Origin of pixel-specific energy-calibration drift during long-term operation

Determine the origin of the up-to-10 eV drift observed in the fitted Mn-Kα peak position for four pixels of the TRISTAN 166‑pixel silicon drift detector module during a 63‑hour stability measurement with a 55Fe source in the x‑ray bench test setup operated at −34.1 °C with ≤1 K temperature variation, and characterize the underlying mechanism to assess its impact on energy calibration stability.

Background

The paper reports a 63-hour stability measurement of a TRISTAN silicon drift detector module using a 55Fe x‑ray source in a vacuum and light-tight bench setup. During this test the detector was held at −34.1 °C with at most 1 K temperature variation, and stability of count rate, energy calibration, and energy resolution was evaluated.

For most pixels, the Mn-Kα peak position varied by less than 5 eV over time. However, four pixels exhibited a larger drift of up to 10 eV. The authors explicitly note that the cause of this pixel-specific drift is currently not understood and requires further investigation, making the determination of its origin an explicit open question relevant to calibration stability in long-duration measurements.