Hydrogen-bonded perylene bisimide J-aggregate aqua material

A water-soluble perylene bisimide dye self-assembles in aqueous media into thermoresponsive aqua materials with photoluminescence within the biological transparency window.


Materials and methods
Solvents and reagents were purchased from commercial suppliers (Sigma-Aldrich, ACROS, Alfa Aesar, Merck) and used as received, without further purification, with the exception of N,N'dicyclohexylcarbodiimide (DCC), which was distilled in prior, and solvents were distilled and dried by standard procedures. All reactions were carried out under nitrogen atmosphere. Column chromatography was performed with commercial glass columns using silica gel 60M (particle size 0.04  0.063 mm) from Macherey-Nagel as stationary phase. 1 H and 13 C NMR spectra were generally recorded on a Bruker Avance III HD 400 spectrometer operating at 400 MHz ( 1 H) or 100 MHz ( 13 C), with the residual protic solvent used as the internal standard. The chemical shifts are reported in parts per million (ppm). Multiplicities for proton signals are abbreviated as s, d, and m for singlet, doublet and multiplet, respectively. DOSY NMR spectra were recorded on Bruker DMX 600 spectrometers equipped with a BGPA 10 gradient generatore, a BGU II control unit and a conventional 5 mm broad band ( 15 N, 31 P)/ 1 H probe with z axis gradient coil capable of producing pulsed magnetic field gradients in the z direction of 52 G cm 1 . The spectral data were acquired using the longitudinal eddy current delay sequence with bipolar gradient pulse pairs for diffusion (BPP-LED) 1 and additional sinusoidal spoil gradients after the second and fourth 90° pulses were used. The temperatures were calibrated with a probe of 4% MeOH in CD 3 OD. The fluctuation of the temperature was less than 0.1 K during the measurements. The strength of the pulsed magnetic field gradients was calibrated by 1 H DOSY experiments with a sample of 1% H 2 O and 99% D 2 O, doped with GdCl 3 (0.1 mg mL 1 ) to achieve short spin-lattice relaxation times, using known value of the diffusion coefficient for H 2 O at 295 K in this H 2 O/D 2 O mixture. Diffusion decay signals were fitted by using Eq. S1: (Eq. S1) where γ is the gyromagnetic ratio of the proton, g is the gradient field strength, Δ and δ are experimental parameters. The fitting considered a log-normal distribution of size for the supramolecular polymers and was performed by utilizing the Matlab fitting routine as in the reference. 2 A 1 H NMR spectrum was measured in H 2 O/D 2 O 1:9 mixture by using a 5 mm 13 C/ 1 H cryoprobe at 295 K and calibrated to the residual solvent signals.
High resolution mass spectra (HRMS) were recorded on an ESI micrOTOF focus spectrometer (Bruker Daltonic GmbH, Germany).  The aqueous solution samples of Aggregate 1 were prepared by dissolving   the suitable amount of solid material in distilled water and kept it at room temperature for 2 -3 weeks. For Aggregate 2, the samples were prepared freshly following the same procedure but keeping the sample 2 days at room temperature (r.t.). The final state was checked by UV-Vis spectroscopy.
Preparation of the hydrogels and lyotropic liquid crystals: Hydrogels and lyotropic liquid crystal samples were prepared by weighting MEG-PBI in flasks and subsequent addition of the appropriate amount of water. The flasks were closed, sealed with para-film and stored at r.t. until the mixtures became homogeneous (ca. 24 h). In the case of the lyotropic samples, the mixtures were treated with a spatula to ensure the homogenous distribution of water in the PBI material. Prior to the measurement, a thermal equilibration time of 30 min was ensured for each sample. The stability of the system was checked by UV-Vis spectroscopy.
The hydrogel samples for the XRD measurements were prepared as follows: the viscous PBI hydrogels (c = 10 wt%) 1 were placed inside a glass Mark-tube (Ø = 1.5 mm). The tube, open at both sides, was placed vertically in a vial to allow the sample to reach the middle part of the tube by gravity (≈ 12 h). Then, the tube was sealed by melting both sides. This tube was inserted into another Mark-tube (Ø = 2 mm) which was sealed by melting. This sample preparation prevented the evaporation of water during the measurements that were performed under vacuum.  Optical textures of the liquid-crystalline materials were examined with a Nikon Eclipse LV100Pol optical polarizing microscope equipped with a Linkam LTS420 heating stage and a Linkam T95-HS system 1 Unless specified, the concentrations for the mixtures of the hydrogel and the lyotropic LC samples are always reported as wt% referred to the MEG-PBI content.

S4
controller. Thermal analyses by differential scanning calorimetry were performed on a TA instrument DSC Q1000 with a DSC refrigerated cooling system. Temperature-dependent and polarized FT-IR spectra were recorded with an AIM-8800 infrared microscope connected to a Shimadzu IRAffinity FT-IR spectrometer. The sample was prepared as a thin film on a KBr plate (thickness 2 mm) which was placed on a THMS600 heat stage with a Linkam TP94 controller. Polarization-dependent FT-IR spectra were measured by using a precision automated polarizer (ZnSe) from PIKE Technologies. This includes the PIKE Technologies Motion Control Unit and AutoPro.
Atomic force microscopy (AFM) measurements were performed under ambient conditions using a Bruker Multimode 8 SPM system operating in tapping mode in air. Silica cantilevers (OMCL-AC200TS, Olympus) with a resonance frequency of 170 kHz and a spring constant of 10 Nm 1 were used.
Polarized optical microscopy (POM) images were taken with a Zeiss Axio instrument (Zeiss Axiocam 503color, 3Mpx) with temperature-controlled stage Linkam LTS420 with LNP95. UV/NIR absorption and emission spectra were obtained using Zeiss CCD detector.

Cryogenic scanning electron microscopy (Cryo-SEM) measurements were performed using a Zeiss Ultra
Plus Field Emission SEM operating at 0.7  1.5 kV with an aperture size set up to 30 µm to avoid excessive charging and radiation damage of imaged areas. Sample preparation consisted of placing a small drop of the hydrogel onto a copper stub sample holder. The specimen was then plunged into liquid nitrogen slush (mixture of solid/liquid nitrogen) at -210 °C. The sample was then transferred under vacuum using the loading transfer rod into the high vacuum cryo-preparation chamber (Quorum PP2000T) at -150 °C, fractured and then transferred into a SEM sample chamber maintained at about -150 °C.
Differential scanning calorimetry (DSC): Thermal analyses by differential scanning calorimetry were performed on a TA instrument DSC Q1000 with a DSC refrigerated cooling system. The samples of MEG-PBI with the appropriate amount of water were freshly prepared one day before by mixing the two components and keeping the mixture in the fridge overnight. DSC measurements were performed with 10 °C/min heating rate and no equilibration.

Wide-and middle angle X-ray scattering (WAXS, MAXS): Temperature-dependent WAXS and MAXS
investigations were performed on a Bruker Nanostar (Detector Vantec2000, Microfocus copper anode Xray tube Incoatec). Aligned samples were prepared by fibre extrusion using a home-made mini-extruder.

S5
The fibres were transferred into Mark capillaries (Hilgenberg) and assembled in the heating stage of the Nanostar. WAXS experiments were performed at a sample-detector distance of 21 cm, with the detector tilted by 14° upwards in order to study the angular range of 2 = 0.8° -28°. Silver behenate was used as calibration standard for WAXS and MAXS studies. All X-ray data were processed and evaluated with the program data squeeze (http://www.datasqueezesoftware.com/).

Temperature-dependent studies
Fig . S2 shows the temperature-dependent spectra of a 1.0 × 10 5 M MEG-PBI aqueous solution obtained after equilibration at each temperature upon increasing the temperature in the range 24 -36 °C in steps of 1 °C. It is interesting that the aggregation kinetics follow a pseudo first order rate and last < 10 min at 45 °C and >3 weeks at 25 °C (Fig. S2a). The formation of the J-aggregate was monitored at  max = 634 nm.
In order to investigate the J-aggregation, the final spectra, i.e. the spectra that do not show time evolution any longer, are shown in Fig. S2b and the normalized formation of the J-aggregate is shown in

Turbidity measurements
A turbidity measurement consists in monitoring the transmission at 800 nm, far from the absorption of the PBI dye. The measurement revealed a sudden precipitation at the cloud point. The CST was measured by monitoring the transmittance of the solution at 800 nm where there was no absorption, neither from MEG-PBI nor from its aggregate. Heating/cooling rate is 0.01 °C/min to ensure thermal equilibration under 100 rpm stirring. Diffusion decay curves of the DOSY experiment of MEG-PBI in water solution 2.5 x 10 4 M with water signal suppression at 25 °C (red) and after heating at 50 °C (blue). The data were fitted with a Matlab fitting routine that consider a log-normal size distribution of the supramolecular aggregates, as described in the text. The resulting lognormal distribution of the diffusion coefficients is shown in the inset.

IR spectroscopy
In order to investigate the structure of the aggregates, the IR spectra of MEG-PBI in various conditions are here compared: the spectrum in chloroform (Fig. S8b) is in good agreement with the calculated one (A). Calculated spectrum allowed us to unambiguously assign the vibrational modes carrying most structural information, as displayed in the spectrum. In Figure S8c, the spectrum in the liquid state (195 °C) also matches well with the one in the monomeric state. The spectrum of the J-aggregate as lyotropic liquid crystal formed upon cooling is shown in blue. The spectrum shows the formation of two more intense N-H stretching signals at 3182 and 3067 cm 1 . As a proof of the intermolecular hydrogen bonds involving PBI molecules, the band of the imidic C=O stretching at 1702 cm 1 decreases and a different band appears at lower energy (1677 cm 1 ). This can be taken as a direct evidence of the hydrogen bonds involving the imidic N-H of one PBI with one of the two imidic C=O of the second PBI.
The ATR-IR spectra in water (d) also shows a clear indication for hydrogen bonds in the J-aggregate (blue), that are missing in the red aggregate at room temperature (red).

Polarized spectroscopy on sample alignment
The aligned sample of MEH-PBI showed higher polarized absorption at 631 nm (most intense peak) along the parallel direction of the alignment (Fig. S15). The broad band (500-700 nm) is assigned to the excitonically coupled transition corresponding to the S 0 -S 1 electronic transition of the PBI monomer, in which the main transition dipole moment (µ tr1 ) lies along the long molecular axis. Conversely, the band at 442 nm corresponds to the electronic S 0 -S 2 transition of the PBI, where the transition dipole moment (µ tr2 ) lies along the short molecular axis. The pronounced absorption at 631 nm with parallel polarized light proves the orientation of the PBI molecules with the long axis parallel to the shearing direction.
The quality of the alignment can be described by the dichroic ratio and the order parameter . Both

Calculation of the number of molecules per unit cell
The number of molecules Z in the unit cell was calculated using Eq. S4: 10 where δ is the density, M the molecular mass, NA the Avogadro's constant and V unit cell the volume of the unit cell. The density was assumed to be 1 g/cm 3 .
The volume of the unit cell was calculated according to Eq. S5: Z was calculated based on the X-ray cell parameters: a = 37.3 Å, c = 15.4 Å (M = 3185.39 g/mol). Thus, the value of Z corresponds to 3 molecules.

Lyotropic liquid crystal
Polarized optical microscopy