JAXA’s High energy-resolution X-ray spectroscopy

An image to illustrate x-ray spectroscopy

JAXA’s Kazuhisa Mitsuda showcases some of the Japanese Space Agency’s recent achievements in the field of X-ray spectroscopy and looks towards an M pixel class X-ray integral field.

Energy spectra (or, equivalently, wavelength spectra) deliver plenty of useful information about the X-ray emitting material and emission mechanism. In material analysis, an Energy Dispersive X-ray Spectrometer (EDS) and a Wavelength Dispersive X-ray Spectrometer (WDS) are standard apparatus attached to electron microscopes for elemental analysis. The semiconductor detectors used for EDS cover a wide energy area, e.g. 0.5 to 10 keV, and thus most elements can be detected.

However, the energy resolution is limited to 100 to 200 eV in full width half maximum (FWHM), and emission lines often overlap each other, which introduces systematic errors in element-concentration estimation. Grating spectrometers used for WDS have energy resolutions as good as 1 eV FWHM. However, the energy range coverage is limited (< a few keV with a single grating). Thus, X-ray spectrometers which have both a high energy resolution to separate emission lines from different elements (< 10 eV) and wide energy range coverage are highly desired. X-ray astronomy Energy spectra have long been one of the most important key measurements in astronomy.

In optical and infrared astronomies, integral field spectroscopy, namely imaging spectroscopy is becoming important and its enabling technologies are rapidly developed. In X-ray astronomy, high energy-resolution X-ray imaging spectroscopy is also highly demanded. However, only medium energy-resolution imaging spectroscopy and non-imaging high-resolution spectroscopy have been realised. In X-ray astronomy, reflecting X-ray focusing mirrors are used to collect photons and spectrometers are placed at the focal plane or in the X-ray path after the X-ray mirrors.

Energy-dispersive semiconductor detectors with imaging capability, such as Charge Coupled Devices (CCD), have become standard focal plane instruments since Japan’s X-ray astronomy satellite, ASCA (inserted into the orbit in 1993).1 Although its imaging capability is very good (> 1 M pixel), the energy resolution is limited to 100 to 200 eV. High energy resolution spectroscopy (FWHM ~1 eV at ~1 keV) in X-ray astronomy was obtained with wavelength-dispersive spectrometers, the Reflective Grating Spectrometer (RGS) on board ESA’s XMM-Newton observatory (1999)2 and the Low-Energy and High-Energy Transmission Gratings (LETG and HETG) on board NASA’s Chandra observatory (1999).3

The high-energy resolution is obtained only for point sources with both instruments.
An X-ray microcalorimeter, which is operated at a cryogenic temperature and detects an X-ray photon as heat,4 is presently the only promising technology for integral field spectroscopy at X-ray wavelengths. However, as an imaging device it is still in a primitive phase. For example, the Soft X-ray Spectrometer (SXS) onboard JAXA’s ASTRO-H observatory (2016)5 had only 36 imaging pixels. The key technology to increase the number of pixels is the signal multiplexing at a cryogenic temperature, since the X-ray microcalorimeters are operated at temperature lower than 100 mK. For the X-IFU (X-ray Integral Field Unit)6 on board ESA’s Athena mission, which will be inserted into the orbit early 2030’s, more than 3,000 pixels are planned by utilising the next generation X-ray microcalorimeter, TES (Transition-Edge sensor) type,7 and the frequency division multiplexing (FDM) at the cryogenic temperature.

Response time

TES X-ray microcalorimeter is also promising technology for an EDS of material analysis. However, its slow response time is problematic. It is faster than conventional X-ray microcalorimeters, but not as fast as semiconductor detectors, such as Silicon Drift Detectors (SDD). The response time limits the maximum counting rate and thus the minimum exposure time to obtain a statistically meaningful spectrum. As an EDS for an electron microscope, the imaging capability is not required because elemental maps can be obtained by correlating the X-ray photon detection time with the electron beam position. However, multi-pixel detectors similar to imaging devices are highly demanded in order to increase the counting rate.


In the Institute of Space and Astronautical Science (ISAS) of Japan Aerospace Exploration Agency (JAXA), we have been developing TES X-ray microcalorimeters and the related technologies for future space missions. In parallel, we have also worked on the ground applications. We have collaborated with several institutions in Japan for the research and development. In this article, we would like to show some recent achievements.

We developed an EDS system for Scanning Transmission Electron Microscope (STEM), Hitachi model HD-2700, in collaboration with National Institute of Material Science (NIMS), Kyushu University, Hitachi High-Tech Science Corporation, Hitachi High-Technologies Corporation, National Institute of Advanced Industrial Science and Technology (AIST), and Taiyo Nippon Sanso Corporation with financial support from Japan Science and Technology Agency (JST-Sentan). In the group, JAXA was mainly responsible for the development of the cold detector heads, including the TES X-ray microcalorimeter detector array and the data analysis.

The requirement for the EDS is an FWHM energy resolution of < 10 eV in energy range of 0.5 to 10 keV and the maximum counting rate of 5 k cps. These requirements are different from those for space missions, where energy resolution is typically 2 eV, but the maximum counting rate can be an order of 100 cps or even smaller. In space missions, the typical operating temperature of the TES is between 50 mK to 100 mK. To obtain faster response for each pixels, we selected relatively high operating temperature of ~150 mK with the sacrifice of energy resolution.8

Then, in order to fulfill the counting rate requirement with margin, we adopted 8×8 format 64 pixel array. In Figure 1, we show the picture of the TES microcalorimeter array (a), the STEM system (b), and a spectrum (c) and an image (d) of a test sample. Since the number of pixels is only 64, we did not adopt signal multiplexing. The cold detector head on which the TES microcalorimeter array was mounted was installed in a cryogen-free 100mK dilution-fridge. The mechanical micro vibrations from mechanical refrigerators, which are used to pre-cool the Helium working gas, often disturb the performance of microcalorimeters. For this STEM EDS system we therefore developed a vibration-free system.9

The spectra shown in panel (a) is close up near the silicon K emission. The tungsten M emission exists only with a 15 eV separation. In the figure, spectra for the TES microcalorimeter and the semiconductor detector are presented. We clearly see that our TES microcalorimeter can separate these two emission lines while the semiconductor detector cannot. Using the TES microcalorimeter, we can construct silicon and tungsten elemental maps separately. This is not possible with semiconductor detectors.

The elemental analysis of minerals

One of the applications of this STEM-EDS system will be the elemental analysis of minerals including astromaterials, such as meteorites and returned samples from space missions. Figure 2 shows the spectrum of olivine (silicate) obtained with the TES-microcalorimeter compared with the solid-state detector, SDD. The superior performance of the TES microcalorimeter is obvious from the figure. The TES can not only detect weak spectral features such as the Cu L line, the Mg and Si absorption edges, but also separate O, Si, and Mg emission lines from the continuum emission. The latter point is essential for the quantitative estimation of the elemental concentrations.10

The future

High-energy resolution imaging spectroscopy with ~1-M pixels is a goal of future X-ray astronomy beyond Athena X-IFU. The most promising enabling technology towards this goal is the microwave (GHz) FDM, which utilises microwave resonators coupled with TESs instead of biasing each TESs with MHz bias voltages. This technology is being studied at several institutions in the world; we are one of them and we are collaborating with AIST11 towards this goal. In the near future, we predict that an M pixel class X-ray integral field unit will be realised.


1 Tanaka, Y., Inoue, H. & Holt, S.S. 1994, Publ. of Astro. Soc. of Japan, 46, L37
2 Den Herder, J. W. et al. 2001, Astro. & Astrophys.,
365, L7)
3 Canizares, C.R., et al. 2005, Publ. of Astro, Soc. of Pacific, 117,1144
4 Moseley, S.H., Mather, J.C., McCammon, D. 1984, J. Appl. Phys. 56, 1257
5 Kelley, R. et al. 2016, Proc. SPIE. 9905
6 Barret, D. et al. 2018, Proc. SPIE. 10699
7 Irwin K.D. 1995, Appl. Phys. Lett. 66, 1998
8 Mitsuda, K. 2016, Physica C-Superconductivity & its Applications, 530, 93
9 Maehata, K. et al. 2015, J. of Superconductivity and Novel Magnetism, 28, 1161
10 Hayashi et al. 2019, submitted to IEEE Trans. on
Applied Superconductivity
11 Nakashima et al. 2018, J. of Low Temp. Phys., 193, 618

Kazuhisa Mitsuda
Astrophysics Institute of Space and Astronautical Science (ISAS)
Japan Aerospace Exploration Agency (JAXA)
+81 50 3362 3621

Laboratory Supplies Directory - Now Live


Please enter your comment!
Please enter your name here