We have fabricated and characterized 10,000 and 20,440 pixel Microwave Kinetic Inductance Detector (MKID) arrays for the Dark-speckle Near-IR Energy-resolved Superconducting Spectrophotometer ...(DARKNESS) and the MKID Exoplanet Camera (MEC). These instruments are designed to sit behind adaptive optics systems with the goal of directly imaging exoplanets in a 800-1400 nm band. Previous large optical and near-IR MKID arrays were fabricated using substoichiometric titanium nitride (TiN) on a silicon substrate. These arrays, however, suffered from severe non-uniformities in the TiN critical temperature, causing resonances to shift away from their designed values and lowering usable detector yield. We have begun fabricating DARKNESS and MEC arrays using platinum silicide (PtSi) on sapphire instead of TiN. Not only do these arrays have much higher uniformity than the TiN arrays, resulting in higher pixel yields, they have demonstrated better spectral resolution than TiN MKIDs of similar design. PtSi MKIDs also do not display the hot pixel effects seen when illuminating TiN on silicon MKIDs with photons with wavelengths shorter than 1 µm.
Microwave Kinetic Inductance Detector (MKID) devices offer inherent spectral resolution, simultaneous read out of thousands of pixels, and photon-limited sensitivity at optical wavelengths. Before ...taking observations the readout power and frequency of each pixel must be individually tuned, and if the equilibrium state of the pixels change, then the readout must be retuned. This process has previously been performed through manual inspection, and typically takes one hour per 500 resonators (20 h for a ten-kilo-pixel array). We present an algorithm based on a deep convolution neural network (CNN) architecture to determine the optimal bias power for each resonator. The bias point classifications from this CNN model, and those from alternative automated methods, are compared to those from human decisions, and the accuracy of each method is assessed. On a test feed-line dataset, the CNN achieves an accuracy of 90% within 1 dB of the designated optimal value, which is equivalent accuracy to a randomly selected human operator, and superior to the highest scoring alternative automated method by 10%. On a full ten-kilopixel array, the CNN performs the characterization in a matter of minutes — paving the way for future mega-pixel MKID arrays.
Optical and near-infrared Microwave Kinetic Inductance Detectors, or MKIDs, are low-temperature detectors with inherent spectral resolution that are able to instantly register individual photons with ...potentially no false counts or readout noise. These properties make MKIDs transformative for exoplanet direct imaging by enabling photon-statistics-based planet-discrimination techniques as well as performing conventional noise-subtraction techniques on shorter timescales. These detectors are in the process of rapid development, and as such, the full extent of their performance enhancing potential has not yet be quantified. MKID Exoplanet Direct Imaging Simulator, or MEDIS, is a general-purpose end-to-end numerical simulator for high-contrast observations with MKIDs. The simulator exploits current optical propagation libraries and augments them with a new MKIDs simulation module to provide a pragmatic model of many of the degradation effects present during the detection process. We use MEDIS to demonstrate how changes in various MKID properties affect the contrast-separation performance when conventional differential imaging techniques are applied to low-flux, short duration observations. We show that to improve performance at close separations will require increasing the maximum count rate or pixel sampling when there is high residual flux after the coronagraph. We predict that taking pixel yield from the value achieved by current instruments of 80% and increasing it to 100% would result in an improvement in contrast of a factor of \(\sim\) 4 at 3\(\lambda/D\) and \(\sim\) 8 at 6\(\lambda/D\). Achieving better contrast performance in this low flux regime would then require exploiting the information encoded in the photon arrival time statistics.
We have fabricated and characterized 10,000 and 20,440 pixel Microwave Kinetic Inductance Detector (MKID) arrays for the Dark-speckle Near-IR Energy-resolved Superconducting Spectrophotometer ...(DARKNESS) and the MKID Exoplanet Camera (MEC). These instruments are designed to sit behind adaptive optics systems with the goal of directly imaging exoplanets in a 800-1400 nm band. Previous large optical and near-IR MKID arrays were fabricated using substoichiometric titanium nitride (TiN) on a silicon substrate. These arrays, however, suffered from severe non-uniformities in the TiN critical temperature, causing resonances to shift away from their designed values and lowering usable detector yield. We have begun fabricating DARKNESS and MEC arrays using platinum silicide (PtSi) on sapphire instead of TiN. Not only do these arrays have much higher uniformity than the TiN arrays, resulting in higher pixel yields, they have demonstrated better spectral resolution than TiN MKIDs of similar design. PtSi MKIDs also do not display the hot pixel effects seen when illuminating TiN on silicon MKIDs with photons with wavelengths shorter than 1 um.
Abstract
We report the direct imaging discovery of a low-mass companion to the nearby accelerating A star, HIP 109427, with the Subaru Coronagraphic Extreme Adaptive Optics (SCExAO) instrument ...coupled with the Microwave Kinetic Inductance Detector Exoplanet Camera (MEC) and CHARIS integral field spectrograph. CHARIS data reduced with reference star point spread function (PSF) subtraction yield 1.1–2.4
μ
m spectra. MEC reveals the companion in
Y
and
J
band at a comparable signal-to-noise ratio using stochastic speckle discrimination, with no PSF subtraction techniques. Combined with complementary follow-up
L
p
photometry from Keck/NIRC2, the SCExAO data favors a spectral type, effective temperature, and luminosity of M4–M5.5, 3000–3200 K, and
log
10
(
L
/
L
⊙
)
=
−
2.28
−
0.04
+
0.04
, respectively. Relative astrometry of HIP 109427 B from SCExAO/CHARIS and Keck/NIRC2, and complementary Gaia–Hipparcos absolute astrometry of the primary favor a semimajor axis of 6.55
+3.0
−0.48
au, an eccentricity of
0.54
−
0.15
+
0.28
, an inclination of
66.7
−
14
+
8.5
degrees, and a dynamical mass of
0.280
−
0.059
+
0.18
M
⊙
. This work shows the potential for extreme AO systems to utilize speckle statistics in addition to widely used postprocessing methods to directly image faint companions to nearby stars near the telescope diffraction limit.
Abstract
We present the direct imaging discovery of a low-mass companion to the nearby accelerating F star, HIP 5319, using SCExAO coupled with the CHARIS, VAMPIRES, and MEC instruments in addition ...to Keck/NIRC2 imaging. CHARIS
JHK
(1.1–2.4
μ
m) spectroscopic data combined with VAMPIRES 750 nm, MEC
Y
, and NIRC2
L
p
photometry is best matched by an M3–M7 object with an effective temperature of
T
= 3200 K and surface gravity log(
g
) = 5.5. Using the relative astrometry for HIP 5319 B from CHARIS and NIRC2, and absolute astrometry for the primary from Gaia and Hipparcos, and adopting a log-normal prior assumption for the companion mass, we measure a dynamical mass for HIP 5319 B of
31
−
11
+
35
M
J
, a semimajor axis of
18.6
−
4.1
+
10
au, an inclination of
69.4
−
15
+
5.6
degrees, and an eccentricity of
0.42
−
0.29
+
0.39
. However, using an alternate prior for our dynamical model yields a much higher mass of
128
−
88
+
127
M
J
. Using data taken with the LCOGT NRES instrument we also show that the primary HIP 5319 A is a single star in contrast to previous characterizations of the system as a spectroscopic binary. This work underscores the importance of assumed priors in dynamical models for companions detected with imaging and astrometry, and the need to have an updated inventory of system measurements.
PDF
