Oligo(indenoindene) (OInIn): Carbon Ladder & Magnetism

• OInIn is a quasi-one-dimensional π-conjugated ladder polymer featuring pentagon-ring radicals that localize unpaired electrons and induce open-shell magnetism.
• It displays multiple isomers with distinct magnetic exchange patterns and topological phases analyzed using tight-binding, Hubbard models, and DMRG.
• Termination-dependent edge states emerge in double-isomer systems, resulting in tunable SSH-like topological phases with localized magnetic modes.

Searching arXiv for OInIn / oligo(indenoindene) papers to ground the article in the latest available work. Oligo(indenoindene) (OInIn) denotes finite segments of poly(indenoindenes), described as $\pi$-conjugated ladder carbon polymers and as a quasi-one-dimensional, non-alternant, multiradical $\pi$-conjugated hydrocarbon built from fused five- and six-membered carbon rings. In the open-shell limit, each five-membered ring hosts one unpaired electron, so an OInIn can be viewed as an effective chain of localized spins; at the same time, the pentagon rings make the carbon lattice non-bipartite, so Lieb’s theorem does not apply. Across recent work, OInIn has been analyzed as a platform in which pentagon-centered radical states, quasi-zero modes, correlated magnetism, frustration, and termination-dependent topological edge physics can all arise within a single carbon-based ladder architecture (Ortiz et al., 2022, Ortiz, 3 Sep 2025, Agirre et al., 21 Dec 2025).

1. Molecular architecture and low-energy electronic structure

OInIn is built from alternating hexagons and pentagons. Chemically, each repeating unit is an indenofluorene-like motif containing pentagons embedded in a conjugated backbone; electronically, these pentagons host unpaired $\pi$ electrons whose low-energy coupling defines the physics of the chain. The unpaired electrons are mainly localized around the pentagons and adjacent hexagons, so the magnetism is dominated by the exchange between these radical centers. Geometry optimizations further show that the molecules remain essentially planar, so the $\pi$-only picture is justified (Ortiz, 3 Sep 2025, Ortiz et al., 2022).

A central spectral feature is a gapped, weakly dispersing manifold of $P$ quasi-zero modes near the Fermi energy, where $P$ is the number of pentagons. These modes are not perfectly flat zero modes, but they are close enough to midgap and sufficiently localized in energy that, at half filling, they favor one electron per mode and thus generate effective local moments. This identifies OInIn as a controlled open-shell carbon ladder where spin physics emerges from a correlated $\pi$-electron manifold rather than from atomic spins (Agirre et al., 21 Dec 2025).

The structural role of the pentagons is also central to how the electronic problem is organized. One analysis treats OInIn as a finite conjugated ladder structure with a tunable closure hopping $t'$ that completes the pentagons. In the limit $t'=0$, the system is easier to interpret as a bipartite network with zero modes associated with sublattice imbalance; turning on $t'$ introduces the pentagon-induced frustration and hybridizes those zero modes (Ortiz et al., 2022).

2. Isomerism and magnetic classification

OInIn has multiple isomers, and these are classified by the dominant magnetic exchange patterns of their unpaired electrons. One study identifies six possible OInIn isomers and groups them into two classes of three: Class I comprises isomers $\pi$0, and Class II comprises isomers $\pi$1. This classification is based on the magnetic interactions inferred from the underlying sublattice structure and hybridization pattern (Ortiz et al., 2022).

Class I is characterized by frustration-driven behavior. These isomers have sublattice imbalance in the $\pi$2 picture, so neighboring zero modes overlap in a way that produces competing exchanges. Their effective magnetic description is a frustrated $\pi$3 Heisenberg chain with nearest-neighbor exchange that can be ferromagnetic or effectively competing with AF tendencies, and next-nearest-neighbor exchange that is antiferromagnetic. In the simplified form used in the literature,

$\pi$4

with $\pi$5 and $\pi$6 in the relevant OInIn regime (Ortiz et al., 2022).

Class II is characterized by more conventional antiferromagnetic order. These isomers have $\pi$7 in the $\pi$8 picture, so the unpaired electrons are connected mainly by first-neighbor hopping without the same frustrating overlap structure. In the terminology of the topological study, the relevant building blocks are Class-II isomers, for which the dominant coupling is antiferromagnetic first-neighbor exchange. This matters because Class-II units behave, at low energy, like effective “sites” in a one-dimensional tight-binding chain formed by the localized pentagon-centered radical orbitals (Ortiz, 3 Sep 2025).

3. Model Hamiltonians and reduced descriptions

The noninteracting carbon-$\pi$9 problem is introduced through a nearest-neighbor tight-binding Hamiltonian with one $\pi$0-orbital per carbon atom,

$\pi$1

with $\pi$2 eV, and the crystal Hamiltonian is then built through Bloch’s theorem as

$\pi$3

For the isolated radical-derived low-energy degrees of freedom, the problem is further reduced to an effective chain of hybridized unpaired electrons with hopping $\pi$4. When two radical orbitals $\pi$5 and $\pi$6 interact, their splitting is

$\pi$7

This reduction underlies the mapping of double-isomeric Class-II OInIn onto an SSH-like chain (Ortiz, 3 Sep 2025).

For the interacting problem, two distinct but complementary Hubbard-based descriptions appear. A mean-field Hubbard description introduces

$\pi$8

with local magnetization

$\pi$9

and total energy

$\pi$0

A separate treatment uses the full Fermi-Hubbard Hamiltonian

$\pi$1

with representative nanographene parameters $\pi$2 and $\pi$3, and solves it by DMRG in an MPS framework (Ortiz, 3 Sep 2025, Agirre et al., 21 Dec 2025).

The many-body reductions are not purely local. In the DMRG-based spin construction, the low-energy sector is reduced to $\pi$4 emergent spin-$\pi$5 degrees of freedom, one per pentagon, built from optimized delocalized fermionic modes

$\pi$6

with normalization $\pi$7, and corresponding spin operator

$\pi$8

The single-occupancy character is quantified by the “spin fidelity”

$\pi$9

and the coefficients $P$0 are optimized using nonlinear constrained optimization (NLopt/COBYLA) (Agirre et al., 21 Dec 2025).

4. Topological edge states in double-isomeric Class-II OInIn

When the chain is made by alternating two different Class-II OInIn isomers, the effective coupling alternates between two values, directly paralleling the Su-Schrieffer-Heeger model,

$P$1

In this construction, the two isomer orientations generate two distinct inter-radical hoppings, $P$2 and $P$3, and the ratio $P$4 controls the topological phase (Ortiz, 3 Sep 2025).

The band-structure consequences follow directly from this mapping. Pure periodic Class-II OInIn forms, built from a single isomer, remain essentially gapless: the effective radical chain has equivalent couplings and therefore crossing bands near the Fermi level, analogous to the SSH model at $P$5, where a Dirac-like crossing appears. In contrast, the double-isomeric polymer develops a direct band gap because the alternating isomer orientations make the effective hopping alternate, just as in the dimerized SSH chain. For finite fragments, this produces in-gap boundary states whose presence depends strongly on the termination. One termination yields localized edge modes, while the other does not, even though the bulk polymer is the same (Ortiz, 3 Sep 2025).

Topology is characterized through the Zak phase, the 1D Berry phase of the occupied band(s),

$P$6

With inversion or mirror symmetry, $P$7 is quantized to $P$8 or $P$9 modulo $P$0, making it a $P$1 invariant. In the SSH analogy, $P$2 identifies the trivial dimerization pattern and $P$3 the non-trivial one. The double-isomeric OInIn exhibits a non-zero Zak phase for one termination or topological embedding and zero for the other, confirming that the two terminations correspond to different bulk topological sectors. This is the bulk-boundary correspondence in action: the non-trivial phase predicts edge states, and finite-chain calculations show in-gap states strongly localized at the chain ends (Ortiz, 3 Sep 2025).

The localization physics is termination-sensitive. In the non-trivial termination, the HOMO-like in-gap state is exponentially concentrated at the termini and its splitting $P$4 from the partner boundary state decreases rapidly as the number of pentagons $P$5 grows, tending to zero in the long-chain limit. By contrast, the trivial termination lacks such protected boundary states, so the frontier orbitals are more bulk-like and the finite-size splitting behaves differently. In this sense, termination is not a minor structural detail; it decides whether the finite OInIn fragment realizes topological edge physics (Ortiz, 3 Sep 2025).

5. Correlated magnetism, frustration, and emergent spin chains

The earliest magnetic classification emphasizes that OInIn magnetism is controlled by the interplay of pentagon-induced radicalization, sublattice imbalance, and exchange frustration. For isomer $P$6, the ground state is a valence-bond solid of ferromagnetic dimers. Each dimer behaves as an effective $P$7 object, and these dimers order in an alternating pattern consistent with a VBS. The work argues that this state is topologically analogous to the Affleck-Kennedy-Lieb-Tasaki model because the ferromagnetic dimers act as effective spin-1 units, these spin-1 units form a valence-bond solid, such a state is known to possess a hidden topological order and a spin gap, and on an open chain it supports fractional $P$8 edge states. For isomer $P$9 with $\pi$0, local magnetism appears for $\pi$1, and the FM dimer state is lower in energy than broken-dimer, closed-shell, and fully FM states over a substantial interval (Ortiz et al., 2022).

The DMRG study reaches a related but distinct low-energy description for OInIn ladders in terms of a frustrated $\pi$2-$\pi$3 Heisenberg chain,

$\pi$4

with $\pi$5 and $\pi$6. The couplings arise from two distinct processes: ferromagnetic exchange from finite overlap of the delocalized modes via the Hubbard term, and antiferromagnetic superexchange from higher-order hopping processes. The fitted parameters typically satisfy $\pi$7, placing the system in the regime associated with a gapped Haldane-dimer phase with AKLT-like topological character. The low-energy excitation spectrum, entanglement entropy profiles, and spin-spin correlation functions of the full electronic problem are well reproduced by this frustrated spin chain (Agirre et al., 21 Dec 2025).

The emergent spin description is quantitatively refined by optimized delocalized fermionic modes. These modes are nearly singly occupied in the relevant low-energy states, are robust across system sizes, are largely transferable between different OInIn lengths and eigenstates, and are much more accurate than using only the pentagon-tip site. The optimized modes typically place about $\pi$8 of their weight on the pentagon tip and distribute the rest mainly over a small set of neighboring hexagon sites, reaching spin fidelities around $\pi$9 to $t'$0 (Agirre et al., 21 Dec 2025).

Interactions also reshape the topological Class-II case. Mean-field Hubbard calculations show that finite double-isomeric OInIn chains become magnetic once the HOMO-LUMO gap is sufficiently small; crucially, the non-trivial termination, with its smaller $t'$1, becomes magnetic at smaller $t'$2 and shorter length than the trivial one. The local moments are concentrated mainly on the pentagon vertices, and whenever the system orders magnetically, the ground state is antiferromagnetic rather than ferromagnetic, indicating that the dominant interaction is hopping-mediated first-neighbor antiferromagnetic exchange, as expected for Class-II radicals (Ortiz, 3 Sep 2025).

6. Computational evidence, terminations, and interpretive boundaries

The OInIn literature combines several levels of theory. Tight-binding analysis is used to inspect the single-particle spectrum and zero modes in the noninteracting limit. Hubbard-based calculations are carried out by collinear mean-field approximation and by exact diagonalization or CAS-Hubbard with a complete active space $t'$3. DFT benchmarks for isomer $t'$4 use PBE-GGA and PBE0, while the topological study uses unrestricted Kohn–Sham orbitals, and multireference quantum chemistry is represented by CASSCF. The DMRG study treats the full Fermi-Hubbard Hamiltonian directly in an MPS framework rather than starting from a pre-assumed spin model (Ortiz et al., 2022, Ortiz, 3 Sep 2025, Agirre et al., 21 Dec 2025).

These methods yield several convergent results. DFT calculations reproduce the same termination-dependent localization of the HOMO at the edges in the non-trivial case and show that the open-shell antiferromagnetic solution becomes favorable much earlier for the non-trivial termination, while the trivial termination remains non-magnetic over a longer length range. CASSCF calculations reveal that odd-$t'$5 chains have $t'$6 ground states with a single singly occupied natural orbital, while even-$t'$7 chains can be closed-shell or open-shell depending on termination. For the non-trivial termination, the singlet-triplet gap

$t'$8

decreases exponentially with $t'$9, reflecting the approach to degenerate topological edge states; for the trivial termination, $t'=0$0 decreases much less strongly (Ortiz, 3 Sep 2025).

Open issues are also stated explicitly. Isomer $t'=0$1 is described as more hybridized, so its effective open-shell description is less clean; the stronger hybridization can keep the nearest-neighbor exchange antiferromagnetic enough that the frustrated-dimer picture is less robust, and more work is needed to fully settle its ground state (Ortiz et al., 2022). Mild asymmetry effects are also noted in the optimized spin-chain description: because OInIn is non-bipartite and the effective modes are not perfectly symmetric, small coupling asymmetries can mix nearly degenerate triplet states and slightly modify some observables, although these are described as minor corrections (Agirre et al., 21 Dec 2025).

Taken together, these results identify OInIn as a carbon ladder in which the same fused pentagon-hexagon architecture supports several low-energy organizations: effective radical tight-binding chains, SSH-like topological phases in double-isomeric Class-II systems, frustrated $t'=0$2-$t'=0$3 spin chains in open-shell ladders, and AKLT-like or Haldane-dimer interpretations of the correlated magnetic sector. A plausible implication is that OInIn is best understood not through a single universal reduced model, but through a family of reductions whose applicability depends on isomer class, termination, and whether the dominant question concerns band topology, open-shell magnetism, or low-energy many-body spin structure.