Organic diradical molecules have emerged in recent years as highly promising candidates for next generation optoelectronic and quantum information technologies. As efforts continue to synthesise diradicals with unique photophysical properties, developing our theoretical understanding of their complex electronic structure is crucial to enable rational design strategies. However, existing computational tools struggle to properly describe the multi-reference character of diradical electronic states while also preserving a low computational scaling that enables fast calculations on large emissive diradicals. This work presents a new computational method, termed Diradical Extended Restricted Open-shell PPP (D-ExROPPP) theory, which we show yields accurate and spin-pure electronic states at a very low computational cost relative to existing approaches. This is achieved by implementing a novel, spin-adapted configuration interaction approach within the semi-empirical Pariser-Parr-Pople framework. D-ExROPPP is then used to study the electronic structure of three prototypical organic diradicals, and benchmarked against results from computationally intensive, post-Hartree-Fock methods. Excitation energies for these molecules are predicted that show a similar level of accuracy as high-level ab-initio calculations, while requiring up to five orders of magnitude less computational time. We further apply D-ExROPPP to a number of recently reported stable, emissive, organic diradicals. In the majority of cases, electronic transitions measured in experimental UV-vis spectra are reproduced with qualitative accuracy and the spin states involved can be reliably assigned. Together, these results establish D-ExROPPP as a promising new tool for efficiently predicting and interpreting the photophysics of organic diradical molecules.
Introduction
Organic diradical molecules have gained increasing relevance in the past several years, emerging as a promising class of molecules for a variety of applications. While the first stable organic diradical was discovered over 100 years ago[1], interest in the field has exploded since the discovery of synthetic pathways to highly stable, luminescent, organic diradicals[2, 3, 4, 5]. Hosting two unpaired electrons, these large organic diradicals often have unique spin properties that allow for highly unconventional photophysical properties, including high internal photoluminescence quantum efficiencies and the control of emission properties by temperature and magnetic fields[6]. This has drawn interest towards the development of diradical OLED technologies[7], and in particular molecular qubits that make use of organic diradical molecules[8, 9].
Despite this rapidly growing interest, a significant gap remains between experimental advances and our theoretical understanding of diradical molecules and their emission properties. Existing computational methods face major challenges in modelling the highly-correlated electronic structure of these molecules, which typically are made up of 80 - 150 atoms. State-of-the-art multireference methods such as CASSCF and CASPT2 do well to capture electron correlation, but their poor scaling can make them prohibitively expensive for large molecules. Conversely, density-based methods such as time dependent density functional theory (TDDFT) are much more efficient but results often feature a high degree of spin contamination, making it difficult to assign spin states and obtain accurate energies. As research continues into the synthesis and characterisation of organic diradicals, there is a growing need for computational methods that can efficiently and reliably provide insight into their photophysical properties and so help to guide rational design efforts.
This report introduces a novel computational method capable of predicting spin-pure excited electronic states of diradical molecules. We show that this method, termed Diradical Extended Restricted Open-Shell PPP theory (D-ExROPPP) can achieve an accuracy comparable to high-level multireference calculations in some cases, but at a fraction of the computational cost. We ensures spin-purity by constructing configuration state functions (CSFs) which form a basis for an extended configuration interaction approach that includes all single excitations with select double excitations across the π orbitals (XCIS-D). The use of machine-learned Pariser-Parr-Pople (PPP) parameters further ensures the method maintains a low computational cost relative to ab-initio methods, while implicitly capturing features of electronic correlation. After establishing the methodology, D-ExROPPP is tested on a number of organic diradical molecules. We first run calculations on several small diradicals and compare results against those obtained with Complete Active Space Self-Consistent Field (CASSCF) and Generalised Multiconfigurational Quasi-Degenerate Perturbation Theory (GMC-QDPT). We then study a set of 14 recently presented large emissive diradicals, computing UV-vis absorption spectra and comparing predictions against experimental data and results from Spin-Flip Time-Dependent Density Functional Theory (SF-TDDFT).
Previously Published Work
Electronic Properties of Diradicals
Chemical Structure and Bonding
Diradical molecules are widely defined as open-shell molecules, possessing two degenerate or nearly degenerate singly occupied molecular orbitals (SOMOs)[10, 11]. The presence of two unpaired electrons in diradical molecules lends itself to more complex spin states in both ground and excited electronic states. Diradical electronic structure has long been understood through the two-orbital two-electron model (TOTEM), which looks at the electronic configurations that arise from distributing of two electrons among two spatial orbitals (or four spin-orbitals)[10, 11]. This model yields six possible electronic configurations, of which four have spin magnetic quantum number (MS) of 0 (see Figure 1). Of these four, two are zwitterionic configurations and behave like closed-shell singlets, while the remaining two open-shell configurations are not eigenstates of the S2 operator. Spin-adapted linear combinations of these four configurations yield the zwitterionic ∣1ZW±⟩ states and the open-shell singlet ∣1OS⟩ and triplet ∣3OS⟩ states shown in Figure 1.
Figure 1.
Figure 1. Representations of the electronic configurations and spin-adapted electronic states that arise from the two-electron two-orbital model for describing diradical electronic structure. K00′ represents the exchange energy between the two SOMOs, 0 and 0′. The energy difference between the ∣1OS⟩ and ∣1ZW−⟩ states is given by ΔE=EOS1−EZW−1=J00′+2K00′−21(J00+J0′0′), where Jab is the Coulomb energy between the MOs a and b. This figure shows the case where ΔE>0, however, this depends on the molecule and it is possible to have ΔE<0 or ΔE=0.
It is also important to consider the nature of the SOMOs themselves. In many cases, SOMOs are delocalised across the diradical molecule, meaning that the electron density associated with each unpaired electron is distributed over many atoms in the conjugated π-system. However, it is possible to rotate SOMO coefficients such that each unpaired electron is mostly or fully localised on different atoms in the molecule. SOMOs that are localised in this way are often called generalised valence bond (GVB) orbitals[12, 13]. Molecules in which GVB orbitals do not share electron density between any of the same atoms can be classified as disjoint, whereas molecules without this property are classed as non-disjoint[14]. Salem and Rowland showed in 1972 that the choice of delocalised versus localised SOMOs alters the description of the low-lying TOTEM electronic states[10]. For a given molecule, the ∣1OS⟩ state formed from delocalised orbitals maps exactly to the zwitterionic state ∣1ZW−⟩ when formed from localised orbitals. Figure 2 highlights this effect in the case of the widely studied cyclobutadiene (CBD) molecule. The exchange interaction K00′ will typically be very small between localised SOMOs, meaning that ∣1OS⟩ and ∣3OS⟩ will be close in energy. However, when SOMOs are delocalised K00′ is expected to be much greater, meaning that ∣1ZW−⟩ will typically be the lowest energy singlet state.
Figure 2. Energy level diagram highlighting the transformation between ∣1ZW+⟩ and ∣1OS⟩ that occurs when transforming from the delocalised to the localised SOMOs of cyclobutadiene. ϕ and χ represent localised and delocalised molecular orbitals respectively, and the algebraic transformation between the two sets is shown in grey. As in Figure 1, the ∣1ZW−⟩ state is depicted as being lower in energy than ∣1OS⟩ in the delocalised case, however, the relative energies of these states varies depending on the diradical molecule.
A closely related class of molecules to diradicals are diradicaloids, which are loosely defined as molecules with an electronic structure that is intermediate between pure open-shell and closed-shell molecules[11]. This arises when a relatively small energy gap exists between the two molecular orbitals, meaning they behave both like the degenerate SOMOs of a pure diradical and the HOMO and LUMO of a pure closed-shell molecule.
Diradicaloids typically possess a singlet ground state which is somewhere in-between the ∣1ZW−⟩ state of a pure-diradical and the doubly-occupied S0 state of a closed-shell molecule (Figure 3). In a pure diradical, 0 and 0’ are degenerate meaning ∣ψ0ψˉ0⟩ and ∣ψ0′ψˉ0′⟩ configurations have equal weighting in ∣1ZW−⟩. As the 0’ orbital rises in energy relative to the 0 orbital, its weighting (C0′) slowly decreases until it eventually contributes very little to the ∣1ZW−⟩ state, which is now effectively described by the single ∣ψ0ψˉ0⟩ configuration. We find C0C0′ indirectly characterises the degree of open- vs. closed-shell character, where C0C0′=1 indicates a pure-diradical and C0C0′=0 suggests that a molecule can be described as fully closed-shell.
Figure 3. Schematic to represent the transition from diradical to diradicaloid to closed shell molecule, inspired by that presented by Stuyver et al.[11]. ϵ0 and ϵ0′ represent the energies of orbitals 0 and 0’ respectively, where orbital 0 is lower in energy and orbital 0’ is higher in energy. Energies are shown relative to that of the open-shell triplet in order to highlight how the singlet energy becomes much lower than the triplet energy.
Since the 1970s, there have been many attempts to define a formal metric for a molecules ’diradical character’, which is typically expressed using the index y∈[0,1], where y=0 indicates purely closed-shell character and y=1 means that a molecule is a pure-diradical[15, 16, 17, 18, 19, 20]. One of the most widely used formulations is given by Yamaguchi’s equation, presented in eq:1 where nHONO and nLUNO are the occupation numbers of the highest occupied and lowest unoccupied natural orbitals respectively calculated with a broken-symmetry unrestricted Hartree-Fock approach[21].
y0=1−1+T022T0(1)
T0=2nHONO−nLUNO(2)
Despite the number of different approaches developed, estimating a molecules diradical character remains a theoretical challenge, meanwhile obtaining estimates from experiment is harder still. This may in part explain why the distinction between diradical and diradicaloid molecules is largely ill-defined across the literature, with the two terms sometimes being used interchangeably.
Photophysical Properties
To date, emissive diradical molecules are most commonly synthesised by coupling two monoradical molecules together with an aromatic linker[6]. The most widely used radical precursor is the tris(2,4,6-trichlorophenyl)methyl (TTM) radical, from which a number of diradicals have been designed using a wide variety of linkers[8, 22, 23, 24, 25, 26, 27, 28, 29, 30]. Closely related radicals M2TTM as well as the PyBTM (pyridyl bis(2,4,6-trichlorophenyl)methyl) have also seen widespread use[2, 4, 5, 31, 32, 33, 34].
The open-shell electronic structure of organic diradicals typically gives these molecules unique photophysical properties that have the potential to be harnessed in a number of novel technologies. One avenue that has garnered attention is the use of organic diradicals in a new generation of OLED devices[2, 6, 22]. In conventional closed-shell molecules, the excitation of a single electron will yield singlet and triplet excited states in a 1:3 ratio, meaning radiative decay (a.k.a. fluorescence) back to the ground singlet state is typically limited to a maximum internal quantum efficiency (IQE) of 25%[35]. In contrast, the near-degeneracy of low-lying singlet and triplet electronic states in diradicals means that there exist many more spin-allowed, radiative pathways back to both singlet and triplet states. Research is ongoing to develop diradical OLEDs with much improved IQEs, and more widely to develop optoelectronic technologies that take advantage of favourable spin-statistics. However, most organic diradicals synthesised to date have shown modest photoluminescence quantum yields of around 1 − 2%[8, 22, 25, 32, 36], highlighting the need for design strategies that are guided by an understanding of diradical electronic structure.
The property that has perhaps attracted the most interest in organic luminescent diradicals is their capacity for magnetoluminescence and thermally activated emission[6]. Many emissive diradicals feature very small singlet-triplet energy gaps (∣ΔES−T∣≈1−10meV), which is comparable to or smaller than kBT at 298K. This means that the thermal populations of these states are highly sensitive to environmental conditions and can be manipulated with relative ease[2, 4, 22, 33, 34, 37].
In the case of magnetoluminescence, external magnetic fields can be applied that couple directly to the open-shell triplet state, lifting the degeneracy between states with different Ms and decreasing the relative energy the Ms=−1 state[38]. The result is an increase in the equilibrium population of the ∣3OS⟩ state, meaning the intensity of emissive transitions arising from triplets is increased compared to those associated with singlet states. In 2023, this was observed by Kusamoto et. al after their synthesis of the THDBA-(PyBTM)2 molecule. By applying a magnetic field, they were able to enhance the intensity of a 645nm emission peak by 22% while decreasing the intensity of a broad 750nm peak by 40%[33]. A similar phenomenon that several diradicals have been found to display is thermally activated emission, whereby temperature can be used to influence the relative populations of S0 and T0 states and thus change the intensities of emissive transitions noticeably[4, 23].
Beyond their potential in optoelectronic devices, organic luminescent diradicals have recently emerged as promising candidates for molecular qubits in quantum information science. Solid-state qubit platforms, such as nitrogen-vacancy centres in diamond, have been heavily researched in recent years, but are limited by the restricted tunability of their optical and spin characteristics[9]. Organic diradicals have attracted attention as potential alternatives, due to their high tuneability and relatively low cost[8]. Near degenerate singlet and triplet low-lying states provide a convenient framework for spin-optical interfaces, enabling optical initialisation and read-out of spin states often through spin-selective intersystem crossing pathways[9]. In 2023, Evans and co-workers exploited this in the diradical molecule (TTM-1Cz)2-An, where the near-degeneracy of the singlet and triplet manifolds facilitated the generation of a high-spin quintet state upon photoexcitation, which could subsequently be optically read out[24]. Building on this, Wasielewski and co-workers established meta-TTM-Ph-TTM as an optically addressable molecular qubit in 2024, demonstrating that the spin-polarised ground state arising from the diradical electronic structure could be coherently manipulated and read out optically[8]. Together, these results highlight the growing potential of organic diradicals as a chemically tuneable platform for quantum sensing technologies, where once again functionality relies on the precise characterisation of the energies and multiplicities of electronic states.
Existing Computational Methods for Diradicals
Density-Based Methods
Density functional theory (DFT) has long been the most widely used computational method across the chemical sciences. The use of exchange-correlation functionals (vxc[ρ]) to describe electron correlation decreases the computational cost and scaling of DFT compared to most other ab-initio methods, while still yielding sufficient qualitative accuracy for most applications. Time-dependent density functional theory (TD-DFT) can be seen as an extension of DFT that allows for prediction of excited electronic states through linear response theory. The computational cost of TD-DFT is also relatively low compared to other methods for predicting excitation energies, which largely explains its widespread use in the study of emissive diradical molecules and their photophysical properties[2, 4, 5, 8, 22, 23, 25, 26, 27, 30, 33, 34, 37]. However, the single-reference nature of DFT and TD-DFT limits the ability of these methods to reliably describe the electronic structure of diradical molecules. In almost all cases, DFT and TD-DFT are used with an unrestricted reference, meaning that the molecule’s α and β electrons are allowed to occupy orbitals with different spatial distributions. While this approach typically yields better energy estimates, the electronic states predicted, and particularly the excited states, are highly susceptible to spin contamination, whereby states of different multiplicity are mixed to yield wavefunctions of ill-defined total spin[39]. Consequently, even in cases where TD-DFT excitation energies are in good agreement with experimental results, the spin multiplicities of the underlying states often cannot be reliably determined and the interpretive value of results is limited. Additionally, the results of DFT and TD-DFT calculations are known to be sensitive to the choice of exchange-correlation functional, with no universally optimal choice having been established for studying diradicals[40, 41]. A vast number of functionals have been developed over recent decades, with UB3LYP being the most widely adopted for use with emissive diradical molecules[2, 4, 5, 8, 22, 23, 25, 26, 27, 30, 33, 34, 37].
A wide variety of approaches have been developed to add multiconfigurational character into DFT and address the issue of spin-contamination. Created over forty years ago, broken symmetry DFT (BS-DFT) offers a way to mimic the open-shell singlet state in diradicals by localising α and β electrons in the HOMO onto opposite sites of the molecules[42, 43]. This creates an artificial ’symmetry broken’ singlet state which can approximate the 1OS⟩ state but is still spin-contaminated meaning it is coupled to triplet and higher multiplicity states[44]. Reasonably accurate ΔES−T estimates can often be obtained regardless using spin-projection corrections such as that presented by Yamaguchi (see (3)), where ⟨ST2⟩, ⟨SBS2⟩ are the expectation values of the total spin operator for the triplet and broken symmetry singlet states respectively[45]. Owing to their relative simplicity and low computational cost, broken-symmetry approaches remain a very popular way of approximating the multireference states of diradicals[4, 8, 22, 23, 25, 27, 30, 33]. However, approximating spin-pure solutions through spin-projections of this kind has been shown to be unreliable in a variety of contexts including for organic molecules with high dynamic correlation[44, 46, 47, 48, 49].
ΔES−T=ES−ET≈⟨ST2⟩−⟨SBS2⟩⟨ST2⟩(EBS−ET)(3)
Spin-flip time dependent density functional theory (SF-TDDFT) is another method that aims to approximate multi-reference diradical states within the single-reference formalism of TD-DFT[50]. SF-TDDFT applies spin flips to a high-spin triplet reference determinant (∣↑↑⟩) to generate additional configurations (∣↑↓⟩, ∣↓↑⟩) that are then used in a typical TD-DFT treatment. This approach can recover a decent amount of the multi-configurational character of low-lying diradical electronic states (often referred to as static correlation), and can yield better estimates of ΔES−T than broken symmetry approaches[51, 52]. However, the addition of spin-flips increases the computational cost of the method and typically struggles to produce spin-purity in the higher-lying electronic states of diradicals, meaning spin-contamination and accuracy remain an issue[40].
This was addressed later by Ho Choi and colleagues with their creation of mixed-reference spin-flip TDDFT (MRSF-TDDFT), which includes a second reference determinant ∣↓↓⟩ from which spin-flipped excitations can be generated[53, 54]. This has been shown to generally improve the accuracy of ΔES−T predictions for diradical molecules compared to SF-TDDFT and also mostly removes spin-contamination from low-lying electronic states[55]. The MRSF-TDDFT method in particular shows promise for characterising the excited states of large emissive diradicals, but has not yet widely been adopted, perhaps due in part to its increased computational cost compared to conventional TD-DFT or its broken symmetry equivalent.
Wave function-Based Methods
Looking beyond density-based methods, there exist a number of mature wavefunction-based methods which extend the single-reference Hartree-Fock formalism. Perhaps the most widely used of these are multiconfigurational self-consistent field (MCSCF) approaches, which express electronic states as a linear combination of many configurations, which are formed by distributing a defined number of electrons across a defined set of active MOs (Eq. 3). The coefficients of the linear expansion (cI) and the molecular orbital coefficients are simultaneously optimised to minimise the energy of one or more electronic states, depending on whether a state-specific or state-averaged approach is used[56]. The most widely used MCSCF method is the complete active space self consistent field (CASSCF) method[57], although others have been developed such as RASSCF (restricted active space SCF) which offer slightly more efficient computational scaling[58]. MCSCF methods are ideal for capturing multireference character, which has led to their frequent adoption for studying the electronic structure of small and medium-sized diradical and diradicaloid molecules[9, 41, 59, 60].
References
[1]J. Thiele and H. Balhorn, “Ueber einen chinoiden kohlenwasserstoff”, Berichte der deutschen chemischen Gesellschaft 37, 1463–1470 (1904).DOI
[2]Y. Hattori, E. Michail, A. Schmiedel, M. Moos, M. Holzapfel, I. Krummenacher, H. Braunschweig, U. Müller, J. Pflaum, and C. Lambert, “Luminescent mono-, di-, and triradicals: bridging polychlorinated triarylmethyl radicals by triarylamines and triarylboranes”, Chemistry – A European Journal 25, 15463–15471 (2019).DOI
[3]Z. Feng, Y. Chong, S. Tang, Y. Fang, Y. Zhao, J. Jiang, and X. Wang, “A stable triplet diradical emitter”, Chemical Science 12, 15151–15156 (2021).
[4]A. Mizuno, R. Matsuoka, S. Kimura, K. Ochiai, and T. Kusamoto, “Spin-correlated luminescence of a carbazole-containing diradical emitter: single-molecule magnetoluminescence and thermally activated emission”, Journal of the American Chemical Society 146, 18470–18483 (2024).DOI
[5]R. Chowdhury, P. Murto, N. A. Panjwani, Y. Sun, P. Ghosh, Y. Boeije, C. D. Cordeiro, V. Derkach, S.-J. Woo, O. Millington, D. G. Congrave, Y. Fu, T. B. E. Mustafa, M. Monteverde, J. Cerdá, G. Londi, J. Behrends, A. Rao, D. Beljonne, A. Chepelianskii, H. Bronstein, and R. H. Friend, “Bright triplet and bright charge-separated singlet excitons in organic diradicals enable optical read-out and writing of spin states”, Nature Chemistry 17, 1410–1417 (2025).DOI
[6]Y. Zhu, Z. Zhu, S. Wang, Q. Peng, and A. Abdurahman, “Stable luminescent diradicals: the emergence and potential applications”, Angewandte Chemie International Edition 64, e202423470 (2025).DOI
[7]P. Murto and H. Bronstein, “Electro-optical π-radicals: design advances, applications and future perspectives”, Journal of Materials Chemistry C 10, 7368–7403 (2022).DOI
[8]S. M. Kopp, S. Nakamura, B. T. Phelan, Y. R. Poh, S. B. Tyndall, P. J. Brown, Y. Huang, J. Yuen-Zhou, M. D. Krzyaniak, and M. R. Wasielewski, “Luminescent organic triplet diradicals as optically addressable molecular qubits”, Journal of the American Chemical Society 146, 27935–27945 (2024).DOI
[9]Y. R. Poh, D. Morozov, N. P. Kazmierczak, R. G. Hadt, G. Groenhof, and J. Yuen-Zhou, “Alternant hydrocarbon diradicals as optically addressable molecular qubits”, Journal of the American Chemical Society 146, 15549–15561 (2024).DOI
[10]L. Salem and C. Rowland, “The electronic properties of diradicals”, Angewandte Chemie International Edition in English 11, 92–111 (1972).
[11]T. Stuyver, B. Chen, T. Zeng, P. Geerlings, F. De Proft, and R. Hoffmann, “Do diradicals behave like radicals?”, Chemical Reviews 119, 11291–11351 (2019).DOI
[12]W. Carl Lineberger and W. Thatcher Borden, “The synergy between qualitative theory, quantitative calculations, and direct experiments in understanding, calculating, and measuring the energy differences between the lowest singlet and triplet states of organic diradicals”, Physical Chemistry Chemical Physics 13, 11792 (2011).DOI
[13]W. T. Borden, “Diradicals – a fifty year fascination”, in The foundations of physical organic chemistry: fifty years of the james flack norris award (American Chemical Society, 2015), pp. 251–303.
[14]W. T. Borden and E. R. Davidson, “Effects of electron repulsion in conjugated hydrocarbon diradicals”, Journal of the American Chemical Society 99, 4587–4594 (1977).DOI
[15]E. F. Hayes and A. K. Q. Siu, “Electronic structure of the open forms of three-membered rings”, Journal of the American Chemical Society 93, 2090–2091 (1971).
[16]K. Takatsuka, T. Fueno, and K. Yamaguchi, “Distribution of odd electrons in ground-state molecules”, Theoretica Chimica Acta 48, 175–183 (1978).DOI
[17]M. Nakano, “Open-shell-character-based molecular design principles: applications to nonlinear optics and singlet fission”, The Chemical Record 17, 27–62 (2016).
[18]M. Nakano, K. Fukuda, S. Ito, H. Matsui, T. Nagami, S. Takamuku, Y. Kitagawa, and B. Champagne, “Di-radical and ionic characters of open-shell singlet molecular systems”, The Journal of Physical Chemistry A 121, 861–873 (2017).
[19]V. Bachler, G. Olbrich, F. Neese, and K. Wieghardt, “Theoretical evidence for the singlet diradical character of square planar nickel complexes containing two o-semiquinonato type ligands”, Inorganic Chemistry 41, 4179–4193 (2002).
[20]M. Head-Gordon, “Characterizing unpaired electrons from the one-particle density matrix”, Chemical Physics Letters 372, 508–511 (2003).
[21]K. Yamaguchi, “The electronic structures of biradicals in the unrestricted hartree-fock approximation”, Chemical Physics Letters 33, 330–335 (1975).DOI
[22]A. Abdurahman, J. Wang, Y. Zhao, P. Li, L. Shen, and Q. Peng, “A highly stable organic luminescent diradical”, Angewandte Chemie International Edition 62, e202300772 (2023).
[23]A. Abdurahman, L. Shen, J. Wang, M. Niu, P. Li, Q. Peng, J. Wang, and G. Lu, “A highly efficient open-shell singlet luminescent diradical with strong magnetoluminescence properties”, Light Sci. Appl. 12, 272 (2023).DOI
[24]S. Gorgon, K. Lv, J. Grüne, B. H. Drummond, W. K. Myers, G. Londi, G. Ricci, D. Valverde, C. Tonnelé, P. Murto, A. S. Romanov, D. Casanova, V. Dyakonov, A. Sperlich, D. Beljonne, Y. Olivier, F. Li, R. H. Friend, and E. W. Evans, “Reversible spin-optical interface in luminescent organic radicals”, Nature 620, 538–544 (2023).DOI
[25]X. Chang, M. E. Arnold, R. Blinder, J. Zolg, J. Wischnat, J. van Slageren, F. Jelezko, A. J. C. Kuehne, and M. von Delius, “A stable chichibabin diradicaloid with near-infrared emission”, Angewandte Chemie International Edition 63, e202404853 (2024).DOI
[26]D. Mesto, M. Orza, B. Bardi, A. Punzi, I. Ratera, J. Veciana, G. Farinola, A. Painelli, F. Terziani, D. Blasi, and F. Negri, “Luminescent trityl-based diradicaloids: a theoretical and experimental assessment of charge-resonance in low-lying excited states”, Chemistry – A European Journal 31, e202500749 (2025).DOI
[27]M. E. Arnold, A. Lebzelter, P. Thielert, R. Blinder, J. Schmid, J. Zolg, E. Spatola, F. Jelezko, M. von Delius, S. Richert, and A. J. C. Kuehne, “Photostable triphenylmethyl-based diradicals with a degenerate singlet-triplet ground state and strong photoluminescence”, Chemical Science 16, 14616–14624 (2025).DOI
[28]Z. Feng, J. Zhou, X. He, B. Wang, G. Xie, X. Qiao, L. Liu, Z. Xie, and Y. Ma, “Extremely stable perylene bisimide-bridged regioisomeric diradicals and their redox properties”, Chemistry – A European Journal 30, e202302943 (2023).DOI
[29]Y. Hattori and G. Rapenne, “Facile preparation methods for triarylmethyl-type stable luminescent diradicals and higher multiradicals”, European Journal of Organic Chemistry 29, e70329 (2026).DOI
[30]L. Schöneburg, M. E. Arnold, M. R. Rapp, R. Blinder, M. Gross, J. Zolg, F. Jelezko, H. F. Bettinger, and A. J. C. Kuehne, “A novel [2.2]paracyclophane bridged tris(2, 4, 6-trichlorophenyl)methyl-diradical and its interaction with circularly polarized light”, Journal of Materials Chemistry C 14, 1666–1674 (2026).
[31]R. Chowdhury, A. Inglis, L. E. Walker, P. Murto, C. Delpiano-Cordeiro, C. Morrison, N. A. Panjwani, Y. Fu, Y. Sun, W. Zhou, P. J. Skabara, A. Rao, A. Chepelianskii, H. Bronstein, S. L. Bayliss, and R. H. Friend, Room temperature optical control of spin states in organic diradicals, 2025.
[32]C. P. Yu, R. Chowdhury, Y. Fu, P. Ghosh, W. Zeng, T. B. E. Mustafa, J. Grüne, L. E. Walker, D. G. Congrave, X. W. Chua, P. Murto, A. Rao, H. Sirringhaus, F. Plasser, C. P. Grey, R. H. Friend, and H. Bronstein, “Near-infrared luminescent open-shell π-conjugated systems with a bright lowest-energy zwitterionic singlet excited state”, Science Advances 10, eado3476 (2024).
[33]R. Matsuoka, S. Kimura, T. Miura, T. Ikoma, and T. Kusamoto, “Single-molecule magnetoluminescence from a spatially confined persistent diradical emitter”, en, J. Am. Chem. Soc. 145, 13615–13622 (2023).
[34]Y. Hattori, K. Yamamoto, R. Kitajima, and K. Uchida, “Bright fluorescent p-phenylene-bridged triarylmethyl highly stable diradical”, Chemistry – A European Journal 30 (2024).
[35]A. Abdurahman, T. J. H. Hele, Q. Gu, J. Zhang, Q. Peng, M. Zhang, R. H. Friend, F. Li, and E. W. Evans, “Understanding the luminescent nature of organic radicals for efficient doublet emitters and pure-red light-emitting diodes”, Nature Materials 19, 1224–1229 (2020).
[36]S. Arikawa, A. Shimizu, D. Shiomi, K. Sato, T. Takui, H. Sotome, H. Miyasaka, M. Murai, S. Yamaguchi, and R. Shintani, “A kinetically stabilized nitrogen-doped triangulene cation: stable and nir fluorescent diradical cation with triplet ground state”, Angewandte Chemie International Edition 62, e202302714 (2023).
[37]Z. Tong, S. Zhang, W. Niu, T. Yu, X. Zhang, P. Yao, J. Wang, Y. Han, G. Li, and S. Dong, “Deciphering spin-governed dual emissions in ttm diradicaloids: unraveling the interplay of singlet and triplet fluorescence pathways”, Journal of the American Chemical Society 147, 39232–39246 (2025).
[38]X. Wang, Z. Zhu, Q. Peng, and A. Abdurahman, “Magnetoluminescence in organic radical emitters”, Fundamental Research (2026).
[39]J. A. Pople, P. M. W. Gill, and N. C. Handy, “Spin-unrestricted character of kohn-sham orbitals for open-shell systems”, International Journal of Quantum Chemistry 56, 303–305 (1995).
[40]Q. Sun, J.-L. Brédas, and H. Li, “Reliable diradical characterization via precise singlet–triplet gap calculations: application to thiele, chichibabin, and müller analogous diradicals”, Journal of Chemical Theory and Computation 21, 1194–1202 (2025).
[41]D. Reta Mañeru, A. K. Pal, I. d. P. R. Moreira, S. N. Datta, and F. Illas, “The triplet–singlet gap in the m-xylylene radical: a not so simple one”, Journal of Chemical Theory and Computation 10, 335–345 (2013).
[42]L. Noodleman, “Valence bond description of antiferromagnetic coupling in transition metal dimers”, The Journal of Chemical Physics 74, 5737–5743 (1981).
[43]F. Neese, “Definition of corresponding orbitals and the diradical character in broken symmetry dft calculations on spin coupled systems”, Journal of Physics and Chemistry of Solids 65, 781–785 (2004).
[44]N. Ferré, N. Guihéry, and J.-P. Malrieu, “Spin decontamination of broken-symmetry density functional theory calculations: deeper insight and new formulations”, Physical Chemistry Chemical Physics 17, 14375–14382 (2015).
[45]K. Yamaguchi, F. Jensen, A. Dorigo, and K. Houk, “A spin correction procedure for unrestricted hartree-fock and møller-plesset wavefunctions for singlet diradicals and polyradicals”, Chemical Physics Letters 149, 537–542 (1988).
[46]J. M. Wittbrodt and H. B. Schlegel, “Some reasons not to use spin projected density functional theory”, The Journal of Chemical Physics 105, 6574–6577 (1996).
[47]R. Costa, R. Valero, D. R. Mañeru, I. d. P. R. Moreira, and F. Illas, “Spin adapted versus broken symmetry approaches in the description of magnetic coupling in heterodinuclear complexes”, Journal of Chemical Theory and Computation 11, 1006–1019 (2015).
[48]48P. Rivero, I. d. P. R. Moreira, F. Illas, and G. E. Scuseria, “Reliability of range-separated hybrid functionals for describing magnetic coupling in molecular systems”, The Journal of Chemical Physics 129 (2008).
[49]49G. Trinquier and J.-P. Malrieu, “Kekulé versus lewis: when aromaticity prevents electron pairing and imposes polyradical character”, Chemistry – A European Journal 21, 814–828 (2014).
[50]50Y. Shao, M. Head-Gordon, and A. I. Krylov, “The spin–flip approach within time-dependent density functional theory: theory and applications to diradicals”, The Journal of Chemical Physics 118, 4807–4818 (2003).
[51]51L. V. Slipchenko and A. I. Krylov, “Singlet-triplet gaps in diradicals by the spin-flip approach: a benchmark study”, The Journal of Chemical Physics 117, 4694–4708 (2002).
[52]52R. Valero, F. Illas, and D. G. Truhlar, “Magnetic coupling in transition-metal binuclear complexes by spin-flip time-dependent density functional theory”, Journal of Chemical Theory and Computation 7, 3523–3531 (2011).
[53]53S. Lee, M. Filatov, S. Lee, and C. H. Choi, “Eliminating spin-contamination of spin-flip time-dependent density functional theory within linear response formalism by the use of zeroth-order mixed-reference (mr) reduced density matrix”, The Journal of Chemical Physics 149, 104101 (2018).
[54]54K. Komarov, M. Oh, H. Nakata, S. Lee, and C. H. Choi, “Umrsf-tddft: unrestricted mixed-reference spin-flip-tddft”, The Journal of Physical Chemistry A 128, 9526–9537 (2024).
[55]55Y. Horbatenko, S. Sadiq, S. Lee, M. Filatov, and C. H. Choi, “Mixed-reference spin-flip time-dependent density functional theory (mrsf-tddft) as a simple yet accurate method for diradicals and diradicaloids”, Journal of Chemical Theory and Computation 17, 848–859 (2021).
[56]56A. C. Wahl and G. Das, “The multiconfiguration self-consistent field method”, in Methods of electronic structure theory, Vol. 3, edited by H. F. Schaefer III, Modern Theoretical Chemistry (Plenum Press, New York, 1977), pp. 51–78.
[57]57B. O. Roos, P. R. Taylor, and P. E. Sigbahn, “A complete active space scf method (casscf) using a density matrix formulated super-ci approach”, Chemical Physics 48, 157–173 (1980).
[58]58J. Olsen, B. O. Roos, P. Jo/rgensen, and H. J. A. Jensen, “Determinant based configuration interaction algorithms for complete and restricted configuration interaction spaces”, The Journal of Chemical Physics 89, 2185–2192 (1988).
[59]59L. Matasović, H. Bronstein, R. H. Friend, and F. Plasser, “Classification and quantitative characterisation of the excited states of π-conjugated diradicals”, Faraday Discussions 254, 107–129 (2024).
[60]60A. G. Starikov, M. G. Chegerev, A. A. Starikova, and V. I. Minkin, “Magnetic properties of diradical dihydropyrene derivatives: a computational modeling”, Russian Chemical Bulletin 74, 2702–2712 (2025).
[61]N. W. Moriarty, R. Lindh, and G. Karlström, “Tetramethylene: a caspt2 study”, Chemical Physics Letters 289, 442–450 (1998).
[62]W. T. Borden and E. R. Davidson, “The importance of including dynamic electron correlation in ab initio calculations”, Accounts of Chemical Research 29, 67–75 (1996).
[63]W. T. Borden, E. R. Davidson, and D. Feller, “Rhf and two-configuration scf calculations are inappropriate for conjugated diradicals”, Tetrahedron 38, 737–739 (1982).
[64]K. Andersson, P.-Å. Malmqvist, and B. O. Roos, “Second-order perturbation theory with a complete active space self-consistent field reference function”, The Journal of Chemical Physics 96, 1218–1226 (1992).
[65]C. Angeli, R. Cimiraglia, and J.-P. Malrieu, “N-electron valence state perturbation theory: a spinless formulation and an efficient implementation of the strongly contracted and of the partially contracted variants”, The Journal of Chemical Physics 117, 9138–9153 (2002).
[66]H. Nakano, R. Uchiyama, and K. Hirao, “Quasi-degenerate perturbation theory with general multiconfiguration self-consistent field reference functions”, Journal of Computational Chemistry 23, 1166–1175 (2002).
[67]B. O. Roos and K. Andersson, “Multiconfigurational perturbation theory with level shift — the cr2 potential revisited”, Chemical Physics Letters 245, 215–223 (1995).
[68]G. Li Manni, R. K. Carlson, S. Luo, D. Ma, J. Olsen, D. G. Truhlar, and L. Gagliardi, “Multiconfiguration pair-density functional theory”, Journal of Chemical Theory and Computation 10, 3669–3680 (2014).
[69]S. J. Stoneburner, D. G. Truhlar, and L. Gagliardi, “Mc-pdft can calculate singlet–triplet splittings of organic diradicals”, The Journal of Chemical Physics 148, 064108 (2018).
[70]R. Pariser and R. G. Parr, “A semi-empirical theory of the electronic spectra and electronic structure of complex unsaturated molecules. i.”, The Journal of Chemical Physics 21, 466–471 (1953).
[71]R. Pariser and R. G. Parr, “A semi-empirical theory of the electronic spectra and electronic structure of complex unsaturated molecules. ii.”, The Journal of Chemical Physics 21, 767–776 (1953).
[72]J. A. Pople, “Electron interaction in unsaturated hydrocarbons”, Transactions of the Faraday Society 49, 1375 (1953).
[73]J. A. Pople and D. L. Beveridge, Approximate molecular orbital theory (McGraw-Hill, New York, 1970).
[74]J. D. Green and T. J. H. Hele, “Exropp: fast, accurate, and spin-pure calculation of the electronically excited states of organic hydrocarbon radicals”, The Journal of Chemical Physics 160, 164110 (2024).
[75]D. Maurice and M. Head-Gordon, “On the nature of electronic transitions in radicals: an extended single excitation configuration interaction method”, The Journal of Physical Chemistry 100, 6131–6137 (1996).
[76]76M. Dubbini, F. Bonvini, L. Savi, and F. Di Maiolo, “Turning on organic radical emitters”, The Journal of Physical Chemistry C 128, 18158–18169 (2024).
[77]77L. Savi, M. T. Barreca, M. Bedogni, and F. Di Maiolo, “Organic diradicals bridged by inverted singlet–triplet units for optical–spin interfaces”, Journal of Chemical Theory and Computation, 1465–1475 (2025).
[78]78M. T. Barreca and F. Di Maiolo, “Spin coupling in symmetric and asymmetric allyl and phenalenyl diradicals bridged by an inverted singlet–triplet system”, The Journal of Physical Chemistry A 130, 2399–2410 (2026).
[79]79H. C. Longuet-Higgins and J. A. Pople, “The electronic spectra of aromatic molecules iv: excited states of odd alternant hydrocarbon radicals and ions”, Proceedings of the Physical Society. Section A 68, 591–600 (1955).
[80]80T. Hele, “On the electronic structure of alternant conjugated organic radicals for light-emitting diode applications”, in Physical chemistry of semiconductor materials and interfaces xx (2021), p. 35.
[81]81A. Szabo and N. S. Ostlund, Modern quantum chemistry : introduction to advanced electronic structure theory (Dover Publications, Mineola, N.Y., 1989).
[82]82T. J. H. Hele, E. G. Fuemmeler, S. N. Sanders, E. Kumarasamy, M. Y. Sfeir, L. M. Campos, and N. Ananth, “Anticipating acene-based chromophore spectra with molecular orbital arguments”, The Journal of Physical Chemistry A 123, 2527–2536 (2019).
[83]83A. M. Alvertis, S. Lukman, T. J. H. Hele, E. G. Fuemmeler, J. Feng, J. Wu, N. C. Greenham, A. W. Chin, and A. J. Musser, “Switching between coherent and incoherent singlet fission via solvent-induced symmetry breaking”, Journal of the American Chemical Society 141, 17558–17570 (2019).
[84]84N. Mataga and K. Nishimoto, “Electronic structure and spectra of nitrogen heterocycles”, Zeitschrift für Physikalische Chemie 13, 140–157 (1957).
[85]85K. Jug, “Theoretical basis and design of the ppp model hamiltonian”, International Journal of Quantum Chemistry 37, 403–414 (1990).
[86]86J. Shen, L. E. Walker, K. Ma, J. D. Green, H. Bronstein, K. T. Butler, and T. J. H. Hele, “Learning radical excited states from sparse data”, Chemical Science 16, 17356–17368 (2025).
[87]87T. Helgaker, P. Jorgensen, and J. Olsen, Molecular electronic-structure theory (John Wiley & Sons, Ltd, Chichester, UK, 2000).
[88]88E. Cho, Q. Sun, E. P. McBride, J.-L. Brédas, and V. Coropceanu, “Luminescent radicals based on the tris(trichlorophenyl)methyl acceptor: how to choose the donor component”, ACS Materials Letters 7, 1620–1625 (2025).
[89]I. Shavitt, “The method of configuration interaction”, in Methods of electronic structure theory (Plenum Press, New York, 1977), pp. 189–275.
[90]R. B. Lehoucq, D. C. Sorensen, and C. Yang, ARPACK users’ guide: solution of large-scale eigenvalue problems with implicitly restarted Arnoldi methods, Software, Environments, and Tools (Society for Industrial and Applied Mathematics, Philadelphia, PA, 1998).
[91]G. Maier, “The cyclobutadiene problem”, Angewandte Chemie International Edition in English 13, 425–438 (1974).
[92]W. T. Borden, “Can a square or effectively square singlet be the ground state of cyclobutadiene”, Journal of the American Chemical Society 97, 5968–5970 (1975).
[93]E. Monino, M. Boggio-Pasqua, A. Scemama, D. Jacquemin, and P.-F. Loos, “Reference energies for cyclobutadiene: automerization and excited states”, The Journal of Physical Chemistry A 126, 4664–4679 (2022).
[94]P. G. Wenthold, J. Hu, R. R. Squires, and W. C. Lineberger, “Photoelectron spectroscopy of the trimethylene- methane negative ion. the singlet-triplet splitting of trimethylenemethane”, Journal of the American Chemical Society 118, 475–476 (1996).
[95]L. V. Slipchenko and A. I. Krylov, “Spin-conserving and spin-flipping equation-of-motion coupled-cluster method with triple excitations”, The Journal of Chemical Physics 123, 084107 (2005).
[96]D. A. Hrovat, M. A. Murcko, P. M. Lahti, and W. Thatcher Borden, “The effects of heteroatom substitution on the singlet–triplet energy differences in diradicals—ab initio calculations of ΔEST in metabenzoquinomethane and in 1, 3-naphthoquinomethane”, Journal of the Chemical Society, Perkin Transactions 2, 1037–1044 (1998).
[97]T. Wang and A. I. Krylov, “The effect of substituents on electronic states’ ordering in meta-xylylene diradicals: qualitative insights from quantitative studies”, The Journal of Chemical Physics 123, 104304 (2005).
[98]V. LeJeune, A. Despres, and E. Migirdicyan, “Site selection spectroscopy of the non-kekule m-xylylene biradical and its methylated derivatives in shpolskii matrixes”, The Journal of Physical Chemistry 88, 2719–2722 (1984).
[99]P. Neuhaus, D. Grote, and W. Sander, “Matrix isolation, spectroscopic characterization, and photoisomerization of m-xylylene”, Journal of the American Chemical Society 130, 2993–3000 (2008).
[100]K. W. Haider, E. Migirdicyan, M. S. Platz, N. Soundararajan, and A. Despres, “Recent studies of meta-xylylene and a diaza derivative”, Molecular Crystals and Liquid Crystals Incorporating Nonlinear Optics 176, 85–98 (1989).
[101]E. Migirdicyan and J. Baudet, “Electron spectra of o- and m-xylylenes and their methylated derivatives. experimental and theoretical study”, Journal of the American Chemical Society 97, 7400–7404 (1975).
[102]B. B. Wright and M. S. Platz, “Electron spin resonance spectroscopy of the triplet state of m-xylylene”, Journal of the American Chemical Society 105, 628–630 (1983).
[103]P. G. Wenthold, J. B. Kim, and W. C. Lineberger, “Photoelectron spectroscopy of m-xylylene anion”, Journal of the American Chemical Society 119, 1354–1359 (1997).
[104]B. Bleaney and K. D. Bowers, “Anomalous paramagnetism of copper acetate”, Proceedings of the Royal Society of London. Series A. Mathematical and Physical Sciences 214, 451–465 (1952).
[105]N. Suaud, R. Ruamps, N. Guihéry, and J.-P. Malrieu, “A strategy to determine appropriate active orbitals and accurate magnetic couplings in organic magnetic systems”, Journal of Chemical Theory and Computation 8, 4127–4137 (2012).
[106]M. W. Schmidt, K. K. Baldridge, J. A. Boatz, S. T. Elbert, M. S. Gordon, J. H. Jensen, S. Koseki, N. Matsunaga, K. A. Nguyen, S. Su, T. L. Windus, M. Dupuis, and J. A. Montgomery, “General atomic and molecular electronic structure system”, Journal of Computational Chemistry 14, 1347–1363 (1993).
[107]G. M. J. Barca et al., “Recent developments in the general atomic and molecular electronic structure system”, The Journal of Chemical Physics 152, 154102 (2020).
[108]P. C. Reeves, J. Henery, and R. Pettit, “Further experiments pertaining to the ground state of cyclobutadiene”, Journal of the American Chemical Society 91, 5888–5890 (1969).
[109]F. Neese, “Software update: the ORCA program system—Version 5.0”, WIREs Computational Molecular Science 12, e1606 (2022).