[Paper Review] Skipper-CCD Sensors for the Oscura Experiment: Requirements and Preliminary Tests

Brenda A. Cervantes-Vergara, Santiago Pérez+53 more|arXiv (Cornell University)|Apr 10, 2023

CCD and CMOS Imaging Sensors4 citations

TL;DR

This paper presents the requirements and preliminary testing of Skipper-CCD sensors for the Oscura experiment, a next-generation dark matter direct detection search targeting sub-GeV DM-electron interactions. By optimizing sensor performance to minimize instrumental backgrounds—particularly spurious charge and dark current—it demonstrates that a 30 kg-year exposure with less than one background event per electron bin (2–10 e⁻) is achievable, enabling unprecedented sensitivity to light dark matter.

ABSTRACT

Oscura is a proposed multi-kg skipper-CCD experiment designed for a dark matter (DM) direct detection search that will reach unprecedented sensitivity to sub-GeV DM-electron interactions with its 10 kg detector array. Oscura is planning to operate at SNOLAB with 2070 m overburden, and aims to reach a background goal of less than one event in each electron bin in the 2-10 electron ionization-signal region for the full 30 kg-year exposure, with a radiation background rate of 0.01 dru. In order to achieve this goal, Oscura must address each potential source of background events, including instrumental backgrounds. In this work, we discuss the main instrumental background sources and the strategy to control them, establishing a set of constraints on the sensors' performance parameters. We present results from the tests of the first fabricated Oscura prototype sensors, evaluate their performance in the context of the established constraints and estimate the Oscura instrumental background based on these results.

Motivation & Objective

• To design a multi-kg Skipper-CCD array for direct detection of sub-GeV dark matter via electron recoils.
• To establish stringent performance constraints on sensors to achieve a background goal of <1 event per electron bin (2–10 e⁻) over 30 kg-year exposure.
• To identify and mitigate instrumental background sources, especially spurious charge and dark current, in prototype sensors.
• To evaluate prototype sensor performance against target specifications and estimate instrumental background contributions.
• To enable a science reach that surpasses existing skipper-CCD experiments like SENSEI and DAMIC-M.

Proposed method

• Defined performance constraints for Skipper-CCD sensors based on background goals: <0.025 dru radiation background and <1×10⁻⁶ e⁻/pix/day from instrumental sources.
• Conducted preliminary tests on first Oscura prototype sensors, measuring key parameters including readout noise, dark current, spurious charge, trap density, and charge transfer inefficiency (CTI).
• Used a binomial distribution model to estimate total spurious charge-induced events over 30 kg-year exposure, with λ = κ_SC × N_trans.
• Evaluated the impact of readout mode (1×1 vs. 1×10 binning) on reducing effective transfers and readout time, thereby lowering dark current and spurious charge contributions.
• Calculated instrumental background contributions assuming R_DC,1e⁻ = 1.6×10⁻⁴ e⁻/pix/day and κ_SC = 7.2×10⁻⁷ e⁻/pix/transfer, with N_exp = 1 exposure/day.
• Projected science sensitivity under both stringent and relaxed background assumptions using 90% C.L. limits and QeDark cross-section models.

Experimental results

Research questions

• RQ1 Can Skipper-CCD sensors meet the stringent instrumental background requirements needed for a 30 kg-year exposure in the Oscura dark matter experiment?
• RQ2 What are the dominant instrumental background sources in prototype Skipper-CCDs, and how do they compare to required performance thresholds?
• RQ3 How does readout mode (e.g., 1×1 vs. 1×10 binning) affect the balance between dark current and spurious charge contributions?
• RQ4 To what extent can the current prototype performance support Oscura’s goal of <1 background event per electron bin in the 2–10 e⁻ range?
• RQ5 What is the projected sensitivity of Oscura to sub-GeV dark matter if the background goal is not fully achieved?

Key findings

• The prototype sensors failed to meet the spurious charge constraint (7.2×10⁻⁷ e⁻/pix/transfer vs. target <6×10⁻⁹), making it the dominant instrumental background source.
• Dark current in the prototype was measured at 1.6×10⁻⁴ e⁻/pix/day, below the target of 1.6×10⁻⁴ e⁻/pix/day, indicating it is not the primary contributor.
• Charge transfer inefficiency (CTI) was below the target threshold, with <10⁻⁵ per transfer, consistent with prior successful skipper-CCD performance.
• Spurious charge rate (κ_SC) is proportional to number of effective transfers; binning (e.g., 1×10) reduces effective transfers and thus lowers SC contribution.
• With N_exp = 1 exposure/day and 1×10 binning, the expected number of 4 e⁻ events from spurious charge is 9.5 per 30 kg-year exposure, while dark current contributes 0 events.
• Even if the stringent background goal is not fully achieved, Oscura can still achieve significant science reach, with projected sensitivity remaining competitive under relaxed background assumptions (e.g., 3 e⁻ threshold with zero background events).