ST-E1: Tokamak Fusion & E1 Applications

• ST-E1 is a multi-context designation encompassing Tokamak Energy’s low aspect-ratio fusion concept and varied electric-dipole (E1) applications across nuclear physics, astrophysics, and engineering.
• Disruption modelling using tools like MAXFEA and DIV3D informs critical reactor design decisions and first-wall protection strategies for ST-E1.
• E1 usage spans from nuclear spectroscopy and reaction theory to photon-strength modelling in atomic and quarkonium studies, highlighting its domain-dependent interpretations.

ST-E1 is a context-dependent designation rather than a single universally standardized term. In the literature considered here, the exact label most specifically denotes Tokamak Energy Ltd.’s low aspect-ratio tokamak fusion power plant concept, while the closely related token $E1$ denotes electric-dipole observables and transitions across nuclear structure, nuclear astrophysics, halo effective field theory, quarkonium, and highly charged ions; it also appears in unrelated names such as BioSerenity-E1 and in E1 HEMP grid analysis (Scarpari et al., 18 Dec 2025, Bettinardi et al., 13 Mar 2025, Das et al., 2024). The technical meaning is therefore set by disciplinary context.

1. Exact and extended usages

In the material considered here, the designation splits into three principal usages.

Usage	Technical sense	Representative source
ST-E1	Tokamak Energy fusion power plant concept	(Scarpari et al., 18 Dec 2025)
E1	Electric-dipole response, transition, or strength	(Nesterenko et al., 2014)
E1 as proper-name suffix	Unrelated naming in HEMP and EEG work	(Das et al., 2024, Bettinardi et al., 13 Mar 2025)

The exact string ST-E1 appears as the name of a reactor concept in disruption modelling for a low aspect-ratio tokamak fusion power plant (Scarpari et al., 18 Dec 2025). By contrast, most of the remaining literature uses E1 in its standard spectroscopy and reaction-theory sense: electric dipole strength, electric-dipole transitions, or the $E1$ photon strength function. Those usages span low-lying dipole response in nuclei, astrophysical capture, halo breakup, quarkonium radiative decays, and tungsten-ion spectroscopy (Nesterenko et al., 2014, Ando, 2018, Göbel et al., 2022, Steinbeißer et al., 2017, Ding et al., 2016).

A further layer of ambiguity is purely nominal. The paper on EEG foundation models explicitly states that it does not use the term “ST-E1” anywhere and that the relevant model is BioSerenity-E1, while the HEMP paper uses E1 for the early-time component of HEMP rather than for electric-dipole physics (Bettinardi et al., 13 Mar 2025, Das et al., 2024). This suggests that any encyclopedic treatment of ST-E1 must separate exact nomenclature from broader lexical overlap.

2. ST-E1 as a tokamak fusion power plant concept

In its exact designation, ST-E1 is Tokamak Energy’s fusion power plant concept developed under the U.S. DOE Milestone Program, based on a low aspect-ratio tokamak with HTS magnets. The disruption-modelling paper treats the pre-conceptual design stage and studies two reactor-scale design points used to support ST-E1 development: ST425 and ST500 (Scarpari et al., 18 Dec 2025).

Parameter	ST425	ST500
$R_0$	$4.25\ \mathrm{m}$	$5.0\ \mathrm{m}$
Aspect ratio $A$	$2.15$	$2.30$
$B_T$	$4.0\ \mathrm{T}$	$E1$0
$E1$1	$E1$2	$E1$3
Plasma cross-section	$E1$4	$E1$5
$E1$6	$E1$7	$E1$8
$E1$9	$R_0$0	$R_0$1

The plant architecture already reflects strong disruption sensitivity. The machine includes 5 PF coil pairs and a 10-segment central solenoid, a 36-sector full-height modular breeding blanket, passive stabilising structures on the outboard side, and a vacuum vessel that must carry vacuum, thermal, seismic, and disruption loads. The vacuum vessel is described as a 50 mm double-shell 316LN-grade structure with internal toroidal and poloidal ribs, intentionally arranged to reduce effective toroidal conductivity and thereby limit induced eddy currents during fast transients (Scarpari et al., 18 Dec 2025).

Within this usage, ST-E1 is not merely a machine name. The paper frames disruption analysis as providing “critical design drivers, particularly for the VV design and the placement and protection strategy of PFCs.” A plausible implication is that, at least in this literature, ST-E1 functions as a design-program identifier around which plasma physics, electromagnetic load definition, and first-wall protection are organized.

3. Disruption modelling as the central ST-E1 design tool

For ST-E1 in the fusion sense, disruption modelling is treated as an integrated physics-to-engineering workflow rather than as a post hoc verification step. The paper argues that, although future fusion plants will need disruption avoidance and mitigation, complete avoidance is unattainable, so qualification requires understanding the consequences of unmitigated events. The workflow is split between physics modelling and engineering assessment, with MAXFEA used for free-boundary equilibrium and disruption evolution and DIV3D used for first-wall heat-flux estimation (Scarpari et al., 18 Dec 2025).

The disruption sequence follows the standard decomposition into Thermal Quench (TQ), Current Quench (CQ), and halo-current evolution. The adopted timings are $R_0$2, $R_0$3 for ST425, and $R_0$4 for ST500, with normalized CQ scaling $R_0$5 (Scarpari et al., 18 Dec 2025). The principal force mechanisms are global eddy-current loads from $R_0$6 and local halo-current loads from $R_0$7.

A core result is that configuration asymmetry matters more than eddy-current magnitude alone. Across both ST425 and ST500, vacuum-vessel eddy-current magnitudes remain broadly similar across DN, DDN, and SN, but halo currents and vertical-force asymmetry increase progressively from DN to DDN to SN. The paper identifies the SN VDE toward the null as the worst-case scenario for ST-E1. In ST425, the SN downward VDE gives halo current peak $R_0$8, halo fraction $R_0$9, and peak VV vertical force from induced currents $4.25\ \mathrm{m}$0. In ST500, the corresponding values are $4.25\ \mathrm{m}$1, $4.25\ \mathrm{m}$2, and $4.25\ \mathrm{m}$3 (Scarpari et al., 18 Dec 2025).

The same modelling also constrains plasma-facing component strategy. For ST500 DN upward VDE, DIV3D calculations used $4.25\ \mathrm{m}$4 and $4.25\ \mathrm{m}$5. Under those assumptions, a bare wall sees large upper outer wall areas above $4.25\ \mathrm{m}$6, whereas a limiter configuration with $4.25\ \mathrm{m}$7 and $4.25\ \mathrm{m}$8 nearly eliminates breeder-wall loading, although each limiter can see $4.25\ \mathrm{m}$9 (Scarpari et al., 18 Dec 2025). This led directly to a modular sacrificial-limiter strategy. In this exact ST-E1 usage, disruption modelling is therefore a plant-defining methodology.

4. E1 as electric-dipole response in nuclear structure

Outside the fusion-plant usage, $5.0\ \mathrm{m}$0 most commonly denotes electric-dipole response. In nuclear structure this does not correspond to a single mode. In $5.0\ \mathrm{m}$1Pb, Skyrme-RPA calculations show that the low-lying $5.0\ \mathrm{m}$2 strength usually identified as the pygmy dipole resonance overlaps strongly with the toroidal dipole resonance in the $5.0\ \mathrm{m}$3 region. The key result is that the familiar neutron-surface-dominant transition-density pattern is real, but the summed RPA current transition densities in $5.0\ \mathrm{m}$4 are predominantly isoscalar toroidal, leading to the interpretation that the isoscalar PDR can be understood as a local surface manifestation of collective toroidal motion (Nesterenko et al., 2014).

Systematics from neutron-rich isotopes between $5.0\ \mathrm{m}$5 and $5.0\ \mathrm{m}$6 complicate any simple identification of low-energy $5.0\ \mathrm{m}$7 strength with neutron-skin thickness. In canonical-basis time-dependent Hartree-Fock-Bogoliubov calculations over more than 350 isotopes, the PDR fraction is defined by

$5.0\ \mathrm{m}$8

while the dipole polarizability is

$5.0\ \mathrm{m}$9

The paper finds that the correlation between neutron-skin thickness and PDR fraction strongly depends on neutron number, whereas the correlation with $A$0 is much more stable; for Sn isotopes with $A$1 to $A$2, the PDR–skin slope is about half that found in Ge isotopes, and the PDR component “jumps up again” above $A$3 (Ebata et al., 2013).

In rare-earth nuclei below about 4 MeV, the low-lying $A$4 problem takes a different form. Nuclear resonance fluorescence data and $A$5-IBM calculations point to enhanced low-energy $A$6 strength in addition to the usual octupole-related $A$7 state. The model Hamiltonian

$A$8

and the one-body $A$9 operator

$2.15$0

support an interpretation in terms of a $2.15$1-boson degree of freedom associated with an $2.15$2-cluster mode and local breaking of isospin symmetry (Spieker et al., 2015).

A still different low-energy $2.15$3 enhancement occurs in $2.15$4Ne. There the transition $2.15$5 between weakly bound excited states at 885 keV and 765 keV is explained by a deformed Woods–Saxon calculation that treats deformation and weak binding simultaneously. With $2.15$6 and $2.15$7, the intrinsic matrix elements

$2.15$8

lead to $2.15$9 when $2.30$0, reproducing the observed order of magnitude and showing that deformation, halo-like radial tails, and weak-binding shell evolution can together generate enhanced $2.30$1 strength (Hamamoto, 2019).

5. E1 in reaction theory and effective field theory

In reaction theory, $2.30$2 often labels a specific radiative channel rather than a collective response mode. The clearest example here is the astrophysical $2.30$3 factor for

$2.30$4

In cluster EFT, the E1 astrophysical factor is defined by

$2.30$5

with the Gamow-peak energy $2.30$6. The NLO analysis finds that only two unfixed parameters remain in the amplitudes after the elastic $2.30$7-wave input is fixed, and the resulting extrapolation gives

$2.30$8

about 30% smaller than several recent phenomenological estimates discussed in the paper (Ando, 2018). A related conference paper reports the same framework as a first result and quotes

$2.30$9

while explicitly noting that the error estimate was still under investigation (Ando, 2019).

Halo EFT supplies another technically distinct $B_T$0 problem: breakup of the $B_T$1 halo nucleus $B_T$2Li. There the observable is the differential dipole strength $B_T$3, and the analysis develops a Møller-operator expansion for final-state interactions that preserves the non-energy-weighted cluster sum rule

$B_T$4

The calculation shows that the neutron-neutron FSI is by far the most important contribution and largely determines the maximum of the $B_T$5 distribution, while $B_T$6 FSI mainly shifts the peak to slightly lower energies. It also finds that good agreement with experiment requires the low-energy $B_T$7-$B_T$8Li interaction to be present in both spin channels rather than only the spin-2 channel (Göbel et al., 2022).

These two EFT cases illustrate different roles of the same notation. In one case $B_T$9 indexes a radiative capture component in a stellar reaction; in the other it indexes continuum dipole breakup of a Borromean halo nucleus. The shared label does not imply shared dynamics.

6. E1 in quarkonium, highly charged ions, and photon-strength modelling

In heavy quarkonium, $4.0\ \mathrm{T}$0 denotes radiative electric-dipole transitions between bound states. A weak-coupling pNRQCD calculation provides the first numerical evaluation of the complete set of relativistic corrections of relative order $4.0\ \mathrm{T}$1 for

$4.0\ \mathrm{T}$2

The master width formula is built around

$4.0\ \mathrm{T}$3

with additional recoil, spin-dependent, and wave-function corrections. The final predictions are

$4.0\ \mathrm{T}$4

$4.0\ \mathrm{T}$5

$4.0\ \mathrm{T}$6

while also showing strong renormalization-scale sensitivity driven mainly by radiative corrections to the static potential (Steinbeißer et al., 2017).

In atomic spectroscopy, the same notation labels allowed line arrays. For Ca-like tungsten $4.0\ \mathrm{T}$7, the E1 transitions studied are those between

$4.0\ \mathrm{T}$8

The MCDF calculation finds about 466 possible E1 lines, with the strongest lines concentrated around 2.95–3.25 nm and 1.86–1.96 nm. The paper emphasizes that for the tabulated transitions the relative deviations between Babushkin and Coulomb gauge results are $4.0\ \mathrm{T}$9, and it predicts several strong lines in the 1.86–1.96 nm region with $E1$00-values of order $E1$01, such as the 1.8593 nm line with $E1$02 (Ding et al., 2016).

A more statistical use of $E1$03 appears in photon-strength modelling. The paper on practical expressions for E1 photon strength functions introduces the TSE model, in which the response of a low-energy state (LES) and the giant dipole resonance (GDR) is represented by two coupled damped states. The basic relation

$E1$04

is combined with a susceptibility built from the coupled LES–GDR response. In the limit $E1$05, the model reduces to two independent Lorentzian-like components; with nonzero coupling, it improves the fit quality for spherical nuclei relative to SLO-type descriptions and better reproduces QRPA and QTBA strength distributions (Plujko et al., 2016).

7. Non-physics and infrastructure uses of the designation

Some occurrences of the designation are unrelated to electric-dipole physics. In E1 HEMP grid analysis, $E1$06 denotes the early-time component of a high-altitude electromagnetic pulse. The Bayesian component-failure paper develops statistical fragility models for use in Sandia’s HEMP Transmission Consequence Model, with failure probability expressed as a CDF conditional on insult voltage,

$E1$07

Its specific methodological contribution is to combine sparse laboratory tests, subject matter expert priors, Bayesian optimization, and MCMC or NUTS sampling so that grid-wide E1 consequence studies can use runtime-efficient CDF sampling rather than deterministic per-component damage modelling (Das et al., 2024).

In clinical machine learning, the relevant term is BioSerenity-E1, not ST-E1. The paper explicitly states that it does not use the term ST-E1 anywhere and that BioSerenity-E1 is the pretrained masked-token predictor model produced by a two-stage self-supervised EEG pipeline. The model is pretrained on 4,005 EEG hours, uses spectral tokenization based on log-multitaper spectra, and applies 70% block masking in the masked-token stage. It reports AUROC $E1$08 and Sensitivity $E1$09 for seizure detection, AUPRC $E1$10 on proprietary normal/abnormal classification and 0.910 on TUH-Abnormal, and Weighted F1 $E1$11 for multiclass pathology differentiation (Bettinardi et al., 13 Mar 2025).

Taken together, these non-physics examples show that the lexical string E1 is not semantically stable across research domains. In the exact sense, ST-E1 is a fusion-reactor concept; in the dominant physics sense, $E1$12 denotes electric-dipole operators, strengths, and transitions; and in engineering or machine-learning contexts it may instead label an electromagnetic-pulse regime or a model family identifier. The designation is therefore encyclopedically tractable only when its domain is made explicit.