Microfluidic gratings for X-ray Phase Contrast Imaging

TL;DR

This work tackles the high cost and scalability barriers of X-ray grating fabrication for X-ray Phase Contrast Imaging (XPCI) by introducing a soft-lithography, Hg-filled microfluidic approach. The authors demonstrate a complete fabrication and validation workflow, deploying Hg-filled PDMS channels as absorbing septa in a Beam Tracking imaging setup with a relatively modest, unfocused X-ray source. The results show comparable visibility to Au-based absorbers, enabling attenuation, refraction, and scattering contrast with high-quality phase images, including 3D μ-CT of soft-tissue specimens. By offering a two-step, scalable fabrication path and potential for flexible or curved gratings, this method could substantially lower entry barriers for clinical and industrial deployment of XPCI.

Abstract

Fabrication of X-ray gratings has surged in the last two decades thanks to their vast employment in X-ray Phase Contrast Imaging, an imaging technique able to boost X-ray sensitivity to detect otherwise invisible details. These high aspect ratio devices are usually fabricated by complex, costly, multi-step processes that limit their size and volume scaling. These steps commonly involve UV or X-ray lithography, semiconductor selective etching and high-Z metal plating, usually Au, which require expensive tools and materials. Here we present a proof-of-concept fabrication via soft lithography and Hg infusion of microfluidic X-ray absorption gratings and their performance in biomedical imaging. Such fabrication technique requires fewer, less expensive, and more scalable processes using alternative and more sustainable materials, while showing comparable visibility with their conventional Au-based, solid equivalent. Our results constitute a promising shift in X-ray optics fabrication that could significantly lower barriers to commercialization and accelerate the practical deployment of X-ray Phase Contrast Imaging.

Figures (4)

• Figure 1: Fabrication process. a SU8 layer exposure to UV-photolithography using a photomask, resulting in a b) patterned SU8 mold. c Replication on PDMS via soft-lithography and (d) demolding. e Chip fabrication via bonding with glass and Hg pumping. f Microfluidic chip scheme of the design: a serpentine with two infusion/extraction channels to pump Hg in/out.
• Figure 2: a Optical microscopy of the SU-8 mold. Scale-bar: 50 $\upmu$m. b Profilometry measurement of SU-8 mold inlet channel. The protrusive $\sim$ 21 $\upmu$m channel is later imprinted in the PDMS chip. c Microfluidic chip during Hg pumping from the inlet to the serpentine through the loading channel (highlighted by black arrows). d Optical microscopy of the Hg-filled mask. Scale-bar: 50 $\upmu$m. e Attenuation coefficient profile of Hg channel (light purple shading) sandwiched between PDMS and glass (light blue shading), averaged over 1000 pixels and 10 channels. f Transmission-based radiography across multiple mm of the mask (pixel size = 5 $\upmu$m).
• Figure 3: a Scheme of a BT imaging setup. The purple source emits a conical beam split by the grey mask into individual, well-separated beamlets resolved by the black detector at the back. In presence of a sample, the beamlets are perturbed allowing for analysis and retrieval. The sample's attenuation, refraction, and scattering effects on the beamlets are reported in the inset on the left. b The attenuation-based radiography of three wires i.e., (from left to right) sapphire, boron with a W-core, and nylon. c Corresponding refraction-based image, highlighting the edges and the nylon wire, almost invisible in the attenuation-based image. d Comparison between experimental data (red points) and simulated values (black profiles). The normalized profile on the left represent three unperturbed beamlets. The plots in the middle and the right show transmission and refraction through a 250 $\upmu$m wide sapphire wire.
• Figure 4: a,b Cross-sectional slices of the mouse lung obtained from attenuation- and phase-based reconstructions, respectively. c,d Corresponding attenuation- and phase-based slices of the piglet esophagus. Scale-bars: 2 mm. e,f Volumetric reconstruction of the lung and esophagus, respectively. Scale-bar: 1 mm.