Guest‐Modulated Circularly Polarized Luminescence by Ligand‐to‐Ligand Chirality Transfer in Heteroleptic PdII Coordination Cages

Abstract Multicomponent metallo‐supramolecular assembly allows the rational combination of different building blocks. Discrete multifunctional hosts with an accessible cavity can be prepared in a non‐statistical fashion. We employ our shape‐complementary assembly (SCA) method to achieve for the first time integrative self‐sorting of heteroleptic PdII cages showing guest‐tunable circularly polarized luminescence (CPL). An enantiopure helicene‐based ligand (M or P configuration) is coupled with a non‐chiral emissive fluorenone‐based ligand (A or B) to form a series of Pd2L2L′2 assemblies. The modular strategy allows to impart the chiral information of the helicenes to the overall supramolecular system, resulting in CPL from the non‐chiral component. Guest binding results in a 4‐fold increase of CPL intensity. The principle offers potential to generate libraries of multifunctional materials with applications in molecular recognition, enantioselective photo‐redox catalysis and information processing.


Materials and measurements
Unless otherwise stated, all chemicals were obtained from commercial sources and used as received.
Ligands A, [1] enantiopure helicene M/P [2] and homoleptic cage [1] [Pd2A4] 4+ were prepared according to literature procedures. Gel permeation chromatography (GPC) purification of ligands was performed on a JAI 9210-II NEXT GPC System with a JAIGEL HH-2/HH-1 column combination running with CHCl3 (HPLC grade). High resolution Electrospray Ionization (HR-ESI) mass spectra and trapped ion mobility data were recorded on Bruker ESI-timsTOF and (electrospray ionization-trapped ion mobility-time of flight) Compact mass spectrometers. All samples were diluted with spectroscopic grade CH3CN (1:10) prior to measurement. NMR experiments were measured on Bruker AVANCE III and NEO (500, 600 or 700 MHz) spectrometers. Chemical shifts for 1 H and 13 C are reported in ppm with residual solvent as reference: Acetonitrile (1.94 ppm for 1 H, 1.32 ppm for 13 C), DMSO (2.50 ppm for 1 H, 39.52 ppm for 13 C), DMF (2.75 ppm for 1 H, 29.76 ppm for 13 C). Abbreviations for signal multiplicity of 1 H NMR spectra are shown as following: s: singlet, d: doublet, t: triplet, dd: doublet of doublets; dt: doublet of triplets; m: multiplet, br: broad. UV-vis spectra were recorded on a DAD HP-8453 UV-Vis spectrometer. Circular dichroism spectra were recorded in CD3CN with an Applied Photophysics Chirascan qCD Spectrometer with a temperature-controlled cuvette holder. The spectra were backgroundcorrected and smoothed with a window size of 5. Emission measurements were performed on a JASCO FP-8300, quantum yield determination have been performed on a JASCO ILF-835 integrating sphere as accessory of the JASCO FP-8300, circularly polarized luminescence measurements were performed using a JASCO CPL-300 spectrophotometer, both equipped with a (150 W) Xe lamp as light source. CPL spectra of ligand M and P were recorded with excitation bandwidth of 25 nm and emission bandwidth of 15 nm and averaged over 10 spectra. CPL of heteroleptic cages and host guest complexes were recorded with excitation and emission bandwidth of 30 nm and averaged over 10 spectra. 3,6-dibromo-9H-fluoren-9-one (250 mg, 0.74 mmol, 1 equiv.), 3-(4-(4,4,5,5-tetramethyl-1,3,2-dioxaborolan-2yl)phenyl)pyridine (500 mg, 1.78 mmol, 2.4 equiv.), Na2CO3 (235 mg, 2.22 mmol, 3 equiv.) and Pd(PPh3)4 (25.6 mg, 0.022 mmol, 0.03 equiv.) were suspended in a mixture of DMF/H2O (3:1, 12 ml). The mixture was degassed three times using the Freeze-Pump-Thaw-method, heated to reflux and stirred overnight. After cooling down to room temperature, the reaction mixture was extracted with chloroform. The organic phase was washed consecutively with water and then brine. It was dried over MgSO4, filtered and the solvent was removed under reduced pressure. The crude product was purified by column chromatography (SiO2; CHCl3/methanol; 0 to 10 %) yielding a yellow powder that was further purified by GPC. The yield of the final product was 87 mg (0.179 mmol, 24%).

Ion Mobility Mass Spectrometry
Ion mobility measurements were performed on a Bruker timsTOF instrument combining a trapped ion mobility (TIMS) with a time-of-flight (TOF) mass spectrometer in one instrument.
In contrast to the conventional drift tube method to determine mobility data, where ions are carried by an electric field through a stationary drift gas, the TIMS method is based on an electric field ramp to hold ions in place against a carrier gas pushing them in the direction of the analyzer. Consequently, larger sized ions that experience more carrier gas impacts leave the TIMS units first and smaller ions elute later. This method offers a much higher mobility resolution despite a smaller device size. Normalized I x 10 -3 Wavelength (nm)

S17
Briefly, in trapped ion mobility spectrometry (TIMS) the ions are held by an electric field gradient (EFG) while exposed to a constant N2 flow. The source of the flow is a pressure difference (∆p) between the entrance and the exit of the TIMS-tunnel. Lowering the EFG results in the elution of the analyte according to its inverse mobility (1/K0). Depending on the difference between start-EFG and end-EFG (ΔEFG) as well as the duration of this reduction (ramp time = tr), rather high ion mobility resolutions can be achieved. Resolutions of 200 and higher was targeted in all the experiments using custom mode of measurement. In front of the trapping part of the TIMS-tunnel, a second ramp accumulates the ions for a given period of time (accumulation time = ta) in order to enhance the overall duty cycle.
Since TIMS doesn't operate on first principles, a calibration is necessary. For calibration, Agilent Tune-Mix TM was used as purchased and the reported drift tube DT CCSN2 values published by Stow et al. have been used. [3] TIMS determines (calibrated) inverse mobilities which were converted to the tabulated TIMS CCSN2 values following Mason-Schamp equation: where, q is the charge of the ion, N the number density of the collision gas, μ is the reduced mass of the ion and the N2 collision gas, k the Boltzmann's constant, T the temperature in Kelvin. Both the analyte solution and the calibrants have been electrosprayed in cationic mode with the same experimental conditions. Therefore, the first part of the equation is constant for every ions measured in the same condition. The constant was found by calculating the reduced mobility of the calibrant ions and plotting them against their reported CCS. The plot was fitted using linear fit with fixed intercept at 0. The slope is the constant for the specific experiment. The 1/K0 found from the experiment was then multiplied as per the equation.      DFT geometry optimizations were performed with and without dispersion correction contributions, using ωb97xd or B3LYP functionals and computed using the Gaussian 16 software. [7] This has been done since the flexibility of the systems, especially for the cage and the host-guest system involving ligand A, can potentially lead to several conformations of similar energies. In both computational approaches, the trend followed is the same, with an increase of the helicenes' helical pitch and of the Pd⋯Pd distances when moving from ligand A to ligand B. Guest encapsulation results in a decrease of the Pd⋯Pd distance, particularly pronounced for G@Pd2B2M2 and almost neglectable for G@Pd2A2M2.

S20
Calculation of ligand absorption and emission features: A full theoretical treatment of the observed photophysical features for the herein reported, hundreds of atoms counting coordination cages and host-guest complexes will be part of an extensive computational project including MD-based sampling of host and host-guest co-conformations (requiring development of an adapted force-field, studies of solvent and counter ion effects and analysis of trajectories), Boltzmann-weighted extraction of lowest-energy structures, QM electronic structure calculations with a variety of methods (i.e. different DFT functionals and basis sets to ensure proper treating of Pd cations and intracage and host-guest non-covalent interactions) and excited state calculations (including vibronic effects) to elucidate the nature of the electronic transitions giving rise to the observed absorption and emission features. Here, we give some first insights based on a simplified analysis comprising the organic ligands, only, that already indicate key phenomena of the experimental observations.
1. As shown in Figure S21, the absorption features of the hetero-and homoleptic Pd(II)-coordination cages largely resemble the UV-Vis spectra of the constituting ligands (with the helicene absorption undergoing a slight redshift and a loss of fine structure in particular in case of the heteroleptic cages). Hence, we chose to base our preliminary analysis of photophysical features on the individual, non-coordinated ligands to large simplify the computational efforts here.
2. Full photophysical characterization of a close isomer of helicene ligand M/P (para-instead of meta-pyridyl) has been reported by Autschbach, Crassous and Favereau, including both experimental absorption, emission, CD and CPL spectra as well as computed electronic transitions. [8] Essentially all reported photophysical features of the there reported compound perfectly superimpose with ligand M/P in our system. In particular, the reported emission maxima of the free ligands perfectly match and so do the glum values (there: 2.7 x 10 -2 , here 2.8 x 10 -2 ).
3. As for the fluorenone ligand A, we performed TD-DFT calculations to compute the absorption and emission features as follows: gas-phase geometry optimization on CAM-B3LYP/6-31G+g(d,p) level of theory using Gaussian '16 [7] , followed by frequency calculation to confirm optimization into a minimum on the PES (a conformer was chosen as starting structure with both pyridine N lone pairs pointing in a colinear fashion as they do in the assembled cage as shown below; computations on a conformer with pyridines turned outside confirmed almost identical energy and electronic transitions). Then, a TD-DFT computation was performed to calculate vertical excitations with linear response (nstates=6). The obtained absorption maxima match well with the experimental values, i.e. for the two most intense bands (exp.: 285 nm; calc.: 274 nm and exp.: 300 nm; calc.: 298 nm), while the broader band at longer wavelengths (exp.: 325-375 nm; calc.: 366 nm) is more shifted and lacks fine structure in the computed spectrum (see Fig. S21a). Next, we optimized the first excited state (nstates=6, root=1) on the same level of theory and analyzed the output for the lowest energy (assuming the compound shows regular Kasha behaviour) S1-S0 singlet emission (in accordance with the experimentally observed lack of sensitivity of the emission to the presence of oxygen), yielding an emission wavelength of 483 nm (vertical transition, no vibronic effects considered), matching well to the experimentally observed emission maximum of ligand A at 500 nm (Fig. 5a). Frontier orbital analysis of this transition reveals that the corresponding transition is -as expected for an aromatic ketone -between an excited state MO with π-character spanning the aromatic rings of the backbone and featuring a nodal plane concerning the C=O bond (higher energy; Fig. S30 right) and a π-MO also on the delocalized aromatic backbone system (lower energy; Fig. S30 left). 4. Next, the experimentally observed, strong emission red-shift upon addition of the guest was tried to be reproduced in the simplified computational model. Therefore, the excited state geometry of the ligand as computed above was combined with a model of the guest dianion (ground-state pre-optimized on same level of theory) in a geometry, that puts the guest's carbon and sulfur atoms in plane with the flat ligand with two sulfonate oxygen substituents pointing onto the ligand's pyridine H7 atoms (assignment see above) in a distance of 2.06 Å in accordance with the geometry optimized host-guest complex G@Pd2A2M2 as shown in Fig. 3d. With this model, again TD-DFT calculations of vertical transitions were performed and the lowest wavelength transition, in accordance with Kasha's rule, was found to comprise orbitals of the same geometry as for the free ligand (compare Figs. S30 and S31) and feature a red-shifted emission at 507 nm, hence qualitatively supporting the experimental result ( Fig. S32; exp. red-shift: 50 nm, calc.: 24 nm; computed transitions involving pure guest-localized MO contributions were excluded from the analysis as those would represent guest-ligand charge transfer states not considered relevant here).  5. The question may be raised whether direct electronic interaction (e.g. strong exciton coupling or ligand-to-ligand FRET processes) between the chiral helicenes and the emitting fluorenones play a role in the emission and CPL of the heteroleptic cages and host-guest complexes. While only deeper photophysical characterization studies involving transient absorption and chiroptical analyses together with full-fletched MD/QM treatment of the entire coordination cage assemblies may serve to fully answer this, a preliminary experiment involving a chiral ligand X with an aliphatic backbone (not shown) that shows no absorption > 300 nm and cleanly forms a Pd2A2X2 heteroleptic cage was found to exhibit analogous chirality transfer and CPL emission from the fluorenone ligand (excitation at 335 nm), indicating that strong electronic interaction between the helicenes and the fluorenones does not seem to be a requirement for the herein described phenomena.
In conclusion, our preliminary computational study on ligand A and the guest-ligand combination is able to reproduce the experimental absorption and emission features ascribed to ligand A as well as the red-shifting influence of the bound guest on the emission from the fluorenone (noteworthy is the very strong NMR downfield shift of protons H7 likewise pointing to a close and strong interaction between the sulfonate guest and the pyridine proton of ligand A in the hostguest assembly). In our current hypothesis, the helicene ligand does not play a significant electronic role in the observed chirality transfer, emission, CPL and guest modulation of the latter, but rather imprints its chirality onto the emissive moiety of the heteroleptic assembly via a structural influence on the overall geometry.