Thymine functionalised porphyrins, synthesis and heteromolecular surface-based self-assembly

The synthesis and surface-based self-assembly of thymine-functionalised porphyrins is described.

2 formed. The mixture was then cooled to room temperature, and toluene (50 mL) was added via canula. 3,5-di-tert-butylphenyl-substituted dipyrromethane (0.63 g, 1.9 mmol) was added in one portion, and the mixture was stirred at 90 °C for 2 h under a nitrogen atmosphere. The reaction was then quenched via addition of saturated NaOAc solution (100 mL) and stirred for a further hour at 90 °C. The mixture was cooled to room temperature and the organic layer separated. The aqueous layer was extracted with DCM (3 × 100 mL). The organic washings were combined, washed with water (3 × 100 mL), dried over MgSO 4 and evaporated under reduced pressure to yield crude product. This was  C,82.98;H,9.20;N,3.65. Found: C,83.11;H,9.28;N,3.74.

1-formylphenyl-3-benzoyl-thymine.
3-benzoylthymine (2 g, 8.69 mmol), 4formylphenylboronic acid (2.6 g, 17.4 mmol), Cu(OAc) 2 (2.37 g, 13.0 mmol) and molecular sieves (3 Å, 3 g) were added to a flame-dried flask. Dry DCM (70 mL) was added via canula, followed by the addition of pyridine (1.4 mL, 17.4 mmol) before the mixture was stirred at room temperature for 74 hours in air. Once a TLC showed that the reaction was complete, the mixture was diluted with DCM (50 mL), filtered through celite and the resulting green solution washed with water in the presence of EDTA (100 mg in 500 mL of distilled water). Single crystals suitable for analysis by X-ray diffraction were grown by slow diffusion of MeOH into a solution of the product in CDCl 3 .

Single crystal X-ray diffraction Studies:
Single crystal diffraction data of mono-TP and benzoyl-mono-TP were collected at 120(2)K on either Beamline I19 at Diamond Light Source (DLS-I19) S1 (mono-TP) or an Oxford Diffraction SuperNova using mirror monochromated Cu-Kα radiation (benzoyl-mono-TP).
Using Olex2 S2 , the structure was solved with the Superflip S3 structure solution program using 6 Charge Flipping and refined with the ShelXL S4 refinement package using Least Squares minimisation.
For the refinement of the structure of mono-TP the tert-butyl group C41A-C44B is disordered over two orientations. Equivalent 1,2-and 1,3-distances of the disordered methyls were restrained to be approximately equal. PLATON SQUEEZE was applied to the data to remove the scattering contribution from a disordered MeOH solvent molecule which could not be sensibly modelled and produce a set of solvent-free diffraction intensities for the final cycles of refinement. A total of 20 electrons were removed from the unit cell, equating to approximately 1 MeOH molecule per formula unit: these were included in the unit cell contents.
For the refinement of the structure of benzoyl-mono-TP geometric similarity restraints were applied to the 1,2 and 1,3 distances of all tert-butyl groups and CHCl 3 solvent molecules (SAME). Rigid bond and similarity restraints were applied to the anisotropic thermal displacement parameters of all atoms in the structure (RIGU, SIMU). Orientational disorder was modelled in CHCl 3 molecule C1C/C1D with their occupancies refined with a free variable (C1C occupancy 0.49(1)) and constrained to sum to unity. The locations of the pyrrolic nitrogen bound hydrogen atoms of the porphyrin ring could not be identified in an electron difference map and so were placed geometrically and refined using a riding model at half occupancy as disordered pairs of opposing positions. PLATON SQUEEZE was applied to the data to remove the scattering contribution from several disordered solvent moieties and produce a set of solvent-free diffraction intensities for the final cycles of refinement. A total of 175 electrons were removed from the unit cell, equating to approximately 1.5 CHCl 3 molecules per formula unit: these were included in the unit cell contents.

Binding Studies:
Deuterated chloroform was filtered through activated basic alumina and dried over 4Å molecular sieves. Dilution and titration experiments was repeated twice, reported error are twice the standard error.
Self-association studies were performed separately for mono-TP and 9-propyladenine using the following approach. A high concentration of analyte was prepared in CDCl 3 and 450 μl this solution was transferred to a capped NMR tube. A small volume of pure CDCl 3 was added successively to dilute the sample and 1 H NMR spectra recorded following each dilution. The data was fitted to dimerisation model by solving the equations below using non-linear curve fitting procedure. S5 In each case the NH chemical shift was used for fitting using the following equations.
[   Mono-TP was employed as the host and the solution concentration was kept constant.
Mono-TP was dissolved in CDCl 3 to make a 2 ml solution (4 -5 mM). 500 μl of the mono-TP solution was transferred to a capped NMR tube and a 1 H NMR spectrum was recorded. The 9-propyladenine compound was dissolved in the remaining host solution and titrated into the NMR tube. 1 H NMR spectra were recorded following addition of each titre. The data was fitted to 1:1 binding isotherm by solving the equation below using non-linear curve fitting procedure. S6 The Chemical shift of mono-TP NH on host was used for fitting using the following equations.
[HG] =  Drift correction of high resolution STM images was carried out using the Scanning Probe Image Processor (SPIP) software (Image Metrology ApS, Lyngby, Denmark). This process involved the collection of a high resolution image of a molecular structure immediately followed by collecting an image of the underlying HOPG lattice. The image of the HOPG lattice was collected using identical scanning parameters apart from the tunnel current and bias voltage. By assuming that the level of drift is constant for both images the known lattice dimensions of HOPG can be used to produce a set of correction parameters for the HOPG image. These parameters can them be applied to the high resolution molecular image to produce a drift corrected image. Figure S7   In an analogous manner to the tetra-TP network, the tetra-TP and 9-propyladenine cocrystal also forms homochiral domains. Within a single domain all of the thymine groups on the tetra-TP molecule adopt the same orientation with respect to the porphyrin core. This separation of the prochiral tetra-TP molecules produces two possible mirror image domain structures for the tetra-TP and 9-propyladenine co-crystal, Fig. S9. An example of the molecular arrangement for one of these mirror image structures is given in Fig. S10b.

Molecular Mechanics (MM) Studies:
MM simulations of the tetra-TP and tetra-TP/9-propyladenine surface structures were carried out using the HyperChem software package. A single layer of graphite was fixed in place and used as a substrate. On top of this layer were placed the molecules representing an individual unit cell of the different 2D structures. The starting positions for the molecules were derived using drift-corrected STM images as a guide and maintaining a 0.35 nm vertical distance between the planar porphyrin and adenine cores of the molecules and the 16 underlying graphite layer. Once positioned, these structures were geometry optimised using the MM+ force field. The optimisation process was terminated when a gradient < 0.01 kcal Å -1 mol -1 was reached. All unit cell dimensions and angles are taken from geometry optimised structures. Fig. S10 shows the geometry optimised molecular structures with the underlying layer of graphite: tetra-TP network (Fig. S10a) and the tetra-TP/9-propyladenine co-crystal (Fig. S10b).