interactions, showed that diphenalenyl hosts two antiferromagnetically coupled unpaired spins resulting in an S = 0 ground state and an S = 1 excited state with energy E = J. The calculated singlet–triplet splitting was found to be close to the experimentally observed energies.

Role of Intermolecular Hybridization and Hopping

The study provides evidence for intermolecular hybridization, which is crucial for kinetic interphenalenyl exchange. Theoretical calculations showed that intermolecular hybridization is present because third-neighbor hopping is non-zero, as the zero modes have a null weight on the binding sites if only first-neighbor hopping were considered. The calculation of the total exchange and its relative contribution from intermolecular hybridization and intramolecular addition energies indicated that the dimer has a strong diradical character. The third-nearest-neighbor hopping, denoted as t3, was found to be non-zero and is dominant in driving the intermolecular hybridization between the two phenalenyl units.

Substrate Dependence and Renormalization

The coupling of diphenalenyl to the substrate significantly affects the excitation energies and linewidths. The energy difference between excitations on Au(111) (41 meV) and on NaCl (48 meV) is attributed to renormalization of the spin excitation energies due to exchange with the Au(111) electrons. Ab initio calculations using an Anderson impurity model solved via the one-crossing approximation (OCA) showed that hybridization with Au(111) leads to a substantial renormalization compared to NaCl. Specifically, DFT calculations indicated that for Au(111), hybridization is appreciable, whereas for the NaCl monolayer, it is very weak, leading to negligible renormalization of the excitation energy by Kondo exchange coupling on NaCl.

Bias Asymmetries and Orbital Contributions

The analysis of dI/dV spectra revealed position-dependent bias-asymmetry in the height of inelastic steps. This asymmetry is an intrinsic property of the diphenalenyl molecules and independent from the underlying substrate. The contribution to step heights was analyzed using the projected density of states (PDOS) for HOMO and LUMO orbitals. For example, the HOMO has a significantly larger step for negative bias than for positive bias, while for the LUMO it is exactly the opposite. Furthermore, on Au(111), the spectra show characteristic triangular overshoots induced by Kondo exchange with the conduction electrons, which are especially pronounced for the positive-bias step in the LUMO spectral function. The LDOS maps confirmed these asymmetries, showing that over the center of the molecule, the negative bias step is significantly larger than the positive bias step.

Conclusion and Outlook

The study concludes that diphenalenyl acts as an open-shell molecule exhibiting strong interphenalenyl antiferromagnetic exchange. The findings establish a framework for understanding how substrate coupling modifies these spin excitations, providing essential insights for the design of platforms exploiting phenalenyl and other planar nanographene radicals in quantum technologies. The results underscore the importance of considering both intermolecular hybridization and substrate effects when modeling spin interactions in coupled nanographene systems.

Based on the provided scientific paper, here are specific improvements for AI systems and what those improved systems could achieve:

The core of this research lies in understanding and manipulating spin interactions (exchange interactions) in open-shell nanographene dimers (like diphenalenyl). The proposed AI improvements focus on leveraging the deep quantum mechanical insights derived from these calculations to create more sophisticated materials science simulators and quantum computing platforms.

Here are the specific improvements:

1. ---

2. The paper details a comprehensive theoretical framework involving:

3. • Extended Hubbard models (S1, S3) incorporating third-nearest-neighbor hopping and Coulomb interactions (including Coulomb exchange/CDE).
4. • The Anderson Impurity Model (AIM) coupled to a substrate bath (S4).
5. • The One-Crossing Approximation (OCA) for solving the AIM to obtain spectral functions and excitation energies.

An AI system trained on this framework could be used for:

1. ---

2. The improved AI system could perform highly accurate, first-principles simulations of spin dynamics in molecular systems, specifically predicting the singlet–triplet excitation energy (E = J) and its dependence on substrate coupling strength (e.g., Au(111) vs. NaCl).

3. ---

4. The improved AI system could predict the renormalization of these spin energies caused by substrate hybridization, allowing researchers to design molecules where the exchange interaction is robust against environmental changes or engineered for specific quantum phases (e.g., topological order).

5. ---

6. The paper explicitly links the electronic structure (HOMO/LUMO contributions) to experimental observables like dI/dV spectra asymmetries and bias-dependence at different spatial locations within the dimer.

7. ---

8. The improved AI system could be used for inverse design of nanographene dimers, allowing researchers to computationally design molecular structures that exhibit specific, desired spectroscopic signatures (like a large negative bias step asymmetry) before synthesizing them experimentally.

9. ---

10. The paper shows that the role of third-neighbor hopping (t3) is crucial for intermolecular exchange in these systems, which is often ignored in graphene modeling.

11. ---

12. The improved AI system could serve as a tool to identify and prioritize structural motifs (like specific bond geometries or hopping pathways) that drive quantum phenomena (like interphenalenyl exchange), guiding the synthesis of materials with desired spin lattices for quantum magnetism exploration.

13. ---

14. The paper provides detailed, substrate-dependent coupling parameters (hybridization strengths, Γ values) derived from DFT and NEGF calculations.

15. ---

16. The improved AI system could serve as a predictive tool for designing molecular platforms that optimize spin qubit properties by precisely tuning the molecule-substrate interface to maximize or minimize Kondo exchange effects, thereby engineering robust spin qubits.

17. ---

18. The paper demonstrates how coupling to the substrate modifies spectral linewidths (broadening) and Kondo exchange effects, which are crucial for understanding experimental limitations in STM/STS measurements.

19. ---

20. The improved AI system could be used to predict the expected experimental resolution and line-shape of spin excitation steps for a given molecule-substrate combination, ensuring that experimental setups are optimized to resolve subtle quantum features accurately.

21. ---

22. The paper explores the distinction between different coupling mechanisms (intermolecular hybridization vs. Coulomb-mediated exchange).

23. ---

24. The improved AI system could act as a diagnostic tool for experimentalists, helping them determine whether the observed spin dynamics are dominated by kinetic exchange or Coulomb interactions, thereby interpreting complex spectroscopic data accurately and distinguishing between competing physical mechanisms in real-time.

Phenalenyl is a radical nanographene with triangular shape that hosts an unpaired electron with spin S = 1/2. The open-shell nature of phenalenyl is expected to be retained in covalently bonded networks. Here, we study a first step in that direction and report the synthesis of the phenalenyl dimer by combining in-solution synthesis and on-surface activation and its characterization both on Au(111) and on a monolayer of NaCl on top of Au(111) by means of inelastic electron tunneling spectroscopy (IETS). IETS shows inelastic steps that, together with a thorough theoretical analysis, are identified as the singlet-triplet excitation arising from interphenalenyl exchange. Two prominent features of our data permit to shed light on the nature of spin interactions in this system. First, the excitation energies with and without the NaCl decoupling layer are 48 and 41 meV, respectively, indicating a significant renormalization of the spin excitation energies due to exchange with the Au(111) electrons. Second, a position-dependent bias-asymmetry of the height of the inelastic steps is accounted for by an interphenalenyl hybridization of the singly occupied phenalenyl orbitals that is only possible via third neighbor hopping. This hybridization is also essential to activate kinetic interphenalenyl exchange. Our results set the stage for future work on the bottom-up synthesis of spin S = 1/2 spin lattices with large exchange interaction.