XRISM/Resolve X-ray Spectrometer

Updated 25 October 2025

XRISM/Resolve is a high-resolution non-dispersive microcalorimeter using a 36-pixel HgTe/Si array cooled to ~50 mK to directly measure X-ray photon energies.
It integrates advanced cryogenic cooling, real-time calibration with 55Fe and MXS sources, and a robust, modular data pipeline for precise event processing.
The instrument enables detailed astrophysical plasma diagnostics, supporting studies on turbulence, kinematics, and chemical composition in high-energy environments.

The XRISM/Resolve X-ray Spectrometer is a high-resolution non-dispersive X-ray microcalorimeter, operating as the flagship instrument on the X-Ray Imaging and Spectroscopy Mission (XRISM). Building on the technological and scientific heritage of the Hitomi/Soft X-Ray Spectrometer (SXS), Resolve delivers fine-energy-resolution (FWHM < 7 eV) spectroscopy over the 0.3–12 keV range, targeting observations of hot plasmas in astrophysical environments. Its core is a 36-pixel array of HgTe/Si microcalorimeters cooled to ~50 mK, enabling direct measurement of photon energies via transient thermal signals. The instrument’s design integrates advanced cooling systems, time-resolved event processing, and precise onboard calibration, together with a distributed, rigorously validated science data pipeline. XRISM/Resolve is a crucial resource for probing plasma diagnostics, kinematics, turbulence, and chemical composition across a range of high-energy astrophysical sources, including galaxy clusters, supernova remnants, X-ray binaries, and active galactic nuclei.

1. Instrument Design and Operational Principles

Resolve employs a non-dispersive microcalorimeter architecture, in which X-ray photons are absorbed by a HgTe film thermally linked to an ion-implanted silicon thermistor. The incident energy $E$ produces a temperature rise $\Delta T \simeq E/C$ in a sensor of heat capacity $C$ . The energy resolution is described by

$\Delta E \propto \sqrt{\frac{k_{\rm B} T^2 C}{\alpha}}$

where $T$ is the operating temperature, $k_{\rm B}$ is Boltzmann’s constant, and $\alpha = d\,{\rm ln}R / d\,{\rm ln}T$ is the logarithmic sensitivity of resistance to temperature. The pixel array consists of 36 elements (6 × 6): 35 pixels form the primary 3′×3′ field-of-view, with one calibration pixel illuminated by a $^{55}$ Fe source for real-time gain tracking (Sato et al., 2023).

Photon absorption is registered as a thermal pulse, whose shape is digitized and analyzed for amplitude, timing, and quality attributes. Signal filtering and event grading algorithms enable discrimination of isolated X-ray events (“Hp” grade) from pile-up or cross-talk conditions, optimizing the tradeoff between throughput and energy resolution (Sato et al., 2023).

The instrument’s cryogenic system uses a multistage chain: mechanical cryocoolers (four Stirling, one Joule–Thomson) cool from ambient to ~4 K, followed by adiabatic demagnetization refrigerators (ADR) to reach the sensor’s base temperature of ~50 mK. Mechanical vibration noise (microphonics), including low-frequency beat phenomena between cooler harmonics, is mitigated via a vibration isolation system (VIS) and flexible control of driver frequencies (Imamura et al., 2023).

2. Calibration, Event Screening, and Data Processing Pipeline

Accurate energy scale calibration across the detector array is maintained with both distributed $^{55}$ Fe sources (for Mn Kα/β fluorescence lines) and an active modulated X-ray source (MXS), generating pulsed reference lines (e.g., Cr, Cu) with precise spacecraft-clock synchronization (Shipman et al., 19 Aug 2025). The filter wheel system includes positions for neutral density (ND), optical blocking (OBF), beryllium (Be), and calibration sources, facilitating high-dynamic-range operations and protection against optical contamination (Shipman et al., 19 Aug 2025).

Background reduction—essential for detecting faint spectral features—is realized via event screening utilizing 19 discrete criteria, categorized as pulse-shape, relative timing, and operational interval signals (Mochizuki et al., 11 Jan 2025). Core metrics include RISE_TIME, DERIV_MAX, and TICK_SHIFT, with empirically optimized correlation boundaries:

$f_1({\rm DERIV\_MAX}) \leq {\rm RISE\_TIME} \leq f_2({\rm DERIV\_MAX})$

Initial screening yields background rates as low as $1.8 \times 10^{-3}$ cts s $^{-1}$ keV $^{-1}$ over 0.3–12 keV, reduced further to $1.0 \times 10^{-3}$ cts s $^{-1}$ keV $^{-1}$ for optimized pulse shape selection, outperforming prior ASTRO-H/SXS thresholds.

The data processing system is a networked, daemon-driven workflow that ingests FITS-formatted Level-1 data post-telemetry conversion in Japan, verifies, calibrates, and screens events, and produces science-ready Level-2 products (Doyle et al., 2022). Processing is modular:

Pre-fetch/Archive Daemons: Automated data ingestion, integrity checks, and archive synchronization between GSFC, ISAS, and HEASARC
Calibration/Screening: Using FTOOLS and XRISM-specific tasks, it executes gain corrections, PHA-to-PI conversion, bad pixel flagging (e.g., rslflagpix), and previews (images, lightcurves, spectra)
Integration/Testing: Comprehensive suite of end-to-end tests, unit tests (Comprehensive / Limited Functional Tests via HEASoft’s aht tool), regression, and continuous integration ensures reliability and reproducibility

Documentation covers command-line help (fhelp), in-code annotation (Doxygen format), and an auditable change log for traceability (Doyle et al., 2022).

3. Spectral Performance, High Count Rate Effects, and Event Processing

Resolve achieves a composite FWHM energy resolution of ∼4.5 eV at 6 keV (Collaboration, 1 Nov 2024, Kosec et al., 8 Oct 2025, Sato et al., 2023). Under nominal count rates, the absolute energy scale is stable to ±0.3 eV above 5.4 keV (Collaboration, 1 Nov 2024).

For bright point sources, high photon influx induces two primary systematic effects (Mizumoto et al., 7 Jun 2025):

Energy scale offset: At elevated count rates, energies are shifted negatively (Mn Kα at 6 keV by ∼–1 eV), primarily in heavily illuminated pixels. This shift is sufficient to mimic Doppler velocities of 50 km s $^{-1}$
Resolution degradation (“cross talk”): FWHM broadening is approximately linear with per-pixel count rate:

${\rm FWHM} = a R + b$

Application of a nearest-neighbor coincidence cut (excluding events within ∼25 ms in electrically linked pixels) restores the instrument to its intrinsic resolution, with a → 0 (Mizumoto et al., 7 Jun 2025).

Screening for these high-count-rate effects is essential for studies requiring detailed velocity and width diagnostics (e.g., wind kinematics, turbulence). Correction for these systematics is required for spectral analysis pipelines, particularly when observing galactic binaries, AGN, and cluster cores.

4. Data Products, Quick-Look Tools, and Community Archives

Pipeline output includes fully calibrated event files, standardized auxiliary files (attitude, orbit, compositions), and quicklook preview data (spectra, images, lightcurves) in FITS format (Doyle et al., 2022). Data and metadata are transferred to geographically redundant archives (GSFC, HEASARC, ISAS/DARTS).

The XSLIDE tool (“X-ray Spectral Line Identifier and Explorer”) provides a graphical interface for interactively rebinned spectra, continuum fitting, automated line detection, atomic line identification via AtomDB/XSTAR, and rapid diagnostic line ratio analysis (Braun et al., 2022). XSLIDE bypasses full forward-fitting by assuming a diagonalized response: $C(I) = T \int R_\mathrm{RMF}(I,E)R_\mathrm{ARF}(E) S(E) dE$ can be discretized to $S(E) = C(I)/[R_\mathrm{ARF}(E)T\Delta E]$ . Diagnostic ratios for H-like and He-like ions (e.g., $G = (x+y+z)/w$ , $R = z/(x+y)$ ) provide electron density and temperature estimates. The tool streamlines triage and proposal preparation by enabling rapid, model-independent screening for features of interest.

5. Scientific Applications and Calibration Advances

Resolve's energy coverage, combined with its low background and high spectral resolution, underpins several breakthroughs:

Astrophysical plasma diagnostics: Fe K-shell line measurements in high-temperature plasmas (e.g., GT Mus, SN1987A, Cyg X-3) have enabled direct temperature structure and velocity field mapping via line ratios and line width measurements (Kurihara et al., 9 Apr 2025, Collaboration, 12 May 2025, Collaboration, 1 Nov 2024).
ISM/IGM absorption features: Resolve’s fine energy calibration enables direct detection and quantification of weak ISM absorption lines, such as S II Kβ, and dust-phase S features; systematic offset corrections of +7–8 eV are applied to atomic templates for consistency with XRISM’s spectral scale (Corrales et al., 10 Jun 2025).
Multi-instrument cross-calibration basis: High-resolution “spline” fits to Resolve spectra provide a model-independent reference for cross-calibration against Chandra, NuSTAR, and XMM-Newton. Empirical correction functions $f(E)$ adjust for effective area and redistribution systematics; e.g., $f(E) = 1 / [1 - 0.00333 \exp((4 - E)/0.582)]$ (Collaboration, 10 Sep 2025).
Velocity and turbulence analyses: Bulk velocity differences, such as $\Delta v = 656 \pm 35$  km s $^{-1}$ in Abell 754, and line-of-sight dispersions up to ∼500 km s $^{-1}$ in clusters, are measured via Fe–K line centroids and widths, supporting direct quantification of ICM dynamics (Omiya et al., 18 Oct 2025).

6. Lessons from Hitomi and Evolution Beyond

Resolve is fundamentally derived from Hitomi/SXS but incorporates key enhancements:

Refined dewar and baffle structures for debris/micrometeoroid (MMOD) protection
Modulated X-ray source (MXS) for real-time gain calibration over a broader energy range (Shipman et al., 19 Aug 2025)
Improved filtering and automated flagging of bad/pile-up events to support high-resolution performance at enhanced throughput
A “cryogen-free” cooling mode retaining full performance post–liquid-helium exhaustion (Sato et al., 2023)
Comprehensive, modular, and well-documented pipeline, with improved regression and functional testing

Resolve’s technological advances—optimized event processing, robust calibration strategies, and high-fidelity software infrastructure—address previous limitations in calibration robustness, high-count-rate systematics, and data product uniformity. The distributed, rigorously validated processing infrastructure and documentation set a modern standard for space-based X-ray spectroscopy missions (Doyle et al., 2022).

Summary Table: XRISM/Resolve Technical Highlights

Aspect	Key Specs/Features	Reference Paper(s)
Detector	36-pixel HgTe/Si array, 50mK	(Sato et al., 2023, Doyle et al., 2022)
Energy Resolution	<7 eV FWHM at 6 keV	(Sato et al., 2023, Collaboration, 1 Nov 2024)
Cooling Chain	4 Stirling, 1 JT, ADR stages	(Imamura et al., 2023, Sato et al., 2023)
Calibration	$^{55}$ Fe, MXS, OBF, Be, ND	(Shipman et al., 19 Aug 2025, Doyle et al., 2022)
Event Screening	19 optimized parameters	(Mochizuki et al., 11 Jan 2025)
High Count Rate Mitigation	Nearest-neighbor cut	(Mizumoto et al., 7 Jun 2025)
Data Pipeline	Modular daemons + FTOOLS	(Doyle et al., 2022)
Cross-calibration Approach	Spline model on Resolve	(Collaboration, 10 Sep 2025)