Novel approach to enhance X-ray reflectivity for liquid surface studies

Investigating atomic and molecular processes on liquid surfaces is crucial for advancing fundamental surface science and practical applications in physics, chemistry, and biology. However, achieving the sub-nanometre precision required for such studies is challenging with conventional experimental methods, making synchrotron-based X-ray scattering indispensable.

XRR is a key technique that measures the intensity of X-rays reflected from a surface versus grazing angle variation. This data helps determine the surface’s electron-density profile, enabling researchers to characterize liquid interfacial roughness, mimic cell membranes at the air-aqueous interface, and study layered liquid-metal surfaces.

Performing XRR on liquid surfaces is particularly challenging because neither the liquid sample nor the synchrotron source can be tilted. This limitation complicates the adjustment of the X-ray beam’s grazing angle (μ), which is necessary to vary the scattering vector component perpendicular to the surface, denoted as q_z = 4π λ^-1sin(μ) (where λ is the X-ray wavelength).

Traditional methods employ mirrors or single/double crystal deflectors to tilt the X-ray beam [1]. While mirrors restrict the maximum achievable q_z value to several critical angles, single-crystal deflectors (SCD) significantly extend this range but require moving the sample, which risks disturbing the liquid surface.

A more stable solution involves a double-crystal deflector (DCD) [2], which uses double Bragg reflection from two crystals. This setup allows the crystals to rotate as a rigid body around the incident beam axis without moving the sample, thereby enabling more stable measurements (Figure 1). However, even with the standard use of DCDs, the maximum achievable incident grazing angle is often insufficient for studying certain liquid metals, such as copper, which require higher q_z values to observe features like surface layering peaks.

Click image to enlarge

Fig. 1: A geometric sketch of the side view (vertical plane) of the DCD crystal assembly and a sample at the DCD position for μ_max.

The new technique developed at beamline ID10 addresses this limitation by employing three-times higher-order reflections from the same crystal set, Ge(333)/Ge(660), instead of conventional Ge(111)/Ge(220), and by tripling the X-ray energy. This approach effectively triples the maximum q_z value, significantly extending the range of measurable surface structures without requiring any modifications to existing DCD setups.

Although the X-ray beam intensity decreases when using Ge(333)/Ge(660) reflections, the powerful ESRF-EBS beam compensates for this loss. The narrower Bragg peaks from Ge(333)/Ge(660) necessitate precise DCD alignment, a process that is thoroughly detailed in the study.

To demonstrate the technique’s capabilities, researchers recorded XRR curves from bare liquid copper and liquid copper covered with a graphene monolayer in situ at 1400 K inside a customized portable CVD reactor in a CH₄/H₂/Ar atmosphere [3,4]. Figure 2a shows the total scattering signal plotted as a function of q_z.
 

Click image to enlarge

Fig. 2: a) A plot of the total scattering intensity (diffuse scattering and scattering from the bulk of liquid copper) as a function of q_z recorded in the new DCD configuration in situ from bare liquid copper (orange curve) and graphene-covered liquid copper (blue curve) at 1400 K inside a portable CVD reactor in a CH₄/H₂/Ar atmosphere, compared with similar measurements performed in the standard configuration (grey solid and dotted curves, respectively). b) Specular rod (00q_z), obtained after subtracting the diffuse background from the total scattering intensity of bare liquid copper (orange symbols) and graphene-covered liquid copper (blue symbols) at 1400 K, compared with the corresponding data obtained with the conventional DCD setup (grey symbols).

The data show distinct features: a first-order peak at q_z = 30 nm^-1 and a broad second-order peak up to 70 nm^-1 for bare copper. These two broad peaks arise from the bulk liquid structure. For the graphene-covered copper, the curve is measured only for q_z < 40 nm^-1 because from 25 nm^-1 onwards, the measured signal is dominated by scattering from the bulk of liquid copper. After subtracting the diffuse background, the reconstructed specular rod intensity is plotted in Figure 2b. In contrast to bare copper, the graphene-covered sample exhibits a pronounced minimum at q_z = 8 nm^-1, consistent with previous studies [4].

The ability to measure q_z values up to three times higher enables detailed studies of surface layering in liquid metals, previously unachievable with older methods. This advancement significantly enhances the potential for exploring liquid surfaces and interfaces across various scientific fields, offering a powerful tool for detailed material characterization. The technique’s compatibility with existing synchrotrons equipped with DCDs makes it accessible for immediate use in advanced research settings.

Principal publication and authors
Tripling of the scattering vector range of X-ray reflectivity on liquid surfaces using a double-crystal deflector, O. Konovalov (a), V. Rein (a,b), M. Saedi (c), I.M.N. Groot (d), G. Renaud (b), M. Jankowski (a), J. Appl. Cryst. 57, 258-265 (2024); https://doi.org/10.1107/S1600576724000657
(a) ESRF
(b) Univ. Grenoble Alpes, CEA, IRIG/MEM/NR, Grenoble (France)
(c) Physics Department, Shahid Beheshti University, Tehran (Iran)
(d) Leiden Institute of Chemistry, Leiden University, Leiden (The Netherlands)

References
[1] P.S. Pershan & M. Schlossman, Liquid Surfaces and Interfaces: Synchrotron X-ray Methods, Cambridge University Press (2012).
[2] V. Honkimaki et al., J. Synchrotron Rad. 13, 426-431 (2006).
[3] M. Saedi et al., Rev. Sci. Instrum. 91, 013907 (2020).
[4] M. Jankowski et al., ACS Nano 15, 9638-9648 (2021).