A Chirally Locked Bis-perylene Diimide Macrocycle: Consequences for Chiral Self-Assembly and Circularly Polarized Luminescence

Macrocycles containing chiral organic dyes are highly valuable for the development of supramolecular circularly polarized luminescent (CPL) materials, where a preorganized chiral framework is conducive to directing π–π self-assembly and delivering a strong and persistent CPL signal. Here, perylene diimides (PDIs) are an excellent choice for the organic dye component because, alongside their tunable photophysical and self-assembly properties, functionalization of the PDI’s core yields a twisted, chiral π-system, capable of CPL. However, configurationally stable PDI-based macrocycles are rare, and those that are also capable of π–π self-assembly beyond dimers are unprecedented, both of which are advantageous for robust self-assembled chiroptical materials. In this work, we report the first bay-connected bis-PDI macrocycle that is configurationally stable (ΔG⧧ > 155 kJ mol–1). We use this chirally locked macrocycle to uncover new knowledge of chiral PDI self-assembly and to perform new quantitative CPL imaging of the resulting single-crystal materials. As such, we discover that the chirality of a 1,7-disubstituted PDI provides a rational route to designing H-, J- and concomitant H- and J-type self-assembled materials, important arrangements for optimizing (chir)optical and charge/energy transport properties. Indeed, we reveal that CPL is amplified in the single crystals of our chiral macrocycle by quantifying the degree of emitted light circular polarization from such materials for the first time using CPL-Laser Scanning Confocal Microscopy.


Bis-PDI macrocycle 1
Compound 4, K 2 CO 3 1 eq, high dilution Cu(NCCH 3 ) 4 PF 6 TBTA DCM rt, 72 h 23% TMS-protected bis-alkyne PDI 4 (50 mg, 53 µmol) was dissolved in DCM (20 mL).To this was added K 2 CO 3 (20 mg) in MeOH (10 mL).The reaction was monitored by TLC (1:99 MeOH:DCM).Upon completion the reaction mixture was thoroughly washed with water in a separating funnel (3 x 100 mL) and dried with MgSO 4 to yield crude deprotected bis-alkyne PDI in DCM, which was used immediately without further purification due to its tendency to aggregate and crash out of solution over time.This was added to a flask, along with acyclic bis-triazole PDI 5 (62 mg, 53 µmol, 1 equiv), tris((1-benzyl-4-triazolyl)methyl)amine (TBTA) (12 mg, 21 µmol, 0.4 eq) and a further 350 mL of DCM.The reaction mixture was thoroughly de-gassed with N 2 .The copper catalyst Cu(CH 3 CN) 4 PF 6 (8 mg, 21 µmol, 0.4 equiv) was then added and the reaction mixture was thoroughly de-gassed again.The reaction was stirred at rt for 36 h and monitored by TLC (2:98 MeOH-DCM).The solvent was then removed in vacuo.C NMR spectrum of macrocycle 1 (TCE-d 2 , 373 K, 400 MHz).The poor signal to noise ratio is due to the large number of quaternary carbons and the somewhat low solubility of macrocycle 1.  Figure S3: Chiral HPLC chromatogram (Phenomenex i-Amylose-1, 250 x 4.6 mm) of a sample of enantiopure 1-MM, previously purified by chiral HPLC, that was heated at 180 °C for 24 h in 1,2-dicholorbenzene.After heating, the solvent was removed and the sample was re-dissolved in DCM and reinjected into the chiral HPLC column and eluted with 9:1 (v/v) DCM:methanol.No racemisation to form the other enantiomer 1-PP can be detected in this chromatogram, proving that macrocycle 1 is chirally locked at temperatures at least as high as 180 °C.

X-ray crystallography a) Macrocycle 1-rac
Purple, needle-like crystals of macrocycle 1-rac, suitable for single crystal X-ray diffraction, were grown by slow diffusion of methanol into a chloroform solution.
Single crystal X-ray diffraction experiments were performed by the UK National Crystallography Service on a Rigaku 007HF diffractometer with HF Varimax confocal mirrors, an UG2 goniometer and HyPix 6000HE detector.
The crystals were kept at 100(2) K during data collection.The structures were solved by direct methods using ShelXT 3 and refined with ShelXL 4 using a least squares method.Olex2 software was used as the solution, refinement and analysis program. 5e crystal diffracted weakly with a low-resolution diffraction limit; the data was truncated to a resolution of 1.00 A. Many attempts were made to grow stronger diffracting crystals, and this sample was sent to the synchrotron on several occasions.This dataset from the UK National Crystallography Service (NCS) gave the best refinement.
All non-hydrogen atoms were refined anisotropically.All hydrogen atoms were geometrically placed and refined using a riding model.Owing to the weak data, all methyl groups were placed in eclipsed conformations (AFIX 33) rather than having their torsion angles refined against a search of the Fourier map.This method aided convergence of the refinement.Restraints and constraints were applied to the structure to aid refinement.The anisotropic displacement parameters of tertiary carbon atoms C32, C43, C78, and C89 in the t-butyl ester groups were constrained to be identical (EADP).The anisotropic displacement parameters of carbon atoms C90, C91, and C92, which from a t-butyl group with tertiary carbon C86, were constrained to be identical (EADP) and the C-C bond lengths of the bonds carbon atoms C90, C91, C92 form with carbon atom C89 were restrained to have similar distances (SADI).The anisotropic displacement parameters of CH3 carbon atoms C35 C34 C33 C45 C44 C46 C79 C80 C81 of the remaining three t-butyl groups were restrained to be similar (SIMU).The anisotropic displacement parameters of oxygen atoms O1, O2, O3, O4, O9, O10, O11, and O12 were restrained to be similar (SIMU).
Due to poor data quality residual solvent could not be sensibly modelled.A solvent mask was calculated and 469 electrons were found in a volume of 1757 cubic angstrom in 1 void per unit cell.This is consistent with the presence of 13 methanol molecules per Asymmetric Unit which account for 468 electrons per unit cell.

b) Enantiopure macrocycle 1-PP
Needle-like crystals of enantiopure macrocycle 1-PP suitable for single crystal X-ray diffraction were grown from an enantiopure sample of 1-PP by slow diffusion of hexane into a 1:1 chloroform:1,2dicholobenzene solution.
Single crystal X-ray diffraction experiments were performed at the UK Diamond Light Source I19-1 3circle diffractometer ( = 0.6889 Å). [6][7][8] A suitable single crystal was selected and mounted using fomblin film on a micromount.Data were collected on a dectris-CrysAlisProabstract goniometer imported dectris images diffractometer.The crystals were kept at 100(2) K during data collection.The structures were solved by direct methods using ShelXT, 3 and refined with ShelXL 4 using a least squares method.Olex2 software was used as the solution, refinement and analysis program. 5Figures were produced using CrystalMakerX.
All non-hydrogen atoms were refined anisotropically.Positional disorder is modelled for carbon atoms C44A and C44B, C45A and C45B, C46A and C46B.The occupancies of the two sites were refined and constrained to sum to unity.Their occupancies refined to 0.5 so their occupancies were then set to 0.5.The anisotropic displacement parameters of carbon atoms C44A, C44B, C45A, C45B, C46A, and C46B were constrained to be similar (SIMU).The C-C bond lengths of the C43 with C44, C45, and C46 were restrained to be the same (EADP).
All hydrogen atoms were geometrically placed and refined using a riding model.Methyl hydrogens on carbon atoms C44A, C44B, C45A, C45B, C46A, and C46B were placed in eclipsed conformations (AFIX 33) rather than having their torsion angles refined against a search of the Fourier map.This method aided convergence of the refinement.
The majority of disordered solvent molecules could not be sensibly modelled.Carbon atom C117 is the pivot point (occupancy 1) for several possible orientations of one disordered chloroform model.Chlorine atoms Cl1, Cl2, and Cl3 are modelled with occupancy 0.5.Although further Cl orientations for this disordered molecule could be observed in the electron density map, they could not be sensibly modelled.A solvent mask was calculated using the Olex implementation of SQUEEZE and 394 electrons were found in a volume of 1218 cubic angstroms in 1 void per unit cell.This is consistent with the presence of 3.4[CCl3H] per Asymmetric Unit which account for 394 electrons per unit cell.

Chiroptical studies a) Circular dichroism
Circular dichroism (CD) spectra were recorded on a Jasco J-1500 CD spectrophotometer with a wavelength accuracy ± 0.2 nm (250 to 500 nm), ± 0.5 nm (500 to 800 nm) and a CD root mean square noise < 0.007 mdeg (500 nm).A quartz cuvette with 1 mm path length was used.The spectra were recorded at a concentration of 10 µM.The enantiomers were assigned by corroboration of the samples with the crystal structure of enantiopure 1-PP, as well as by comparison of their CD spectra to the TD-DFT computationally predicted CD spectra (Table S4).

b) Circularly polarised luminescence
CPL was measured with a home-built (modular) spectrometer. 9The excitation source was a broad band (200 -1000 nm) laser-driven light source EQ 99 (Elliot Scientific).The excitation wavelength was selected by the incorporation of an Acton SP-2155 monochromator (Princeton Instruments); the collimated light was focused into the sample holder (1 cm quartz cuvette).Emission was collected perpendicular to the excitation direction with a lens (f = 150 mm).The emission was fed through a photoelastic modulator (PEM) (Hinds Series II/FS42AA) and through a linear sheet polariser (Comar Optics).The light was then focused into a second scanning monochromator (Acton SP-2155) and subsequently on to a photomultiplier tube (PMT) (Hamamatsu H10723 series).The detection of the CPL signal was achieved using the field modulation lock-in technique.The electronic signal from the PMT was fed into a lock-in amplifier (Hinds Instruments Signaloc Model 2100).The reference signal for the lock-in detection was provided by the PEM control unit.The monchromators, PEM control unit and lock-in amplifier were interfaced to a desktop PC and controlled by a custom-written Labview graphic user interface.The lock-in amplifier provided two signals, an AC signal corresponding to (I L -I R ) and a DC signal corresponding to (I L + I R ). Background subtraction was achieved post data collection The emission dissymmetry factor was, therefore, readily obtained from the experimental data as 2 AC/DC.Spectral calibration of the scanning monochromator was performed using a Hg-Ar calibration lamp (Ocean Optics HL-3P-CAL).A correction factor for the wavelength dependence of the detection system was constructed using a calibrated lamp (Ocean Optics HL-3_CAL).The measured raw data was subsequently corrected using this correction factor.The validation of the CPL detection systems was achieved using light emitting diodes (LEDs) at various emission wavelengths.The LED was mounted in the sample holder and the light from the LED was fed through a broad band polarising filter and /4 plate (Comar Optics) to generate circularly polarised light.Prior to all measurements, the /4 plate and a LED were used to set the phase of the lock-in amplifier correctly.The emission spectra were recorded with 0.5 nm step size and the slits of the detection monochromator were set to a slit width corresponding to a spectral resolution of 0.25 nm.CPL spectra (as well as total emission spectra) were obtained through an averaging procedure of several scans.

e) Optical rotation measurements
The specific optical rotations of 1-MM/PP were measured on a Bellingham+Stanley ADP450 digital polarimeter fitted with a 10 cm path length sample holder and a light source with 589 nm wavelength.Measurements were carried out at 20°C and at a concentration of 3 x 10 -5 g/mL.We measured specific optical rotations of +1333° and -1333° for 1-MM and 1-PP [] 20 D respectively.

Photophysics a) UV-vis-NIR absorption and emission spectra in solution
All steady state electronic absorption and emission spectra were recorded at a concentration of 10 µM (unless otherwise stated) at 298 K.For UV-vis-NIR spectroscopy a Shimadzu UV-3600i Plus spectrophotometer was used, with a wavelength accuracy ± 0.2 nm in the UV-vis range and absorbance accuracy ± 0.002 Abs.For fluorescence spectroscopy a Jasco FP-8500 was used with emission and excitation wavelength accuracies ± 1.0 nm.The detector base sensitivity is 8500:1.Quartz cuvettes with 1 cm path length were used.

b) Calculation of the exciton coupling energy of macrocycle 1
The exciton coupling energy of macrocycle 1-rac was calculated following the method originally reported by Spano, 21 and used by Würthner to investigate long-and short-range coupling in PDI dimers. 22r small rotational displacements of chromophores (< 45°, as seen in macrocycle 1 in the xray crystal structure in Figure S7) the rotational displacement of transition dipole moments of two chromophores can be neglected and the ratio of intensities of the 00 and 01 vibronic absorption bands of a dimer aggregate exhibiting exciton-vibrational coupling can be calculated using the equation: Where is the Huang-Rhys factor, is the exciton coupling energy, and is the vibrational  2   0 frequency.

The vibrational function
is given by: ( t , 2 )
From the UV-vis absorption spectrum of 1-rac in toluene we measure: We obtained the Huang-Rhys factor by gaussian fitting of the vibronic peaks in the UV-vis spectrum of 1-rac (Figure S16), considering only the area of the spectrum arising from the S 0 -S 1 transition (< 23360 cm -1 ).We calculated the Huang-Rhys factor as: Gaussian  (0 -0) Gaussian = 1.598 [4]   The frequency was set as 1709 cm -1 , corresponding to the energy difference between the  0 maxima of the Gaussian functions of the 00 and 01 absorption bands.The exciton coupling energy was then calculated using equation 1.We obtained = 420.4cm -1 . 

c) Quantum yields
Absolute fluorescence quantum yields were obtained on an Edinburgh Instruments FLS1000 photoluminescence spectrometer fitted with an integrating sphere.All samples were recorded in toluene at a 1 µM with a 7 -8 nm excitation slit and 0.1 -0.2 nm emission slit width.Experiments were carried out in solution using 1 cm path length quartz cuvettes with four transparent polished faces.
Table S2: Quantum yields for compounds 1 and 3 in toluene.

d) Fluoresence lifetime measurements
Fluorescence lifetimes measurements were carried out on the emission band of compounds 1-rac (635 nm) and 3 (604 nm) in toluene using a Horiba Fluorolog-3 Time Correlated Single Photon Counting (TCSPC) module controlled by Datastation software.The sample was excited using a 373-nm Delta Diode 373 nm laser attenuated with a neutral density filter and analysed using DAS6 software.Fluoresence lifetimes were obtained by fitting the resulting data to exponential decay curves.

e) UV-vis-NIR absorption (reflectance) and emission spectra from crystals
UV-vis-NIR absorption (reflectance) spectra from crystals of 1-rac and 1-PP were measured on a Shimadzu UV3600i spectrometer fitted with a ISR-603 integrating sphere with a powder sample holder.Absorption spectra were calculated from the resulting reflectance spectra using the Kubelka-Munk function: Abs = (1-R 2 )/2R where R is reflectance. 23The reflectance spectrum of BaSO 4 was measured as a baseline with 100% reflectance.
Emission spectra from crystals were measured on a Horiba Fluorolog-3 (L-configuration) equipped with a 450 W Xenon light source, R928P photomultiplier tube and double monochromators.The crystals were placed on a quartz slide on the instrument's slide-holder attachment.Figure S20: Solid state absorption spectra for crystals of 1-rac and 1-PP.Spectra were calculated from raw reflectance spectra using the Kubelka-Munk formula: Abs = (1-R 2 )/2R where R is reflectance.The spectra were normalised to give an absorbance of 0 at 650 nm.

Self-assembly studies
UV-vis spectra were measured as described in Section 5a.Quartz cuvettes with 0.1, 1 and 10 cm path lengths were used to enable accurate spectra to be measured over large concentration ranges (two orders of magnitude).All experiments were carried out in 3:2 CH 2 Cl 2 :n-hexane solution at 298 K.For each experiment, a sample at high concentration (black traces with exact concentration given in Figures S22, S24
In the modified isodesmic (nucleation-elongation), dimerization (with a distinct dimerization constant K d = K 2 ) is followed by isodesmic aggregation, such that This can be described by the cubic equation: Where p = K 2 / K. Therefore, p < 1 implies cooperative binding when the aggregate forms after the initial dimerization nucleation event.
is the mole fraction of the monomer (i.e., the  mon unaggregated macrocycle) and is the total concentration of the sample.This equation   cannot be solved for analytically for general cases.However, it is possible to calculate  mon KC T as a function of KC mon for specific values of p using the equation: From this we can calculate and (the mole fraction of macrocycle 1-rac in the  mon  agg aggregate state) because: is related to the extinction coefficients of the aggregate ( , monomer ( ) and  agg  agg )  mon intermediate state ( ) by the equation: Hence, Hence the data can be plotted against and fitted manually to the parameters , ( T )  Figure S25: Change in extinction coefficient at 596 nm of macrocycle 1-rac upon changing the concentration from 0.7 μM to 76 μM, manually fitted to the modified isodesmic (nucleation-elongation) model for various values of p = K 2 /K.The best manual fit (red trace) was obtained for K = 90,000 M -1 , p = 0.001, ε agg = 33,000 cm -1 M -1 and ε mon = 18,750 cm -1 M -1 .From this we can deduce K 2 = 90 M -1 .We estimate an error of < 10% based on the quality of other fits that we tested.

d) Attempt to disrupt the intramolecular dimer in macrocycle 1
We attempted to disrupt the macrocycle intramolecular dimer by adding methanol to break any hydrogen bonds between the PDI units.However, no significant change to either the UVvis or NMR spectra is observed on going up to 15% MeOH in TCE.

CPL microscopy a) Circular Polarisation Luminescence Laser Scanning Confocal Microscope (CPL-LSCM)
CPL-LCSM 26 was enabled by adapting a commercial LSCM (SP5 II, Leica Microsystems) with excitation provided by a fibre coupled 80 mW variable power 355 nm Nd:YAG CW laser.The CPL analysis module was external to an output port and all elements were mounted in a 30 mm cage mount system for optimal alignment (assorted 30 mm components, Thorlabs).First, light from the sample focal plane excites the mirror controlled X1 emission port and passes through a selectable high transmission 570 nm longpass filter (FGL570, Thorlabs) mounted in a switchable filter selector apparatus (CFS1/M, Thorlabs).The circularly polarised emission is then converted to linearly polarised light by an achromatic quarter wave plate (AQWP05M-600, Thorlabs) and is separated into two detection arms by a simple 50/50 beam-splitter cube (BS013, Thorlabs).In each arm, the linearly polarised light is selectively analysed by linear polarisers (LPVISE100-A, Thorlabs) mounted within ultra-high precision computer-controlled rotation mounts with ± 60 µrad unidirectional repeatability (K10CR1/M, Thorlabs), orientated to select for left or right CPL states via computer control software (Kinesis, Thorlabs).The intensity of emission in each path was quantified by fibre-coupled (200 micron) high performance matched tandem avalanche photodiodes (Leica ADPs, Becker & Hickl ID-120).
The two detection arms were aligned to achieve matched sensitivity to enable rapid simultaneous acquisition of left and right CPL images.Calibration of the linear polarisers for enantioselective localisation was executed based upon the procedure reported by Mackenzie et.al. 27

b) Enantioselective differential chiral contrast imaging
We performed enantioselective differential chiral contrast (EDCC) imaging of single crystals of 1-rac and 1-PP/MM using CPL-LSCM.In CPL-LSCM, right-and left-circularly polarised photons are collected simultaneously from the sample, generating independent right and left CPL images rapidly.EDDC is used to quantify the difference between the amount of lefthanded-and right-handed-circularly polarised photons that are emitted, one of the two images can be subtracted from the other and the brightness of the resulting image determines the chirality-induced helicity dominance of each enantiomer.Having distinct enantiopure and racemic single crystals provided us with the opportunity to quantify the degree of circularly polarized emitted light from single crystals for the first time using CPL-LSCM, by calculating an EDCC dissymmetry factor (g EDCC ), a value analogous to the luminescence dissymmetry factor obtained from CPL spectroscopy (g lum ).In doing so, it is critical to correct for the inherent CPL bias arising from orientation induced reflection and helicity inversion of light.Therefore, we calculated a bias factor B as follows: From the EDCC images of 1-rac we calculate a (left-handed) contrast transfer function (CTF): Where is the left-handed EDCC average 8-bit pixel value (Left CPL -Right CPL) and  ( -) is the right-handed EDCC average 8-bit pixel value (Right CPL -Left CPL).

𝐼 (𝑅 -𝐿)
As 1-rac is racemic, it will emit equal amounts of Left-and Right-handed light.From this, the bias factor B is half of the CTF as the bias is present equally in the Left and Right channels, so B = CTF (1-rac) / 2 = 0.0093.
We then define g EDCC as: EDCC = CTF  ( + ) -B =  ( -) - ( -) Where is the total image average 8-bit pixel value (Left CPL + Right CPL) and B is the  ( + ) calculated bias factor.From this we obtain g EDCC values of +0.0582 and -0.0643 for the 1-PP and 1-MM crystals respectively.We can validate this analysis by also calculating the bias factor B from the uncorrected CTFs for the enantiomeric crystals 1-PP/MM: B = CTF (-)  ( + ) + CTF (-)  ( + )

[12]
From equation [12] we obtain a B value 0.0084, which is within 10% of the B value of 0.0093 calculated from equation [10].

Computational studies
Conformer searches for macrocycle 1 were performed using the combination of the CREST code 28 and the GFN2-xTB semiempirical tight-binding method. 29The lowest energy conformers found using CREST were subsequently reoptimized by means of density functional theory using the B97-3c 30 composite scheme.Solvation effects in the DFT calculations were described using the COSMO 31 (toluene,  r 2.83) implicit solvation models.Vertical excitation and circular dichroism spectra of the DFT optimised conformers were calculated by single point calculations using the combination of the B97x 32 density functional and the def2-SVP basis-sets. 33 The conformer spectrum of macrocycle 1 was found to be very dense in part because of the degrees of freedom associated with the sidechains on the imide nitrogen.We reoptimized the lowest 40 structures with DFT (B97-3c) and found the structure shown in Figure S34 to be the lowest in energy.As discussed in the main text, it is clear from the structures of the lowest energy conformers that the combination of PDI and the two phenyl rings on either side is too large for them to somersault through the macrocycle cavity.Next, we reoptimized the structure of the lowest energy conformer in toluene and predicted the optical absorption spectrum (lowest energy vertical singlet excitations, Table S3) and circular dichroism spectrum (Table S4) in toluene using TD-DFT (wB97x).As expected with use of wB97x the spectrum is blue shifted compared to experiment but can be used to assign the experimental circular dichroism spectrum in Figures 2d and S9 to the MM and PP enantiomers of macrocycle 1.

Figure S1 :
Figure S1: Chiral HPLC chromatogram (Phenomenex i-Amylose-1, 250 x 4.6 mm) of 1-rac dissolved in DCM and eluted with 9:1 (v/v) DCM:methanol.The peaks were assigned as the MM and PP enantiomers by single crystal Xray crystallography of the enantiopure crystals of 1-PP (the second peak).The two peaks have equal integral areas.

Figure S4 : 1 .
Figure S4: Chiral HPLC chromatogram (Phenomenex i-Amylose-1, 250 x 10 mm) of the crude reaction mixture of macrocycle 1.The sample was microfiltered prior to injection into the HPLC column to remove insoluble oligomeric side-products, but no other purification was carried out prior to injection.Peaks A-E were collected and analysed by ESI-MS.Macrocycle 1 was only present in peaks B and E (m/z = 1958) which, as identified by CD-spectroscopy and assigned by TD-DFT, are 1-MM and 1-PP respectively.

Figure S5 :
Figure S5: HPLC chromatogram (COSMOSIL Buckyprep 250 x 10 mm) of the crude reaction mixture of macrocycle 1.The sample was microfiltered prior to injection into the HPLC column to remove insoluble oligomeric sideproducts, but no other purification was carried out prior to injection.Peaks A-D were collected and analysed by ESI-MS.Macrocycle 1 was only present in peak B (m/z = 1958) which, when resolved by chiral HPLC (Figure S1), contains only the enantiomers 1-MM and 1-PP.

Figure S6 :
Figure S6: Molecular unit of macrocycle 1 from the 1-rac crystal structure b with all non-hydrogen atoms represented by ellipsoids at the 25% probability level.Hydrogen atoms omitted for clarity (C, black; O, red; N, blue).

Figure S7 :
Figure S7: Crystal structure of macrocycle 1 from the 1-rac crystal structure depicting the length of the imide groups and the rotational displacement of the PDI cores.

Figure S8 :
Figure S8: Molecular unit of macrocycle 1 from the 1-PP crystal structure with all non-hydrogen atoms represented by ellipsoids at the 50% probability level.Hydrogen atoms omitted for clarity (C, black; O, red; N, blue).

Figure
Figure S15: UV-vis absorption and fluorescence emission spectra of acyclic PDI 3 in toluene and TCE (10 µM).

Figure S16 :
Figure S16: Normalised UV-vis absorption spectrum of macrocycle 1 in toluene (10 µM, black trace), along with the cumulative fit of the spectrum as the sum of three gaussian functions for the 0-0, 0-1 and 0-2 vibronic peaks.

Figure S18 :
Figure S18: Exponential decay of emission for compound 3 (toluene, 5 μmol) measured at 604 nm, from which a fluoresence lifetime of 8 ns is obtained.

Figure S19 :
Figure S19: Solid state reflectance spectra for crystals of 1-rac and 1-PP as well as BaSO 4.

Figure S23 :
Figure S23: Change in extinction coefficient of acyclic PDI 3 at 551 nm upon changing the concentration from 5 μM to 418 μM, fitted to the monomer-dimer model, giving a dimerization constant K d = 1678 ± 318 M -1 .
Figure S28: UV-vis absorption spectrum of macrocycle 1-PP upon changing the concentration from 0.9 μM (blue trace) to 94 μM (black trace).The change in the spectrum is negligible compared to 1-rac.

Figure S30 :
Figure S30: NMR spectra of macrocycle 1-rac in different MeOH:TCE solvent mixtures.No significant change is observed upon increasing the methanol content of the solution.

Figure S31 :
Figure S31: Enantioselective differential chiral contrast (EDCC) images for crystals of 1-rac along with the average 8-bit (0-255 greyscale) pixel values for the highlighted areas in the images.

Figure S32 :
Figure S32: Enantioselective differential chiral contrast (EDCC) images for crystals of 1-PP along with the average 8-bit (0-255 greyscale) pixel values for the highlighted areas in the images.

Figure S33 :
Figure S33: Enantioselective differential chiral contrast (EDCC) images for crystals of 1-MM along with the average 8-bit (0-255 greyscale) pixel values for the highlighted areas in the images.

Figure S34 :
Figure S34: Structures of the MM (left) and PP (right) stereoisomers of the lowest energy conformer of macrocycle 1 after B97-3c optimisation (hydrogens omitted for clarity).

Table S1 :
Absorption and emission dissymmetry factors g abs and g lum for macrocycle 1 enantiomers in toluene and TCE (both at 10 μM).

Table S3 :
Lowest singlet excitations in toluene as predicted with TD-wB97x for the lowest energy conformer after B97-3c optimisation.

Table S4 :
Circular dichroism spectrum in toluene as predicted with TD-wB97x for the lowest energy conformer after B97-3c optimisation.