Skip to content

Focusing, collimation, and beam expansion of high-energy X-rays with diamond refractive lenses

FieldValue
JournalJournal of Synchrotron Radiation
Published2026-07-24
AuthorsS. D. Shastri, Sergey Antipov
DOI10.1107/s1600577526006995

A diamond compound refractive lens produced a 30 μm vertical focal spot at 50 keV on beamline 1-ID. At 42 keV, diamond collimation increased post-monochromator flux by a factor of 8.5. A convex prototype expanded a 0.39 mm beam to 0.80 mm at 50 keV. The work covers focusing, collimation, and expansion on two APS undulator beamlines.

At X-ray energies, the refractive index is slightly below one. Concave elements therefore focus, while convex elements diverge. Diamond’s high density provides enough refraction for shorter lens assemblies than beryllium above 30 keV. Single-crystal diamond also reduces scattering and decoherence.

Pulsed laser ablation made all focusing and collimating elements. CVD-2A and HPHT-grown single-crystal plates were 1 mm thick and 1 to 2 mm in diameter. Each plate was press-fitted into a machined CuCrZr frame before laser machining. The two-dimensional frames were disks 12.00 mm in diameter and 2.00 mm thick.

Two-dimensional elements used a paraboloid, while one-dimensional elements used a planar parabola. A 90° V-block and the frame edges aligned their optical axes. A tangential flat constrained the cylinder axes of one-dimensional elements. The block rotated the focusing direction by 90°.

Stacked elements used guide rods, compression springs, and a leaf spring to seat every frame edge. This allowed one-dimensional and two-dimensional elements to share one self-aligning mount. Half- and quarter-strength elements with 200 mm and 400 mm radii allowed focal-length fine-tuning.

CuCrZr offered hardness and thermal conductivity, but the tests used only low-power monochromatic beams. Its performance under high heat load was not measured.

At beamline 1-ID before the APS upgrade, 50 keV X-rays were focused with a CRL 38 m from the source. The focus was 70 m from the source, giving f = 17.37 m. Ten two-dimensional bi-concave elements had R = 100 mm. The calculated element count was N = 9.9, and the lens illuminated a 0.44 mm by 0.44 mm beam.

Profiles were measured with a scintillator detector, a 90° reflection mirror, a 5x objective, and a planar sensor. The vertical FWHM was 30 μm and the horizontal FWHM was 430 μm. The expected vertical size was 22 μm, so the measured focus was 36% larger. A saw-tooth refractive lens confirmed the vertical focus at the same location.

The horizontal size was 430 μm versus an expected 535 μm from source demagnification. The two-dimensional case gave a flux-density gain of 27, compared with an expected 36. One-dimensional focusing gave 16 against an expected 23 because it lacked horizontal focusing.

Polishing from Sa below 300 nm to below 30 nm produced no detectable change at 50 keV within detector sensitivity. CVD-2A and HPHT-2A elements also showed no difference. Five elements with R = 50 mm gave a nearly indistinguishable 30 μm vertical focus.

At beamline 20ID-HEXM after the upgrade, a 42 keV beam came from the fifth harmonic of a 4.7 m undulator. A Si(111) high-energy monochromator and a four-reflection Si(440) high-resolution monochromator followed. The first diffracted horizontally, and the second vertically. The high-resolution monochromator reduced ΔE/E to 4.8 × 10^-6 and had 740 nrad vertical angular acceptance.

A vertical collimating CRL sat 31 m from the source with f = 31 m. It used four one-dimensional R = 100 mm elements, while calculation gave N = 3.9. Measured transmission was 0.965, close to the predicted 0.960.

An HRM rotation scan measured a 1.2 mrad angular width, close to the expected 1.1 mrad. The reference combined the 740 nrad HRM acceptance with 770 nrad residual source divergence. Removing the collimator reduced post-HRM flux by a factor of 1/8.5.

Flux through the HRM fell by 1/113 relative to the collimated beam. This was consistent with the bandwidth reduction and indicated effective vertical collimation without an additional angular-acceptance penalty. The surviving thin beam had a vertical size of 147 μm, close to 134 μm from the 740 nrad acceptance at 181 m.

Final focusing used a CRL at 75 m to deliver a beam to 181 m. The horizontal focal length was 44 m. The vertical focal length was 106 m because the collimator had already removed vertical divergence.

Four elements formed the astigmatic stack. They combined a two-dimensional element with full, half, and quarter-strength vertical elements. The half and quarter elements had R = 200 mm and R = 400 mm. A planned 0.15-strength correction was omitted because the collimator slightly over-refracted.

Measured FL transmission was 0.90, compared with 0.92 predicted. The measured focus was 49 μm by 66 μm, versus 41 μm by 80 μm expected. Both differences were about 20%.

The larger horizontal size could come from the high-energy monochromator or the two undulator sections. The smaller vertical size implies a smaller electron-beam size or beta function than assumed.

Low-emittance radiation reached 181 m as a 2.1 mm by 2.4 mm beam. Tomography and grain mapping can benefit from beams up to 4 mm.

The proposed one-dimensional convex CRL sat at 33 m. At 65 to 70 keV, it used R = 100 mm and N = 11 to 12 bi-convex elements, with f = -27 m. Calculations at 70 keV predicted B = 1.88, a 4 mm beam, and 58% integrated efficiency.

A concave alternative used N = 18 to 19 R = 50 mm elements and had 60% efficiency. Its profile was less flat because off-axis attenuation narrowed the expanded beam. The convex design used a 0.4 mm aperture at 33 m, leaving 0.05 mm unilluminated at each edge. These were calculations, not prototype measurements.

The proof-of-principle test used beamline 1-ID before the upgrade. The optic was 38 m from the source, and measurements were made 70 m downstream. The target expansion required f = -17 m. Five bi-convex R = 50 mm elements were machined in one monolithic single-crystal device.

The expansion extent was 0.31 mm, and the pillar length was 0.78 mm. The illuminated vertical aperture was 0.21 mm after excluding 0.05 mm at each edge. Horizontal illumination was 0.6 mm.

Energy (keV)No expander FWHM (mm)With expander FWHM (mm)Measured BExpected B
411.002.562.502.50
500.390.802.052.00
600.671.721.721.70

Measured expansion fell as energy rose because diamond refraction weakened. The measured factors agreed with predicted values across 41 to 60 keV.

Profiles showed non-uniform intensity in the expansion direction. The authors linked this to residual correlated surface discontinuities from laser machining. SEM confirmation was not performed, so the attribution remained provisional.

Laser-machined surfaces still limited prototype image quality. Process parameters, polishing, and separate framed elements were proposed improvements. A monolithic two-dimensional convex device may not be feasible, while two orthogonal expanders would combine their on-axis attenuation.

Short focal lengths were not tested because high energy requires many CRL elements. Saw-tooth lenses were planned for focal lengths of 0.7 to 2.5 m. Manual stacking could also be replaced by actuator-driven insertion and removal.

For long-beamline systems, concave elements suit focusing and collimation, while convex elements suit expansion. Mixed-strength elements give practical control of astigmatic focal length. The illuminated aperture must account for each lens profile’s attenuation.

The paper’s laser-machined refractive lenses require single-crystal diamond plate stock. It tested 1 mm CVD-2A and HPHT-grown plates. 6CCVD’s diamond materials catalog lists MPCVD-grown single-crystal plates up to 125 mm, with thicknesses from 0.1 to 500 μm, so the paper’s 1 mm thickness requires a custom fabrication review through engineering support.

View Original Abstract

Diamond is extremely well suited for synchrotron X-ray refractive optics. The effectiveness of diamond refractive lenses, fabricated by pulsed laser ablation, is demonstrated here for the focusing, collimation, and expansion of high-energy, undulator-source X-ray beams in the 40-70 keV photon energy range. The design of the lens elements and their mounting scheme allow the convenient feature of stacking both one-dimensionally and two-dimensionally focusing elements together in a self-aligning manner within the same holder for astigmatic manipulation, when needed. The CuCrZr alloy frames embedding the diamonds are suitable for thermal management, while possessing hardness. Collimating lenses, placed following a high-heat-load monochromator, are employed to increase the throughput of a subsequent narrow-angular-acceptance high-energy-resolution monochromator, after which the X-rays are focused to deliver a beam to coherent diffraction experiments on a long beamline. Also motivated and proposed is the concept of a convex beam-expander, which might seem counterintuitive given that it has higher attenuation on-axis. However, for high-energy X-rays from a low-emittance source on a long beamline, convex diamond beam-expanders could be practical. A proof-of-principle test result of this idea is presented.