Resonant Soft X-ray Scattering Reveals the Distribution of Dopants in Semicrystalline Conjugated Polymers

The distribution of counterions and dopants within electrically doped semicrystalline conductive polymers, such as poly(3-hexylthiophene-2,5-diyl) (P3HT), plays a pivotal role in charge transport. The distribution of counterions in doped films of P3HT with controlled crystallinity was examined using polarized resonant soft X-ray scattering (P-RSoXS). The changes in scattering of doped P3HT films containing trifluoromethanesulfonimide (TFSI−) and 2,3,5,6-tetrafluoro-7,7,8,8-tetracyanoquinodimethane (F4TCNQ•−) as counterions to the charge carriers revealed distinct differences in their nanostructure. The scattering anisotropy of P-RSoXS from doped blends of P3HT was examined as a function of the soft X-ray absorption edge and found to vary systematically with the composition of crystalline and amorphous domains and by the identity of the counterion. A computational methodology was developed and used to simulate the soft X-ray scattering as a function of morphology and molecular orientation of the counterions. Modeling of the P-RSoXS at N and F K-edges was consistent with a structure where the conjugated plane of F4TCNQ•− aligns perpendicularly to that of the P3HT backbone in ordered domains. In contrast, TFSI− was distributed more uniformly between domains with no significant molecular alignment. The approach developed here demonstrates the capabilities of P-RSoXS in identifying orientation, structural, and compositional distributions within doped conjugated polymers using a computational workflow that is broadly extendable to other soft matter systems.

Methods
Materials and Processing Conditions. Low number-average molecular mass regioregular (RRe) P3HT [4.5 kg/mol, dispersity (Đ) = 1.6], high molecular mass RRe P3HT (23 kg/mol, Đ = 1.8), and regiorandom (RRa) P3HT (16.3 kg/mol, Đ = 2.5) (Figure 2), were dissolved in an equal-volume mixture of chlorobenzene and dichlorobenzene. P3HT solutions of varying blend composition were drop cast onto substrates to form (700 to 1000) nm thick films. P3HT films were doped in an inert nitrogen glovebox atmosphere, enclosing the film and ≈3 mg of F4TCNQ crystals within a jar in the orientation depicted in Figure 3.36 The jar was heated to 200 °C for 45 min, allowing for sublimation of F4TCNQ to oxidize the P3HT film. TFSI−-containing samples were formed by anion exchange from immersion in a concentrated LiTFSI solution (0.03 mass fraction in acetonitrile) for 120 min at 60 °C. Table S1 in the corresponding article summarizes the varying P3HT blend compositions utilized in this study, resultant levels of aggregation, and dopant counterion concentrations.

Atomic Force Microscopy. An Asylum MFP-3D atomic force microscope was used in tapping mode to examine the surface roughness of the P3HT blend films. The AFM tip was a silicon cantilever with a resonance frequency of approximately 61 kHz and a spring constant of about 1.6 N/m.

UV−Vis Absorbance Spectroscopy. All ultraviolet−visible (UV−vis) spectra were acquired using an Agilent Technologies Cary 60 UV−vis spectrometer. Samples were drop cast from solution onto quartz substrates to form optically transparent films. Spectra for P3HT films of varying composition were fit to the Spano model via a custom Python script to quantify aggregate mole fractions.

X-ray Photoelectron Spectroscopy. X-ray photo-electron spectroscopy (XPS) measurements were performed using an Escalab Xi+ Spectrometer from ThermoFisher Scientific. The spectrometer operated under a high vacuum condition of 10−6 Pa and utilized a monochromatic aluminum Kα X-ray source. To stabilize charge during the measurements, we used a dual ion-electron low-energy flood source. For acquiring survey spectra, we set the pass energy to 100 eV and conducted five scans at intervals of 0.25 eV, each with a dwell time of 50 ms. Depth profiling was done using an ion gun with a 1000-atom Ar+ cluster and an ion energy of 6000 eV. Ion sputtering covered a square region measuring (1.5 × 1.5) mm2. Within this area, we collected photoexcited electrons from the inner (400 × 400) μm2 region to selectively isolate signal from crater centers. All spectra are presented in the Supporting Information along with an AFM images of the topography of the films.

Grazing Incidence Wide Angle X-ray Scattering. Grazing incidence wide-angle X-ray scattering (GIWAXS) was performed at the BioPACFIC MIP user facilities at UC Santa Barbara and experimental station 11-3 at the Stanford Synchrotron Radiation Lightsource. Angle-resolved GIWAXS scans were acquired with 120 s exposures at grazing incidence angles of 0.05°, 0.10°, and 0.13° using an X-ray energy of 12.7 keV. 2D detector images were remapped to q-space using Nika and the WAXSTools Igor packages. Partial pole figure analysis was done using GIWAXS_Tools, a custom open-source software. 

Resonant Soft X-ray Scattering. Polarized resonant soft X-ray scattering (P-RSoXS) experiments were performed at the Spectroscopy Soft and Tender (SST-1) beamline funded and operated by the National Institute of Standards and Technology (NIST) at the National Synchrotron Light Source II (NSLS-II). Data reduction was performed using PyHyperScattering, an open source package for hyperspectral scattering reduction and analysis. Thin film samples on transparent silicon nitride windows were mounted normal to the incident X-ray beam with samples measured in transmission mode under high vacuum conditions.

Computation of Near Edge X-ray Absorption Fine Structure. Near edge X-ray absorption fine structure (NEXAFS) simulations were carried out using the PWscf and XSpectra software packages of the Quantum ESPRESSO distribution. The simulation process consists of (1) sourcing equilibrated or equilibrating atomic coordinates for a given molecule, (2) obtaining the electronic structure for each core-hole configuration of the molecule, and calculating the polarization-dependent X-ray absorbance spectra for each core-hole configuration. The configuration-specific spectra are offset by their relative total energies. The sum of spectra for each polarization direction are offset to experimental absorption onsets (e.g., the C 1s → π*C=C peak measured at 285.25 eV) to obtain oriented NEXAFS. The NEXAFS are normalized to the bare atom scattering factors to obtain the imaginary component of the refractive indices, β, which can be used to solve for the real portion, δ, using the Kramers−Kronig relations. To calculate P3HT NEXAFS, atomic coordinates for unit cells of low-energy crystalline polymorphs of P3HT were sourced from literature. We adopt the approach of using a supercell consisting of 3-hexylthiophene 8-mer with periodic boundary conditions to represent a single polymer chain. This is consistent with prior work demonstrating that 6 repeat units is sufficient to isolate adjacent core-hole excitons. Our tests also confirmed that π-stacking effects are minimal and that k-point sampling density variations produce negligible spectral changes (Figures S10−S13). This further confirms that the chosen supercell is sufficient to capture key attributes of the simulated NEXAFS. 

For the simulations involving dopant counterions, atomic coordinates were geometrically optimized through a relaxation calculation in Quantum ESPRESSO. A single dopant molecule within a sufficiently large cubic lattice was used to ensure the isolation of core-hole exciton effects. The optimization process employed the generalized gradient approximation (GGA), following the Perdew−Burke−Ernzerhof (PBE) scheme, and utilized a plane-wave cutoff energy of 30 Ry. 

P-RSoXS Simulations. P-RSoXS simulations were carried out using the NIST RSoXS Simulation Suite (NRSS) which incorporates tools to validate input models and CyRSoXS, a virtual beamline instrument. Simulated morphologies were generated using DopantModeling, a custom open-source software developed specifically for this work.