Truncated conjugation in fused heterocycle-based conducting polymers: when greater planarity does not enhance conjugation

One of the main assumptions in the design of new conjugated polymer materials for their use in organic electronics is that higher coplanarity leads to greater conjugation along the polymer backbone. Conventionally, a more planar monomer structure induces a larger backbone coplanarity, thus leading to a greater overlap of the carbon π-orbitals and therefore a higher degree of π-electron delocalisation. However, here we present a case that counters the validity of this assumption. Different diselenophene-based polymers were studied where one polymer possesses two selenophene rings fused together to create a more rigid, planar structure. The effects of this greater polymer coplanarity were examined using Raman spectroscopy and theoretical calculations. Raman spectra showed a large difference between the vibrational modes of the fused and unfused polymers, indicating very different electronic structures. Resonance Raman spectroscopy confirmed the rigidity of the fused selenophene polymer and also revealed, by studying the excitation profiles of the different bands, the presence of two shorter, uncoupled conjugation pathways. Supported by Density Functional Theory (DFT) calculations, we have demonstrated that the reason for this lack of conjugation is a distortion of the selenophene rings due to the induced planarity, forming a new truncated conjugation pathway through the selenophene β-position and bypassing the beneficial α-position. This effect was studied using DFT in an ample range of derivatives, where substitution of the selenium atom with other heteroatoms still maintained the same unconventional conjugation–planarity relationship, confirming the generality of this phenomenon. This work establishes an important structure–property relationship for conjugated polymers that will help rational design of more efficient organic electronics materials.


Experimental Section
Films were fabricated by drop-casting a 5 mg/mL chloroform solution of the HPMI and HPPI polymers. HPMI and HPPI were synthesized following the procedure described in our previous work. 1 Ground state absorbance. Ground state absorbance was obtained with a Perkin Elmer Lambda 365.
Raman. Raman spectra were recorded using a custom-made back scattering setup with an Andor iDus 416 CCD camera attached to an Andor Shamrock 500i spectrograph. Raman signal was generated via excitation with a 6 ns, 10 Hz Nd:YAG laser (Spectra-Physics, INDI-  for the excitation pulse. The intensity of the Raman excitation was decreased with the use of neutral density filters to maintain it at 0.1 mW, measured with an ES111C sensor (Thorlabs). The excitation wavelength was selected with a versaScan L-532 OPO and the appropriate notch filters were used in front of the spectrograph slits (200 mm).

Theoretical Section
The density functional theory (DFT) method was chosen to optimise the electronic ground state of the trimers HPMI and HPPI with the global hybrid B3LYP functional as well as long range corrected functionals such as CAM-B3LYP and ωB97X-D3. We performed optimisations with the 6-31G** and 6-311** basis sets. Frequencies were calculated at the B3LYP/6-31G** level of theory.
To decrease the calculation time and expense, the comparison between dimer, trimer and tetramer have been performed at the B3LYP/6-31G* level (Fig. S2).
Potential energy scans around the Se-C-C-Se dihedral bond that connects two monomeric units were carried out using B3LYP/6-31G** in both HPMI and HPPI trimers. For each manually fixed Se-C-C-Se torsion angle (from -180 to 180 degrees) of one of the terminal monomeric units, a geometry optimisation has been performed at the B3LYP/6-31G** level of theory, letting the other degrees of freedom reach their minimal energy conformation.
Vertical energies at the optimised geometries were computed with linear response time dependent density functional theory (LR-TDDFT) using the B3LYP and the 6-31G** and 6-311** basis sets.
The simulations were carried out with QChem5.4.   S2. Calculated orbitals for HOMO and LUMO levels for HPPI (top) and HPMI (bottom) dimer, trimer and tetramers using B3LYP/6-31G* level TD-DFT calculations. Energy LUMO (purple) and HOMO (green) levels for those molecules are also displayed. Table S2. Se-C-C-Se torsional angles in degree from trimer optimised structures calculated with different functionals and basis sets. TZ corresponds to the triple zeta Pople basis set 6-311G** and DZ to 6-31G**. CAM is an abbreviation for CAM-B3LYP. There are two torsional angles per trimer, corresponding to the two Se-C-C-Se torsional angles that link the three maleimide monomer units.

Raman Characterisation
The most intense HPPI experimental Raman band (1528 cm -1 ) corresponds to the calculated vibrational modes at 1512 and 1519 cm -1 . These calculated bands correspond to the C=C stretch along the selenophene backbone. Furthermore, the calculated 1519 cm -1 mode also shows a strong contribution from the central fused benzene ring CC stretching corresponding to the bond linking the selenophene and the maleimide, while the 1512 cm -1 mode contains terminal benzene stretches. The latter is less likely in a long-chain polymer, and thus we expect the experimental band to be dominated by the central benzene stretch combined with the selenophene C=C stretch. The bands in the range from 1200 to the 1300 cm -1 belong to vibrational modes associated with the benzene ring in the HPPI. Particularly, the experimental band at 1272 cm -1 was assigned to the bond that fuses the selenophenes (C3-C11), which is obviously not present in the HPMI polymer. The experimental band at 1575 cm -1 was assigned to the calculated band seen at 1543 cm -1 , which related to CC stretching of the benzene bonds that connect the maleimide moiety with the selenophenes. This vibrational mode was also contributing to the most intense HPPI band.
HPMI showed a very different Raman spectrum due to the absence of the fused-ring CC bond (C3-C11), thereby negating benzene-related vibrational contributions to the modes. HPMI shows its most intense band at 1390 cm -1 , associated with the calculated selenophene C-C stretch at 1389 cm -1 . The band experimentally seen experimentally at 1207 cm -1 , corresponding to the calculated vibrational mode at 1183 cm -1 , was associated to the C-C stretch of the bond that connect HPMI monomers (C13-C14). This mode has similar position and intensity to HPPI (experimentally seen at 1184 cm -1 ).
A remarkable difference between HPPI and HPMI Raman spectra are the bands in the 1500-1600 cm -1 region. The HPPI has high and medium intensity bands that were associated with the C=C stretch of the selenophene conjugation and the bonds connecting the maleimide and selenophene (C8-C9 and C5-C6), respectively. The HPMI band seen experimentally in this region at 1509 cm -1 was associated with the stretch of the maleimide double bond (C5-C9 and C18-C18b). The band associated with this same vibration in the HPPI appears in the calculated spectrum at 1468 cm -1 , largely displaced to lower frequencies with negligible intensity. This large difference in both intensity and displacement means that fusing the selenophenes has a huge impact in the maleimide vibrational force constant.
Another difference with the HPPI polymer is the band seen at 1348 cm -1 in the HPMI experimental Raman spectrum, associated with the selenophene breathing. There was no equivalent vibrational mode for HPPI in the spectrum for the calculated range. This lack of vibrational breathing mode in the HPPI suggests a lack of aromaticity on the HPPI selenophene ring in comparison with the HPMI.
In addition, a new band rises at 1254 cm -1 associated with the maleimide stretches of the terminal monomers when exciting at shorter wavelengths. At shorter wavelengths the shorter length polymers are going to be excited, and therefore this vibrational mode has higher intensity because the terminal monomers are going to represent a larger portion of the smaller chain polymers in comparison to their proportion in longer chain polymers. Scheme S1. Atom labels used for the HPPI and HPMI Raman band assignment.     Calculations were performed at B3LYP/6-31G** theory level.

Calculation of the HOMA value
To calculate the HOMA value the following formula was used: 2 (1) where n was the number of members of the ring and R ij the calculated bond length for each of the bonds of the ring.  and R opt were obtained from literature, 2 where the authors calculated them using the following formulae: Where k is the force constant for the bonds, 2 and 1 for k (d) and k (s) , respectively. R (s) and R (d) are the reference single and double bonds.