Broadband Optical Modulation and Control at Millikelvin Temperatures

TL;DR

This work tackles the challenge of precisely calibrating radiation responses in cryogenic photon detectors by introducing a broadband optical beam-steering system that delivers μs-scale photon pulses via fiber to targeted detector locations. The authors develop a two-mirror MEMS-based architecture with a reflective broadband focuser and a fixed output-fiber plate, enabling beam mapping from 365 nm to 970 nm down to cryogenic temperatures with minimal heating. Comprehensive beam-size and pulse-width measurements at room temperature and ~10 mK demonstrate sub-100 μm spots across the band and ~4 μs pulses, with cryogenic improvements in scan speed and a clear path toward single-mode operation and longer-wavelength extension. The system provides a versatile, low-dissipation tool for calibrating and understanding cryogenic detectors, including superconducting sensors, quasiparticle dynamics, and phonon transport, with clear directions for future miniaturization and wavelength expansion.

Abstract

A universal experimental challenge when studying radiation effects on cryogenic devices is to precisely and accurately characterize the position-dependent device response very near the energy detection threshold. We have developed a compact cryogenic optical beam steering system that can be used to generate O(μs) pulses of small numbers of photons over the energy range of 1.2 - 4.5eV at room temperature, and deliver those photons via fiber optic to any specified location on the surface of a detector operating at cryogenic temperatures. This new system will allow for robust calibration of any photon-sensitive detector, including supercondcting devices. The system can be used efficiently to explore the physics of target materials, quantify the position sensitivity of different sensor designs, measure phonon transport, and study the effects of quasiparticle poisoning on detector operation. We describe the design of this pulsed calibration method and present first results obtained with a second-generation system operated at room temperature and sub-Kelvin temperatures.

Figures (10)

• Figure 1: Left: Optical chopping test setup (without its copper enclosure "lid"), shown on the test bench. The reflective focusing unit brings monochromatic photons entering the package via fiber optic onto a stationary mirror that directs the beam onto a tilt-able MEMS mirror of diameter $d_r$ = 4.6 mm. The voltage-controlled MEMS mirror is then used to scan the photon beam over an exit slit or fiber with a diameter of 105 um mounted to a removable "output plate". The result is a focused, chopped beam of light on a device-under-test (not shown). Right: Simplified diagram (not to scale) of the optical path inside the MEMS scanner unit. The focusing unit includes a reflective collimator unit of focal length $f_c$ = 33 mm and a separate off-axis parabolic mirror (OAP) of focal length $f_0$ = 203 mm. The center of the MEMS mirror is located a distance $l_a$ = 84 mm away from the exit port.
• Figure 2: Top Row: Example images of the focused beam spot taken with a 4.8 $\mu$m-pixel CCD camera and $\lambda$ = 625 nm photons. The setup included only the reflective focusing unit (OAP and collimator) and not the MEMS mirror. The central image shows the best focus, where the spot is round with a (4$\sigma$) full width of 90 $\mu$m. The left and right images show the effect on spot size and shape of defocusing the beam by $\pm$1.2 mm, illustrating the slight astigmatism in our system. The source of this astigmatism was later found to be caused by slight angular misalignments within the focusing unit. See Appendix \ref{['app:newFocuser']} for details on how this issue was mitigated. Bottom Left: Measurements of the spot size from the reflective focusing unit (only) at 625 nm as a function of distance from the expected focal length, based on the calculated spot size for $M=6.16$, NA=0.1, and mean fiber diameter of 10 $\mu$m. The gray bar is the range of spot sizes expected within the quoted range of Thorlabs fiber diameter, which is only accurate to the 3 $\mu$m level. Astigmatism in the focusing unit is evident in the top plots, where the best focus is located where X and Y have the same width, yet each (X and Y) has a focus shifted relative to this nominal focal point. For a system with astigmatism corrected, we thus expect this system to be diffraction limited at 625 nm to 65 $\mu$m, which is in agreement with the lower range estimate of the theory band. We have demonstrated this correction to the astigmatism and present initial results in Appendix \ref{['app:newFocuser']}. Bottom Right: Spot size vs wavelength for the range of optical diodes used in this study, where the mean radius, semi-major and semi-minor axes are as-measured values, obtained in the full MEMS system during warm checkouts. The absolute minimum spot size, set by the magnified size of the 10 micron fiber, is shown as a horizontal dashed line. The best focus shows the spot size at the respective X and Y focuses from the left figure, which are much closer to the divergence limit. The gray points show the imputed spot size using a system with a larger magnification, which reduces diffraction and astigmatism effects. Finally, the black points show the major and minor axes of the spot produced by a newer collimator design that largely corrects the astigmatism seen in this system. The grey band shows the expected diffraction-limited performance assuming a 4$\sigma$ fiber diameter of 10$\mu$m. The fact that the imputed spot size at $1\mu$m is below this band reflects the fact that the fiber is approaching the single-mode limit.
• Figure 3: Top: Room temperature image of beam coupled to 105 $\mu$m diameter output fiber. Intensity through the output fiber was measured using voltage out of an APD. The conversion from 2-axis MEMS tilt controller settings to physical units ($\mu$m) assumes the intensity profile along each principle component of the spot is equivalent to a Gaussian convolved with a 105 $\mu$m square pulse. Bottom Left: The intensity along one principal component fit to a Gaussian convolved with a 105 $\mu$m square pulse. Bottom Center: Same as Bottom Left for the second principle component. Bottom Right: The relevant legend, including 4$\sigma$ widths of the relevant Gaussian fits.
• Figure 4: Left: Room-temperature image of 660 nm beam coupled to 105 $\mu$m diameter output fiber. Individual figures comparable to those in Figure \ref{['fig:APDspotSizeWarm']}. Right: Cryogenic (10 mK) 660 nm image.
• Figure 5: Left: Pulse-width measurement from the MEMS chopping system at warm temperature with a 660 nm LED. Right: Pulse width from the MEMS chopping system at cold temperature.