Merging Supramolecular and Covalent Helical Polymers: Four Helices Within a Single Scaffold

Supramolecular and covalent polymers share multiple structural effects such as chiral amplification, helical inversion, sergeants and soldiers, or majority rules, among others. These features are related to the axial helical structure found in both types of materials, which are responsible for their properties. Herein a novel material combining information and characteristics from both fields of helical polymers, supramolecular (oligo(p-phenyleneethynylene) (OPE)) and covalent (poly(acetylene) (PA)), is presented. To achieve this goal, the poly(acetylene) must adopt a dihedral angle between conjugated double bonds (ω1) higher than 165°. In such cases, the tilting degree (Θ) between the OPE units used as pendant groups is close to 11°, like that observed in supramolecular helical arrays of these molecules. Polymerization of oligo[(p-phenyleneethynylene)n]phenylacetylene monomers (n = 1, 2) bearing L-decyl alaninate as the pendant group yielded the desired scaffolds. These polymers adopt a stretched and almost planar polyene helix, where the OPE units are arranged describing a helical structure. As a result, a novel multihelix material was prepared, the ECD spectra of which are dominated by the OPE axial array.


Materials and Methods
NMR experiments have been recorded in a Bruker spectromer ( 1 H frequency 300 mHz).

H NMR T 2 relaxation experiments and H NMR STD experiments have been measured at 300
K in a Bruker NEO-750 spectrometer ( 1 H frequency 750 mHz).
CD measurements were done in a Jasco-720 and UV spectra were registered in a Jasco V-630. The amount of polymer used is indicated in the corresponding section.
VT-CD were measured in a Jasco-1100.
Raman spectra were carried out in a Reinshaw confocal Raman spectrometer (Invia Reflex model) with 785 nm diode laser.
GPC studies were carried out in a Waters Alliance equipped with Phenomenex GPC columns (10 3 Å, 10 4 Å and 10 5 Å). The amount of polymer used was 0.5 mg mL -1 . THF was used as eluent (flow rate: 1 mL min -1 ) and as inner standard, polystyrene narrow standards (PSS) were used.
For the metal ions titrations, poly-2 and poly-3 were dissolved in CHCl 3 and the correspondent metal salt solution was then added. The amount of polymer used is indicated in the corresponding section.
TGA traces were obtained in a TGA Q5000 (TA Instruments, New Castle, UK) using a platinium pan. SAXS experiments were performed in a SAXSPOINT 5.0 X-ray diffractometer. Primux 100 micro-Cu was the microfocus X-ray source with an ASTIX optics. The detector was an HPC Pilatus3 R aM model. The sample was monted in a plate for solid samples sealed with Kapton windows and placed at 1300 mm away from the detector.
Next triethylamine (Et 3 N, 15.00 mL) and ethynyltrimethylsilane (3.20 mL, 23.10 mmol, 1.00 equiv) were added and the mixture was stirred for three hours at room temperature. After removing the solvent, the crude product was chromatographed on silica gel (70-230 mesh) using hexane as eluent (7.00 g, 96% yield).

General Procedure for Polymerization
The polymers were synthesized in a flask (sealed ampoule) previously dried under vacuum and flushed with Ar for three times. The monomers were added as a solid and dissolved in dry THF. Next a solution of rhodium norbadiene chloride dimer [Rh(nbd)Cl] 2 was added and the mixture was stirred overnight. The resulting polymers were diluted in DCM and precipitated in a large amount of MeOH, centrifuged (twice) and reprecipitated in hexane and centrifuged again. 1 H NMR spectroscopy in CDCl 3 indicates that poly-2 adopts a cisconfiguration in the polyene backbone (vinylic proton,  = 5.7-5.8 ppm). However, integration of the peak signals do not match with the expected value. This is related to the difference in mobility of the protons within the molecule -a highly flexible chiral moiety and a more rigid core made up of the OPE units and the polyene backbone-that will need a different relaxation times for the correct signal integration.
A more complex situation is occurred when the 1 H NMR for poly-2 is recoreded in CCl 4 . In this case, not only the integration area of the peak do not match with the expected value, but also the highly stretched scaffold adopted by the polymer in solution results in a broadening of the signals bacause of its large aggregation tendency. Interestingly, in these conditions, the vinylic proton is not observed. The dissapereance of the vinylic peak is confirmed by registering the 1 H NMR spectra of poly-2 at different CDCl 3 /CCl 4 ratios. By increasing the CCl 4 fraction, the vinyl signal at 5.93 ppm (100% CDCl 3 ) is shifted to 6.02 ppm (75% to 25% CDCl 3 ) and becomes null when no ammount of CDCl 3 is present ( Figure S12).
This finding clearly indicates that the vanishing of this signal is directly related to the scaffold adopted by poly-2 in solution. The structure displayed by the polymer will have a large impact on the polymer mobility, affecting directly to the relaxation time during NMR measurements and resuting in a broadening or dissapereance of NMR peaks.
A similar scenario to that observed for poly-2 in CCl 4 occurs for poly-3. This polymer adopts a highly stretched and rigid scaffold in all the tested solvents and, as a consequence, no signal in the vinylic region is observed when registering the NMR spectra. As occurred before, the relaxation time during the NMR experiments will be affected by the high rigidity of the polymer scaffold, affecting to the peak integration area.

STD NMR Experiments
In order to corroborate that fractions of NMR signals remain hidden, Saturation-Transfer Difference (STD) measurements were carried out at different chemical shifts -5.80 ppm (vinylic band) and 7.00 ppm (aryl ring closer to the polyene backbone)-. By using this technique, a proton signal can be selectively saturated, followed by an effective spin diffussion through out the protons network directly attached to the irradiated signal. Therefore, if there are hidden protons in the baseline, the magnetization is propagated across the entire network of protons in large macromolecules and will become visible. On the other hand, if no protons are hidden at the irradiated frequencies, no signal is observed as the saturation cannot be difussed. S2 For poly-2 saturation at 7.00 ppm reveals that the aryl group closer to the polyene backbone is connected to the vinyl proton and the signals adscribed to the pendant moiety show a reduce in intensity, as the protons are placed far away for an effective spin difussion. The proximity between the vinyl proton and the phenyl group is confirmed by irradiating at 5.80 ppm, observing that the more intense peaks are for the benzene groups. In an analogous way, STD 1 H NMR experiments were carried out for poly-2 in different CDCl 3 /CCl 4 mixture ratios and in pure CCl 4 . In all cases the recorded spectra confirmed the proximity between the vinyl proton and the aryl ring. Similar to poly-2, poly-3 was saturated at two different NMR shifts. By selectively irradiating the signal at 7.40 ppm, a signal not previously observed at the 1 H NMR STD off appears at 6.15 ppm, which can be ascribed to the vinylic proton. STD experiments at 6.10 ppm reveal that this signal is directly related to the benzene rings, indicating that the proton observed at 6.15 ppm corresponds to the vinylic one. This peak is not observed due to the broadening of the signal, which makes it remain hidden in the baseline.

Raman Experiments
The bands observed by Raman resonance confirmed the cis configuration. The peak at highest wavelength corresponds to the C=C bond stretching and overlaps with that of the phenyl ring. The band at 1350-1340 cm -1 arises from the cis C-C bond coupled with the single bond connecting the main chain and the phenyl ring. The peak at lowest wavelength corresponds to the C-H bond of the cis form. The disappearance of the alkyne band (ca. 2110 cm -1 ) confirms the formation of the conjugated double bonds of the polymer backbone.

GPC Studies
The molecular weight was estimated by GPC using THF (flow rate: 1.0 mL min -1 ) as eluent and polystyrene narrow standards (PSS) as calibrants.

Polymers Stability Studied by GPC
To corroborate the stability of poly-2 and poly-3, GPC experiments were mesured before and after one week. The obtained traces are almost coincident to those initially obtained, indicating that these polymers are stable in solution.

TGA Studies
The thermal stability was evaluated by TGA. The degradation starts at 300 ºC for poly-2 and for poly-3 at 310 ºC. 6. Dynamic Beahviour 6.1 Studies for poly-2

Analysis in Different Solvents
The CD spectra for poly-2 were recorded in different solvents at a 0.44 mM concentration.
The different scaffolds adopted -classical c-t helix, ω 1 =165º, and matryoshka-like helix, ω 1 = 170º-can be easily tracked by UV-Vis as well as by CD spectroscopy.

Raman Experiments for the Elongated and Stretched Scaffolds
The stretching observed by UV-Vis for low-polar solvents is confirmed by Raman spectroscopy. If poly-2 is dissolved in CHCl 3 (stretched polymer) and in CCl 4 (elongated polymer), a shift of 7 nm in the vibrational band associated to the cis C-H stretching (ca., 1000 nm -1 ) is observed. The addition of metal ions produces no chiral enhancement or helical inversion, which indicates that, although the metal ions can attach to the chiral moiety, no effect is produced in the helical scaffold. This is probably due to the supramolecular interactions established between pendants that forbid any changes in the polymer.

Analysis in Different Solvents
To prepare the polymer samples, due to the difficulty of solubilizing the polymer, small ammounts of poly-3 were weighted and dissolved in the corresponding solvent at the desired concentration (0.36 mM). The resulting mixture was heated overnight at 40 ºC while being magnetically stirred.
Depending on the dielectric constant of the solvent in which poly-3 is dissolved, it will adopt an M or P helix. The change in the conformation of the chiral moiety is produced when (ε-1)/(ε+1)> 0.8.

Formation of SP-3
Polymer solutions of SP-3 were prepared by weighting the monomer and adding the necessary ammount of MCH to obtain the desired concentration (0.15 mM). To completetly solubilize the monomer the mixture was heated up to 80 ºC for two hours.
A similar protocol was used to obtain m-3. In this case, the monomer presents no solubility problems in medium polar or polar solvents, such as DCM, and, as a consequence, the sample does not need to be heated.
The large solubility of m-3 in polar solvents prevents the formation of the aggregate and this can be easily observed by ECD spectroscopy. If the monomer is dissolved in DCM a null ECD trace is obtained, whereas if it is dissolved in MCH an active ECD is produced, indicative of the formation of a chiral aggregate (Figure 24a). The formation of the aggregate (SP-3) is also confirmed by UV-Vis spectroscopy (Figure 24b).

SAXS Experiments
The obtained 1D SAXS profile for poly-2 shows a single peak at around 4.0 nm. This value is in close to that obtained by AFM for the helical pitch, confirming that the 3D model propossed is a good approximation.
In the case of poly-3, four peaks can be observed. The smallest values -2.859 and 1.858 nm-correspond to the non covalent interactions established between pendants, mainly hydrogen bonding and π-π stacking. The largest values -5.560 and 12.928 nm-are ascribed to the helix width and the helical pitch, respectively. This data is coincident with the one obatined for the 3D model of poly-3.

Theoretical Calculations
Considering the difficulties to carry out ECD theoretical calculations on large polymers, representative oligomers were used.
On the one hand, in the case of poly-2 an oligomer with n= 8 -where n denotes the number of monomer repeating units (mru)-was employed and, in order to reduce the computational demands, the long alkyl chains were replaced by methyl groups. The number of monomer units was selected considering the results of previous studies, S3-S5 where we evaluated the spectra for a series of poly(phenylacetylene) (PPA) oligomers obtained through systematic increase of monomer units, and concluded that 8-10 monomers were enough to describe the n+2 polymer ECD spectra. The starting structure of poly-2 was built through adjustment to the experimental data obtained from structural techniques, such as AFM and UV-Vis spectroscopy, defining the four different dihedral angles needed to build up the helical scaffold (ω 1 , ω 2 , ω 3 and ω 4 ; see Figure S25). Additionally, the pendant groups were introduced in the most stable conformation (ap). The oligomer geometry was optimized using the DFT method S6 together with the B3LYP-D3 functional S7 and the 6-31G** basis set. S8 Figure S27. Main dihedral angles involved in the helical structure for PPAs and derivatives.
On the other hand, for poly-3 the chiral moiety was removed from the pendant units in order to achieve a longer oligomer (n= 20, 504 atoms). In this case the length of the oligomer is essential to visualize the helices described by the OPE pendants (helix 3 and helix 4).
The ECD computational methodology was selected according to the size of polymers under investigation. Taking this into account, to evaluate the theoretical spectra time dependent density functional theory (TD-DFT), S9 in combination with the CAM-B3LYP functional S10 and the 3-21G basis set, S11 have been used. The ECD calculations were performed with the ORCA program (including 80 excitations). S12 The Gabedit S13 code was used to plot the spectra and the 39 density differences were displayed with Avogadro. S14 Furthermore, the full width at half height (FWHM) was fixed to 20.0 nm and the ECD were plotted with Gaussian curves.
For an efficient comparison and taking into account the tendency of the TD-DFT method to overestimate the excitation energies, the wavelength and intensity at the maximum/minimum Cotton effect correspondent to the polyene backbone in the theoretical spectra were adjusted to the experimental spectra. Employing the same correction factors, the lambdas were shifted and the intensities rescaled. The resulting ECD spectra are in good agreement with the experimental ones.
Poly-2 (ω 1 ca. 165º) displays a classical ECD trace with a positive Cotton effect at ca. 405 nm, dominated by the S 0 to S 1 excitation and ascribed to the polyenic backbone ( Figure S29). To get more insight into these spectral bands, the electron density differences for the corresponding transitions were evaluated at the same level of theory ( Figure S30).  Figure S30. Electron density differences with respect to the ground state for (a) S0 to S1, (b) S0 to S4, (c) S0 to S5, (d) S0 to S8, (e) S0 to S9, (f) S0 to S11, (g) S0 to S15 and (h) S0 to S16. An isovalue of 0.0005 was selected.

41
Poly-3 (ω 1 ca. 170º) ECD shows a highly intense positive Cotton effect at ca. 340 nm, dominated by the S 0 to S 1 excitation and ascribed to the polyenic backbone. The excited states that mostly contribute to the Cotton bands are depicted in Figure S31. To gain insight into the spectrum, the corresponding electron density differences were evaluated at the same level of theory ( Figure S32). 42 a) b) c) d) Figure S32. Electron density differences with respect to the ground state for (a) S0 to S1, (b) S0 to S4, (c) S0 to S9, and (d) S0 to S25. An isovalue of 0.0005 was selected.