Muon-Induced X-Ray Emission (MIXE)

• Muon-Induced X-ray Emission (MIXE) is a non-destructive method leveraging negative muons for precise, depth-resolved elemental and isotopic analysis.
• Advanced tracking and calibration techniques, such as time projection chamber trackers and high-purity Ge arrays, enable sub-micron spatial resolution and high-energy muonic X-ray spectroscopy.
• Innovative imaging methods like sphere encoding and tomographic reconstruction extend MIXE applications to fields including cultural heritage, battery research, and fusion diagnostics.

Muon-Induced X-ray Emission (MIXE) is a non-destructive analytical approach for depth-resolved, element- and isotope-specific characterization of materials. MIXE leverages the atomic capture of negative muons ($\mu^-$), their subsequent radiative cascade, and the detection of high-energy muonic X-rays ($\mu$X) together with gamma rays from muon nuclear capture. By controlling the incident muon beam momentum, the stopping depth of $\mu^-$ in target materials can be precisely tuned from microns to centimeters, enabling spatially resolved composition analysis in fragile, valuable, or operando samples. Systematic advances in tracking, calibration, detector hardware, and simulation methodologies have substantially enhanced the resolution, sensitivity, and applicability of MIXE across fields ranging from battery research and cultural heritage to fusion diagnostics.

1. Physical Basis and Key Mechanisms

Negative muons lose energy in matter via ionization and excitation, governed by the Bethe–Bloch formalism. Upon thermalization, $\mu^-$ is captured into high-$n$ atomic orbits of nuclei, forming muonic atoms with reduced mass $\mu = m_\mu M / (m_\mu + M)$ ($m_\mu \simeq 207\, m_e$). The non-relativistic Bohr model gives muonic level energies:

$$E_n = \frac{\mu e^4}{2 \hbar^2 n^2} \sim Z^2 \mu,\quad E_{\rm transition}^{\mu} \simeq 207 E_{\rm transition}^{e}$$

Muonic X-ray transitions, particularly K- and L-series, emerge from the radiative cascade ($n\to n'$ transitions), emitting high-energy X-rays (e.g., $\mu$C–K$\mu$0 keV, $\mu$1Cu–K$\mu$2 MeV). After descending to the 1s state, the muon may either decay or undergo nuclear capture ($\mu$3), giving rise to prompt $\mu$4-rays characteristic of the daughter nucleus. The yield per incident $\mu$5 is

$\mu$6

with $\mu$7 the capture probability (scaling $\mu$8 for low $\mu$9) and $\mu^-$0–$\mu^-$1 (probability of radiative de-excitation) (Li et al., 2024).

2. Instrumentation and Calibration Methodologies

Advances in tracking and calibration have substantially increased MIXE spatial and compositional resolution. The ultra-low material budget twin GEM-based Time Projection Chamber (TPC) tracker (Zhao et al., 17 Jan 2025) enables precise 3D trajectory reconstruction for incident muons. Each TPC drift chamber contains:

• Drift region: uniform $\mu^-$2 ($\mu^-$3 V/cm for Ar/CO$\mu^-$4 (75:25), $\mu^-$5 V/cm for He/CO$\mu^-$6 (90:10)).
• Triple-GEM amplification stack (gain $\mu^-$7 per stage).
• 2D micropattern readout (strip/pad, pitch $\mu^-$8–$\mu^-$9m).

Drift velocity $\mu^-$0 is calibrated via a custom scintillating-fiber detector with SiPM readout, mounted upstream of the tracker. The drift velocity from fiber timing:

$\mu^-$1

where $\mu^-$2 mm (fiber spacing), $\mu^-$3 is the peak time offset. The permille-level accuracy ($\mu^-$4) enables $\mu^-$5m Y-resolution; X/Z resolution from charge centroiding is $\mu^-$6–$\mu^-$7m. Repeated calibrations maintain $\mu^-$8 drift $\mu^-$9 over hours.

A high-purity Ge array (e.g., GIANT setup at PSI (Gerchow et al., 2022)) provides energy and time-resolved $n$0X/$n$1 detection (energy resolution $n$2 keV at 1 MeV, timing $n$3 ns). Baseline and ELET timing corrections reduce pile-up and resolve prompt/delayed signals.

3. Beam Source, Simulation, and Depth Profiling

Beam momentum tuning is essential for depth-resolved MIXE. At PSI, the continuous $n$4E1 beamline delivers negative muons with $n$5–$n$6 MeV/c (kinetic energies $n$7–$n$8 MeV) at fluxes up to $n$9 kHz (Biswas et al., 2022). The continuous time structure ($\mu = m_\mu M / (m_\mu + M)$0 duty factor) achieves pile-up probability $\mu = m_\mu M / (m_\mu + M)$1 at $\mu = m_\mu M / (m_\mu + M)$2 kHz rates for $\mu = m_\mu M / (m_\mu + M)$3s-acquisition systems, with SNR$\mu = m_\mu M / (m_\mu + M)$4 for characteristic lines (e.g., Cu $\mu = m_\mu M / (m_\mu + M)$5 at $\mu = m_\mu M / (m_\mu + M)$6 MeV in 30 min acquisition).

Muon stopping profiles as a function of momentum ($\mu = m_\mu M / (m_\mu + M)$7) are modeled with SRIM, GEANT4, and PHITS. For multilayer targets, GEANT4 and PHITS agree on stopping depth to $\mu = m_\mu M / (m_\mu + M)$8; SRIM gives fast estimates with layer-dependent biases ($\mu = m_\mu M / (m_\mu + M)$9) (Lamotte et al., 15 Jan 2026). Muonic X-ray spectra simulated via PHITS capture cascade intensities but have K-line energy offsets ($m_\mu \simeq 207\, m_e$0–$m_\mu \simeq 207\, m_e$1 for $m_\mu \simeq 207\, m_e$2); hybrid analysis using MUDIRAC-formulated energies restores sub-keV accuracy.

Profiling allows extraction of stopping depth variations from $m_\mu \simeq 207\, m_e$3–$m_\mu \simeq 207\, m_e$4m up to $m_\mu \simeq 207\, m_e$5 mm. Elemental sensitivity reaches $m_\mu \simeq 207\, m_e$6–$m_\mu \simeq 207\, m_e$7 wt% in 30 min with momentum scanning; detection limits scale as $m_\mu \simeq 207\, m_e$8.

4. Advanced Imaging: Sphere Encoding and Tomographic MIXE

Coded-aperture and sphere-encoded imaging methodologies extend MIXE to depth-resolved, element-specific tomography. Sphere encoding employs a high-$m_\mu \simeq 207\, m_e$9 sphere (radius $$E_n = \frac{\mu e^4}{2 \hbar^2 n^2} \sim Z^2 \mu,\quad E_{\rm transition}^{\mu} \simeq 207 E_{\rm transition}^{e}$$0) with a pattern of small aperture holes or zone plates. The target (e.g., ICF capsule) sits upstream; a pixelated detector array (CdTe, CdZnTe) records the modulated X-ray signal (Li et al., 2024). The curvature preserves uniform magnification and minimizes tilt-induced blurring. The spatial resolution for spherical coded-imaging systems:

$$E_n = \frac{\mu e^4}{2 \hbar^2 n^2} \sim Z^2 \mu,\quad E_{\rm transition}^{\mu} \simeq 207 E_{\rm transition}^{e}$$1

where $$E_n = \frac{\mu e^4}{2 \hbar^2 n^2} \sim Z^2 \mu,\quad E_{\rm transition}^{\mu} \simeq 207 E_{\rm transition}^{e}$$2 is detector pixel size, $$E_n = \frac{\mu e^4}{2 \hbar^2 n^2} \sim Z^2 \mu,\quad E_{\rm transition}^{\mu} \simeq 207 E_{\rm transition}^{e}$$3 magnification, $$E_n = \frac{\mu e^4}{2 \hbar^2 n^2} \sim Z^2 \mu,\quad E_{\rm transition}^{\mu} \simeq 207 E_{\rm transition}^{e}$$4 wavelength, $$E_n = \frac{\mu e^4}{2 \hbar^2 n^2} \sim Z^2 \mu,\quad E_{\rm transition}^{\mu} \simeq 207 E_{\rm transition}^{e}$$5 penetration blur, $$E_n = \frac{\mu e^4}{2 \hbar^2 n^2} \sim Z^2 \mu,\quad E_{\rm transition}^{\mu} \simeq 207 E_{\rm transition}^{e}$$6 manufacturing error. For $$E_n = \frac{\mu e^4}{2 \hbar^2 n^2} \sim Z^2 \mu,\quad E_{\rm transition}^{\mu} \simeq 207 E_{\rm transition}^{e}$$7C-K$$E_n = \frac{\mu e^4}{2 \hbar^2 n^2} \sim Z^2 \mu,\quad E_{\rm transition}^{\mu} \simeq 207 E_{\rm transition}^{e}$$8 at $$E_n = \frac{\mu e^4}{2 \hbar^2 n^2} \sim Z^2 \mu,\quad E_{\rm transition}^{\mu} \simeq 207 E_{\rm transition}^{e}$$9 keV, resolutions $n\to n'$0m are achievable.

Encoded images $n\to n'$1 are reconstructed via Wiener filtering or iterative Richardson–Lucy deconvolution. Geant4 simulations confirm analytic projections, and experimental setups yield layer-by-layer maps with $n\to n'$2m depth resolution and elemental sensitivity to $n\to n'$3 wt% W dopant.

5. Detector Technologies: Calorimetric and Germanium Arrays

Dedicated detector development has enabled absolute nuclear charge radii measurements and high-fidelity elemental analysis in MIXE. The QUARTET collaboration at PSI applies metallic magnetic calorimeter (MMC) arrays (maXs-30 modules) for low-energy ($n\to n'$4–$n\to n'$5 keV) muonic atom X-ray spectroscopy (Unger et al., 2023). The detector integrates 64 gold absorbers ($n\to n'$6m thick, $n\to n'$7 area), thermally coupled to Au:Er paramagnetic sensors read out via SQUID multiplexing, baseline resolution $n\to n'$8 eV at $n\to n'$9 mK.

Detection efficiency is determined by window transmission and absorber quantum efficiency:

$\mu$0

Where

$\mu$1

With $\mu$2 layers, efficiencies span $\mu$3–$\mu$4 for $\mu$5–$\mu$6 keV X-rays. Resolving power $\mu$7 exceeds $\mu$8–$\mu$9 (a $\mu$00 gain over semiconductors). For a $\mu$01 kHz beam, count rates for the primary $\mu$02 line in $\mu$03Li reach $\mu$04 s$\mu$05. Statistical precision on line centers is $\mu$06 eV (ppm level) in hours.

In contrast, the GIANT HPGe array at PSI targets heavy/medium-Z elements, delivering sub-keV resolution and timing for bulk analysis, with minimum detection limits $\mu$07 at% in 1 h, and isotope separation for $\mu$08 (Gerchow et al., 2022).

6. Applications and Future Prospects

MIXE's ability to spatially resolve element/isotope distributions non-destructively is being deployed for:

• Archaeometry: Elemental and isotope-specific mapping in bronze artifacts (Al, Cu, Fe, Ni, Pb).
• Battery research: Depth-profiling of Li and transition metals in operando cells (e.g. Li-L lines at $\mu$09 keV).
• Fusion diagnostics: Sub-$\mu$10m elemental mapping in ICF targets, with sphere-encoded imaging (Li et al., 2024).
• Cultural heritage: Tomographic mapping of paint cross-sections, fragile layered objects.
• Meteorite/environmental science: Bulk composition and isotope ratios for provenance.
• Industrial applications: Quality control for alloys, coatings.

In the context of simulation and user access, hybrid PHITS+MUDIRAC analysis tools (DEEP$\mu$11 project) enable predictive design for end-users of MIXE instruments (Lamotte et al., 15 Jan 2026). Automation (real-time beam tuning, sample changing) and expanded detector arrays are underway at PSI (Gerchow et al., 2022). Dynamic (time-resolved) and machine-learning-based reconstruction techniques are in development for advanced tomography.

MIXE now enables micron-scale, 3D, element-resolved, and isotope-specific imaging in a variety of multi-disciplinary contexts, combining continuous high-rate muon sources, advanced tracking/calibration, coded imaging, and spectrometer arrays for quantitative, non-destructive materials analysis.