Steering the Self‐Assembly Outcome of a Single NDI Monomer into Three Morphologically Distinct Supramolecular Assemblies, with Concomitant Change in Supramolecular Polymerization Mechanism

Abstract Noncovalent self‐assembly creates an effective route to highly sophisticated supramolecular polymers with tunable properties. However, the outcome of this assembly process is highly dependent on external conditions. In this work, a monomeric naphthalene diimide (NDI), designed to allow solubility in a wide range of solvents, can assemble into three distinct noncovalent supramolecular species depending on solvent composition and temperature. Namely, while the self‐assembly in chlorinated solvents yields relatively short, hydrogen‐bonded nanotubes, the reduction of solvent polarity changes the assembly outcome, yielding π–π stacking polymers, which can further bundle into a more complex aggregate. The obtained polymers differ not only in their global morphology but—more strikingly—also in the thermodynamics and kinetics of their supramolecular self‐assembly, involving isodesmic or two‐stage cooperative assembly with kinetic hysteresis, respectively. Ultimately, three distinct assembly states can be accessed in a single experiment.


General
Chemicals were purchased from commercial suppliers (mainly Sigma Aldrich and Merck) and used as received. Solvents used for spectroscopic analysis were purchased at ACS spectroscopic grade (Sigma Aldrich). Methylcyclohexane and THF were used as received. Chloroform was purified and dried prior to use as follows: Commercial CHCl3 was extracted 3 times with equal volumes of deionized water, pre-dried with anhydrous potassium carbonate, filtered, and then dried with the activated Al2O3 (Basic Brockmann ≈ 15% of the CHCl3 volume) for 24 h in an amberized bottle.
NMR solvents were purchased from Deutero GmbH (Germany). CDCl3 was purchased with silver foil stabilizer, and dried with the activated Al2O3 (Basic Brockmann, ≈ 15% of the CDCl3 volume) prior to use. MCH-d14 was purchased in single-use ampoules and used as received.
NMR spectra were recorded on a Bruker Fourier 300 MHz, Bruker Avance III HD 600 MHz, or Bruker Avance III 700 MHz spectrometers and referenced on TMS δ = 0.00 ppm.
ESI-MS spectra were recorded on Waters Synapt G2-S Q-TOF mass spectrometer in negative ion mode.
The UV-Vis spectra were recorded on Agilent Cary 100 spectrophotometer equipped with water-loop temperature controller, using 10 mm optical path quartz cuvettes.
The CD spectra were recorded on a Jasco J-710 or Jasco J-810 spectropolarimeter equipped with Peltier-type temperature controller, using 10 mm optical path quartz cuvettes.
The DLS spectra were recorded on Malverin Pananalytical Zetasizer equipped with Peltier-type temperature controller, at 173° scattering angle using 10 mm optical path quartz cuvettes. FT-IR spectra were recorded on Bruker IFS 66v/S vacuum FT-IR spectrometer in the airtight Si-crystal cuvette, and then processed in Bruker OPUS software. Spectra of the pure solvents were used for subtraction.

UV spectroscopy
All spectra were replotted and rescaled from Cary WinUV files, using OriginPro 8.0. No data smoothening or noise reduction protocols were applied. After normalization, VT-UV data were fitted to the isodesmic [S1] and nucleation-elongation [S2] models.

CD spectroscopy
All spectra were replotted and rescaled from Jasco Spectra Manager files using, OriginPro 8.0.
After normalization, VT CD data were fitted to the isodesmic [S1] and nucleation-elongation [S2] models. For clarity of the figures, spectra and plots given in main text have been smoothen using Savitzky-Golay algorithm, however the thermodynamic parameters were calculated based on the raw data, prior to smoothening.

DLS
Dynamic viscosity of MCH at specified temperatures were obtained from the from Celsius company datasheet (http://www.celsius-process.com), and are listed in Table S1.