Electroluminescence (S2) in Dual-Phase TPCs

Updated 22 April 2026

Electroluminescence (S2) region is the gaseous gap above liquid argon where ionization electrons generate secondary scintillation, enabling precise 3D vertex and energy reconstruction.
It encompasses both VUV excimer emission above a sharp field threshold and broadband neutral bremsstrahlung, with its yield modulated by electric fields and minor impurities such as nitrogen.
Optimized design of the S2 region through controlled gap geometry and photodetector placement enhances light collection, spatial resolution, and low-threshold detection in rare-event searches.

Electroluminescence (S2) Region

The electroluminescence (S2) region in dual-phase argon time projection chambers (TPCs) is a well-defined gaseous gap above the liquid argon, in which ionization electrons extracted from the liquid are accelerated by a strong electric field and cause secondary scintillation (proportional electroluminescence). The S2 region enables precise reconstruction of deposited energy, three-dimensional vertex localization, and background rejection via the spatial and temporal pattern of light emission. S2 production in argon involves both vacuum ultraviolet (VUV) excimer emission and broadband visible/near-infrared electroluminescence via neutral bremsstrahlung; its performance is strongly dependent on geometrical configuration, electrodynamical fields, photosensor placement, and even minor impurity dopants such as nitrogen. The architecture and modeling of the S2 region are thus fundamental for the sensitivity, threshold, and background rejection of next-generation low-threshold rare-event experiments employing dual-phase argon TPCs (Yin et al., 9 Feb 2026, Buzulutskov et al., 2018, Bondar et al., 2015).

1. Physical Mechanisms of S2 Signal Production

Two microscopic processes produce electroluminescence (EL) photons in the argon S2 region:

Excimer (VUV) Emission: Electrons with energy above the 11.55 eV argon excitation threshold collisionally excite Ar atoms (Ar*). These then form high-lying excimer states (Ar $_2^*$ ), which decay with photon emission at λ ≈ 128 nm (VUV). The VUV yield is sharply thresholded at a reduced field $E/N \approx 4.0$ Td, rising linearly above threshold (Buzulutskov et al., 2018, Aoyama et al., 2021, collaboration et al., 2020).
Neutral Bremsstrahlung (NBrS): Below the excitation threshold, electrons scatter elastically off neutral Ar atoms and emit photons, generating a continuous spectrum from UV through NIR (∼200–1000 nm) with no field threshold; NBrS operates for all nonzero E/N, though with lower yield than VUV EL (Buzulutskov et al., 2018, collaboration et al., 2020, Aoyama et al., 2021).

The relative importance of these processes depends on field and impurity content. At high field, excimer VUV dominates; at lower field, or for wavelength ranges beyond ∼300 nm, NBrS is substantial. In doped argon, e.g., with ∼50 ppm N $_2$ , Ar* transfers energy to N $_2$ , producing strong NUV emission via the N $_2$ Second Positive System (260–430 nm), enhancing S2 detectability (Bondar et al., 2016, Bondar et al., 2015).

2. Geometrical and Electrostatic Modeling of the S2 Region

Gas Pocket Geometry and Emission Modeling

In state-of-the-art dual-phase TPCs, the S2 region is implemented as a cylindrical argon gas gap above the liquid interface, with typical thicknesses of 5–20 mm. For instance, in one compact TPC, the S2 region is a cylinder of 90 mm diameter and 7 mm thickness, matching the DarkSide-50 configuration to ensure field uniformity and photon yields of order $10^3$ per drifting electron (Yin et al., 9 Feb 2026).

To model continuous photon emission along the vertical (z) axis, the EL gap is subdivided into K equal-thickness (e.g., 1 mm) slices. Each slice at depth $z_k$ emits a uniform number of photons per ionization electron, enabling calculation of the light-response function at each (x, y, z) using semi-analytic solid-angle relations.

Geometrical Solid-Angle (GSA) Response

The GSA method models the fraction of emitted S2 photons collected by each photosensor (e.g., PMT) by integrating the solid angle subtended by the sensor over all emission slices:

$\Omega_i(x,y;z_k) = 4\left[ \omega(x_2-x, y_2-y; z_k + h_{\rm ref}) -\omega(x_1-x, y_2-y; z_k + h_{\rm ref}) -\omega(x_2-x, y_1-y; z_k + h_{\rm ref}) +\omega(x_1-x, y_1-y; z_k + h_{\rm ref}) \right]$

$\omega(x, y; z) = \arctan\left(\frac{x\,y}{z\,\sqrt{x^2 + y^2 + z^2}}\right)$

Total solid angle per PMT is then $\bar\Omega_i(x,y) = \sum_{k=1}^K \Omega_i(x,y;z_k)$ for event vertex (x, y) (Yin et al., 9 Feb 2026).

The normalized probability that a photon is detected by channel $E/N \approx 4.0$ 0 at position (x, y) is: $E/N \approx 4.0$ 1 XY position reconstruction is performed by minimizing the difference between observed and expected photoelectron fractions via a $E/N \approx 4.0$ 2 statistic.

3. Optimization of Photosensor Placement and Gap Thickness

The performance of S2-based event reconstruction depends critically on the vertical placement of the photodetector plane above the S2 region. Scanning the PMT height (h) from 0 to 50 mm produces strongly non-monotonic dependence of both reconstruction bias and spatial resolution, as shown in the following data (Yin et al., 9 Feb 2026):

h (mm)	⟨Δr⟩ (41.5 keV) [mm]	σ(Δr) (41.5 keV) [mm]	⟨Δr⟩ (1.0 keV) [mm]	σ(Δr) (1.0 keV) [mm]
0	2.5 ± 0.4	1.9 ± 0.3	4.2 ± 0.5	3.1 ± 0.4
5	1.6 ± 0.2	1.4 ± 0.2	2.0 ± 0.3	1.8 ± 0.3
10	1.1 ± 0.2	1.1 ± 0.2	2.2 ± 0.3	1.9 ± 0.3
50	1.8 ± 0.3	1.6 ± 0.3	3.4 ± 0.4	2.9 ± 0.4

At very small h, light is highly localized, degrading position sensitivity.
At large h, light is nearly uniform across sensors, again washing out XY discrimination, with photon statistics deteriorating due to lower PE yield.
There exists an optimum (h ≈ 10 mm for ∼40 keV events; h ≈ 5 mm for ∼1 keV events) where both light-sharing gradient and photon statistics are favorable.

Gas-gap thickness of 5–10 mm provides ample S2 gain while keeping vertical extent compatible with simple GSA-based models.

4. Spectroscopic Composition and Enhancement Mechanisms

VUV and Non-VUV Components

VUV emission ( $E/N \approx 4.0$ 3nm): Excimer decay; sharp threshold at E/N ≈ 4.0 Td.
Neutral Bremsstrahlung (NBrS): Broadband continuum 200–1000 nm; present at all E/N; dominant below VUV threshold and for detectors employing bare-photosensor readout (Buzulutskov et al., 2018, collaboration et al., 2020, Aoyama et al., 2021).
Nitrogen Doping ( $E/N \approx 4.0$ 4 ppm N₂): Facilitates two-body energy transfer, producing N₂*(C $E/N \approx 4.0$ 5), which emits via SPS at 260–430 nm. NUV photons enjoy ∼20× higher light-collection efficiency in PMTs compared to TPB-shifted VUV, resulting in a factor ∼3 increase in S2 pulse size when compared to pure argon (Bondar et al., 2016, Bondar et al., 2015).

At 87 K and 1 atm with 56 ppm N₂, the total amplification parameter is measured at 109 ± 10 photons/e/kV, with 51 ± 6% emitted in the N₂ UV band. NBrS contribution in pure Ar is of order 0.2–0.5 photons/e/cm in 300–600 nm at E/N = 4–8 Td, and can account for ~50% of S2 emission at operating fields typical for dark-matter detectors (Aoyama et al., 2021, Buzulutskov et al., 2018).

Spectral Dependence and Field Thresholds

The S2 spectrum in pure argon is accurately modeled as a sum of ordinary excimer EL and NBrS: $E/N \approx 4.0$ 6 with field-dependent scale factors, providing high-fidelity prediction of spectral and field dependences (Aoyama et al., 2021).

5. S2 Pulse Shape and Electron Diffusion

The S2 time profile is dominated by the convolution of a uniform electron arrival distribution and two-component excimer decay (singlet τ₁ ∼ 11 ns, triplet τ₂ ∼ 3.2 μs), further smeared by longitudinal diffusion in the liquid:

$E/N \approx 4.0$ 7

(Agnes et al., 2018)

Additional slow (∼4–6 μs) and long (∼40 μs) S2 components, growing with field strength, have been identified and attributed to metastable negative ion trapping and delayed self-detachment, rather than excimer or interface emission. These components become significant at E/N > 4 Td, require multi-exponential fitting, and must be accounted for in S2-only analyses and drift-field optimization (Bondar et al., 2020).

Electron diffusion over the drift length produces longitudinal broadening of the initial cloud: $E/N \approx 4.0$ 8 with measured $E/N \approx 4.0$ 9 cm $_2$ 0/s for E_d = 200 V/cm (Agnes et al., 2018).

6. Implications for Detector Design and Rare-Event Searches

Configuration Optimization: A 5–10 mm gas gap with 5–15 mm PMT array height is optimal, balancing light-sharing for XY reconstruction and photon statistics for triggering at low energy thresholds (Yin et al., 9 Feb 2026).
Light Collection Strategies: Employing direct readout of the S2 region using bare SiPMs or PMTs (without TPB) leverages the sizable NBrS yield, simplifying construction and boosting total S2 photon detection (Buzulutskov et al., 2018, collaboration et al., 2020).
N₂ Doping: Controlled nitrogen doping ( $_2$ 1 ppm) enhances NUV emission and S2 signal yield, at the cost of requiring precise impurity control to prevent S1 and drift degradation (Bondar et al., 2016, Bondar et al., 2015).
Pulse-Shape Modeling: S2 waveforms must be modeled with multi-exponential fits to accurately integrate over all emission components, especially for z-positioning and electron diffusion measurements (Bondar et al., 2020, Agnes et al., 2018).
Low-Threshold Operation: Enhanced photon and PE yield from optimized geometry and spectral collection, combined with improved XY and z resolution via modeling, directly lowers the threshold for nuclear-recoil and neutrino-coherent scattering detection.

Future S2 region designs should:

Enable fine mechanical adjustment of the PMT (or SiPM) height to access the optimal trade-off between light-sharing and photon statistics for the entire event energy range.
Incorporate slice-dependent EL yields $_2$ 2 to accommodate field nonuniformities and improve spatial reconstruction fidelity.
Exploit visible and NIR photon detection (NBrS and N₂ SPS) with fully wavelength-shifter-free, high-coverage SiPM readouts to simplify scaling and mitigate aging effects in very large TPCs (collaboration et al., 2020).
Develop pulse-shape discrimination and z-reconstruction algorithms robust against slow emission components.
Extend S2 pattern modeling to include real-world effects: electron diffusion, field gradients, impurity-induced spectral features, and multiple EL mechanisms.

Collectively, the precise engineering of the S2 region—including geometry, field configuration, emission modeling, and spectral optimization—forms the foundation for high-performance fiducialization, background rejection, and single-electron sensitivity in dual-phase argon TPCs for rare-event searches (Yin et al., 9 Feb 2026, Buzulutskov et al., 2018, Bondar et al., 2016, Bondar et al., 2015, Agnes et al., 2018, collaboration et al., 2020, Aoyama et al., 2021, Bondar et al., 2020).