Tuning the donor–acceptor interactions in phase-segregated block molecules

The assembly of donor–acceptor molecules via charge transfer (CT) interactions gives rise to highly ordered nanomaterials with appealing electronic properties. Here, we present the synthesis and bulk co-assembly of pyrene (Pyr) and naphthalenediimide (NDI) functionalized oligodimethylsiloxanes (oDMS) of discrete length. We tune the donor–acceptor interactions by connecting the pyrene and NDI to the same oligomer, forming a heterotelechelic block molecule (NDI-oDMSPyr), and to two separate oligomers, giving Pyr and NDI homotelechelic block molecules (Pyr-oDMS and NDI-oDMS). Liquid crystalline materials are obtained for binary mixtures of Pyr-oDMS and NDI-oDMS, while crystallization of the CT dimers occurred for the heterotelechelic NDI-oDMS-Pyr block molecule. The synergy between crystallization and phase-segregation coupled with the discrete length of the oDMS units allows for perfect order and sharp interfaces between the insulating siloxane and CT layers composed of crystalline CT dimers. We were able to tune the lamellar domain spacing and donor–acceptor CT interactions by applying pressures up to 6 GPa on the material, making the system promising for soft-material nanotechnologies. These results demonstrate the importance of the molecular design to tune the CT interactions and stability of a CT material.


Materials and Methods
All chemicals were purchased from commercial sources and used without further purification. The discrete length oligodimethylsiloxanes (oDMS) dihydride with a length of 8, 24 or 40 repeating units were synthesized according to literature procedure. [1] NDI-3 and 10-undecen-1-amine were synthesized according to literature procedures. [2] Dry solvents were obtained with an MBRAUN solvent purification system (MB-SPS). Oven-dried glassware (120 °C) was used for all reactions carried out under argon atmosphere. Reactions were followed by thin-layer chromatography (TLC) using 60-F254 silica gel plates from Merck and visualized by UV light at 254 nm. Automated column chromatography was conducted on a Biotage Isolera One system using Biotage Sfär Silica Flash Cartridges. NMR spectra were recorded on Bruker 400 MHz Ultrashield spectrometers (400 MHz for 1 H NMR, 100 MHz for 13 C NMR). Deuterated solvents used are indicated in each case. Chemical shifts (δ) are expressed in ppm and are referred to the residual peak of the solvent. Peak multiplicity is abbreviated as s: singlet; d: doublet, q: quartet; p: pentet; h: heptet; m: multiplet; dd: double doublet; dt: double triplet; ddt: double doublet of triplets. Matrix assisted laser absorption/ionization-time of flight mass spectra (MALDI-ToF) were obtained on a Bruker Autoflex spectrometer using α-cyano-4-hydroxycinnamic acid (CHCA) or trans-2-[3-(4-tertbutylphenyl)-2-methyl-2-propenylidene]-malononitrile (DCBT) as matrix. Polarized Optical Microscopy (POM) samples were placed on glass substrates and imaged using Nikon Xfinity1 Lumenera microscope with 5x magnification. Differential scanning calorimetry (DSC) data were collected on a DSC Q2000 from TA instruments, calibrated with an indium standard. The samples (4-8 mg) were weighed directly into aluminium pans and hermetically sealed. The samples were initially heated to 180 °C and then subjected to two cooling/heating cycles from -50 °C to 180 °C with a rate of 10 K min -1 . The data that is presented, represents the second heating and/or cooling cycle. Bulk small angle X-ray scattering (SAXS) was performed on an instrument from Ganesha Lab. The flight tube and sample holder are all under vacuum in a single housing, with a GeniX-Cu ultra low divergence Xray generator. The source produces X-rays with a wavelength (λ) of 0.154 nm and a flux of 1 × 108 ph s -1 . Scattered X-rays were captured on a 2-dimensional Pilatus 300K detector with 487 × 619 pixel resolution. The sample-to-detector distance was 0.084 m (WAXS mode) or 0.48 m (MAXS mode). The instrument was calibrated with diffraction patterns from silver behenate. High-pressure powder X-ray diffraction (HP-XRD) was performed using a Rigaku R-axisIV ++ imaging plate diffractometer with a MicroMax-007 x-ray generator (MoKα, λ = 0.07107 nm) and a Varimax-Mo confocal mirror optics. The incident x-ray was collimated by a single pinhole colimator with a hole diameter of 0.3 mm. Powder crystal samples with small ruby balls as pressure markers were loaded into clamped diamond anvil cells made of cupper-bellyrium alloy without any pressure-transmitting medium. The load was controlled by high-pressure He gas via a membrane, or by 3 x M3 screws for higher load. A pair of Boehler-Almax type conical diamonds with 0.8 mm culet were used as anvils. Copper-bellyrium or Inconel alloy plates with a thickness of 0.1-0.2 mm and a hole diameter of 0.3-0.4 mm were used as gasket. Applied pressure was estimated by ruby fluorescence method. [3] The obtained 2D diffraction images were converted into 1D diffraction pattern using a IPAnalyzer and PDIndexer softwares. [4] High-pressure fluorescence (HP-FL) measurements were performed on an OLYMPUS BX51 microscope. Samples were loaded in diamond anvil cells as well as HP-XRD. The samples were irradiated by a Laser Quantum Gem 532 DPSS laser (5mW on the stage of microscope), and the fluorescence was filtered by a Edmund optics 532 nm notch filter, and the spectra were obtained by an Ocean Optics USB2000 spectrometer. Exposure time was 0.1 sec, and the spectra were measured 5 times and then averaged. Ultraviolet-visible (UV-vis) absorbance spectra were recorded on a Jasco V-650 UV/Vis spectrometer at 293 K with a Jasco ETCT-762 temperature controller. UV-vis measurements were performed using spincoated quartz plates or quartz cuvettes (1 cm) with a Teflon cap. For the thin film samples, the quartz substrates were cleaned by subsequent sonication of the substrate in ethanol and acetone for 10 minutes per solvent. The substrates were dried by a stream of air and the thin film samples were prepared by spincoating (800 rpm, 1 minute) 50 μL of a 10 mg/mL solution containing the material onto a cleaned quartz substrate. The solution samples were prepared by solvation of the material in the solvent. The samples prepared in oDMS15 solvent were heated to 80 °C and subsequently cooled to room temperature before the measurement.

1-(Pent-4-en-1-yloxy)pyrene (3)
To a suspension of K2CO3 (1.60 g, 11.6 mmol, 2 eq) in dry DMF (0.2 M), 1-hydroxypyrene (1), 5-bromo-1pentene (2) (1.73 g, 11.6 mmol, 2 eq) and KI (0.1 g, 0.56 mmol, 0.1 eq) were added. The reaction was stirred at 80 ˚C for 24 h. The reaction mixture was cooled to room temperature and 50 mL of acetone was added. The mixture was filtered by vacuum filtration using a Buchner funnel and paper filter. The filtrate was collected, and the acetone was removed by rotary evaporation. Subsequently, the crude mixture was precipitated in 300 mL water, dropwise and while stirring. The solids were isolated by filtration using vacuum filtration with a Buchner funnel and paper filter. The solid residue was dried in vacuo, yielding pure olefin functionalized pyrene 3 as a dark green solid (1.48 g, 89%). 1

Pyr-3
Pyrene 3 (0.34 g, 1.2 mmol, 2.4 eq) and Si40 dihydride (0.29 g, 0.50 mmol) were dissolved in DCM (1 mL). The reaction was stirred at room temperature for 2 h. After full conversion of the hydride, the crude mixture was purified by automated column chromatography using heptane/DCM (gradient 95/5 to 80/20) as eluent. The pure product Pyr-Si40-Pyr was obtained as a dark-green oil (0.48 g, 83 %). 1 isoquinoline-6,7-dicarboxylic acid (6) NTCDA (4) (0.6 g, 2.24 mmol, 2 eq) and n-pentylamine (5) (0.10 g, 1.12 mmol) were loaded into a 20 mL microwave reaction vessel equipped with a magnetic stirrer. DMF (7 mL) was added, the vessel was sealed, and the suspension was sonicated for 15 minutes. The mixture was heated in the microwave (300 W) at 75 °C for 5 minutes, followed by heating at 140 °C for 15 minutes. The suspension was poured into 200 mL NaOH (1 M), precipitates formed, and the solids were removed by vacuum filtration using a Buchner funnel and paper filter. The filtrate was acidified with concentrated HCl (37%) until pH ∼7. Then, the mixture was acidified to pH ∼3 more gently using a 3M HCl solution resulting in the formation of precipitates. The mixture was filtered by vacuum filtration using a Buchner funnel and paper filter, and the solid residue was dried in vacuo at 80 °C overnight. Pure naphthalene mono-imide (NMI) 6 was obtained as a light-brown solid (86 mg, 22 %). 1 [3,8] phenanthroline-1,3,6,8(2H,7H)-tetraone (7) NMI 6 (92.5 mg, 0.26 mmol) and 4-pentene-1-amine (34 mg, 0.39 mmol, 1.5 eq) were loaded into a 5 mL microwave reaction vessel equipped with a magnetic stirrer. DMF (4 mL) was added, the vessel was sealed and the suspension was sonicated for 10 minutes. The mixture was heated in the microwave (300 W) at 75 °C for 5 minutes, followed by heating at 140 °C for 15 minutes. The suspension was poured into 100 mL water, precipitates formed, and the solids were isolated by vacuum filtration using a Buchner funnel and paper filter. The solids were washed thoroughly with water. Subsequently, the solid residue was dried in vacuo at 100 °C overnight. The crude product was dissolved in CHCl3, filtered over a silica plug (4 cm) and eluted with CHCl3. The product was dried in vacuo, yielding pure NDI 7 as a light pink solid (50 mg, 48%).

2-Pentyl-7-(undec-10-en-1-yl)benzo[lmn][3,8]phenanthroline-1,3,6,8(2H,7H)-tetraone (11)
NMI 10 (0.5 g, 1.14 mmol) and n-pentylamine (0.5 g, 5.71 mmol, 5 eq) were loaded into a 5 mL microwave reaction vessel equipped with a magnetic stirrer. DMF (4 mL) was added, the vessel was sealed, and the suspension was sonicated for 10 minutes. The mixture was heated in the microwave (300 W) at 75 °C for 5 minutes, followed by heating at 140 °C for 15 minutes. The suspension was poured into 100 mL water, precipitates formed, and the mixture was filtered under vacuum using a Buchner funnel and paper filter. The filter cake was washed with water and dried in vacuo at 100 °C overnight. The crude product was dissolved in DCM, filtered over a silica plug (4 cm) and eluted with DCM. The solvent was removed in vacuo and pure NDI 11 was obtained as a light pink solid (130 mg, 28%). 1

Bulk co-assembly of homotelechelic NDI-1 with Pyr-oDMS, having a varying siloxane oligomer volume fraction
Mixing Pyr-1 with NDI-1 gives a mixture in which the two components have equal siloxane linker lengths (Si8). To investigate the influence of the siloxane oligomer length on the bulk morphology and CT properties of the two-component material, mixtures of NDI-1 with Pyr-2 and Pyr-3 were formed, denoted as Pyr-2:NDI-1 and Pyr-3:NDI-1, respectively. The thermal properties of the individual components ( Figure S6) and mixtures were evaluated using DSC measurements ( Figure S7A-C and Table 1). The DSC traces of the three mixtures are very comparable, showing a broad endothermic transition upon heating with a relatively low enthalpic energy (< 3.6 kJ mol -1 ) compared to the melting temperatures of the individual components (Table 1)   We investigated the liquid crystalline order in the mixtures by medium-and wide-angle X-ray scattering (MAXS and WAXS) to determine the morphology and the extent of mixing ( Figure S8). For the mixtures, co-assembly of the two oligomers can be concluded when a new, single nanostructured morphology is obtained. In contrast, a linear combination of the two components indicates a self-sorted, macrophase-segregated system. Lamellar nanostructures were formed for all single components as follows from the scattering peaks at integer multiples of q* in the transmission scattering profiles ( Figure S8). This indicates the presence of a single lamellar structure for both Pyr-2:NDI-1 and Pyr-3:NDI-1 mixtures. Similar S19 to the Pyr-1:NDI-1 mixture, the sharp scattering peaks in the high-q region of the transmission scattering profile of the single components disappear upon mixing and a single scattering peak at 18.3 nm -1 appears. Hence, the NDIs and pyrenes in the mixtures order by means of CT and πstacking interactions in combination with microphase segregation. The presence of CT complexes in the nanostructure was confirmed by UV-Vis spectrometry ( Figure S9). Interestingly, the scattering peaks of the mixtures in the low-q region were much sharper than those of the individual Pyr-1, Pyr-2 and Pyr-3 block molecules. This suggests that a better-defined lamellar morphology is obtained for the mixtures compared to the Pyr-oDMS block molecules. Furthermore, the lamellar domain spacings of the mixtures were approximately the average of the domain spacing of the NDI-1 and the pyrene block molecules (Table 1). A larger siloxane volume fraction of the Pyr-oDMS block molecule in the mixture gave rise to a larger domain spacing as the siloxane layer increases ( Table 1, entries 5-7). Hence, we were able to tune the feature sizes of the CT material by changing the siloxane length in just one of the components in the mixture, while maintaining a high degree of order.

Influence of alkyl spacer on the morphology and CT properties of mixtures of Pyr-1 with NDI block molecules
The siloxane linker length of Pyr-oDMS in the mixtures had no effect on the order of the lamellar morphology and solely co-assembled structures were obtained, even with a difference of 32 siloxane repeating units between the NDI-and Pyr-oDMS block molecule. We also evaluated the effect of the carbon linker length on the co-assembly properties and morphology of the mixtures. For this, we mixed Pyr-1, having a C5-alkyl linker, with reference molecules NDI-2 and NDI-3, that have a C11-alkyl linker, forming mixtures denoted as Pyr-1:NDI-2 and Pyr-1:NDI-3 ( Figure S11A). The two mixtures both formed a macrophase segregated morphology, observed by a linear combination of the individual components in the 1D transmission scattering profiles ( Figure S11). Thus, in order to obtain co-assembled nanostructures in the bulk, the alkyl linker lengths of the NDI-and Pyr-oDMS block molecules have to match to prevent self-sorting. These bulk assembly results are in accordance with results obtained in solution by Ghosh and coworkers, showing selfsorting of donor and acceptor chromophores due to an alkyl linker length mismatch. [5]  Linear correlation in Figure 4E: To the set of experimental data consisting of the photon energy corresponding to the maximum in CT emission intensity at pressure P and the inverse stacking distance 1/dπ2 (P) also at pressure P , the following linear relation has been fitted using linear regression: The fit is shown as the continuous red line in Figure 4E, with intercept ECT(dπ2 = ∞) = 4.4 eV and slope S = -8 eV Å as optimized fit parameters.