Muhammad Imran‡
ab,
Lin Yang‡ab,
Jin-Jiang Zhangb,
Zhen-Lin Qiua,
Yubin Fuab,
Noel Israelc,
Evgenia Dmitrievac,
Andrea Lucottid,
Gianluca Serrad,
Matteo Tommasinid,
Ji Ma*e and
Xinliang Feng*ab
aCenter for Advancing Electronics Dresden (cfaed) & Faculty of Chemistry and Food Chemistry, Technische Universität Dresden, 01062 Dresden, Germany. E-mail: xinliang.feng@tu-dresden.de
bMax Planck Institute of Microstructure Physics Weinberg 2, 06120 Halle, Germany
cLeibniz Institute for Solid State and Materials Research, 01069 Dresden, Germany
dDipartimento di Chimica, Materiali e Ingegneria Chimica “G. Natta”, Politecnico di Milano, Piazza Leonardo da Vinci 32, 20133 Milano, Italy
eCollege of Materials Science and Opto-Electronic Technology & Center of Materials Science and Optoelectronics Engineering, University of Chinese Academy of Science, 100049 Beijing, P. R. China. E-mail: maji@ucas.ac.cn
First published on 20th December 2024
Concealed non-Kekulé (CNK) nanographenes have recently gained attention as promising non-Kekulé model systems due to their distinctive antiferromagnetic electronic spins, which offer potential applications in spintronics and quantum information science. However, synthesizing CNK nanographenes in solution remains a significant challenge because of their strong biradical character and high reactivity. In this study, we report the successful synthesis of a novel CNK nanographene with two phenalene units fused in a cis configuration to perylene (c-CNK), which exhibits persistent stability under ambient conditions, with a half-life (t1/2) of 59 minutes. The formation of c-CNK is confirmed using in situ UV-Vis-NIR spectroscopy, Raman spectroscopy, and high-resolution mass spectrometry. The open-shell character of c-CNK is supported by electron paramagnetic resonance (EPR) spectroscopy by observing an isotropic signal with a g-value of 2.0026. Quantum chemical simulations reveal a high biradical character (y0 = 0.97) and a singlet open-shell ground state with a small singlet–triplet energy gap (ΔES–T) of 0.4 kcal mol−1. This work presents a solution synthesis of a next-generation concealed non-Kekulé nanographene with intrinsic antiferromagnetic electronic spins, highlighting its potential as a promising material for future quantum technologies.
Non-Kekulé NGs can be categorized into (i) obvious non-Kekulé NGs and (ii) concealed non-Kekulé NGs based on the difference in their sublattice and net spin in their lowest energy ground state. In contrast to the obvious non-Kekulé NGs, concealed non-Kekulé NGs (CNKs) have two unpaired electronic spins aligned anti-parallel (anti-ferromagnetic) to each other in the ground state (Fig. 1). The topological frustration of π-bonds in a CNK renders it impossible to assign a classical Kekulé structure without leaving unpaired electrons, driving the system into a magnetically non-trivial open-shell ground state.17–19 CNK-NGs mostly remained an interesting subject of mathematical chemistry owing to their structural peculiarity.20 For example, the smallest CNK can be possible in a special arrangement of eleven fused benzene rings, resulting in only 8 possible arrangements.
Early synthetic efforts towards CNK-NGs by Eric Clar date back to 1972 which resulted in the unsuccessful realization of targeted molecules.21 The representative bow-tie-shaped CNK-NG known as Clar's Goblet was successfully achieved by our group in 2020 in collaboration with Fasel et al., through a combined in-solution and on-surface synthesis approach.22 The discovery of antiferromagnetically coupled spins in a CNK (Clar's Goblet), with an effective exchange parameter (Jeff) of 23 meV, exceeding the dissipation limit, has sparked interest in their potential for fault-tolerant spin-logic operations near room temperature for quantum computing technologies. Surface-assisted synthesis has undoubtedly become a promising alternative for producing challenging molecular systems like non-Kekulé NGs.4,23–28 However, this approach encounters several fundamental limitations, including low yield, poor selectivity, limited scalability, and the necessity for a planar structure. As a result, there is an increasing need to develop robust solution synthesis methods and design structures that offer both scalability and enhanced stability.
Herein, we report the solution synthesis of a novel nonplanar CNK nanographene named c-CNK with very high biradical character (y0 = 0.97) by building two phenalene units in a cis configuration at the 3,10-positions of the perylene core through multi-step synthesis (Fig. 1). The key precursor 2H-c-CNK is achieved by the installation of naphthalene with a benzylic ester functionality at the 3,10-positions of dibromoperylene via double Suzuki–Miyaura coupling. The final c-CNK is synthesized through oxidative dehydrogenation of the key precursor 2H-c-CNK by treatment with tetrachloro-1,4-benzoquinone (TCBQ). The two mesityl substituents at high spin positions of the CNK provide optimum steric protection, ensuring a reasonable stability with a half-life (t1/2) of 59 min in solution under ambient conditions. The in situ formation of c-CNK is confirmed by UV-Vis-NIR spectroscopy, high-resolution (HR) matrix-assisted laser desorption/ionization time-of-flight (MALDI-TOF) mass spectrometry and Raman spectroscopy. c-CNK shows a narrow optical energy gap (Eopt.) of 1.07 eV as revealed by UV-Vis-NIR absorption data. The open-shell character of c-CNK is supported by electron paramagnetic resonance (EPR) spectroscopy. This study provides a synthetic approach for accomplishing novel CNK systems where two spins are anti-ferromagnetically coupled in a single molecule in the ground state.
Subsequently, compound 2 was reacted with potassium acetate at 100 °C to obtain 3. Finally, palladium catalyzed borylation was performed on 3 to get the precursor 4 with a boronic ester functionality for subsequent double Suzuki–Miyaura coupling with 3,10-dibromoperylene. Next, the key precursor 2H-c-CNK was achieved in four steps. After successful double Suzuki–Miyaura coupling, 3,10-dibromoperylene afforded compound 5 with a benzylic ester functionality which was subjected to hydroxylation with KOH without further purification to afford compound 6 in 72% yield over two steps. Compound 6 was oxidized by Dess–Martin oxidation to obtain compound 7 in 41% yield. Afterward, compound 7 was treated with an excess of 2-mesitylmagnesium bromide to produce the diol in situ, which was then subjected to Friedel–Crafts alkylation promoted by BF3·OEt2 to afford the dihydro-precursor 2H-c-CNK with a yield of 68% in two steps. Notably, the key precursor 2H-c-CNK was confirmed by single crystal X-ray diffraction (XRD) analysis, MALDI-TOF, and NMR spectroscopy (Fig. S4†). All of the intermediates were purified and fully characterized by HR MALDI-TOF mass spectrometry and NMR (1H and 13C) spectroscopy.
We further monitored the oxidative dehydrogenation reaction of the precursor 2H-c-CNK using different oxidation conditions. The most successful results were obtained when tetrachloro-1,4-benzoquinone (TCBQ) was used as an oxidant in a toluene solution of 2H-c-CNK at 100 °C as evident from UV-Vis-NIR spectroscopy and MALDI-TOF mass analysis of the reaction mixture. HR MALDI-TOF mass analysis of the reaction mixture in positive mode revealed that the intense signal at m/z = 764.3451 of the precursor 2H-c-CNK ([M+]), calculated for C64H44 as 764.3443, disappeared and a new signal emerged at m/z = 762.3288 (Scheme 1), which is fully consistent with the expected molecular mass of c-CNK ([M+]), calculated for C64H42 as 762.3281, validating the successful formation of the targeted c-CNK. Notably, the isotopic distribution observed for the obtained c-CNK agrees with the simulated pattern for C64H42. Due to the high biradical character and high reactivity of c-CNK, conventional structural characterization by NMR spectroscopy and single-crystal structural analysis or attempts to isolate and purify c-CNK did not succeed. Therefore, we turned to in situ UV-Vis-NIR spectroscopy which is a powerful tool to characterize open-shell NGs due to the characteristic low energy absorption bands in the NIR region associated with SOMO to LUMO transitions. As shown in Fig. 2a, the dihydro-precursor 2H-c-CNK only shows UV-Vis absorption below λ = 560 nm. In contrast, the UV-Vis-NIR absorption spectrum of c-CNK displays the characteristic low-energy absorption bands with absorption wavelengths (λ) extending up to 1150 nm resulting from electronic excitation from SOMO to LUMO with vibrionic features related to absorption bands at 940 nm, 866 nm, 809 nm, and 727 nm. A narrow optical energy gap (Eopt.) of 1.07 eV is revealed by UV-Vis-NIR absorption data of c-CNK. The spectral changes accompanied by colour changes from yellow to greenish-brown can obviously be attributed to the in situ formation of c-CNK.
The in situ generated c-CNK was persistent in solution under ambient conditions, as shown in Fig. 2b. Time-depended UV-Vis-NIR absorption spectra of c-CNK were obtained to test its stability. The spectroscopic results reveal that the intensity of the NIR absorption bands diminishes with the concomitant increase of the peaks in the visible region suggesting the possible oxidative degradation of c-CNK. During the stability test, the solution of compound c-CNK in dry dichloromethane was exposed to ambient air and light. By plotting the change in absorption intensity with the exposure time of ambient conditions, according to the linear regression and extrapolation to % change = 0.50, the half-life time (t1/2) of c-CNK was estimated to be 59 min.
Next, EPR spectroscopy was used to examine the paramagnetic nature of c-CNK first at 293 K. The open-shell character of c-CNK is supported by observing a strong isotropic signal with a g-value of 2.0026 in the EPR spectrum in deuterated dichloromethane (Fig. 3a). The isotropic EPR signal arises from efficient spin delocalization and weak spin–spin coupling, as further supported by DFT calculations. We further performed variable-temperature EPR measurements, which show an increase in the signal intensity by increasing the temperature from 225 to 310 K (Fig. 3b and S2a†). This behaviour is typical of singlet biradicals and is the result of the thermal population of triplet states due to a smaller singlet–triplet energy gap.29 In addition, at low temperatures, intermolecular interactions within the perylene framework lead to enhanced π–π stacking, resulting in singlet state formation due to spin–spin coupling between the interacting molecules resulting in pronounced decrease in the EPR signal.30 The radical is not stable at high temperatures and the EPR signal intensity decreases over time (Fig. S2b†); the latter is in agreement with time-dependent UV-Vis-NIR spectroscopic measurements. Overall, these results confirm the singlet biradical character of the c-CNK generated in situ in dichloromethane solution.
Fig. 3 (a) EPR spectrum of c-CNK measured at 293 K in d-dichloromethane solution. (b) Variable temperature EPR spectra of c-CNK in d-dichloromethane solution. |
To further support the successful synthesis of c-CNK, we measured the Raman spectra, which is also an important tool for in situ investigation of redox reactions in NGs.31 We compared the experimental results with DFT calculation outcomes, as shown in Fig. 4a. The best experimental conditions for Raman spectroscopy of these compounds were reached with the excitation wavelength of 405 nm. Based on the UV-Vis-NIR absorption spectra of 2H-c-CNK and c-CNK (Fig. 2a), the excitation at 405 nm wavelength yields resonance Raman conditions. In the DFT calculation of the resonance Raman spectra of c-CNK and 2H-c-CNK, we obtain the best agreement with the experiments considering excitation wavelengths in the range of 405–415 nm (Fig. S3†). Understandably, these values are not identical to the experimental excitation wavelength (405 nm) because of the differences between the local molecular conditions of the sample and those of the isolated molecular model, which lacks intermolecular interactions and crystal packing effects. Nevertheless, all the spectra show good agreement with their computational counterparts. The precursor 2H-c-CNK and the in situ generated c-CNK were analyzed with excitation at 405 nm. For the precursor 2H-c-CNK, the experimental Raman bands at 1357 and 1383 cm−1 are assigned to the collective CC stretching modes of the graphene moiety. We also assign the intense feature at 1582 cm−1, as well as the shoulder at 1607 cm−1, to G modes. In the spectrum of c-CNK, we assign the band at 1383 cm−1 to collective CC stretching in the phenalene moiety and the peak at 1582 cm−1, together with its shoulder at 1607 cm−1, to the G mode of the perylene moiety. It is remarkable that the positions of the strongest features in the Raman spectrum of 2H-c-CNK and c-CNK do not differ much, in agreement with DFT results. Indeed, the analysis of the nuclear displacements of the strongest modes in the G and D regions reveals similar displacement patterns in the two molecules, irrespective of the dehydrogenation. The presence/absence of the two hydrogen atoms in 2H-c-CNK and c-CNK, respectively, has a weak influence on the collective vibrational dynamics of these CC stretching modes.
Fig. 4 Top panel: experimental (black lines) and simulated (red lines) Raman spectra of (a) 2H-c-CNK and (b) c-CNK. Each spectrum has been normalized to unity. The simulated spectrum of 2H-c-CNK was computed with 410 nm excitation and that of c-CNK with 406 nm excitation resonance. Bottom panel: representation of the nuclear displacements assigned to the strongest G and D peaks observed in the experimental spectra of (c) 2H-c-CNK and (d) c-CNK; the green and blue segments represent stretching and contracting bonds, respectively, and the red lines represent nuclear displacement vectors (see the starred modes in Tables S1 and S2† for details). |
Quantum chemical simulations were performed to gain a deeper understanding of the electronic structure of c-CNK. DFT calculations were performed using the Gaussian 09 package. The geometry of the CNK was relaxed at the UB3LYP/6-311G level of theory. The DFT-optimized structures of compound c-CNK have a slightly curved backbone geometry with the mesityl substitutions perpendicular to the core structure (Fig. 5a). A comparison of the optimized singlet and triplet ground states of c-CNK shows that the singlet state, where the two electronic spins are antiparallel, is the most stable state. The calculated singlet–triplet energy gap (ΔEST) is small with a value of −0.42 kcal mol−1. The biradical character (y0) was calculated using UB3LYP/6-31G, according to the occupation numbers of the lowest singly occupied natural orbitals (SOMO) and lowest singly unoccupied natural orbitals (LUMO). The calculated high biradical yield (y0) of 0.97 agrees with the non-Kekulé structure with two unpaired spins.32 Furthermore, the spin density distribution of c-CNK reveals that spin is mostly concentrated over the phenalene units with the high spin density at the carbon centers attached to the mesityl-substituents (Fig. 5b). The calculated singly occupied molecular orbital (SOMO) profiles of the α and β electrons are separated from each other (Fig. 5c and d), which is typical of open-shell NGs.
Footnotes |
† Electronic supplementary information (ESI) available. CCDC 2390595. For ESI and crystallographic data in CIF or other electronic format see DOI: https://doi.org/10.1039/d4qo02019g |
‡ These authors contributed equally to this work. |
This journal is © the Partner Organisations 2025 |