NEWAGE

NEWAGE is a direction-sensitive dark matter search experiment with a gaseous time-projection chamber. We improved the direction-sensitive dark matter limits by our underground measurement. After the first underground run, we replaced the detector components with radio-pure materials. We also studied the possibilities of head-tail recognition of nuclear tracks in the surface laboratory. For the future large volume detector, we are developing a pixel ASIC named QPIX. In this paper, these recent R&D activities are described.

💡 Research Summary

**
The NEWAGE (NEw generation WIMP‑search with Advanced Gaseous tracking device Experiment) collaboration focuses on a direction‑sensitive dark‑matter search using a gaseous time‑projection chamber (TPC). This paper reports on the series of R&D activities carried out after the first underground run in 2010, with the goal of improving the experiment’s sensitivity by reducing background, enhancing head‑tail discrimination, and developing a pixelated readout ASIC for future large‑volume detectors.

Detector configuration and baseline performance
NEWAGE employs a micro‑patterned gaseous detector (µ‑PIC) with a 400 µm pitch as the main charge‑amplification and readout element, complemented by a GEM stage for additional gain. Three TPC prototypes are operated: a small 10 × 10 cm² device (NEWAGE‑0.1a) for surface tests, a medium 30 × 30 cm² detector (NEWAGE‑0.3a) used underground for dark‑matter data taking, and a larger 30 × 30 cm² version (NEWAGE‑0.3b) for volume‑scaling studies. Two dedicated radon monitors (≈22 cm diameter) are also installed to quantify radon emanation from detector components.

Radon background mitigation
Radon emanating from uranium and thorium impurities in detector materials was identified as the dominant internal background. By measuring the relative radon contribution of each component, the authors found that the glass‑reinforced fluoroplastic used for the TPC cage contributed the most, followed by PTFE, GEM frames, polyimide+Cu, resistors, and the µ‑PIC itself. To suppress this source, the cage material was replaced with radio‑pure PTFE, which reduced the radon contribution to roughly one‑third of its original level. In addition, a charcoal filter (≈100 g) and a Teflon‑bellows pump were installed in a closed‑loop gas circulation system, achieving a ten‑fold reduction in overall radon concentration. These measures allowed a new dark‑matter run to commence on 3 August 2011 with a substantially lower background rate.

Head‑tail (directional) discrimination studies
The ability to determine the “head” versus “tail” of a nuclear recoil track is crucial for enhancing directional sensitivity. The original FPGA firmware required simultaneous X‑Y hits; the upgraded firmware records the rising and falling edges of every hit, providing full two‑dimensional timing information (TPC‑mode5). Using a 252Cf neutron source placed at four positions (±X, ±Y) relative to the detector, the authors constructed a skewness parameter γ based on the pulse‑duration‑weighted third moment of the charge distribution along each axis. The measured skewness distributions show a statistically significant separation between +X and –X runs (up to 7.7 σ in the 200–400 keV band) while +Y and –Y runs are consistent with zero difference. This demonstrates that head‑tail discrimination is achievable down to recoil energies of about 70 keV, although the current skewness definition is not yet optimal for event‑by‑event track reconstruction.

Pixel ASIC development – QPIX
Conventional strip or MWPC readouts limit angular resolution to roughly 30° for sub‑millimeter tracks, which is insufficient for precise three‑dimensional reconstruction. To overcome this, the collaboration is developing a CMOS pixel ASIC named QPIX. The first prototype, QPIX‑ver1, features a 20 × 20 pixel matrix with a 200 µm pitch. Each pixel integrates a 14‑bit time‑of‑flight (TOF) counter, an 8‑bit time‑over‑threshold (TOT) counter, and a 10‑bit ADC, fabricated in a 0.18 µm SMC process. Two mounting schemes were explored: conventional wire‑bonding (which introduces dead areas along the edges) and flip‑chip bonding (which minimizes dead zones but requires a perfectly flat charge‑collection PCB). Preliminary tests of wire‑bonded chips show good TOF linearity up to 2 µs and ADC linearity up to 1.5 pC, but the threshold is about ten times higher than the design target. Flip‑chip attempts were hampered by PCB surface non‑uniformity; a new charge‑collection PCB is under development. Future test‑element‑groups will focus on lowering the threshold and improving the flip‑chip process.

Conclusions and outlook
The NEWAGE team has made substantial progress in three key areas: (1) internal background reduction through material substitution (PTFE cage) and active radon removal, (2) demonstration of head‑tail discrimination at recoil energies as low as 70 keV, and (3) initiation of a pixel ASIC program (QPIX) that promises true three‑dimensional track imaging with fine angular resolution. The upcoming dark‑matter run, now operating with the low‑background PTFE cage, is expected to set more stringent limits on WIMP‑nucleon cross sections. Continued development of QPIX and optimization of head‑tail algorithms will be essential for scaling to the multi‑cubic‑meter detectors required for competitive directional dark‑matter searches.