Highly Entangled Polyradical Nanographene With Coexisting Strong Correlation And Topological Frustration

Result of the Month

Author: Shaotang Song, Andrés Pinar Solé, Adam Matěj, Guangwu Li, Oleksandr Stetsovych, Diego Soler, Huimin Yang, Mykola Telychko, Jing Li, Manish Kumar, Qifan Chen, Shayan Edalatmanesh, Jiri Brabec, Libor Veis, Jishan Wu , Pavel Jelinek & Jiong Lu Institute: ''National University of Singapore'' Nature Chemistry
URL: https://doi.org/10.1038/s41557-024-01453-9
Date: 4/2024
Instruments: LT STM Lab

Open-shell nanographenes exhibit unconventional π-magnetism arising from topological frustration or strong electron–electron interaction. However, conventional design approaches are typically limited to a single magnetic origin, which can restrict the number of correlated spins or the type of magnetic ordering in open-shell nanographenes. Here we present a design strategy that combines topological frustration and electron–electron interactions to fabricate a large fully fused ‘butterfly’-shaped tetraradical nanographene on Au(111). We employ bond-resolved scanning tunnelling microscopy and spin-excitation spectroscopy to resolve the molecular backbone and reveal the strongly correlated open-shell character, respectively. This nanographene contains four unpaired electrons with both ferromagnetic and anti-ferromagnetic interactions, harbouring a many-body singlet ground state and strong multi-spin entanglement, which is well described by many-body calculations. Furthermore, we study the magnetic properties and spin states in the nanographene using a nickelocene magnetic probe. The ability to imprint and characterize many-body strongly correlated spins in polyradical nanographenes paves the way for future advancements in quantum information technologies. 

In this paper, we introduce a new conceptual design of polyradical nanographene that features multiple strongly entangled many-body quantum spins arising from the interplay of strong e-e correlation and topological frustration. Our design involves fusing four [3]triangulene motifs onto the edges of a [3]rhombene to create a butterfly-shaped nanographene with a carefully crafted geometry that renders both sublattices (A and B) topologically frustrated. According to the hexagonal graphs theorem, the number of zero-energy eigenstates in the tight-binding model for graphene nanoflakes can be determined by the term of ‘‘nullity’’ (η), which is equal to the difference between the maximum numbers of nonadjacent vertices (α) and edges (β). The sum of α and β equals to the total number of carbon atoms (N = 118) in this large PAH. The topologically frustrated butterfly geometry in our design yields a nullity of two (η = α – β = 2), which predicts the presence of two radicals. Additionally, this design creates a sufficiently large size (the largest size of fully fused open-shell nanographene by far) to trigger spin-symmetry breaking of occupied frontier orbitals through strong e-e interaction that dominates over the hybridization energy, yielding two more radicals. Combining these two magnetic origins results in a designer tetraradical nanographene with both ferromagnetic and antiferromagnetic coupling of correlated spins that is rarely found in fully aromatic PAHs possessing only a single magnetic origin. 


The on-surface synthesis of the butterfly molecule. Constant-height BRSTM image of magnetic butterfly molecule taken with a CO-functionalized tip at Vs = 20 mV. And the spin-excitation map of the molecule (It = 1 nA, Vac = 2 mV, and f = 439 Hz). 

We further employed the SPM technique with a nickelocene (NiCp2) functionalized probe to directly verify the magnetic properties of the butterfly molecule by obtaining inelastic electron tunneling spectroscopy (IETS) spectra at different tip-sample distances. The NiCp2 molecule has a triplet ground state (S = 1, mS = 0) and doubly degenerate triplet excited state (S = 1, mS = ±1) due to the presence of a positive magnetic anisotropy parameter D ~ 4 meV defining easy magnetization plane along the z-axis with. The spin-excitation from the ground state (mS = 0) to one of the doubly degenerated excited states (mS = ±1) of the NiCp2 probe is manifested as the corresponding peak/dip features at ±4 mV symmetrically around EF in the IETS spectrum taken on a bare Au(111). It is noteworthy that, as the tip approaches the Au(111) surface, the energetic positions of the peak or dip features remain constant while their intensities increase. In contrast, the IETS spectra taken over the corner of the butterfly moleucle display additional features at ±13 mV at a large tip-sample distance. As the tip-sample distance decreases, the IETS signal at ±4 mV moves towards EF,while the new features at ±13 mV shift away from EF, accompanied by a monotonic broadening of peak and dip features. The observed modification of the IETS spectra can be attributed to the magnetic exchange interaction between the spin states of the NiCp2 probe and the tetraradical molecule, which provides direct experimental evidence of the presence of Π-magnetism in the butterfly molecule. Combining a NiCp2 probe study with theoretical calculations allows us to unravel the exotic magnetic coupling with multiple-entangled spin interactions in product 1, which will be discussed in more detail in the following context.  


Probe the Π-magnetism of ‘butterfly’ 1 with a NiCp2 functionalized tip. a. Conceptual illustration of the measurement of molecule 1 with a NiCp2 functionalized tip. b. IETS spectra plotted in a color scale taken over the bare Au(111) overlaid with the IETS spectra acquired at different tip-sample distances with a 25 pm decrease step. c. IETS spectra plotted in a color scale taken over the corner of ‘butterfly’ 1 overlaid with the corresponding IETS spectra taken at different tip-sample distances with the same 25 pm decrease step. IETS spectra lock-in parameters: Vac = 0.5 mV and f = 723 Hz. d. Calculated IETS spectra plot of butterfly 1 as a function of the coupling strength J. States A and C (both red and blue) correspond to the experimental peak/dip signals shown in c, whereas state B (grey) is not visible.