Microfluidic devices for small-angle neutron scattering

Abstract

A comparative examination is presented of materials and approaches for the fabrication of microfluidic devices for small-angle neutron scattering (SANS). Representative inorganic glasses, metals, and polymer materials and devices are evaluated under typical SANS configurations. Performance criteria include neutron absorption, scattering background and activation, as well as spatial resolution, chemical compatibility and pressure resistance, and also cost, durability and manufacturability. Closed-face polymer photolithography between boron-free glass (or quartz) plates emerges as an attractive approach for rapidly prototyped microfluidic SANS devices, with transmissions up to ∼98% and background similar to a standard liquid cell (I ≃ 10^-3 cm^-1). For applications requiring higher durability and/or chemical, thermal and pressure resistance, sintered or etched boron-free glass and silicon devices offer superior performance, at the expense of various fabrication requirements, and are increasingly available commercially.

Keywords: closed-face polymer photolithography; lab-on-a-chip; microfluidic devices; small-angle neutron scattering.

Figure 1

Configurations of microfluidic SANS for phase mapping (top row) and flow processing (bottom row). (a) Schematic of a continuous-flow and a droplet mixer, where a large (∼10 mm) neutron beam (shown in red) illuminates a number (10–20) of microchannels. (b) Patterns of radially averaged scattering intensity of a surfactant solution at acquisition times down to 0.1 s (data shifted vertically for clarity). (c) Rapid contrast variation with a nanoparticle suspension in 120 steps of 10 s at a flow rate of 0.1 ml min⁻¹ (Adamo et al., 2017 ▸). (d) Serial dilution in droplet microfluidics of a nanoparticle suspension (Ludox HS-40) in 120 steps of 5 s at a flow rate of 0.075 ml min⁻¹ with a fluorinated oil carrier [adapted from Adamo et al. (2018 ▸) by permission of The Royal Society of Chemistry]. (e) Contraction–expansion geometry with a 500 µm neutron beam scanning the flow field. (f) Scattering intensity of a CTAC/pentanol/D₂O mixture at 5 min and 1 s acquisitions. (g) Two-dimensional and radially averaged scattering of SDS/brine/octanol in an opposing jet (extensional) geometry and (h) through the first constriction of the illustrated device; adapted with permission from Lopez et al. (2015 ▸) under a Creative Commons Attribution 4.0 International License, https://creativecommons.org/licenses/by/4.0/.

Figure 2

(a) Schematic of microdevice and beam configurations for (left) single-channel mapping with a beam of ∼100–1000 µm footprint and (right) overillumination of several channels, with a beam of ∼10 mm footprint. Representative configurations for the beam footprint and channel are given in panels (b)–(e), in both front view (top row) and cross section (bottom row).

Figure 3

The range of neutron transmissions for different classes of material. The top panel shows the transmission (T) per 1 mm of material, and the bottom panel shows the estimated transmission of a microdevice with a likely thickness t _D (detailed in the text).

Figure 4

Representative scattering profiles for three classes of material, inorganics, polymers and metals. The top row [panels (a)–(c)] shows experimental data in absolute units of cm⁻¹, while the bottom row [panels (d)–(f)] shows I multiplied by the typical thickness t _D (in centimetres) required for a device of the given material. For reference, the scattering from a Hellma cell (orange + symbols) is also included. The window materials (and thicknesses) considered are quartz (0.44 mm), fused silica (1.1 mm), soda lime glass (1.1 mm), borosilicate glass (0.14 mm), steel (0.1 mm) (Shin et al., 2010 ▸), nickel (10 µm × 2) (Yoo et al., 1982 ▸), aluminium (1 mm) and vanadium (1 mm) (Imae et al., 2011 ▸), shown as a horizontal line.

Figure 5

Schematic of microfabrication procedure (method 1) based on a closed-face lithography approach (Cabral et al., 2004; Harrison et al., 2004 ▸) and employed in flow-SANS (Lopez et al., 2015 ▸). Photopolymerization is carried out between two ∼1 mm (boron-free) glass sheets (left). If a thin (∼100 µm) slide is used instead, for instance to enable SAXS measurements as well as SANS, the sandwich is reinforced with a thick glass plate to ensure mechanical integrity (right). The channel geometry is imposed by a photomask and development (see text for details).

Figure 6

Schematic of PDMS–thiolene hybrid microdevice fabrication (method 2), combining method 1 with PDMS fabrication and sealing to enable higher density and smaller-scale patterning, prior to measurement.

Figure 7

(a) Crown glass chip (Dolomite microreactor) installed on the SANS beamline D22 at ILL, operating in (b) continuous and (c) droplet modes. (d) A top view of the setup, showing the beam and detector tubes. (e) A fused silica microchip fabricated by selective laser etching [reproduced with permission from LightFab (2018 ▸), copyright (2018) LightFab GmbH, Aachen, Germany].

Figure 8

System compatibility with microfluidic SANS. Acquisition times are plotted for representative systems based on their characteristic scattering intensity (cm⁻¹), sample volumes of 1 nl and 10 µl, and the q-range/configuration required [defined as high (0.05–0.5 Å⁻¹), mid (0.01–0.2 Å⁻¹) and mid/low (0.005–0.05 Å⁻¹) q]. Adapted and expanded from Lopez et al. (2015 ▸) under a Creative Commons Attribution 4.0 International License, https://creativecommons.org/licenses/by/4.0/.