Chandrayaan-2 CLASS: Lunar XRF Instrument

• The Chandrayaan-2 CLASS instrument is a cutting-edge XRF system that significantly improves lunar elemental detection using a 64 cm² detector array and refined calibration methods.
• It employs advanced energy and spatial resolution techniques, including a gold-coated copper collimator and GEANT4-based background modeling, to isolate key elemental signals.
• CLASS data processing integrates a multi-stage pipeline with Gaussian mixture models to deliver global, high-fidelity maps that enhance our understanding of lunar petrology and geochemistry.

The Chandrayaan-2 Large Area Soft X-ray Spectrometer (CLASS) is the primary X-ray fluorescence (XRF) instrument on the Indian Space Research Organisation’s Chandrayaan-2 lunar orbiter. Designed as a successor to C1XS on Chandrayaan-1, CLASS employs major instrumental, methodological, and analytical advances to perform global, high-resolution mapping of lunar surface chemistry. By integrating a vastly expanded detector area, enhanced energy and spatial resolution, and advanced calibration and background modeling, CLASS measures the elemental composition of the Moon’s uppermost regolith with quantitative fidelity that enables new petrological and geochemical insights.

1. Instrumentation and Technical Advancements

CLASS incorporates 16 swept charge devices (CCD-236), each with 4 cm² active area, yielding 64 cm² total, compared to C1XS’s 24 cm². Its gold-coated copper collimator restricts the field of view to ~14°, enabling a spatial footprint of 25 km × 25 km FWHM from 200 km orbit—double the spatial resolution of C1XS.

The instrument targets an energy resolution better than 250 eV at 1.48 keV and below 200 eV at 5.9 keV, and uses a single 0.2 μm Al filter to minimize filter-origin Al Kα contamination. Extensive ground calibrations have been performed, and simulations with tools such as GEANT4 quantify the impact of particle-induced X-ray emission (PIXE), addressing shortcomings experienced by C1XS during high-background conditions (notably in geotail passages).

The design enables detection of key major elements (O, Na, Mg, Al, Si, Ca, Ti, Fe) in the low-energy (soft) X-ray regime, with sensitivity reaching elements at ~1 wt% relative abundance. This is particularly critical for accurate mapping of petrologically important elements such as Na, which has previously eluded detection in global remote sensing surveys.

2. X-ray Fluorescence Measurement Principle

Transmission of solar X-rays to the lunar surface excites characteristic K- and L-shell fluorescence from regolith atoms. The measured line intensity $I_x$ for element $x$ is given by

$$I_x = \frac{G(E)}{4\pi R^2} \cdot \sigma_x(E) \cdot \omega_x \cdot A_x \cdot \epsilon(E)$$

where $G(E)$ is the incident solar flux at energy $E$, $R$ is the spacecraft’s altitude, $\sigma_x(E)$ the photoelectric cross-section, $\omega_x$ the fluorescence yield, $A_x$ the elemental abundance, and $\epsilon(E)$ the detector efficiency.

Accurate inversion of line intensity to derive $A_x$ depends critically on simultaneous knowledge of $G(E)$, requiring real-time measurements of the solar spectrum, as provided by the co-mounted Solar X-ray Monitor (XSM). The approach enables quantitative, time-resolved determination of absolute and relative abundances under variable solar activity.

3. Data Processing and Quantitative Mapping

CLASS data analysis employs a rigorous multi-stage pipeline (Kumar et al., 21 Aug 2025):

• Background estimation: Night-side (dark) data are selected using geometric filtering (e.g., $R_M (csc\,\theta - 1) \geq h$, with $R_M$ the lunar radius, $\theta$ the local solar elevation angle, $h$ orbit altitude). Backgrounds are modeled as a function of lunar phase and are subtracted from sunlit spectra.
• Signal detection and Gaussian fitting: Binned intervals (~96–296 s) are scanned for statistically significant XRF peaks (signal-to-noise metrics $\mathcal{N}_t$). Peaks are modeled with Gaussian profiles to extract amplitude $A$, centroid $\mu$, and width $\sigma$, constrained by physical expectations for line energies and instrument resolution.
• Line ratio computation: Line intensity ratios are used as robust geochemical proxies, defined as

$$\mathcal{R}_L^M = \frac{A_M \sigma_M}{A_L \sigma_L}$$

for element $M$ relative to base element $L$ (e.g., Si).

• Spatial mapping: Area-weighted averages, with proper error propagation, are compiled into global maps at 5.3 km/pixel resolution. Statistical tools, including Gaussian mixture models, are applied to line-ratio parameter spaces (e.g., Mg/Si vs. Al/Si) to reveal geochemically distinct clusters corresponding to major lunar terranes.

CLASS’s methodology enables direct, model-minimal mapping of elemental line ratios. Linear correlations with independently derived abundance maps (e.g., soil samples, earlier remote sensing campaigns) are used for calibration and validation. For example,

$$(\mathrm{Mg \, wt\%}/\mathrm{Al \, wt\%})_\mathrm{CLASS} = 0.88 \times \mathcal{R}^{(\mathrm{Mg})}_{\mathrm{Al}} - 0.34,$$

confirming the line intensity ratio as a reliable proxy for weight fraction ratios.

4. Scientific Outcomes: Lunar Petrology and Global Geochemistry

CLASS has produced the highest spatial resolution (~5 km) global XRF maps of O/Si, Mg/Si, Al/Si, Mg/Al, Ca/Si, and Fe/Si ever achieved (Kumar et al., 21 Aug 2025). Key findings include:

• Terrane discrimination: The Mg/Al map distinctly separates low-Al, Mg-rich basaltic mare from Al-rich, plagioclase-dominated highlands. These results are consistent with Apollo 15/16 XRF datasets and validate global heterogeneity inferred from sample and orbital gamma-ray spectroscopy.
• Correlation with abundances: The CLASS line ratio maps correlate linearly with existing elemental abundance maps, confirming their reliability. For some elements (notably O), the uniformity of surface abundance is verified (40–45 wt%), providing an internal consistency check.
• Advanced clustering: Gaussian mixture modeling in Mg/Si–Al/Si space partitions the lunar surface into statistically distinct geochemical domains. These correspond to compositional provinces including feldspathic highlands, basaltic maria, South Pole–Aitken basin, and transitional regions, delineating both classical and previously ambiguous boundaries.

This capability enables refinement of lunar magma ocean differentiation models and the identification of landing sites optimized for sampling petrologic diversity.

5. Mineral Physics and Plagioclase Composition Inference

A pivotal application is the direct derivation of anorthite content (An#) in plagioclase feldspar. CLASS exploits its sensitivity to Na and Ca, inaccessible to gamma-ray spectroscopy and challenging for IR remote sensing. Using:

$$\mathrm{An\#} \approx \frac{\mathrm{Ca}}{\mathrm{Ca} + \mathrm{Na}}$$

where Ca and Na are measured directly from XRF line intensities, spatially resolved An# mapping across the lunar highlands becomes possible. Classical approaches, which require nearly pure anorthosites, are superseded by the ability to sample mixed or intermediate compositions, enabling a more complete assessment of crustal petrogenesis.

6. Integration with Solar and Particle X-ray Inputs

CLASS operates in tandem with XSM, which provides contemporaneous measurements of the incident solar X-ray spectrum (1–15 keV; energy resolution ~180 eV at 5.9 keV; 1 s cadence) (Shanmugam et al., 2019). Dynamic calibration using XSM is essential due to variability in $G(E)$, especially during solar flares.

Furthermore, CLASS’s design accommodates particle-induced X-ray emission (PIXE) events, with plans for a complementary onboard particle detector. This enables the separation of solar- and particle-driven XRF, unlocking the potential for night-side observations and expanding the mapping envelope.

7. Relevance and Implications for Planetary Science

CLASS’s recent datasets represent a significant advance in the global, quantitative mapping of lunar surface geochemistry. The combination of expanded spatial resolution, robust statistical methodologies for quantifying line intensities, and advanced geostatistical modeling has enabled robust delineation of compositional domains and their boundaries.

This high-fidelity mapping underscores the Moon’s petrologic complexity and provides critical constraints on models of planetary differentiation and crust formation. The methodologies and instrumental concepts demonstrated by CLASS have clear applicability for future missions to other airless bodies (e.g., Mercury, asteroids), where similar XRF approaches can be used to decode planetary surface evolution and resource potential.

In summary, the Chandrayaan-2 CLASS instrument’s integration of instrumental advances, rigorous data analysis, and geochemical modeling establishes a new benchmark in remote X-ray spectrometry for planetary science, expanding both the spatial and compositional resolution of global lunar chemistry maps and enabling a spectrum of studies into early Solar System crustal processes and planetary differentiation mechanisms (Narendranath et al., 2013, Kumar et al., 21 Aug 2025).