Design of an open-shell nitrogen-centered diradicaloid with tunable stimuli-responsive electronic properties

Organic diradicaloids usually display an open-shell singlet ground state with significant singlet diradical character (y0) which endow them with intriguing physiochemical properties and wide applications. In this study, we present the design of an open-shell nitrogen-centered diradicaloid which can reversibly respond to multiple stimuli and display the tunable diradical character and chemo-physical properties. 1a was successfully synthesized through a simple and high-yielding two-step synthetic strategy. Both experimental and calculated results indicated that 1a displayed an open-shell singlet ground state with small singlet-triplet energy gap (ΔES−T = −2.311 kcal mol−1) and a modest diradical character (y0 = 0.60). Interestingly, 1a was demonstrated to undergo reversible Lewis acid-base reaction to form acid-base adducts, which was proven to effectively tune the ground-state electronic structures of 1a as well as its diradical character and spin density distributions. Based on this, we succeeded in devising a photoresponsive system based on 1a and a commercially available photoacid merocyanine (MEH). We believe that our studies including the molecular design methodology and the stimuli-responsive organic diradicaloid system will open up a new way to develop organic diradicaloids with tunable properties and even intelligent-responsive diradicaloid-based materials.


Supplementary methods
All reagents and starting materials were obtained from commercial suppliers and used without further purification. All air-sensitive reactions were carried out under inert N2 atmosphere. The 1 H NMR, 11 B NMR, 13 C NMR spectra were recorded in solution of CD2Cl2 and DMSO-d6 on Bruker 400 MHz, 500 MHz spectrometer. 2D NMR (NOESY) and variable temperature NMR were measured in solution of CD2Cl2 on Agilent DD2 600 MHz. Coupling constants (J) are denoted in Hz and chemical shifts (δ) are denoted in ppm. Multiplicities are denoted as follows: s = singlet, d = doublet, b = broaden and m = multiplet. UV-vis-NIR spectra were recorded in a quartz cell (light path 10 mm) on a Shimadzu UV2700 UV-visible spectrophotometer. Fluorescence spectra and photoluminescence quantum yields (Ф) were recorded on HORIBA Duetta. Cyclic voltammetry was recorded on a Bio-Logic SAS SP-150 spectrometer in anhydrous DCM containing n-Bu4NPF6 (0.1 M) as supporting electrolyte at a scan rate of 20 mV/s at room temperature. The CV cell has a glassy carbon electrode, a Pt wire counter electrode, and an Ag/Ag + reference electrode. The potential was externally calibrated against the ferrocene/ferrocenium (Fc/Fc + ) couple. The highest occupied molecular orbital (HOMO) and lowest unoccupied molecular orbital (LUMO) energy levels were calculated based on the equations: EHOMO/LUMO = -(4.80 + Eonset ox /Eonset red ) eV. EPR spectra for radicals were obtained on Bruker EMX instrument EMXPLUS-10/12. EPR spectra simulation was conducted on the Bruker SpinFit software. For SQUID measurement, magnetic susceptibility of powder sample (30 mg) was measured in a polycarbonate capsule fitted in a plastic straw as a function of temperature in heating (2 K→ 330 K) mode with 30 seconds of temperature stability at each temperature (1 K increment in a range 2-10 K, 2 K increment in a range 10-20 K,5 K increment in a range 20-100 K, 10 K increment in a range 100-330 K,) at 1.0 T using a SQUID magnetometer (Quantum MPMS3). The data was corrected for both sample diamagnetism (Pascal's constants) and the diamagnetism of the sample holder (polycarbonate capsule). [S1] The single crystals of this work were measured on Bruker Apex duo equipment with Cu Kα radiation (λ = 1.54184 Å). The HR-ESI mass spectra were performed on Q Exactive Focus (Thermo Scientific, USA). The singlet-triplet energy gap (∆ES-T or 2J/kB) was determined by fitting the susceptibility data using the Bleaney-Bowers equation, where -2J is correlated to the excitation energy from the ground state to the first excited state, ρ is the content of paramagnetic impurities, T is the temperature, kB is Boltzmann constant, N is Avogadro constant.

Synthesis and characterization Compound 4a
To a n-heptane solution (30 mL) of compound 3 (2.22 g, 10 mmol), compound 2a (0.92 g, 5.0 mmol) and triethylamine (0.05 g, 0.5 mmol) were added slowly. Then the mixture was heated to reflux for 12 h. After cooling to the room temperature, the mixture was kept at 0 °C for 1.0 h. The precipitate was collected by vacuum filtration and washed by cold n-heptane for three times. The residue was dried in vacuo to give compound 4a (2.67 g, 4.5 mmol) as a grey powder in 90% yield. 1

Compound 4b
To a n-heptane solution (30 mL) of compound 3 (2.22 g, 10 mmol), compound 2b (1.06 g, 5.0 mmol) and triethylamine (0.05 g, 0.5 mmol) were added slowly. Then the mixture was heated to reflux for 12 h. After cooling to the room temperature, the mixture was kept at 0 °C for 1.0 h. The precipitate was collected by vacuum filtration and was washed by cold n-heptane for three times. The residue was dried in vacuo to give compound 4b (2.79 g, 4.5 mmol) as a grey powder in 90% yield. 1

Compound 4c
To a n-heptane solution (30 mL) of compound 3 (2.22 g, 10 mmol), compound 2c (1.06 g, 5.0 mmol) and triethylamine (0.05 g, 0.5 mmol) were added slowly. Then the mixture was heated to reflux for 12 h. After cooling to the room temperature, the mixture was kept at 0 °C for 1.0 h. The precipitate was collected by vacuum filtration and was washed by cold n-heptane for three times. The residue was dried in vacuo to give compound 4c (2.79 g, 4.5 mmol) as a reddishbrown powder in 90% yield. 1

Compound 4d
To a n-heptane solution (30 mL) of compound 3 (2.22 g, 10 mmol), compound 2d (1.2 g, 5.0 mmol) and triethylamine (0.05 g, 0.5 mmol) were added slowly. Then the mixture was heated to reflux for 12 h. After cooling to the room temperature, the mixture was kept at 0 °C for 1.0 h. The precipitate was collected by vacuum filtration and was washed by cold n-heptane for three times. The residue was dried in vacuo to give compound 4d (2.92 g, 4.5 mmol) as a reddishbrown powder in 90% yield. 1

Compound 1a
Compound 4a (592 mg, 1.0 mmol) was dissolved in DCM (20 mL), then lead dioxide (480 mg, 2.0 mmol) was added into the solution. The mixture was heated to reflux for 12 h. After the mixture cooled to the room temperature, precipitate in solution was filtered out, and the solution was collected. The solvent was removed under reduced pressure, and the residue was washed by methanol for several times to give compound 1a (528 mg, 0.9 mmol) as a black purple solid in 90% yield. The target compound was further purified by recrystallization from MeOH/CH2Cl2 for magnetic properties test. 1

Compound 1b
Compound 4b (620 mg, 1.0 mmol) was dissolved in DCM (20 mL), then lead dioxide (480 mg, 2.0 mmol) was added into the solution. The mixture was heated to reflux for 12 h. After the mixture cooled to the room temperature, precipitate in solution was filtered out, and the solution was collected. The solvent was removed under reduced pressure, and the residue was washed by methanol for several times to give compound 1b (528 mg, 0.9 mmol) as a black purple solid in 90% yield. 1           In-situ 11 B NMR spectra showed that after added 1a to the TPFB in CD2Cl2, the boron signal of TPFB shifted from -1.17 ppm to -3.54 ppm, indicating the coordination between 1a and TPFB. BF3· Et2O as internal standard was used to guarantee the accuracy of the change in chemical shift.

Theoretical calculation
Theoretical calculations were performed with the Gaussian16 program suite 2 . All molecules geometry optimizations were performed by using ωB97XD exchange-correlation functional in conjunction with 6-31G(d,p) basis set 3,4 . Initial guess broken-symmetry (open-shell singlet) wavefunction was found with Guess=Mix, Nosymm and Stable=Opt keywords, then the Guess=Read keyword was used to optimize the broken-symmetry state geometry structure at the UωB97XD/6-31G(d,p) level. Frequency calculations were conducted to ensure that these structures were indeed local minima. Transition state structures were verified by frequency calculations and only one imaginary frequency was found in the transition state. Closed-shell wavefunction has an RHF→UHF instability and open-shell wavefunction is stable under the perturbations considered in stability test of wavefunction [5][6][7] . Single point energy was performed at the level of ωB97XD/def2-TZVP 8 . Time-dependent density functional theory (TD-DFT) calculations were performed at the level of ωB97XD/6-31G(d,p). Nucleus-independent chemical shift values (NICS(1)ZZ) were calculated using the standard GIAO 9,10 at the level of ωB97XD/6-31G(d,p). Spin population calculations based on becke method divide atomic space. Electronic structure analyses were performed with the Multiwfn 3.8 (dev) code. The isosurface maps of spin density and were rendered by means of Visual Molecular Dynamics (VMD 1.9.3) software 11 based on the files exported from Multiwfn.
The singlet-triplet energy gap − was calculated as: where the and were the energy of local minima structure in broken-symmetry state and triplet state, respectively.
Unrestricted natural orbital (UNO) occupation number calculations were done by unrestricted hartree-fock (UHF) and unrestricted density functional theory (UDFT), the diradical character ( 0 ) was calculated according to Yamaguchi's scheme as 12 : Where indicates the occupation number of the highest occupied natural orbital (HONO) and the lowest unoccupied natural orbital (LUNO), a molecule with 0 = 0 means a closed-shell structure, whereas a molecule with 0 = 1 means a pure diradical structure. Supplementary Fig. 25: Calculated geometry structure of 1d.