Solar Low-Energy X-ray Spectrometer (SoLEXS)

• SoLEXS is a Sun-as-a-star soft X-ray spectrometer on the Aditya-L1 mission that provides uninterrupted 2–22 keV observations with 1-second cadence using dual silicon drift detectors.
• It employs dual-aperture architecture and precise energy calibration (≈170 eV resolution at 5.9 keV) to capture solar events ranging from quiet A-class to intense X-class flares.
• The instrument’s rigorous calibration framework, forward-folded spectral analysis, and high-duty-cycle L1 operations enable actionable insights into flare evolution, coronal heating, and flare–CME coupling.

Searching arXiv for SoLEXS and related Aditya-L1 calibration/observations papers. The Solar Low-Energy X-ray Spectrometer (SoLEXS) is a Sun-as-a-star soft X-ray spectrometer on board India’s Aditya-L1 mission at the Sun–Earth L1 Lagrange point. It provides continuous solar spectroscopy over 2–22 keV with approximately 170 eV full width at half maximum (FWHM) at 5.9 keV, 1 s spectroscopic cadence, and parallel 0.1 s timing-chain light curves, using two silicon drift detectors (SDDs) and a dual-aperture configuration that spans solar activity from A-class to X-class flares (Sarwade et al., 30 Sep 2025). Its scientific role is quantitative soft X-ray spectroscopy of flare evolution, coronal thermal structure, elemental abundances, and flare–coronal mass ejection (CME) coupling, with the uninterrupted L1 vantage providing essentially 100% observational duty cycle.

1. Mission setting and developmental trajectory

SoLEXS was originally proposed as an additional payload for Aditya-1 to complement the Visible Emission Line Coronagraph (VELC). In that formulation, the instrument’s science case emphasized four major objectives: understanding the DC heating mechanism, studying flare–CME relations from the same platform, deriving independent estimates of temperature and emission measure at flaring sites, and conducting coronal abundance studies and their variation during flares. The proposal also assigned it an operational role as a flare trigger for the coronagraph, with implications for onboard memory management (Sankarasubramanian et al., 2011).

The flight instrument now operates on Aditya-L1 from the Sun–Earth L1 point. That vantage is central to SoLEXS’s observing concept: unlike low-Earth-orbit configurations, L1 avoids eclipses and South Atlantic Anomaly passages that reduce low-Earth-orbit duty cycles to less than 70%, and therefore enables essentially uninterrupted Sun visibility. The mission timeline is explicitly defined. Launch occurred on 2 September 2023; the first in-space power-on took place on 16 October 2023 with the aperture cover closed; aperture cover deployment and first solar light occurred on 13 December 2023 at 08:00 UTC; within 33 hours, the instrument recorded the X2.9 flare SOL2023-12-14T17:02; and continuous 1 s spectroscopic observations began on 6 January 2024 (Sarwade et al., 30 Sep 2025).

This continuity from proposal to flight is scientifically significant. The original rationale centered on coordinated soft X-ray and coronagraphic observations, especially for reconnection-driven heating and flare–CME timing. The in-flight system preserves that logic while extending it through near-continuous, calibrated spectroscopy. A plausible implication is that SoLEXS realizes the original Aditya concept under observational conditions more favorable than those available to a low-Earth-orbit platform.

2. Instrument architecture and observing modes

The flight SoLEXS payload uses two SDDplus detectors from PNDetector, each with 30 mm$^2$ active area, 450 $\mu$m silicon thickness, and an 8 $\mu$m DuraBeryllium Plus window, covering 2–22 keV (Sarwade et al., 30 Sep 2025). The detectors are operated near $-45^\circ$C using integrated Peltier coolers, while the electronics operate around $\sim 40^\circ$C. This thermal partition is part of the instrument’s low-noise design.

Dynamic range is achieved through two fixed cylindrical apertures machined in the front plate. For SDD1, the large aperture has diameter $3.008 \pm 0.001$ mm and area $7.1063 \pm 0.0095$ mm$^2$, optimized for quiet Sun to C-class flares. For SDD2, the small aperture has diameter $0.368 \pm 0.001$ mm and area $0.1065 \pm 0.0006$ mm$\mu$0, optimized for M–X-class flares. The stated purpose of this dual-aperture scheme is to cover the approximately five orders of magnitude variation from A-class to X-class events without saturating both detectors simultaneously (Sarwade et al., 30 Sep 2025).

Spectral products comprise 340 channels per spectrum. Channels 1–168 have approximately 47.75 eV width up to about 8 keV; channels 169–340 are paired for approximately 94.5 eV width up to 22 keV. This binning preserves line diagnostics at lower energies while capturing the continuum at higher energies. Internal collimation is provided by a Zr collimator and the cylindrical apertures, giving fields of view of $\mu$1 for SDD1 and $\mu$2 for SDD2. The electronics chain consists of a low-noise CSPA followed by FPGA-based digital pulse processing (DPP). Two chains operate in parallel: a spectral chain with peaking time 2 $\mu$3s and triangular pulse shaping time about 4 $\mu$4s, and a timing chain with peaking time 0.35 $\mu$5s and shaping time about 0.7 $\mu$6s. Pile-up rejection uses the faster timing chain to identify unresolved events in the spectral chain, and the CSPA is reset every 2 ms to avoid saturation (Sarwade et al., 30 Sep 2025).

In the 2011 proposal, SoLEXS was described as a non-imaging spectrometer with a variable aperture mechanism having small, large, and opaque-with-calibration-source positions, designed for count rates up to approximately 500,000 counts per second and low-energy reach down to $\mu$7 keV (Sankarasubramanian et al., 2011). The flown configuration differs in implementation but preserves the same central design principle: extend dynamic range while maintaining spectroscopic integrity across quiet and eruptive solar conditions.

3. Calibration framework and instrument response

Ground calibration established the quantitative spectroscopic basis of SoLEXS. Energy–channel calibration was derived from X-ray fluorescence targets and radioactive lines. For SDD1, JSC-1A lunar simulant plus salt provided Ca, Ti, Cr, and Fe lines. For SDD2, Mu metal supplied Fe and Ni lines, supplemented by Zr lines from the collimator and Pb lines from solder. Line centroids were fitted to determine linear gain and offset, with explicit account for the binning change at channel 168. The resulting linear mapping is

$\mu$8

The pre-commissioning values were $\mu$9 eV/channel and $\mu$0 eV for SDD1, and $\mu$1 eV/channel and $\mu$2 eV for SDD2 (Sarwade et al., 30 Sep 2025).

Spectral resolution was measured across Ti K$\mu$3, Cr K$\mu$4, Fe K$\mu$5/K$\mu$6, Ca K$\mu$7/K$\mu$8, and Mn K$\mu$9/K$-45^\circ$0 from $-45^\circ$1Fe, and modeled by a standard Fano-plus-electronics expression,

$-45^\circ$2

with Fano factor $-45^\circ$3 for Si and pair creation energy $-45^\circ$4 eV. The fit gave $-45^\circ$5 eV (Sarwade et al., 30 Sep 2025).

The response formalism is explicitly forward-folding:

$-45^\circ$6

where $-45^\circ$7 is counts in channel $-45^\circ$8, $-45^\circ$9 the incident photon flux, $\sim 40^\circ$0 the ancillary response function (ARF), and $\sim 40^\circ$1 the spectral redistribution function (SRF). The SRF is parameterized in HYPERMET form as the sum of a main Gaussian, a Si escape peak 1.74 keV below the main peak, an exponential low-energy tail, and a low-energy shelf. Measured monoenergetic beam data from 6.5–16 keV established the energy dependence of these components; at 6 keV the main peak dominates with probability greater than 99%, so a Gaussian-only SRF is sufficient at high count rates, whereas the full HYPERMET SRF is used for low-count, long integrations (Sarwade et al., 30 Sep 2025).

The ARF is modeled by Beer–Lambert attenuation through the entrance window and silicon layers:

$\sim 40^\circ$2

Here $\sim 40^\circ$3 is set by the measured aperture area, $\sim 40^\circ$4 is taken from the 8 $\sim 40^\circ$5m DuraBeryllium Plus window datasheet, and silicon absorption cross sections are taken from Elam et al. (2002). Because SDD1 and SDD2 have aperture areas of 7.1063 mm$\sim 40^\circ$6 and 0.1065 mm$\sim 40^\circ$7, respectively, their ARFs differ primarily by geometry (Sarwade et al., 30 Sep 2025).

Thermo-vacuum testing in the URSC 1.6 m chamber subjected the payload to five short hot/cold soaks of 2 h each and extended 24 h hot and cold soaks, with shroud temperatures between $\sim 40^\circ$8C and $\sim 40^\circ$9C. Mn K$3.008 \pm 0.001$0 FWHM remained stable within $3.008 \pm 0.001$1 eV across qualification ranges. Gain depends weakly on detector temperature, whereas offset shows a clear quadratic dependence on electronics temperature; these dependencies are applied in orbital corrections (Sarwade et al., 30 Sep 2025).

4. In-flight performance, deadtime, and radiometric validation

On-board calibration uses an internal $3.008 \pm 0.001$2Fe and Ti fluorescence source providing Mn K$3.008 \pm 0.001$3/K$3.008 \pm 0.001$4 at 5.898/6.490 keV and Ti K$3.008 \pm 0.001$5/K$3.008 \pm 0.001$6 at 4.507/4.932 keV. In pre-commissioning measurements with the cover closed and 48 h integration, post-launch spectral resolution was $3.008 \pm 0.001$7 eV at 5.9 keV for SDD1 and $3.008 \pm 0.001$8 eV at 5.9 keV for SDD2, with gains and offsets matching ground values at similar temperatures (Sarwade et al., 30 Sep 2025).

Deadtime behavior is treated with a paralyzable model,

$3.008 \pm 0.001$9

where $7.1063 \pm 0.0095$0 is observed rate, $7.1063 \pm 0.0095$1 true rate, and $7.1063 \pm 0.0095$2 deadtime. Ground measurements yielded $7.1063 \pm 0.0095$3s for the spectral chain and $7.1063 \pm 0.0095$4s for the timing chain. On orbit, reset-related ringing from the 2 ms CSPA reset introduces about 500 counts s$7.1063 \pm 0.0095$5 of spurious signal in the timing chain, leading to the empirical correction formulas

$7.1063 \pm 0.0095$6

$7.1063 \pm 0.0095$7

Timing-chain rates are corrected for $7.1063 \pm 0.0095$8s and spurious counts, then used to correct spectral-chain deadtime and efficiency. The spectral deadtime remains paralyzable, and a modest efficiency reduction of about 11% arises from pile-up rejection rejecting events coincident with resets (Sarwade et al., 30 Sep 2025).

Operationally, during maximal activity from December 2023 to May 2024, SDD2 was the primary detector. SDD1 saturated above about $7.1063 \pm 0.0095$9 counts s$^2$0, whereas SDD2’s small aperture remained suitable for intense flares. The timing chain preserves accurate rate recovery through deadtime correction lookup even when the spectral chain saturates, and pile-up rejection limits spectral pile-up to about 1% at $^2$1 counts s$^2$2 on the ground (Sarwade et al., 30 Sep 2025).

Radiometric accuracy has been assessed through cross-calibration against GOES-16 XRS and Chandrayaan-2/XSM. Against XRS-A, after deadtime and ARF corrections, 1-minute fluxes show linear correlation over $^2$3 to $^2$4 W m$^2$5. Systematically, SoLEXS SDD2 is about 15% lower than XRS-A at lower fluxes, while during flare peaks SoLEXS fluxes exceed XRS-A, consistent with GOES flat-spectrum assumptions. Against XSM, normalized count-rate scatter shows near-unity ratios at high count rates, about 0.95 at lower rates, and a persistent low-energy dip to about 0.9 in the 2–2.3 keV range, indicating residual ARF uncertainties. Direct flare comparisons gave overall ratios of 1.06 for an M-class flare and 0.91 for a C-class flare above the low-energy dip. The stated radiometric consistency is within about 10% after corrections (Sarwade et al., 30 Sep 2025).

Data handling is correspondingly response-driven. Level-0 telemetry is downlinked to ISSDC; higher-level spectra, light curves, and calibrated responses are generated at the SoLEXS Payload Operations Centre at URSC and distributed through the Pradan portal. Public data release began in July 2024, and the SoLEXS_Tools package provides calibration databases, ARF/SRF construction, and analysis utilities. Typical products include 1 s accumulated 340-channel spectra, 0.1 s timing-chain light curves in three coarse energy bands, temperature-dependent gain/offset files, deadtime correction parameters, detector-specific ARFs, and Gaussian-only or HYPERMET SRFs (Sarwade et al., 30 Sep 2025).

5. Spectroscopic methodology and derived observables

SoLEXS spectroscopy is performed by forward-fitting rather than inversion, since the inverse problem is ill-posed. Model spectra are convolved with ARF and SRF and fitted by minimizing

$^2$6

For thermal plasma analysis, the stated workflow uses isothermal thermal bremsstrahlung plus line emission from CHIANTI, implemented as f_vth_abun in sunkit-spex, and fitted with Sherpa. The fit solves for temperature $^2$7, emission measure $^2$8, and abundances such as Fe, Ca, Ar, and S. In schematic form, the photon spectrum obeys $^2$9, with continuum behavior approximately $0.368 \pm 0.001$0. If a hard tail is present, a power-law component $0.368 \pm 0.001$1 may be added, although the 2–22 keV band is described as primarily thermal for most flares (Sarwade et al., 30 Sep 2025).

The early science return already reflects the original science goals. SoLEXS recorded flares across B–X classes, including the second-strongest flare of Solar Cycle 25, the X8.7 event of 14 May 2024. Isothermal fitting yields temperature and emission measure evolution through flare phases, and element-specific line complexes such as the Fe-line region constrain abundances and temperature structure. Example fits gave reduced $0.368 \pm 0.001$2 with variable Fe, Ca, Ar, and S abundances (Sarwade et al., 30 Sep 2025).

A particularly distinctive application is iron fluorescence in X-class flares. Using SDD2 data for 47 X-class flares in 2024, SoLEXS measured both the Fe K$0.368 \pm 0.001$3 fluorescence line at 6.40 keV and the exciting flux above the neutral Fe K-edge at 7.11 keV. This simultaneous continuum-plus-line capability is central: the 6.40 keV Fe K$0.368 \pm 0.001$4 feature and the hot thermal Fe XXV complex near 6.7 keV are separated by about 300 eV, nearly twice the FWHM, allowing reliable decomposition during fitting. The fluorescence line is modeled with a Gaussian at 6.40 keV and Fe K$0.368 \pm 0.001$5 at 7.06 keV, with the K$0.368 \pm 0.001$6/K$0.368 \pm 0.001$7 flux ratio fixed to 0.134. The integrated photon flux is

$0.368 \pm 0.001$8

and the energy flux is

$0.368 \pm 0.001$9

The exciting flux is defined by

$0.1065 \pm 0.0006$0

and operationally estimated by summing energy-weighted counts above 7.11 keV and dividing by effective area and exposure time (Sarwade et al., 21 May 2026).

The observed fluorescence efficiency is then

$0.1065 \pm 0.0006$1

with $0.1065 \pm 0.0006$2 the intrinsic fluorescence efficiency, $0.1065 \pm 0.0006$3 the photospheric iron abundance, $0.1065 \pm 0.0006$4 the coronal source height, and $0.1065 \pm 0.0006$5 the geometric factor for heliocentric angle $0.1065 \pm 0.0006$6. Across the 47 flares, peak $0.1065 \pm 0.0006$7 spans approximately 0.014–0.060 for most events, with one outlier at $0.1065 \pm 0.0006$8. Disk-center events typically show $0.1065 \pm 0.0006$9–0.055, while limb events show $\mu$00–0.02. For a representative near-disk-center X1.0 flare with $\mu$01, $\mu$02 peaked at $\mu$03 W m$\mu$04, $\mu$05 peaked at $\mu$06 W m$\mu$07, and the mean $\mu$08 near peak was $\mu$09. For a representative limb event at $\mu$10, $\mu$11 peaked at approximately $\mu$12 W m$\mu$13 but $\mu$14, showing strong suppression of Fe K$\mu$15 toward the limb (Sarwade et al., 21 May 2026).

An important methodological clarification follows from the closed-door spectra: pre-commissioning data show instrumental fluorescence such as Ni and Pb, but no Fe K$\mu$16 or Fe K$\mu$17, confirming that Fe fluorescence during solar observations is solar in origin (Sarwade et al., 21 May 2026). This addresses a natural instrumental concern and strengthens the interpretation of the 6.4 keV feature as photospheric fluorescence during the thermal phase of X-class flares.

6. Scientific scope, comparative context, and limitations

The scientific uses of SoLEXS are broad but internally coherent. Continuous L1 spectroscopy underpins studies of coronal heating by small-scale events during quiet periods, flare energetics through temperature and emission measure time profiles, abundance evolution during flares, and flare–CME relationships through uninterrupted tracking of thermal evolution. In historical terms, these applications directly continue the 2011 science rationale: DC heating, flare–CME timing, independent thermal diagnostics, and abundance studies (Sankarasubramanian et al., 2011).

Relative to related instruments, the role of SoLEXS is sharply defined. GOES XRS is a broadband photometer and does not perform spectroscopy; SoLEXS adds approximately 170 eV resolution at 5.9 keV, detailed line and continuum diagnostics, and 1 s cadence. RHESSI provided hard X-ray to gamma-ray coverage with Ge detectors, roughly 1 keV resolution at 6 keV, and strong imaging capabilities, but had limited soft X-ray abundance diagnostics and reduced duty cycle in low Earth orbit. Chandrayaan-2/XSM is an SDD-based soft X-ray spectrometer with similar resolution and science goals, but its lunar orbit imposes seasonal duty-cycle constraints and off-axis effects. SoLEXS’s L1 location and dual apertures therefore define its niche as continuous, calibrated soft X-ray spectroscopy of the full Sun across a large flare dynamic range (Sarwade et al., 30 Sep 2025).

The iron fluorescence results further clarify this niche. Classical crystal spectrometers such as SMM/BCS and Yohkoh/BCS resolved Fe K$\mu$18/K$\mu$19 at much higher spectral resolution but did not simultaneously measure the ionizing continuum above 7.11 keV in a broadband passband. SoLEXS, despite having approximately 170 eV resolution at 6 keV rather than approximately 10 eV, can directly determine both $\mu$20 and $\mu$21, enabling empirical efficiency estimates without model-dependent extrapolation of the exciting flux (Sarwade et al., 21 May 2026).

The limitations are correspondingly well characterized. The operational band is 2–22 keV, with a low-energy threshold around 2 keV used to suppress low-energy noise; sensitivity below about 2 keV is therefore reduced. Spectral-chain deadtime is paralyzable, giving rate degeneracy at very high true rates. SDD1 saturates during high activity, and its final on-orbit deadtime characterization awaits quiet A/B-class conditions. Systematic radiometric uncertainty is about 10% after corrections, with residual ARF uncertainty near 2–2.3 keV and deadtime-correction residuals at low count rates around 5% (Sarwade et al., 30 Sep 2025).

For fluorescence diagnostics, a further limitation is fundamental rather than instrumental. With temperature constrained, $\mu$22 can be mapped to source height only by assuming a photospheric Fe abundance. The 2026 analysis found that physically reasonable heights require $\mu$23, whereas lower abundances imply different height solutions. Thus source height and photospheric abundance are degenerate, and $\mu$24 cannot be uniquely determined without an assumed or independently constrained $\mu$25. The mean fluorescence efficiency at flare peak can therefore constrain effective coronal source height, but not uniquely determine it (Sarwade et al., 21 May 2026).

Taken together, these properties place SoLEXS in a specific observational regime: a non-imaging, response-calibrated, high-duty-cycle soft X-ray spectrometer optimized for quantitative thermal and line diagnostics of the unresolved solar disk. Its combination of dual-aperture SDD hardware, forward-folded response modeling, and uninterrupted L1 operations makes it a sustained source of solar soft X-ray spectroscopy for coronal heating studies, flare plasma diagnostics, and geometric probes based on fluorescence.