Topographic measurement of buried thin-film interfaces using a grazing resonant soft x-ray scattering technique

The internal structures of thin films, particularly interfaces between different materials, are critical to system properties and performance across many disciplines, but characterization of buried interface topography is often unfeasible. In this work, we demonstrate that grazing resonant soft x-ray scattering (GRSoXS), a technique measuring diffusely scattered soft x rays from grazing incidence, can reveal the statistical topography of buried thin-film interfaces. By controlling and predicting the x-ray electric field intensity throughout the depth of the film and simultaneously the scattering contrast between materials, we are able to unambiguously identify the microstructure at different interfaces of a model polymer bilayer system. We additionally demonstrate the use of GRSoXS to selectively measure the topography of the surface and buried polymer-polymer interface in an organic thin-film transistor, revealing different microstructure and markedly differing evolution upon annealing. In such systems, where only indirect control of interface topography is possible, accurate measurement of the structure of interfaces for feedback is critically important. While we demonstrate the method here using organic materials, we also show that the technique is readily extendable to any thin-film system with elemental or chemical contrasts exploitable at absorption edges.

Figure 1

Methods to measure buried interfaces. An illustration showing the difficulty of measuring a buried internal interface in a bilayer by established techniques. Red highlighted areas illustrate the locations of sensitivity. (a) X-ray and neutron reflectivity only reveal composition gradients normal to the interface and are insensitive to in-plane correlation lengths. (b) Many grazing-incidence probes can isolate the surface and probe the bulk, but cannot readily distinguish individual layers. (c) Neutrons probe with a single fixed contrast for a given sample, but through methods such as deuteration, sensitivity to the interface can be greatly increased. (d) Using GRSoXS, on a single sample, each of the signals from different sites can be controlled by choice of x-ray energy.

Figure 2

Optical constants and contrast for Experiment 1. (a) Real deviation from 1 and imaginary components of the index of refraction for PMMA and PEG. Inset are molecular structures of (left) PEG and (right) PMMA. (b) Contrast for the top surface and interface.

Figure 3

Experiment 1 XEFI calculation and schematic. (a) A 3D semitransparent isosurface plot of calculated XEFI vs energy, incident angle, and depth for the PMMA PEG system. Isosurface values are 1% (yellow)–100% (black) of the incident intensity. The slices of this space corresponding to the interface (at a depth of 400 nm) and 2.5° incident angle are outlined (see Supplemental Material [28]). (b) Schematic of acquisition geometry with example raw scattering pattern and the PMMA PEG sample. The directly reflected x rays are stopped by the beam stop, while the off-specular scattering is captured in the 2D detector.

Figure 4

Identifying scattering elements. (a) A normalized and background-corrected exposure acquired at 281.6 eV at the high-$q_{xy}$ detector location. The dotted yellow lines correspond to the limits of vertical integration which result in (b) scattering intensity and fit as function of momentum transfer. (Top panel) Scattering intensity at 281.6 eV and fit with Gaussian components corresponding to four dominant ξ's. (Middle panel) Measured scattering intensity at all energies and (bottom panel) resulting fit. (c) Intensity fit at 680 and 37 eV with scaled simulated scattering intensity from the surface and interface. (d) Measured and simulated interface sensitivity. Error bars are calculated from normalization uncertainty (see Supplemental Material [29]). (c) and (d) are displayed with energy on the vertical axis for direct comparison with the data in (b). Dotted red and blue lines are guides to the eye where the interface and surface scattering are highest in energy and momentum transfer. Dashed black lines across the bottom of the figure separate energy regions where the surface or interface sensitivity dominates.

Figure 5

Organic field effect transistor results. (a) Experimental and simulated scattering intensities for the OTFT sample. (b) experimental and simulated interface sensitivities for the OTFT sample. (c) Raw scattering intensities at the index-matched energies for the surface and interface before and after annealing. (d) A schematic of the system illustrating the surface sensitivity at 285.5 eV nm and interface sensitivity at 282 eV.