Soft X-ray tomography

Overview

Soft X-ray (SXR) tomography is a critical diagnostic method in magnetic confinement fusion research for visualizing the internal structure of high-temperature plasmas. The technique reconstructs the two-dimensional (2D) or three-dimensional (3D) spatial distribution of SXR emissivity from a set of line-integrated measurements taken along multiple chords viewing a single poloidal cross-section. Because SXR emission from a fusion plasma is highly sensitive to electron temperature, electron density, and impurity concentration, its spatial profile provides invaluable information about the plasma's state.

The primary application of SXR tomography is the study of magnetohydrodynamic (MHD) instabilities. These instabilities, which manifest as fluctuations in temperature and density, create distinct patterns in the SXR emissivity profile. SXR tomography allows physicists to identify the mode numbers (poloidal and toroidal) of tearing modes, visualize the sawtooth crash mechanism, observe the growth of locked modes that can lead to disruptions, and track the movement of impurity ions within the plasma. This detailed imaging capability is essential for developing control strategies to mitigate or avoid these performance-limiting events, directly impacting the path toward achieving a stable, high-performance fusion plasma.

Physics / Mechanism

The operation of SXR tomography is based on two core principles: the physics of SXR emission from a plasma and the mathematical techniques of tomographic reconstruction.

Emission Physics

The SXR radiation measured by the diagnostic originates from three primary atomic processes within the plasma:

1. Bremsstrahlung (free-free emission): Produced when free electrons are decelerated by collisions with ions. Its emissivity is strongly dependent on electron density (nₑ), electron temperature (Tₑ), and the effective ionic charge (Z_eff), scaling approximately as ε_brems ∝ nₑ² Z_eff / √Tₑ for a given energy range.
2. Recombination radiation (free-bound emission): Occurs when a free electron is captured by an ion. Its intensity also depends on nₑ, Tₑ, and the impurity ion density.
3. Line radiation (bound-bound emission): Emitted when an electron in an impurity ion transitions to a lower energy state. This is highly dependent on the specific impurity species and their charge states, which are in turn a function of Tₑ.

In hot, relatively clean tokamak plasmas, bremsstrahlung is often the dominant source of continuum SXR emission. The strong dependence on temperature (through its effect on the emission spectrum) and density makes SXR emissivity an excellent tracer for perturbations caused by MHD activity.

Tomographic Reconstruction

The hardware consists of one or more "pinhole cameras" positioned around the vacuum vessel. Each camera contains an array of detectors (e.g., silicon photodiodes, silicon drift detectors, or Gas Electron Multiplier (GEM) detectors) that view the plasma through a common aperture or slit. Each detector measures the line-integrated brightness along its unique chord, or line of sight. The measurement for a single detector p is given by the line integral of the local emissivity ε(r,θ) along its path Lₚ:

Iₚ = ∫_Lₚ ε(r,θ) dl

Tomographic reconstruction is the mathematical process of solving this set of integral equations to recover the 2D local emissivity function ε(r,θ) from the collection of measurements {Iₚ}. This is a classic ill-posed inverse problem, as the data are limited and noisy. Several algorithms are used:

• Fourier-Bessel Expansion: The emissivity is expanded as a series of basis functions (typically Zernike polynomials or Bessel functions) that are well-suited to the circular geometry of a tokamak cross-section. The expansion coefficients are determined by a least-squares fit to the experimental data.
• Maximum Entropy (MaxEnt): This method seeks the smoothest, most physically plausible image consistent with the measured data by maximizing an entropy functional. It is robust against noise and can handle sparse data sets.
• Algebraic Reconstruction Technique (ART): The plasma cross-section is discretized into a grid of pixels. The reconstruction becomes a large system of linear equations, which is then solved iteratively.

The quality of the reconstruction depends heavily on the number and geometric arrangement of the detector arrays, the signal-to-noise ratio, and the choice of reconstruction algorithm.

Historical Development

The application of SXR imaging to fusion plasmas began in the 1970s. Early systems on devices like the Princeton Large Torus (PLT) used simple, single-aperture pinhole cameras with a few detectors to observe sawtooth oscillations. These initial measurements, reported by von Goeler et al. in 1974, provided the first direct evidence of the internal disruption associated with the m=1 MHD mode at the core of the plasma [1].

Throughout the 1980s and 1990s, SXR diagnostics evolved significantly. The number of detectors and cameras increased, providing more viewing chords and enabling true 2D tomographic reconstruction for the first time on tokamaks like JET and TFTR. The development of more sophisticated reconstruction algorithms, such as the Fourier-Bessel methods pioneered by Cormack and later adapted for plasmas by Sautenkov and others, allowed for detailed quantitative analysis of MHD mode structures [2].

The 2000s saw major advances in detector technology. The move from simple photodiodes to faster, more sensitive detectors like silicon drift detectors (SDDs) and GEM detectors improved both temporal and spatial resolution. For instance, the SXR system on the Alcator C-Mod tokamak provided high-quality data that was crucial for validating models of sawtooth and tearing mode physics [3]. The development of systems with multiple energy channels also allowed for simultaneous measurement of electron temperature fluctuations, a technique known as two-color SXR tomography.

Current Status

As of 2026, SXR tomography is a standard, mature diagnostic on virtually every major magnetic confinement fusion experiment worldwide, including tokamaks, stellarators, and reversed-field pinches. Modern systems are characterized by high channel counts, high throughput, and advanced data analysis pipelines.

State-of-the-art systems, such as those planned for ITER, are designed with hundreds of viewing chords to provide high-resolution reconstructions of the plasma cross-section. The ITER SXR system, for example, is specified to have a spatial resolution of a/50 (where 'a' is the minor radius) and a temporal resolution of 10 µs, enabling detailed studies of fast MHD events like edge localized modes (ELMs) and disruptions [4].

Detector technology continues to advance. GEM detectors are increasingly used for their radiation hardness, high count rate capability, and large sensitive areas, making them suitable for the harsh nuclear environment of future D-T reactors [5]. Fast data acquisition and real-time processing are also key areas of development. On several devices, SXR data is now being integrated into real-time plasma control systems. For example, signals from SXR detectors can be used to detect the growth of tearing modes, triggering electron cyclotron current drive (ECCD) to stabilize the mode before it leads to a disruption.

Notable Implementations

• JET (Culham, UK): The Joint European Torus has long operated one of the most advanced SXR diagnostic suites. Its two large cameras, each with over 100 lines of sight, have provided benchmark data on sawtooth physics, neoclassical tearing modes (NTMs), and impurity transport for decades [6].
• DIII-D (San Diego, USA): The SXR system on DIII-D is tightly integrated with its plasma control system. It is used for real-time detection and characterization of NTMs, enabling targeted stabilization using its ECCD system. This represents a key step towards active disruption avoidance.
• EAST (Hefei, China): The Experimental Advanced Superconducting Tokamak features a comprehensive SXR system with over 200 channels and multiple energy ranges. It has been instrumental in studying MHD phenomena in long-pulse, high-performance operating scenarios relevant to future reactors.
• Wendelstein 7-X (Greifswald, Germany): Implementing SXR tomography on the complex 3D geometry of a stellarator presents unique challenges. The W7-X SXR system uses a sophisticated camera design and advanced reconstruction algorithms tailored to its non-axisymmetric magnetic field structure, providing crucial insights into stellarator stability [7].
• ITER (Cadarache, France): The planned SXR system for ITER is a flagship example of a next-generation diagnostic. It is designed for extreme radiation hardness and reliability, with extensive coverage of the poloidal cross-section to meet the demanding physics measurement requirements of a burning plasma experiment.

Open Challenges

Despite its maturity, SXR tomography faces several ongoing challenges, particularly in the context of reactor-scale devices like ITER and DEMO.

• Radiation Hardness: The intense neutron and gamma radiation environment in a D-T burning plasma can damage detectors, electronics, and optical components like beryllium windows. Developing and qualifying radiation-hard components that can survive the full operational lifetime of a reactor is a major engineering challenge [4].
• 3D Reconstruction: While 2D tomography is standard, many MHD phenomena (e.g., resonant magnetic perturbations, disruptions) have complex 3D structures. Reconstructing these features requires multiple SXR arrays at different toroidal locations and significantly more complex inversion algorithms. This is an active area of research on devices like MAST-U and KSTAR.
• Integration and Real-Time Control: Using tomographic reconstructions for real-time plasma control requires extremely fast and robust algorithms. The computational time for a full 2D reconstruction can be a bottleneck. Developing simplified, physics-based models that can extract the necessary control parameters (e.g., mode location, amplitude) from raw SXR data in milliseconds is a key goal.
• Interpretation in Complex Scenarios: In plasmas with high impurity content or significant non-thermal electron populations (e.g., from RF heating), interpreting SXR emissivity can be complicated. Disentangling the contributions of Tₑ, nₑ, Z_eff, and non-thermal effects requires integration with other diagnostics and advanced modeling, moving beyond simple tomographic inversion.

Outlook

The 5-15 year trajectory for SXR tomography is focused on addressing the challenges of burning plasma environments and enhancing its role in integrated plasma control. For ITER, the primary focus will be on the commissioning and operation of its robust, radiation-hard SXR system, which will be a cornerstone for understanding alpha particle physics and MHD stability in a burning plasma.

In parallel, research on existing devices will push the boundaries of real-time analysis. We can expect to see SXR tomography routinely used not just to detect instabilities but to provide detailed spatial information to actuators for precise, feedback-controlled stabilization. This will likely involve the application of machine learning and AI techniques to accelerate the inversion process and identify complex patterns in the data that precede disruptive events.

Advances in detector technology, such as energy-resolving pixelated detectors (e.g., Medipix/Timepix), may become more widespread. These would enable "hyperspectral" SXR tomography, providing simultaneous, spatially resolved measurements of the full SXR spectrum. This would allow for a much more detailed and unambiguous determination of electron temperature, impurity distributions, and non-thermal electron dynamics, significantly enhancing the diagnostic's power and utility for understanding and controlling future fusion reactors.

References

1. Observation of m=1 and m=2 Oscillations in the Soft X-Ray Emission from the Princeton Large Torus — Physical Review Letters (1974)
2. Tomography of soft X-ray emission from the T-10 tokamak plasma — Soviet Journal of Plasma Physics (1983)
3. Fast soft x-ray detector arrays for 2-D plasma imaging on Alcator C-Mod — Review of Scientific Instruments (2001)
4. ITER soft x-ray diagnostic system — Review of Scientific Instruments (2016)
5. GEM based soft X-ray diagnostics for fusion plasmas — Journal of Instrumentation (2009)
6. The new soft X-ray diagnostic for JET — Review of Scientific Instruments (1992)
7. Design of the soft X-ray multi-camera tomography system for Wendelstein 7-X — Review of Scientific Instruments (2014)
8. Tomography and Cross-Disciplinary Applications — Fusion Science and Technology (2008)