Click Procedure of Phthalocyanine Star-Shaped Mesogens – the Effect of Size and Spacer Length

ABSTRACT A new Click procedure have been recently presented, in which a mixture of a female piece (star mesogen with shape-persistent conjugated arms and intrinsic free space) and a male building block (the same star mesogen sterically overcrowded with fullerenes attached via spacers to the arms) self-assemble in highly ordered filled liquid crystal structures. This self-assembly is based on the space-filling of the intrinsic free space and the nanosegregation of the fullerenes in helices. Here we highlight the synthesis of larger phthalocyanine star mesogens and fullerenes attached via long spacers and their mixtures as well as mixtures of the components with the stars of intermediate size and spacer length. While the system with long spacers do not show any alignment and π-stacking, the click procedure reappears successfully for the combination of a large female star and the intermediate male star. GRAPHICAL ABSTRACT


1) Materials
All commercial materials employed were used as received, without further purification. P(OEt)3 was freshly distilled under reduced pressure and solvents were distilled and dried via standard procedures.
Several quaternary and ternary signals, especially of the phthalocyanine core and the arms of low intensity could not be identified certainly, although a long time measurement at a 600-MHz-NMR-spectrometer with a highly sensitive cryoprobe was performed. Anyway, the signals at  5, 552.5, 561.2, 573.7, 597.8, 624.8, 660.5, 746.3, 765.6, 799.4 Figure S2 shows a comparison of the 1 H NMR spectra of molecule 2b without (black) and with two drops of pyridine-d5 (red). The aromatic region without pyridine-d5 between 6.5 ppm and 9.5 ppm is extreme broad and individual peaks can almost not be observed. After addition of the pyridine-d5, separate peaks can much better be identified. Also the signal at 4.6 ppm, which can be attributed to half of the diastereotopic protons of the CH2 group within the pyrrolidine ring, is more pronounced in the red spectrum.
Moreover, all relevant protons of the fullerene spacer could be identified when they were compared to the spectrum of fullerene 22 (see Figure S3).  which is further supported by the mass spectra ( Figure S5B).
-23 -  Figure S4: 1 H NMR spectra (CD2Cl2, 400 MHz) of 2b at 295 K between 4.4 ppm and 10.0 ppm. Figure S4 highlights the aromatic region of compound 2b, which could be best resolved in CD2Cl2 with additional pyridine-d5. It confirms the presence of all aromatic and olefinic protons and is in good agreement with the molecular structure. The broad signal at 4.60 ppm with an integration of four protons can be attributed to half of the diastereotopic protons of the CH2 group, belonging to the pyrrolidine ring (see Figure S3).
The presence of four spacers containing fullerenes was unambiguously confirmed, as Figures   S6 and S7 are in good agreement with the molecular structure. Moreover, the successful synthesis is further supported by the mass spectra ( Figure S5B).
-24 -  Figure S5. MALDI-MS spectra of compounds 1b (A) (positive) and 2b (B) (negative). The spectra of 2 clearly shows that four fullerene building blocks are bound to the phthalocyanine since no additional signal at 6982, 5982, 4982 m/z (highlighted with arrows) appears, which would be expected for phthalocyanines with only 3, 2, and only 1 fullerene unit. The additional signal (B) belongs to an unknown decomposition product.
DSC trace of 1b between 25 °C and 300 °C do not show any phase transition.

7) Density measurements by the buoyancy method at 22 °C
The density measurement of compound 1b was carried out in mixtures of deionized water and aqueous calcium chloride (40 °wt%) solution. Before dissolving, the calcium chloride was dried at 140 °C under reduced pressure (1 × 10 -3 mbar). All solvents were degassed by ultra sonication.
The samples were heated to 170 °C and extruded into a thin solid fiber under reduced pressure to avoid inclusion of air. The fiber was cut in small pieces. These samples were put in a vial containing deionized water. Aqueous calcium chloride (40 wt%) solution was added in small portions until the sample started floating. The mixture was allowed to equilibrate between additions. The weight percentage of calcium chloride was determined and the density was calculated according to reference. 12 Note that this method relies on samples which are free from air inclusions, which cannot be completely guaranteed with the present procedure. For materials with much lower clearing temperatures and high thermal stability in the isotropic liquid, the samples can be prepared by keeping them a long time in the isotropic liquid under vacuum, which is supposed the eliminate all air bubbles. [13] However in the present case this is not possible since the star-compounds do not melt. Therefore, the present density can be only a minimum value for star 1b.

8) Modelling
The modelling of the hexagonal unit cell was performed using the program suite "BIOVIA Materials Studio 2017R2" and the Forcite module with the force field COMPASSII.
For the construction of the unit cell, the set-up for 1b started with the dimers (see Figure 3) placed in the unit cell (a = 63.9 Å, c = 48.1 Å). The dimer was then copied, translated by (48.1/16)Å in c-direction and rotated by (360/16)° about the column axis. After 16 dimers, this procedure generated the complete helix. The cell was geometry optimised first by using the atomic summation method and subsequently applying the Ewald summation method until large negative non-bonding interactions (van der Waals and electrostatic interactions) have been obtained, which support the plausibility of the model (Figure 3).