Functionalisation of conjugated macrocycles with type I and II concealed antiaromaticity via cross-coupling reactions

Conjugated macrocycles can exhibit concealed antiaromaticity; that is, despite not being antiaromatic, under specific circumstances, they can display properties typically observed in antiaromatic molecules due to their formal macrocyclic 4n π-electron system. Paracyclophanetetraene (PCT) and its derivatives are prime examples of macrocycles exhibiting this behaviour. In redox reactions and upon photoexcitation, they have been shown to behave like antiaromatic molecules (requiring type I and II concealed antiaromaticity, respectively), with such phenomena showing potential for use in battery electrode materials and other electronic applications. However, further exploration of PCTs has been hindered by the lack of halogenated molecular building blocks that would permit their integration into larger conjugated molecules by cross-coupling reactions. Here, we present two dibrominated PCTs, obtained as a mixture of regioisomers from a three-step synthesis, and demonstrate their functionalisation via Suzuki cross-coupling reactions. Optical, electrochemical, and theoretical studies reveal that aryl substituents can subtly tune the properties and behaviour of PCT, showing that this is a viable strategy in further exploring this promising class of materials.


Synthesis
All reagents were obtained from commercial sources and used as received. With the exception of carbon tetrachloride and methanol, all reaction solvents were obtained from a Grubbs-type solvent purification system and were sparged with nitrogen prior to use. All reactions were performed under nitrogen atmosphere using standard Schlenk line techniques, and work-ups were performed under air. Purification by recycling preparative gel permeation chromatography (GPC) was carried out on a LaboACE LC-5060 (Japan Analytical Industry Co., Tokyo, JAPAN) recycling GPC system equipped with a JAIGEL-2HR column and a TOYDAD800-S detector. CH2Cl2 was used as the eluent for the preparative recycling GPC in all cases, with a flow rate of 10 mL min -1 .
NMR spectra were recorded at room temperature on a Bruker Avance 400 MHz spectrometer and referenced to the residual solvent peaks of CDCl3 or CD2Cl2 at 7.26 or 5.32 ( 1 H NMR) and 77.16 or 53.84 ppm ( 13 C{ 1 H} NMR) respectively. The NMR signals were fully assigned for the separated regioisomers 6-cis and 6-trans (where the assignment was possible) using 2D correlation spectroscopy. Coupling constants are measured in Hz.
The solution of 2-bromo-1,4-dimethylbenzene was added via cannula and the suspension was heated to reflux and stirred for 2 hours. The suspension was then allowed to cool to room temperature, filtered, and the solvent was removed in vacuo. To remove over-brominated species, the crude product was dissolved in THF (300 mL) and diethylphosphite (DEP; 32.8 g, 238 mmol, 2.2 equiv.) and N,N-diisopropylethylamine (DIPEA; 33.5 g, 259 mmol, 2.4 equiv.) were added. The solution was stirred vigorously at room temperature, and the reaction was monitored by taking an aliquot, removing the solvent in vacuo, and collecting a 1 H NMR spectrum, to check for consumption of the over-brominated species (which present indicative singlets at ~6.55 ppm). After 2 days, the reaction was filtered to remove the yellow precipitate. The solvent of the solution was then removed in vacuo, and the residue was purified by silica gel column chromatography, eluting with hexane. The resulting near-pure product was recrystallized from the minimum amount of boiling hexane to give product 1 as white crystals (10.0 g, 29.2 mmol, 27%).
Note: The step involving DEP and DIPEA produces a large amount of an insoluble yellow material. 300 mL of THF (or more) and a large stirrer bar are required to ensure vigorous stirring over the entire reaction time.
Characterisation supported by previously reported synthesis of this molecule. 2
Characterisation supported by previously reported synthesis of this molecule. 2

Dibrominated PCTs 3 (mixture of regioisomers 3-cis and 3-trans)
Prepared by modifying a procedure that was first reported for the synthesis of unsubstituted PCT. 3 Compound 2 (8.20 g, 9.45 mmol, 1.0 equiv.) and terephthaldehyde (1.27 g, 9.45 mmol, 1.0 equiv.) were suspended in DMF (350 mL) and cooled to -40°C (using an acetonitrile/dry ice cooling bath). In a separate flask, LiOMe (1.08 g, 28.4 mmol, 3.0 equiv.) was dissolved in dry MeOH (40 mL). The LiOMe solution was added to the reaction mixture dropwise over the course of 8 hours (using a syringe pump) while maintaining a temperature of -40°C. After the addition, the solution was allowed to warm to room temperature and stirred overnight. The resulting bright yellow suspension was poured into stirring water (300 mL) and subsequently extracted with diethyl ether (4 x 200 mL). The organic phases were combined and washed with water (3 x 200 mL) and brine (200 mL). The solution was then dried over MgSO4, filtered, and the solvent was removed in vacuo to yield a bright yellow solid. This was dissolved/suspended in CH2Cl2 (10 mL) and filtered over a pad of silica, eluting with CH2Cl2. The solvent was removed in vacuo, and the crude product was purified by recycling preparative GPC. The solvent was removed in vacuo to yield product 3 as a bright yellow powder containing a mixture of the two macrocyclic isomers 3-cis and 3-trans (240 mg, 0.42 mmol, 9%).

Thermogravimetric analysis (TGA)
TGA was carried out at a heating rate of 10°C min -1 under a nitrogen flow of 50 mL min -1 . The initial weight loss of 6-cis was attributed to CH2Cl2 inclusions (presumably a result of the porous nature of the macrocyclic material), which cannot easily be removed following purification by recycling GPC. Figure S22. TGA of compounds 3. Figure S23. TGA of compounds 6-cis. The cyclic voltammograms shown in Fig. 3a show the second cycle of the measurement. In line with best practice for solution measurements, 4 the redox potentials shown in Fig. 3a were estimated from the half-wave potential (E 1/2 ) when reversibility was observed and from the inflection-point potential (E i ) when no reversibility was observed. In addition to the estimated redox potentials, Fig. 3a also shows the cathodic and anodic peak potentials (in smaller font size).

Cyclic voltammetry of thin films
For thin film measurements, 5 mg mL -1 solutions were prepared for each macrocycle. These were then dropcast onto the working electrode and the solution was allowed to evaporate to form a film. These were then measured in the same three-electrode configuration as described above for the solution state measurements, using oxygen removed and N2 saturated 0.1 M [n-Bu4N]PF6 in CH3CN as the electrolyte.
The cyclic voltammograms in Fig. 3b show the first cycle of the measurement as the irreversible oxidation at high potential resulted in changes in the second cycle of the measurement. The redox potentials shown in Fig. 3b were estimated from the onset of the reduction.
As the coverage of the working electrode with macrocycle thin films appeared to alter the ferrocene measurements when adding ferrocene as an internal reference, a separate ferrocene measurement with cleaned electrodes was recorded as a reference for all thin-film measurements.

UV-vis absorption and photoluminescence measurements
Solution measurements were obtained using 5 µM (6-cis and 6-trans) or 10 µM (PCT, 3, 4 and 5) solutions in chloroform, and thin film measurements were obtained from cleaned glass slides coated with 40 µL of 5 mg mL -1 solutions using a spin coater at 1000 rpm for 1 min.
UV-vis absorption spectra were recorded on an Agilent Cary 60 UV-Vis spectrophotometer at room temperature under regular lab conditions at a data interval of 0.5 nm.
Photoluminescence (PL) spectra of the macrocycles in solution were acquired on an Agilent Cary Eclipse fluorescence spectrophotometer at a data interval of 1.0 nm, using the same solutions as for the UV-vis absorption measurements. The excitation and emission slits were set to 5 nm, and the detector voltage was set to 'high' (800 V).  Figure S24. Thin-film UV-vis absorption spectra of the macrocycles.

Computational analysis
Geometry optimisations on the singlet ground state (S0), the 2+ and 2-charged states, as well as the triplet (T1), were carried out using the PBE0 functional with def2-SV(P) basis set and D3 dispersion correction. [5][6][7][8] For the singlet states, spin-restricted Kohn-Sham (RKS) computations were performed, and for the triplet unrestricted KS (UKS) was used. Vibrational frequency analyses at the same level were performed to verify the nature of the stationary points as minima and to obtain thermostatistical corrections for the redox potentials. Additional single-point computations were performed at the D3-PBE0/def2-SVPD level, including solvation effects using the SMD model of Truhlar and co-workers 9 to represent DMF. Redox potentials were computed from these D3-PBE0/def2-SVPD energies along with D3-PBE0/SV(P) thermostatistical corrections as described previously. 10 The excited singlet state (S1) was optimised using time-dependent density functional theory (TDDFT) at the LC-BLYP/def2-SV(P) level using a range separation parameter of  = 0.1 a.u. 11 The calculations were performed in Orca 5.0. 12 Nucleus-independent chemical shifts (NICS) 13 were computed at the PBE0/def2-SVP level using gauge including atomic orbitals 14 as implemented in Gaussian 09. 15 NICS tensors were represented graphically using the visualisation of chemical shielding tensors (VIST) method 16 as implemented in TheoDORE 3.0 17 and using VMD as a graphical interface. 18 Changes in the electron density for the charged states and T1 were modelled by computing natural difference orbitals. 19 This analysis was performed using the "analyze_nos" functionality of TheoDORE 3.0 using the molecular orbitals of the charged and neutral states (both computed at the geometry of the charged state) as input. The analysis of S1 states was done using natural transition orbitals. 20 The underlying computational research data is available via a separate repository (DOI: 10.17028/rd.lboro.22306207): Orca input/output files for geometry optimisations, frequency analyses, and solvated single-point computations; Gaussian input/output files for NICS computations.   Figure S27. Analysis of the S0, S1, and T1 states for 4-trans. Centre: VIST plots (using T1 also at the S1 geometry). Left and right: Dominant natural transition/difference orbitals (NTOs/NDOs) for the S1 and T1 states (blue/red for electron detachment; green/orange for attachment). Figure S28. Analysis of the S0, S1, and T1 states for 5-cis. Centre: VIST plots (using T1 also at the S1 geometry). Left and right: Dominant natural transition/difference orbitals (NTOs/NDOs) for the S1 and T1 states (blue/red for electron detachment; green/orange for attachment). Figure S29. Analysis of the S0, S1, and T1 states for 5-trans. Centre: VIST plots (using T1 also at the S1 geometry). Left and right: Dominant natural transition/difference orbitals (NTOs/NDOs) for the S1 and T1 states (blue/red for electron detachment; green/orange for attachment). Table S2. NICS values (in ppm) corresponding to the tensor component perpendicular to the plane of the PCT core at the centre of the molecule (Centre) as well as to the main component of the tensors above the substituted (Ph1) and unsubstituted phenylene subunits (Ph2) of the PCT core, and above the thienyl/thienylene unit directly linked to the PCT core (Thio) and further away (Thio2). Figure S30. ACID plots for 4-cis in the neutral and doubly charged states.  Table S3. Computed redox potentials for the reduction (0 to -2) and oxidation (0 to +2) of thienyland bithienyl-substituted PCTs vs. Fc/Fc + using a voltage of 4.7 V for the reference electrode.