A NASA-sponsored team is advancing single-photon sensing Complementary Metal-Oxide-Semiconductor (CMOS) detector technology that will enable future NASA astrophysics space missions to search for life on other planets. As part of their detector maturation program, the team is characterizing sensors before, during, and after high-energy radiation exposure; developing novel readout modes to mitigate radiation-induced damage; and simulating a near-infrared CMOS pixel prototype capable of detecting individual photons.
Single-photon sensing and photon-number resolving CMOS
image sensors: a 9.4 Mpixel sensor (left) and a 16.7 Mpixel sensor (right).
Credit: CfD, RIT
Are we alone in the universe? This age-old question has inspired scientific
exploration for centuries. If life on other planets evolves similarly to life
on Earth, it can imprint its presence in atmospheric spectral features known
asbiosignatures. They include absorption and emission lines in the spectrum
produced by oxygen, carbon dioxide, methane, and other molecules that could
indicate conditions which can support life. A future NASA astrophysics mission,
the Habitable Worlds Observatory (HWO), will seek to find biosignatures in the
ultraviolet, optical, and near-infrared (NIR) spectra of exoplanet atmospheres
to look for evidence that life may exist elsewhere in the universe.
HWO will need highly sensitive
detector technology to detect these faint biosignatures on distant exoplanets.
The Single-Photon Sensing Complementary Metal-Oxide-Semiconductor (SPSCMOS)
image sensor is a promising technology for this application. These silicon-based
sensors can detect and resolve individual optical-wavelength photons using a
low-capacitance, high-gain floating diffusion sense node. They operate
effectively over a broad temperature range, including at room temperature. They
have near-zero read noise, are tolerant to radiation, and generate very little
unwanted signal—such as dark current. When cooled to 250 K, the dark
current drops to just one electron every half-hour. If either the read noise or
dark current is too high, the sensor will fail to detect the faint signals that
biosignatures produce.
A research team at the Rochester
Institute of Technology (RIT) Center for Detectors (CfD) is accelerating the
readiness of these SPSCMOS sensors for use in space missions through detector
technology maturation programs funded by NASA’s Strategic Astrophysics
Technology and Early Stage Innovations solicitations. These development programs
include several key goals:
- Characterize critical
detector performance metrics like dark current, quantum efficiency, and
read noise before, during, and after exposure to high-energy radiation
- Develop new readout modes
for these sensors to mitigate effects from short-term and long-term
radiation damage
- Design a new NIR version
of the sensor using Technology Computer-Aided Design (TCAD) software
SPSCMOS sensors operate similarly
to traditional CMOS image sensors but are optimized to detect individual
photons—an essential capability for ultra-sensitive space-based observations,
such as measuring the gases in the atmospheres of exoplanets. Incoming photons
enter the sensor and generate free charges (electrons) in the sensor material.
These charges collect in a pixel’s storage well and eventually transfer to a
low-capacitance component called the floating diffusion (FD) sense node where
each free charge causes a large and resolved voltage shift. This voltage shift
is then digitized to read the signal.
Experiments that measure sensor
performance in a space relevant environment use a vacuum Dewar and a
thermally-controlled mount to allow precise tuning of the sensors temperature.
The Dewar enables testing at conditions that match the expected thermal environment
of the HWO instrument, and can even cool the sensor and its on-chip circuits to
temperatures colder than any prior testing reported for this detector family.
These tests are critical for revealing performance limitations with respect to
detector metrics like dark current, quantum efficiency, and read noise. As
temperatures change, the electrical properties of on-chip circuits can also
change, which affects the read out of charge in a pixel.
The two figures show results for SPSCMOS devices. The
figure on the left shows a photon counting histogram with peaks that correspond
to photon number. The figure on the right shows the dark current for a SPSCMOS
device before and after exposure to 50 krad of 60 MeV protons.
Credit: CfD, RIT
The radiation-rich environment for HWO will cause temporary and permanent
effects in the sensor. These effects can corrupt the signal measured in a
pixel, interrupt sensor clocking and digital logic, and can cause cumulative
damage that gradually degrades sensor performance. To mitigate the loss of
detector sensitivity throughout a mission lifetime, the RIT team is developing
new readout modes that are not available in commercial CMOS sensors. These
custom modes sample the signal over time (a "ramp" acquisition) to
enable the detection and removal of cosmic ray artifacts. In one mode, when the
system identifies an artifact, it segments the signal ramp and selectively
averages the segments to reconstruct the original signal—preserving scientific
data that would otherwise be lost. In addition, a real-time data acquisition
system monitors the detector’s power consumption, which may change from the
accumulation of damage throughout a mission. The acquisition system records
these shifts and communicates with the detector electronics to adjust voltages
and maintain nominal operation. These radiation damage mitigation strategies
will be evaluated during a number of test programs at ground-based radiation
facilities. The tests will help identify unique failure mechanisms that impact
SPSCMOS technology when it is exposed to radiation equivalent to the dose
expected for HWO.
Custom acquisition electronics (left) that will
control the sensors during radiation tests, and an image captured using this
system (right).
Credit: CfD, RIT
While existing SPSCMOS sensors are limited to detecting visible light due
to their silicon-based design, the RIT team is developing the world’s first NIR
single-photon photodiode based on the architecture used in the optical sensors.
The photodiode design starts as a simulation in TCAD software to model
the optical and electrical properties of the low-capacitance CMOS architecture.
The model simulates light-sensitive circuits using both silicon and Mercury
Cadmium Telluride (HgCdTe or MCT) material to determine how well the pixel
would measure photo-generated charge if a semiconductor foundry physically
fabricated it. It has 2D and 3D device structures that convert light into
electrical charge, and circuits to control charge transfer and signal readout
with virtual probes that can measure current flow and electric potential. These
simulations help to evaluate the key mechanisms like the conversion of light
into electrons, storing and transferring the electrons, and the output voltage
of the photodiode sampling circuit.
In addition to laboratory testing,
the project includes performance evaluations at a ground-based telescope. These
tests allow the sensor to observe astronomical targets that cannot be fully
replicated in lab. Star fields and diffuse nebulae challenge the detector’s
full signal chain under real sky backgrounds with faint flux levels,
field-dependent aberrations, and varying seeing conditions. These observations
help identify performance limitations that may not be apparent in controlled
laboratory measurements.
In January 2025, a team of
researchers led by PhD student Edwin Alexani used an SPSCMOS-based camera at
the C.E.K. Mees Observatory in Ontario County, New York. They observed star
cluster M36 to evaluate the sensor’s photometric precision, and the Bubble
Nebula in a narrow-band H-alpha filter. The measured dark current and read
noise were consistent with laboratory results.
The team observed photometric
reference stars to estimate the quantum efficiency (QE) or the ability for the
detector to convert photons into signal. The calculated QE agreed with
laboratory measurements, despite differences in calibration methods.
The team also observed the
satellite STARLINK-32727 as it passed through the telescope’s field of view and
measured negligible persistent charge—residual signal that can remain in
detector pixels after exposure to a bright source. Although the satellite briefly
produced a bright streak across several pixels due to reflected sunlight, the
average latent charge in affected pixels was only 0.03 e-/pix – well below both the sky-background and sensor’s read noise.
Images captured at the C.E.K. Mees Observatory. Left:
The color image shows M36 in the Johnson color filters B (blue), V (green), and
R (red) bands (left). Right: Edwin Alexani and the SPSCMOS camera (right).
Credit: : CfD, RIT
As NASA advances and matures the HWO mission, SPSCMOS technology promises
to be a game-changer for exoplanet and general astrophysics research. These
sensors will enhance our ability to detect and analyze distant worlds, bringing
us one step closer to answering one of humanity’s most profound questions: are
we alone?
For additional details, see the
entry for this project on NASA TechPort.
Project Lead(s): Dr. Donald F. Figer, Future Photon Initiative and
Center for Detectors, Rochester Institute of Technology (RIT), supported by
engineer Justin Gallagher and a team of students.
Sponsoring Organization(s): NASA Astrophysics Division, Strategic Astrophysics Technology (SAT) Program and NASA Space Technology Mission Directorate (STMD), Early Stage Innovations (ESI) Program
No comments:
Post a Comment