Table of Contents
Fetching ...

XRISM view of a stellar flare: High-resolution Fe K spectra of HR 1099, an RS CVn-type star

Miki Kurihara, Masahiro Tsujimoto, Michael Loewenstein, Yoshitomo Maeda, Marc Audard, Ehud Behar, Megan E. Eckart, Adam Foster, Liyi Gu, Matteo Guainazzi, Kenji Hamaguchi, Natalie Hell, Shun Inoue, Yukiko Ishihara, Satoru Katsuda, Caroline A. Kilbourne, Maurice A. Leutenegger, Eric D. Miller, Nagisa Nagashima, Frederick Scott Porter, Makoto Sawada, Yohko Tsuboi, Vinay L. Kashyap, David H. Brooks

TL;DR

This study uses XRISM Resolve to observe a large stellar flare on the RS CVn binary HR 1099, marking the first time a stellar flare was captured with an X-ray microcalorimeter. By exploiting the high-resolution Fe XIX–XXVI K-shell lines, the authors reconstruct a differential emission measure (DEM) spanning $1$–$10$ keV and derive elemental abundances without assuming photospheric values. The analysis reveals a bimodal DEM with a hotter component that grows during the flare and shows selective abundance enhancements of Ca and Fe, consistent with chromospheric evaporation and i-FIP physics. The results demonstrate XRISM’s capability to probe coronal heating and chemical fractionation in stellar flares and establish a benchmark for future high-resolution X-ray spectroscopy of active stars.

Abstract

A high-resolution X-ray spectroscopic observation was made of the RS CVn-type binary star HR 1099 using the Resolve instrument onboard XRISM for its calibration purposes. During the $\sim$400 ks telescope time covering 1.5 binary orbit, a flare lasting for $\sim$100 ks was observed with a released X-ray radiation energy of $\sim 10^{34}$ erg, making it the first stellar flare ever observed with an X-ray microcalorimeter spectrometer. The flare peak count rate is 6.4 times higher than that in quiescence and is distinguished clearly in time thanks to the long telescope time. Many emission lines were detected in the 1.7--10 keV range both in the flare and quiescent phases. Using the high spectral resolution of Resolve in the Fe K band (6.5--7.0 keV), we resolved the inner-shell lines of Fe XIX--XXIV as well as the outer-shell lines of Fe XXV--XXVI. These lines have peaks in the contribution functions at different temperatures over a wide range, allowing us to construct the differential emission measure (DEM) distribution over the electron temperature of 1--10 keV (roughly 10--100 MK) based only on Fe lines, thus without an assumption of the elemental abundance. The reconstructed DEM has a bimodal distribution, and only the hotter component increased during the flare. The elemental abundance was derived based on the DEM distribution thus constructed. A significant abundance increase was observed during the flare for Ca and Fe, which are some of the elements with the lowest first ionization potential among those analyzed, but not for Si, S, and Ar. This behavior is seen in some giant solar flares and the present result is a clear example in stellar flares.

XRISM view of a stellar flare: High-resolution Fe K spectra of HR 1099, an RS CVn-type star

TL;DR

This study uses XRISM Resolve to observe a large stellar flare on the RS CVn binary HR 1099, marking the first time a stellar flare was captured with an X-ray microcalorimeter. By exploiting the high-resolution Fe XIX–XXVI K-shell lines, the authors reconstruct a differential emission measure (DEM) spanning keV and derive elemental abundances without assuming photospheric values. The analysis reveals a bimodal DEM with a hotter component that grows during the flare and shows selective abundance enhancements of Ca and Fe, consistent with chromospheric evaporation and i-FIP physics. The results demonstrate XRISM’s capability to probe coronal heating and chemical fractionation in stellar flares and establish a benchmark for future high-resolution X-ray spectroscopy of active stars.

Abstract

A high-resolution X-ray spectroscopic observation was made of the RS CVn-type binary star HR 1099 using the Resolve instrument onboard XRISM for its calibration purposes. During the 400 ks telescope time covering 1.5 binary orbit, a flare lasting for 100 ks was observed with a released X-ray radiation energy of erg, making it the first stellar flare ever observed with an X-ray microcalorimeter spectrometer. The flare peak count rate is 6.4 times higher than that in quiescence and is distinguished clearly in time thanks to the long telescope time. Many emission lines were detected in the 1.7--10 keV range both in the flare and quiescent phases. Using the high spectral resolution of Resolve in the Fe K band (6.5--7.0 keV), we resolved the inner-shell lines of Fe XIX--XXIV as well as the outer-shell lines of Fe XXV--XXVI. These lines have peaks in the contribution functions at different temperatures over a wide range, allowing us to construct the differential emission measure (DEM) distribution over the electron temperature of 1--10 keV (roughly 10--100 MK) based only on Fe lines, thus without an assumption of the elemental abundance. The reconstructed DEM has a bimodal distribution, and only the hotter component increased during the flare. The elemental abundance was derived based on the DEM distribution thus constructed. A significant abundance increase was observed during the flare for Ca and Fe, which are some of the elements with the lowest first ionization potential among those analyzed, but not for Si, S, and Ar. This behavior is seen in some giant solar flares and the present result is a clear example in stellar flares.

Paper Structure

This paper contains 20 sections, 7 figures, 3 tables.

Figures (7)

  • Figure 1: Temperature range at which representative emission lines of different elements in different charge states are formed. Their ion name and the line energy in keV are given for each dot. The dots are placed at the peak of the contribution function (at the low density limit of $10^{10}$ cm$^{-3}$) on the horizontal axis and the line energy on the vertical axis. The horizontal bars indicate the temperature range where the contribution function exceeds half of the peak value. For Fe, selected lines of the lower charge states (Fe XIX--XXII) are predominantly formed by the dielectronic recombination process, while those of the higher charge states (Fe III--XXVI) are by direct excitation. For Si, S, Ar, and Ca, the $w$ line of the He-like and the Ly$\alpha1$ line of the H-like ions are selected. Chianti version 11 dere1997dufresne2024 is used. Alt text: A figure showing temperature ranges that Si, S, Ar, Ca, and Fe lines can cover.
  • Figure 2: (a) Resolve light curve of HR 1099 in the 1.7--10.0 keV with a 1024 s binning after good-time-interval screening. The reference time is 2024-03-06T01:15:32. The binary phase (i.e., the orbital phase of the binary, with the primary star being in front at phase zero) is shown on the top axis. The modes of the observation are shown for the four parts. The flare phase was chosen to be the same with the second part, as it encompasses most of the flare. (b) Temperature of the 50 mK stage of the detector cooling system. The four large deviations at around 25, 125, 225, and 325 ks are due to ADR recycles. Other repeated deviations are mostly due to spacecraft passages of the South Atlantic Anomaly (SAA) region. (c) Status of the MXS on or off. MXS1 is in black, while MXS3 is in green. They were turned on when the source was eclipsed by the Earth. The $^{55}$Fe sources on the filter wheel were also rotated into the aperture during the same intervals. Alt text: A three-panel figure showing (a) X-ray light curve in the 1.7–-10.0 keV energy range, (b) 50 mK stage temperature, and (c) MXS status during the observation.
  • Figure 3: Resolve spectrum in the 1.7--10.0 keV for the flaring (red) and quiescent data (blue) phase as well as the background model spectrum (black). The best-fit model of the fiducial 2$kT$ fit components is also shown. The gray stripes indicate the energy range of the conspicuous He$\alpha$ and Ly$\alpha$ lines. The data were binned to have a minimum of 30 events per bin for this figure. Alt text: Resolve X-ray spectrum in the 1.7--10.0 keV range.
  • Figure 4: (a) Close-up view of the Resolve spectra in the Fe XIX--XXVI (left) and Fe XXVI (right) K-shell complexes of HR 1099 in flare (red) and quiescence (blue) compared to that of GT Mus in quiescence (black; kurihara2025). The intensity is scaled by $\times$20 for the red, $\times$10 for the blue, and $\times$1 for the black spectra to facilitate comparison. The centroid energy of the lines used for the DEM construction is indicated by vertical dashed lines. (b, c) Synthesized spectra with three different plasma codes: SPEX version 3.08.01 (blue), Chianti version 11 (orange), and AtomDB version 3.0.9 (green) for a thermal plasma of (b) 4 keV or (c) 1 keV. The spectra are smoothed with a Gaussian of an FWHM$=4$ eV, normalized at the strongest line of Fe XXVI Ly$\alpha_1$ for $kT=4$ keV and Fe XXII line 6.586 keV for $kT=1$ keV. (d) Synthesized spectra (1 eV binning) with Chianti version 11 decomposed for each ion. Alt text: A four-panel figure of the Fe K complex of the (a) Resolve spectra, (b,c) model spectra from SPEX version 3.08.01, Chianti version 11, and AtomDB version 3.0.9, and (d) model spectra for individual ions.
  • Figure 5: Close-up views of the Resolve spectrum (black) of the Fe K-shell line complexes, the best-fit model with the fiducial 2$kT$ model (green) and phenomenological model (orange), and their residuals. The upper panels (a)--(c) are for the quiescent spectrum, while the lower panels (d)--(f) are for the flare spectrum. Alt text: A pair of three-panel figures showing close-up views of the Resolve spectra in the Fe K-shell line complexes and the best-fit spectral models for the quiescent and flaring phases, respectively.
  • ...and 2 more figures