Molecular Engineering for Enhanced Second-Order Nonlinear Response in Spontaneously-Oriented Evaporated Organic Films

TL;DR

This study demonstrates that molecular engineering of TPA-QCN derivatives can enhance second-order nonlinear responses in spontaneously oriented evaporated films without external poling. By combining design strategies that boost molecular hyperpolarizability with approaches that promote favorable out-of-plane orientation, the authors show that orientation control—enabled by intramolecular bridge-locking and related steric effects—dominates the observed χ^{(2)} improvements. Compound 4 achieves a twofold (off-resonant) to threefold (near-resonant) enhancement relative to the parent molecule, with 41.6–44 pm V^{-1} at 1266 nm, approaching performance of some inorganic materials while remaining poling-free. The work also highlights the importance of thermal stability and deposition conditions, and suggests future optimization of both orientation and stability to enable robust, scalable integrated nonlinear photonic devices. Key contributions include a detailed structure–property analysis combining DFT, surface-potential, and optical anisotropy measurements to decouple orientation effects from intrinsic hyperpolarizability, and a practical demonstration that spontaneous orientation can be steered to maximize χ^{(2)} in poling-free organic films.

Abstract

Materials with large second-order nonlinearities are crucial for next-generation integrated photonics. Spontaneously oriented organic thin films prepared by physical vapor deposition offer a promising poling-free and scalable approach. This study investigates molecular engineering strategies to enhance the second-order nonlinear response of derivatives based on the donor-acceptor molecule 2-(4'-diphenylaminobiphenyl-4-yl)quinoxaline-6,7-dicarbonitrile (TPA-QCN). Four derivatives incorporating modifications designed to increase molecular hyperpolarizability ($β$) or promote favorable orientation were synthesized and characterized. The most successful modification, intramolecular bridge-locking, simultaneously increases hyperpolarizability and enhances spontaneous orientation by reducing detrimental electrostatic interactions during deposition. It leads to a significant enhancement of the second-order nonlinear response, achieving off-resonance $χ^{(2)}_{31} \approx 16$ pm V$^{-1}$ and $χ^{(2)}_{33} \approx 18$ pm V$^{-1}$ at 1550 nm, a twofold improvement over the parent TPA-QCN. Analysis combining nonlinear optical measurements, surface potential measurement, optical anisotropy, and density functional theory calculations indicates that improved molecular orientation, rather than increased $β$ alone, is the primary driver for the enhanced performance in the leading derivatives. These findings demonstrate the effectiveness of targeting molecular orientation via structural design and position spontaneously oriented organic films as compelling poling-free candidates for integrated nonlinear photonic applications where the increased electrode-induced optical losses, fabrication complexity and footprint are a critical limitation.

Figures (6)

• Figure 1: Investigated donor(D)-acceptor(A) molecules in the study. D group is in blue, A group is in red, $\pi$-linker is in green.
• Figure 2: (a) Schematic of a D-A molecule on a substrate defining the tilt angle ($\theta$) and azimuthal angle ($\phi$). The dependence on the azimuthal angle vanishes due to isotropy of the in-plane direction Yokoyama_2011. The black arrow represents the direction of the permanent dipole moment, from a negatively-charged region to a positively-charged region. (b) Pictorial representation of a thin film with D-A molecules and how molecular packing configurations can lead to the presence or absence of a net out-of-plane contribution. In the configuration with zero net contribution, the molecules are arranged in a balanced anti-parallel fashion. A purely horizontal parallel configuration is not observed macroscopically. Due to the in-plane rotational symmetry of the deposition (an amorphous substrate, a homogeneous in-plane environment, and substrate rotation), there is no preferred in-plane direction. Therefore, the in-plane components must average to zero over a macroscopic area, resulting in a film with overall $C_{\infty v}$ symmetry.Yokoyama_2011. In the non-zero contribution scenario, parallel alignment leads to a net out-of-plane contribution(required for $\chi^{(2)}$).The net out-of-plane contribution in the schematic of the film (bottom) is exaggerated to clearly illustrate a non-centrosymmetric arrangement.
• Figure 3: Linear and nonlinear optical properties. (a) Normalized absorption spectra (b, c) SHG polarimetry data (TE and TM pump) at $\lambda_{pump}=1550$ nm, incidence angle 45$^\circ$. Field strengths normalized by $E_{inc}^2 \times L$. Peak intensities: 30-100 GW/cm$^2$. Points: data; Lines: fits from nonlinear transfer matrix. On panel (b), data and fits from compounds 1, 2 and 3 are overlapping. On panel (c), data and fits from compounds 3 and 5 are overlapping.
• Figure 4: Measured second-order susceptibilities at (a) $\lambda_\text{pump}=1266$ nm and (b) $\lambda_\text{pump}=1550$ nm. Plots show $\chi^{(2)}_{33}$ vs. $\chi^{(2)}_{31}$. Points and shaded regions represent weighted means and uncertainties. Dashed line: $\chi^{(2)}_{33}/\chi^{(2)}_{31} = 1$. Note the significant enhancement for 4.
• Figure 5: (a) Second-order nonlinear susceptibility normalized by both the molecular hyperpolarizability, number density and field enhancement factor ($\chi^{(2)}_{ij}/(NF\beta_{zzz,0})$). The dashed line indicates $\chi^{(2)}_{33}/\chi^{(2)}_{31} = 1$. (b) First-order orientational moment, $\langle cos \theta \rangle$, calculated from the measured surface potential and number density (see Supporting Information S8).(c) Optical anisotropy parameter S, derived from the peak values of the unaxial imaginary refractive index components ($k_{\text{ord,max}}, k_{\text{ext,max}}$ as defined in the panel equation) and reflecting the average TDM orientation (S=0: isotropic, S=-0.5: horizontal, S=1: vertical).