Quantitative, orbital-resolved characterization and predictive control of superexchange coupling in cobalt phthalocyanine molecular spin chains

Spin-flip inelastic electron tunneling spectroscopy (IETS) with submolecular spatial mapping on cobalt phthalocyanine (CoPc) chains grown on Pb(111) established in 2008 that one-dimensional antiferromagnetic $S=1/2$ chains couple by superexchange with $J \approx 18$ meV, and qualitatively identified a ligand $\pi(E_g)$-like bridging orbital as the exchange mediator. The experiment itself acknowledged a set of unresolved issues that jointly define one objective: converting this qualitative demonstration into a complete, quantitative, orbital-resolved and predictive characterization of superexchange in molecular spin chains. The open core comprises: (i) direct, orbital-resolved identification of the superexchange-mediating orbital(s) — including adjudication between the ligand-mediated $\pi(E_g)$ pathway proposed experimentally and the $d_{z^2}$-dominated pathway favored by theory; (ii) the microscopic origin and control of the ~10% molecule-to-molecule variation of $J$; (iii) precise measurement of the chain's magnetic anisotropy tensor, of which only an upper bound (<1 meV) exists; (iv) elimination of substrate-induced Kondo screening of the second-layer molecules to access intrinsic chain spin dynamics; and (v) predictive, quantitative structure–property design rules for exchange couplings in molecular magnets. No closure of any of these items is known in the literature to date.

Background

Metal phthalocyanines (MPc) are planar, thermally robust organic molecules in which a central transition-metal ion (here Co$^{2+}$ with spin $1/2$ in cobalt phthalocyanine, CoPc) is coordinated by a rigid conjugated macrocycle. When stacked face-to-face, these molecules form quasi-one-dimensional chains in which the magnetic ions are separated by several angstroms, so direct overlap of the metal $3d$ orbitals is small and the magnetic coupling is dominated by superexchange — an antiferromagnetic effective interaction generated by higher-order virtual hopping of electrons through bridging orbitals, familiar from transition-metal oxides. Identifying which molecular orbitals actually bridge the exchange, and how large the coupling is, is the microscopic basis for any rational design of molecular magnets. Scanning tunneling microscopy (STM) at millikelvin temperatures can address this: in inelastic electron tunneling spectroscopy (IETS), a tunneling electron whose energy exceeds the excitation threshold of a localized spin opens a new inelastic channel, producing a step in the differential conductance $dI/dV$; for spin systems these thresholds are the spin-flip excitation energies. For a finite antiferromagnetic Heisenberg chain $H = J\sum_i \mathbf{S}_i \cdot \mathbf{S}_{i+1}$ with $S_i = 1/2$ operators and $J > 0$ antiferromagnetic, the singlet–triplet excitation energies of chains with $N = 2, 3, 4$ spins are known exactly, so measured thresholds yield the exchange constant $J$. The Kondo effect — screening of a localized spin by conduction electrons of a metal substrate below a characteristic Kondo temperature $T_K$ — appears as a resonance near the Fermi level and competes with the observation of discrete spin excitations; magnetocrystalline anisotropy (a spin–orbit-induced anisotropy of the spin-excitation energies, characterized for an $S=1/2$ Kramers doublet by the anisotropy of the $g$-tensor and by any anisotropic exchange rather than by single-ion zero-field-splitting constants) splits the field-dependent spin-flip steps and is typically far smaller than $J$ for organic $S=1/2$ centers, making it hard to measure. In the source work (Xi Chen, Ying-Shuang Fu, Shuai-Hua Ji, Tong Zhang, Peng Cheng, Xu-Cun Ma, Xiao-Long Zou, Wen-Hui Duan, Jin-Feng Jia, Qi-Kun Xue, Phys. Rev. Lett. 101, 197208 (2008)), CoPc multilayers on superconducting Pb(111) self-assemble into vertical molecular columns: the first layer is magnetically quenched by the substrate, while molecules from the second layer upward form one-dimensional antiferromagnetic chains whose spin-flip thresholds ($J \approx 18$ meV; $g_\perp = 1.88$) match the Heisenberg-chain spectrum for $N = 2$–$4$. By spatially mapping the inelastic channel at the spin-flip threshold and comparing with spin-polarized density-functional calculations, the authors identified a ligand orbital of $E_g$-like symmetry (the $E_g$ label denotes a doubly degenerate irreducible representation of the molecule's approximately square, four-fold-symmetric point group), hybridized with the Co $3d_{z^2}$ orbital, as the superexchange bridge. The paper itself flagged the limitations of this identification and of the extracted parameters: the measured inelastic-channel maps lack the spatial resolution to image the nodal structure of the bridging orbital, so the identification rests on qualitative similarity to calculated charge densities; $J$ varies by roughly 10% between nominally identical molecules because of film inhomogeneity; second-layer molecules retain residual coupling to the metallic substrate, producing Kondo screening with $T_K < 10$ K that obscures isolated-spin behavior; the magnetic anisotropy is only bounded from above (< 1 meV) rather than measured; and structure–property relations in molecular magnetic materials remain too intricate for the spectroscopic capability to serve as a predictive design tool. Later theory (Wei Wu, N. M. Harrison, A. J. Fisher, Phys. Rev. B 88, 024426 (2013)) computed the exchange over the full space of CoPc chain geometries and attributed the coupling mainly to direct hopping between $d_{z^2}$ orbitals rather than the ligand $\pi(E_g)$ bridge, reaching only qualitative agreement with the measured value — so the microscopic pathway itself is still contested. A 2026 review (Fudi Zhou, Can Zhang, Zhaoteng Dong, Mengya Ren, Lili Zhou, Yu Zhang, Yeliang Wang, Adv. Sci. 13, e00003 (2026)) confirms that spin-flip IETS remains the workhorse for spin-excitation spectroscopy of phthalocyanine adsorbates while listing anisotropy and $g$-factor extraction by electron-spin-resonance STM as an unrealized future direction for these systems.

Problem Statement

Take the superexchange-coupled $S=1/2$ CoPc molecular chains on Pb(111) characterized by spin-flip IETS in the source work, and resolve the interlinked issues the experiment itself left open, which together amount to converting its qualitative demonstration into complete quantitative microscopic characterization and predictive control of superexchange in molecular spin chains. The five issues below are interlinked and jointly posed; specifically:

1. Orbital pathway. Establish the superexchange-mediating orbital(s) by direct, discriminating evidence that goes beyond qualitative similarity between measured channel maps and calculated charge densities, and adjudicate between the ligand $\pi(E_g)$-mediated bridging orbital identified in the source and the $d_{z^2}$-dominated (direct hopping) mechanism favored by subsequent theory.

2. Exchange-parameter statistics. Determine the microscopic origin of the molecule-to-molecule variation of the chain exchange constant $J$ (roughly 10% in the source, attributed to film inhomogeneity) and demonstrate a way to control or suppress it.

3. Anisotropy. Measure the full magnetic anisotropy of the chain spins (for an $S=1/2$ Kramers doublet this is the anisotropy of the $g$-tensor together with any anisotropic exchange, since the axial and rhombic zero-field-splitting parameters $D$ and $E$ of an effective spin Hamiltonian vanish identically), of which the source gives only an upper bound (< 1 meV) inferred from unequal field-split IETS steps, and determine its influence on the chain spin excitations.

4. Substrate decoupling. Eliminate or control the residual coupling between the second-layer molecules and the metallic substrate that produces Kondo screening ($T_K < 10$ K), so that the intrinsic low-energy spin dynamics of the molecular chains can be observed, without destroying the chain stacking geometry that sets the exchange.

5. Design rules. Formulate and validate predictive, quantitative structure–property relations (stacking and sliding geometry, molecular chemistry, substrate choice) that map a molecular assembly's structure onto its exchange couplings, turning the demonstrated spectroscopic and imaging capability into a design tool for molecular magnets.

Quantitatively determine the spin Hamiltonian (exchange, anisotropy, $g$-factors) of such chains together with its microscopic orbital mechanism, and demonstrate the control levers (geometry, chemistry, decoupling) that let the coupling be engineered rather than merely measured.

References

1. Xi Chen, Ying-Shuang Fu, Shuai-Hua Ji, Tong Zhang, Peng Cheng, Xu-Cun Ma, Xiao-Long Zou, Wen-Hui Duan, Jin-Feng Jia, Qi-Kun Xue, "Probing Superexchange Interaction in Molecular Magnets by Spin-Flip Spectroscopy and Microscopy", Phys. Rev. Lett. 101, 197208 (2008). DOI: 10.1103/PhysRevLett.101.197208, https://doi.org/10.1103/PhysRevLett.101.197208
2. Ying-Shuang Fu, Tong Zhang, Shuai-Hua Ji, Xi Chen, Xu-Cun Ma, Jin-Feng Jia, Qi-Kun Xue, "Identifying Charge States of Molecules with Spin-Flip Spectroscopy", Phys. Rev. Lett. 103, 257202 (2009). DOI: 10.1103/PhysRevLett.103.257202, https://doi.org/10.1103/PhysRevLett.103.257202
3. Ying-Shuang Fu, Shuai-Hua Ji, Tong Zhang, Xi Chen, Xu-Cun Ma, Jin-Feng Jia, Qi-Kun Xue, "Ultrathin lead oxide film on Pb(111) and its application in single spin detection", Appl. Phys. Lett. 95, 063107 (2009). DOI: 10.1063/1.3205118, https://doi.org/10.1063/1.3205118
4. Wei Wu, N. M. Harrison, A. J. Fisher, "Electronic structure and exchange interactions in cobalt-phthalocyanine chains", Phys. Rev. B 88, 024426 (2013). DOI: 10.1103/PhysRevB.88.024426, https://doi.org/10.1103/PhysRevB.88.024426
5. Michele Serri, Wei Wu, Luke R. Fleet, Nicholas M. Harrison, Cyrus F. Hirjibehedin, Christopher W. M. Kay, Andrew J. Fisher, Gabriel Aeppli, Sandrine Heutz, "High-temperature antiferromagnetism in molecular semiconductor thin films and nanostructures", Nat. Commun. 5, 3079 (2014). DOI: 10.1038/ncomms4079, https://doi.org/10.1038/ncomms4079
6. Manuel Gruber, Fatima Ibrahim, Samy Boukari, Hironari Isshiki, Loïc Joly, Moritz Peter, Michał Studniarek, Victor Da Costa, Hashim Jabbar, Vincent Davesne, Ufuk Halisdemir, Jinjie Chen, Jacek Arabski, Edwige Otero, Fadi Choueikani, Kai Chen, Philippe Ohresser, Wulf Wulfhekel, Fabrice Scheurer, Wolfgang Weber, Mebarek Alouani, Eric Beaurepaire, Martin Bowen, "Exchange bias and room-temperature magnetic order in molecular layers", Nat. Mater. 14, 981-984 (2015). DOI: 10.1038/nmat4361, https://doi.org/10.1038/nmat4361
7. Yu Wang, Xiao Zheng, Jinlong Yang, "Kondo screening and spin excitation in few-layer CoPc molecular assembly stacking on Pb(111) surface: A DFT+HEOM study", J. Chem. Phys. 145, 154301 (2016). DOI: 10.1063/1.4964675, https://doi.org/10.1063/1.4964675
8. Yuqi Wang, Soroush Arabi, Klaus Kern, Markus Ternes, "Symmetry mediated tunable molecular magnetism on a 2D material", Commun. Phys. 4, 103 (2021). DOI: 10.1038/s42005-021-00601-8, https://doi.org/10.1038/s42005-021-00601-8; preprint arXiv:2009.08487, https://arxiv.org/abs/2009.08487
9. Fudi Zhou, Can Zhang, Zhaoteng Dong, Mengya Ren, Lili Zhou, Yu Zhang, Yeliang Wang, "Scanning Tunnelling Microscopy Study of Spin Interactions in Transition-Metal Phthalocyanine Adsorbates: Mechanisms, Signatures, and Control Strategies", Adv. Sci. 13, e00003 (2026). DOI: 10.1002/advs.202600003, https://doi.org/10.1002/advs.202600003