Proceedings Article | 11 May 2011
KEYWORDS: Lenses, Crystals, Laser crystals, Refraction, Rayleigh scattering, Compton scattering, Collimation, Photons, Nuclear physics, Scattering
The upcoming γ facilities MEGa-Ray (Livermore) and ELI-NP (Bucharest) will have a 105 times higher γ flux
F0 = 1013/s and a ~30 times smaller band width (ΔEγ/Eγ = BW ≈ 10-3) than the presently best γ beam
facility. They will allow to extract a small γ beam of about 30 - 100 μm radius 1 m behind the γ production
point, containing the dominant γ energy band width. One can collimate the γ beam down to ΘBW =
√
BW/
γe
,
where γe = Ee/
mec2 is a measure of the energy Ee of the electron beam, from which the γ beam is produced by
Compton back-scattering. Due to the γ energy - angle correlation, the angular collimation results at the same
time in a reduction of the γ beam band width without loss of "good" γ quanta, however, the primary γ flux F0is reduced to about Fcoll ≈ F0 · 1.5 · ΔEγ/Eγ. For γ rays in the (0.1-100) MeV range, the negative real part δ of the
index of refraction n = 1- δ + iβ from coherent Rayleigh scattering (virtual photo effect) dominates over the
positive δ contributions from coherent virtual Compton scattering and coherent virtual pair creation scattering
(Delbrück scattering). The very small absolute value |δ| ≈ 10-6 - 10-9 of the index of refraction of matter for
hard X-rays and γ-rays and its negative sign—in contrast to usual optics—results in a very different γ-ray
optics, e.g. focusing lenses become concave and we use stacks of N optimized lenses. It requires very small radii
of curvature of the γ lenses and thus very small γ beam radii. This leads to a technical new solution, where the
primary γ beam is subdivided into M γ beamlets, which do not interfere with each other, but contribute with
their independent intensities. We send the γ beamlets into a two-dimensional array of closely packed cylindrical
parabolic refractive lenses, where N ≈ 103 lenses with very small radius of curvature are stacked behind each
other, leading to contracted beam spots in one dimension. With a second 1D lens system turned by 900, we can
obtain small spots for each of the beamlets. While focusing the beamlets to a much smaller spot size, we can
bend them effectively with micro wedges to e.g. parallel beamlets. We can monochromatize these γ beamlets
within the rocking curve of a common Laue crystal, using an additional angle selection by a collimator to reach
a strongly reduced band width of 10-4 - 10-6. We propose the use of a further lens/wedge arrays or Bragg
reflection to superimpose the beamlets to a very small total γ beam spot. Many experiments gain much from
the high beam resolution and the smaller focal spot. This new γ optics requires high resolution diagnostics,
where we want to optimize the focusing, using very thin target wires of a specific nuclear resonance fluorescence
(NRF) isotope to monitor the focusing for the resonance energy. With such beams we can explore new nuclear
physics of higher excited states with larger level densities. New phenomena, like the transition from chaotic to
regular nuclear motion, weakly-bound halo states or states decaying by tunneling can be studied. The higher
level density also allows to probe parity violating nuclear forces more sensitively. This γ optics improves many applications, like a more brilliant positron source, a more brilliant neutron source, higher specific activity of
medical radioisotopes or NRF micro-imaging.
