Towards homochiral supramolecular entities from achiral molecules by vortex mixing-accompanied self-assembly

By using a vortex mixing-accompanied self-assembly strategy, homochiral entities with controlled handedness were obtained from exclusively achiral molecules.

handed circularly polarized light, respectively. Experimentally, the value of g CD is defined as g CD = [ellipticity/32980]/absorbance at the CD extremum. The magnitude of CPL can be evaluated by the luminescence dissymmetry factor (g lum ), which is defined as g lum = 2 × (I L -I R )/(I L + I R ), where I L and I R refer to the intensity of left-and right-handed CPL, respectively. Experimentally, the value of g lum is defined as g lum = [elipticity/(32980/ln10)]/total fluorescence intensity at the CPL extremum. For the measurement of CD spectra, the cuvette was placed perpendicularly to the light path of CD spectrometer and rotated within the cuvette plane in or-der to rule out the possibility of the birefringency phenomena and eliminate the possible angle dependence of the CD signals. To estimate the contribution of LD effect on the true CD signal, 36 CD and LD spectra of the samples were measured in steps of 10° by rotating the sample which fixed in the homemade rotator.
Vortex mixing treatment 5 mg BTACA was first dissolved in 0.7 mL dimethylforma-mide (DMF) by sonication in a 5 mL vial, then 0.7 mL Mil-li-Q water was injected with a pipette and instant gels were formed in the mixed solvent. Upon heating (383 K), the gels became into transparent solution. Then vortex mixing at 2500 rpm was applied to the hot solution for 10 minutes. After that, white suspensions with good dispersion were obtained.
After that, these samples were placed in oil bathes at different temperature and stirred at 1000 rpm.

Synthesis and characterization of compounds
The BTA core with three functional cinnamic acid side chains (BTACA) was prepared by the hydrolysis of the ethyl cinnamate-substituted 1,3,5benzenetricarboxamide (BTACE).
The resultant powder was dried to give a white powder with a yield of 88%.
The obtained yellow solution was added 60 mL water and treated with 1 M dilute HCl until pH=1 and solid precipitate was obtained. After filtration, the solid was washed 5 / 47 with water and methanol to get the final product with a yield of 73%. 1    Volume ratio [a] Phase [b] 10-0 S

Supplementary Tables and Figures
The total volume of DMF/H 2 O is 1.4 mL and the concentration of BTACA is 5.54 mM.
[b] S: solution; PG: partial gel; G: stable gel; I: insoluble.   The assemblies of BTACA in DMF/H 2 O through vortex mixing were characterized by means of X-ray diffraction (XRD), UV-Vis and fluorescence spectral measurements. Compared with its solution, UV-Vis spectra of BTACA assemblies broadened with a shoulder peak at 380 nm which suggested the existence of J-like aggregation (Fig. S3a). Meanwhile, the fluorescence intensity was significantly enhanced, accompanying with a drastic red-shift from 434 to 498 nm after assembly ( Fig. S3b). The XRD profile of BTACA xerogel showed a series of peaks at 2.29°, 4.68°, 5.21° and 6.82°, and corresponding distances were estimated to be 3.85, 2.70, 1.88 and 1.69 nm, respectively (Fig. S4a). This indicates that the molecular packing adopts a body-centred cubic structure pattern. Moreover, the in situ investigation of BTACA assemblies in DMF/H 2 O (1:1 v/v) showed identical peaks, which indicated the highly ordered molecular packing (Fig. S4b). Driving force of supramolecular assembly could be further confirmed by Fourier transform-infrared (FTIR) measurement. The peak at 1666 cm -1 for BTACA clearly indicated the formation of hydrogen bond from carboxylic acid groups (Fig. S4c) while the peaks at 1593 cm -1 (amide I band) and 1518 cm -1 (amid II band) suggested the existence of C=O and N-H in the hydrogen bond form. FTIR data demonstrated that the hydrogen bond between carboxylic acid groups played a crucial role in the supramolecular assembly of BTACA.   with the mean spectrum of all 36 LD spectra (dash line).
In Fig. S7a, the CD difference from the maximum value (351 nm) to the minimum value (290 nm) fluctuated around 1600 mdeg. This nonzero value suggested that the observed CD signal represents the authenticity of helical chirality in the suspension system. On the contrary, the values of 36 individual LD spectra fluctuated with the testing angle, suggesting that the angle-dependent LD effect can be eliminated by averaging all of the LD spectra (Fig. S7b). Therefore, the true CD intensity can be obtained by averaging all of these 36 individual spectra as shown in Fig. S7c.
Moreover, the contribution of LD to the CD spectra could be quantitatively estimated by using the following semi-empirical equation (45):

Contamination of CD by LD = LD ×0.02 / CD observed
The contamination of CD by LD in the present case was estimated about 0.05%. The detection areas during CD measurement are also taken into consideration.
Since cuvette of 0.1 mm is used for measuring the CD spectra, 30 μL suspension is enough for each CD measurement. As shown in Fig. S8, we measured 20 times for one BTACA samples obtained by vortex mixing. Therefore, almost 600 μL suspension were measured for one sample. On the other hand, the effective detection areas for CD measurement (JASCO J-1500 spectrometer) is about 0.5 cm 2 (the diameter is 8 mm). Thus, about 10 cm 2 area was measured for one sample.
Clearly, there is no difference among them, indicating the well-dispersed and homochiral of BTACA assembles after vortex mixing process.  SEM images in (a) shows the same SEM images as in (b) but without the markers.
Only P handed nanohelix were observed within the scope of 300 μm × 200 μm. The red boxes in (b) represented that the scopes were carefully analyzed, which were listed in the end of this file.   By carefully looking the UV-vis absorbance spectra of the BTACA assemblies after vortex mixing (Fig. S3a), we could see a small shoulder peak at 380 nm besides the main peak at 328 nm. This shoulder peak can be attributed to the J-like aggregation (pi-pi stacking) of BTACA assembles. After the vortex mixing process, the assemblies were slender. With further ripening operation, the assemblies grew larger, which was reflected by a slight decrease of the main absorption peak and the entanglement of helical nanostrucutres (Fig. S15).
In the CD spectra, due to the formation of the chiral nanostructures, both of these species contributed to the CD signals. For the vortex mixing sample, the CD peak showed at 351 nm, while for the ripening processed assemblies, the CD maximum moved to 380 nm, which was the right position of shoulder absorption peak. In any cases, we saw that the crossover is in the same place at 313 nm (the position of the chromophore), indicating that the exciton coupling is based on the interactions between these chromophores after aggregation.
Considering that the ripening process needed much longer ripening time, these change in CD spectra might be caused by the enhanced J-like aggregation after thermodynamic equilibrium. In addition, tiny nanostructure difference of BTACA assemblies between these two process were observed from the SEM images ( Fig. 3 and S16), which might also lead to the different CD spectra.