InDEx: Indian Dark Matter Search Experiment

• Indian Dark Matter Search Experiment (InDEx) is a direct detection initiative employing superheated droplet detectors at JUSL to probe low-mass WIMPs.
• It utilizes advanced GEANT4 simulations and optimized passive shielding to effectively suppress radiogenic and cosmogenic backgrounds.
• The experiment achieves lower energy thresholds and improved exclusion limits, paving the way for future dark matter research.

The Indian Dark Matter Search Experiment (InDEx) is a direct detection initiative based in India and located at the Jaduguda Underground Science Laboratory (JUSL). InDEx is focused on probing the low-mass regime of Weakly Interacting Massive Particles (WIMPs) using superheated droplet technology and aims to set competitive constraints on both spin-independent (SI) and spin-dependent (SD) WIMP-nucleon cross sections. The experiment’s site-specific background simulation, progressive reduction of energy thresholds, and incrementally increasing exposure underpin its strategic approach to enhance sensitivity and complement global dark matter searches.

1. Experimental Framework and Detector Architecture

InDEx utilizes superheated droplet detectors containing active tetrafluoroethane (C₂H₂F₄) droplets suspended in a gel matrix (Kumar et al., 31 Aug 2025, Das et al., 8 Oct 2025). The technology is comparable to historic bubble chamber and superheated emulsion detectors but is optimized for long-term stability and low recoil energy thresholds. Bubble nucleation is a consequence of localized energy deposition (typically by nuclear recoils from WIMP or neutron interactions), controllable through operational temperature and pressure settings.

Piezo-electric acoustic sensors coupled to FPGA-based DAQ systems digitize and record the bubble nucleation events, extracting features such as acoustic power (Pvar). Signal processing, including fast Fourier transform and amplitude-duration analyses, allows discrimination between genuine particle-induced events and noise artifacts (Kumar et al., 31 Aug 2025).

The detector fabrication occurs at the Saha Institute of Nuclear Physics (SINP) and subsequent deployment at JUSL, where the vertical rock overburden (555 m) ensures significant reduction in cosmic ray background. Each detector batch is calibrated in situ using Am–Be neutron sources to characterize the response and set the nucleation threshold, exploiting the neutron-induced recoil spectrum (Kumar et al., 31 Aug 2025).

2. Radiation Background Simulation and Shielding Strategy

Background suppression in InDEx is predicated on detailed GEANT4 simulations of neutron production and transmission through the experimental cavern and shielding (Banik et al., 2020). The simulation encompasses:

• Radiogenic neutron flux: Predominantly from (α,n) reactions in uranium and thorium decay chains, and spontaneous fission (238U), with calculated yields.
• Cosmogenic neutron flux: Arising from muon interactions (disintegration, capture, showers) with the surrounding rock; Gaisser’s parameterization quantifies muon flux at ∼4.49×10⁻⁷ cm⁻²s⁻¹.
• Combined neutron spectrum: For neutrons above 1 MeV, radiogenic flux is 5.75(±0.58)×10⁻⁶ cm⁻²s⁻¹ while cosmogenic is 7.25(±0.65)×10⁻⁹ cm⁻²s⁻¹.

Optimized passive shielding consists of a multilayer arrangement:

• 40 cm polypropylene (hydrogen-rich moderator for MeV neutrons)
• 30 cm lead (gamma attenuation, noting secondary neutron generation)
• 20 cm additional polypropylene (captures secondary neutrons from lead).

This configuration (CFG-4) was determined to optimally suppress both radiogenic and cosmogenic neutron backgrounds, reducing the rate of neutron-induced nuclear recoils in CsI crystals (reference case) to ~6 events/kg-year (Banik et al., 2020).

3. Data Acquisition, Event Selection, and Statistical Procedures

The InDEx data acquisition workflow integrates acoustic signal analysis and statistical modeling to distinguish nuclear recoil events from backgrounds (Kumar et al., 31 Aug 2025, Das et al., 8 Oct 2025). Acoustic signals are parsed using time-domain and frequency-domain techniques. The sum-of-squares amplitude (Pvar) metric is central for separating neutron-induced (and putative WIMP-induced) nucleations from audio-frequency noise.

Exposure in run1: 2.47 kg-days at 5.87 keV threshold (48.6 days live time).

Exposure in run2: 7.2 kg-days at 1.95 keV threshold (102.48 days live time).

Expected neutron background rates are modeled as:

$$R(E_n, T) = \phi(E_n) V_l \sum_i \left[\epsilon^i(E_n, T) N^i \sigma_n^i(E_n)\right]$$

where $\phi(E_n)$ is the neutron flux, $V_l$ the liquid volume, $N^i$ the atomic density, $\sigma_n^i$ the cross section, $\epsilon^i$ the detection efficiency for element $i$.

Exclusion limits and confidence intervals are derived using a profile likelihood ratio formalism. The likelihood is:

$$L(\mu) = \frac{(\mu s + b)^n}{n!} \exp[-(\mu s + b)]$$

Test statistic:

$$\tilde{t}_\mu = \begin{cases} -2 \ln[L(\mu)/L(\hat{\mu})] & (\hat{\mu} \geq 0) \ -2 \ln[L(\mu)/L(0)] & (\hat{\mu} < 0) \end{cases}$$

with Wald’s approximation for p-values at 90% C.L. Systematic uncertainties—primarily efficiency calibration and liquid volume—are incorporated as Gaussian priors in the fit (Das et al., 8 Oct 2025).

4. Run1 and Run2 Sensitivities and Constraints

Run1 Findings

• SI WIMP-nucleon cross-section for fluorine:

$\sigma_{SI} = [7.939 \pm 0.375_\text{stat} ^{+1.386}_{-0.909}] \times 10^{-39} \ \text{cm}^2$

best achieved at WIMP mass $M_\chi = 30.67$ GeV/c² (Kumar et al., 31 Aug 2025).

• Carbon target sensitivity possible for $M_\chi \gtrsim 4.44$ GeV/c².
• Dominant backgrounds are neutron-induced; observed rates consistent with expectations.
• Acoustic separation methodology validated using neutron source data.

Run2 Findings

• Active mass: 70.4 g C₂H₂F₄ (two detectors) at 1.95 keV threshold, 7.2 kg-days exposure (Das et al., 8 Oct 2025).
• SI WIMP-nucleon cross-section best limit:

$\sigma_{\chi n}^{SI} = (1.55^{+0.62_\text{stat} -0.32_\text{stat} ^{+0.03_\text{sys} -0.02_\text{sys}}) \times 10^{-40} \ \text{cm}^2$

at $M_\chi = 20.4$ GeV/c².

• SD WIMP-proton cross-section (fluorine):

$\sigma_p^{SD} = (7.97^{+3.44_\text{stat} -1.78_\text{stat} ^{+0.15_\text{sys} -0.18_\text{sys}}) \times 10^{-38} \ \text{cm}^2$

at $M_\chi = 21.0$ GeV/c².

• Sensitivity shifts toward lower WIMP masses due to threshold reduction.
• Graphical comparisons (see paper Figures 5–6) indicate that run2 provides improved exclusion limits, especially in the 10–30 GeV/c² range, relative to run1 and similar experiments (SIMPLE, PICASSO, PICO, LUX-ZEPLIN).

5. Comparative Context and Theoretical Landscape

InDEx is methodologically distinct from approaches such as EDELWEISS-II (cryogenic germanium bolometers with interleaved electrodes for active surface rejection) (0912.1196), sodium iodide–based modulation searches (COSINE-100, DAMA) (Adhikari et al., 2019), and noble liquid detectors (XENON, LUX, DEAP360) (Rau, 2011). Its superheated droplet technology enables background discrimination via amplitude-duration acoustic analysis and permits the exploration of lower energy thresholds, pivotal for detecting low-mass WIMPs.

Quenching factor ($QF = NR/ER$) and mass-number coherent scattering enhancement ($\sigma \propto A^2$) govern experimental sensitivity depending on target nuclei. InDEx’s progressive threshold reduction enables hydrogen in C₂H₂F₄ to become a recoil-sensitive target at sub-keV thresholds, with future runs projected to access the MeV mass scale (Das et al., 8 Oct 2025).

Sensitivity improvements in InDEx confine WIMP model parameter spaces, particularly in the region where SI and SD cross-sections overlap with supersymmetric and extra-dimensional scenarios (Rau, 2011).

6. Future Plans, Projected Sensitivity, and Implications

Planned upgrades include:

• Lowering operation thresholds to 0.19 keV by increasing detector temperature to 55°C (broadening sensitivity to hydrogen recoils).
• Scaling up exposures to 1000 kg-days with zero-background operation (Das et al., 8 Oct 2025).
• Advancing passive shielding (polypropylene-lead-polypropylene) and radon suppression.
• Pursuing further purification and volume expansion.

Projected exclusion curves suggest sensitivity orders-of-magnitude improvement, notably in the low-mass regime ($M_\chi \lesssim 10$ GeV/c²), critical for testing current theoretical models and probing putative dark matter candidates overlooked by experiments with higher thresholds (Das et al., 8 Oct 2025).

This strategic shift complements global searches—such as those with cryogenic (EDELWEISS, SuperCDMS), scintillation (NaI/CsI-based DAMA, COSINE-100), noble liquid (ZEPLIN, XENON), and bubble chamber (PICASSO, PICO, SIMPLE) technologies—expanding the reach and cross-validation capabilities in the field of dark matter physics.

7. Significance in the Broader Context of Dark Matter Detection

InDEx’s methodology—prioritizing sub-keV thresholds, background suppression tailored to site-specific neutron and muon fluxes, incremental exposure, and advanced signal characterization—represents a targeted approach to the low-mass WIMP region. Empirical constraints from its early runs (Kumar et al., 31 Aug 2025, Das et al., 8 Oct 2025), when combined with results from complementary technologies and experiments, provide the dark matter community with refined exclusion contours in parameter space and inform theoretical refinements.

A plausible implication is that further threshold reductions and mass increases at InDEx, paired with continued improvements in background modeling, may allow exploration of unexplored parameter space, particularly where traditional cryogenic or scintillation detectors lack sensitivity. The robust acoustic-based background rejection and site-specific simulation set precedents for design choices in future low-background, low-threshold dark matter searches globally.