Embed Size (px)
Citation preview
SOFIA-HIRMES: Looking Forward to the HIgh-ResolutionMid-infrarEd Spectrometer
Samuel N. Richards1,16, Samuel H. Moseley2, Gordon Stacey3, Matthew Greenhouse2,Alexander Kutyrev2, Richard Arendt2, Hristo Atanaso®2, Stuart Banks2, Regis P. Brekosky2,
Ari-David Brown2, Berhanu Bulcha2, Tony Cazeau2, Michael Choi2, Felipe Colazo2,Chuck Engler2, Theodore Hadjimichael2, James Hays-Wehle2, Chuck Henderson4,Wen-Ting Hsieh2, Je®rey Huang5,6, Iver Jenstrom2, Jim Kellogg2, Mark Kimball2,Attila Kov!acs8, Steve Leiter2, Steve Maher2, Robert McMurray5, Gary J. Melnick9,Eric Mentzell2, Vilem Mikula10, Timothy M. Miller2, Peter Nagler2, Thomas Nikola4,Joseph Oxborrow11, Klaus M. Pontoppidan12, Naseem Rangwala13, Alan Rhodes5,
Aki Roberge14, Stefan Rosner7, Karwan Rostem2, Nancy Rustemeyer5, Elmer Sharp2,Leroy Sparr2, Dejan Stevanovic1, Peter Taraschi2, Pasquale Temi15, William D. Vacca1,
Jordi Vila Hernandez de Lorenzo2, Bill Wohler5,7, Edward J. Wollack2
and Shannon Wilks2
1SOFIA Science Center, USRA, NASA Ames Research Center, M/S 232-12P.O. Box 1, Mo®ett Field, CA 94035, USA
2NASA Goddard Space Flight Center8800 Greenbelt Rd, Greenbelt, MD 20771, USA
3Department of Astronomy, Cornell University, Ithaca, NY 14853, USA4Cornell Center for Astrophysics and Planetary Science
Cornell University, Ithaca, NY 14853, USA5NASA Ames Research Center, M/S 232-12, P.O. Box 1, Mo®ett Field, CA 94035, USA
6LOGYX LLC, Mountain View, CA 94043, USA7SETI Institute, Mountain View, CA 94043, USA
8Smithsonian Astrophysical Observatory Submillimeter Array (SMA)MS-78, 60 Garden St, Cambridge, MA 02138, USA
9Harvard-Smithsonian Center for Astrophysics60 Garden St., Cambridge, MA 02138, USA
10American University, Institute for Integrated Space Studies & TechnologyDepartment of Physics, Don Myers Building, 4400, Massachusetts Avenue NW
Washington, DC 20016-8079, USA11Scienti¯c and Biomedical Microsystems, 806 Cromwell Park Drive
Suite R, Glen Burnie, MD 21061, USA12Space Telescope Science Institute, 3700 San Martin Drive, Baltimore, MD 21218, USA
13Space Science and Astrobiology DivisionNASA Ames Research Center, Mo®ett Field, CA 94035, USA
14Exoplanets and Stellar Astrophysics LabNASA Goddard Space Flight Center, Greenbelt, MD 20771, USA
15Astrophysics Branch, NASA Ames Research Center, Mo®ett Field, CA 94035, [email protected]
Received September 15, 2018; Accepted November 2, 2018; Published December 3, 2018
16Corresponding author.This is an Open Access article published by World Scienti¯c Publishing Company. It is distributed under the terms of the CreativeCommons Attribution 4.0 (CC-BY) License. Further distribution of this work is permitted, provided the original work is properlycited.
Journal of Astronomical Instrumentation, Vol. 7, No. 4 (2018) 1840015 (16 pages)#.c The Author(s)DOI: 10.1142/S2251171718400159
1840015-1
J. A
stro
n. In
stru
m. 2
018.
07. D
ownl
oade
d fr
om w
ww
.wor
ldsc
ient
ific.
com
by 7
6.21
8.10
7.31
on
01/1
0/19
. Re-
use
and
dist
ribut
ion
is st
rictly
not
per
mitt
ed, e
xcep
t for
Ope
n A
cces
s arti
cles
.
The HIgh-Resolution Mid-infrarEd Spectrometer (HIRMES) is the 3rd Generation Instrument for theStratospheric Observatory For Infrared Astronomy (SOFIA), currently in development at the NASAGoddard Space Flight Center (GSFC), and due for commissioning in 2019. By combining direct-detectionTransition Edge Sensor (TES) bolometer arrays, grating-dispersive spectroscopy, and a host of Fabry-Perot tunable ¯lters, HIRMES will provide the ability for high resolution (R ! 100; 000), mid-resolution(R ! 10; 000), and low-resolution (R ! 600) slit-spectroscopy, and 2D Spectral Imaging (R ! 2000 atselected wavelengths) over the 25–122!m mid to far infrared waveband. The driving science application isthe evolution of proto-planetary systems via measurements of water-vapor, water-ice, deuterated hydrogen(HD), and neutral oxygen lines. However, HIRMES has been designed to be as °exible as possible to cover awide range of science cases that fall within its phase-space, all whilst reaching sensitivities and observingpowers not yet seen thus far on SOFIA, providing unique observing capabilities which will remainunmatched for decades.
Keywords: SOFIA, astronomy, infrared, instrumentation, high-resolution, spectroscopy, proto-planetary.
1. Introduction
As all astronomical instrumentation should bedriven by the scienti¯c demand, such is the case forthe HIgh-Resolution Mid-infrarEd Spectrometer(HIRMES), planned for commissioning on theStratospheric Observatory For Infrared Astronomy(SOFIA; Temi et al., 2014) in 2019 as a facility-classinstrument. HIRMES is currently undergoing itsIntegration and Testing (I&T) phase of its devel-opment at the NASA Goddard Space Flight Center(GSFC) in partnership with Cornell University,with Samuel H. (Harvey) Moseley as the PrincipleInvestigator. In September 2016, HIRMES was se-lected to be SOFIA's 3rd Generation Instrument.Since then, it has been following an intensive buildschedule, working closely with the NASA/SOFIAScience Instrument Development Team at theNASA Ames Research Center (ARC).
Following the goal of understanding of how theEarth obtained its water, HIRMES will provideanswers to fundamental questions currently beingasked in proto-planetary science. Among thesequestions, HIRMES will tackle: (a) How does themass of the disk evolve during planetary formation?(b) What is the distribution of oxygen, water ice,and water vapor in di®erent phases of planet for-mation? (c) What are the kinematics of water vaporand oxygen in proto-planetary disks? In answeringthese questions, HIRMES will discover where, andin what form, the raw materials for life reside, andhow planetary systems like our own evolve.
HIRMES will quantitatively answer thesequestions by providing low (R ! 600) to very high(R ! 100; 000) spectral resolving power over thecritical spectral range of 25–122!m in the mid- tofar-infrared waveband. HIRMES combines gratingdispersive spectroscopy and Fabry–Perot tunable
narrow-band ¯lters with high e±ciency back-ground-limited direct detectors. The instrumentspectral resolution is designed to match the width ofthe spectral lines, signi¯cantly reducing the back-ground noise, to achieve the maximum possiblesensitivity for mid-infrared spectroscopy withSOFIA. Providing this combination of sensitivityand spectral resolution, HIRMES will open up aunique and useful window on the evolution ofplanetary systems to the astronomical community.HIRMES' order-of-magnitude sensitivity improve-ment to SOFIA's current capabilities is crucial forthis science program, increasing the number of ob-servable Solar mass proto-planetary systems from acouple to hundreds. Furthermore, the instrumenthas utility far beyond the aforementioned investi-gations, providing tools for a range of Galacticstudies, such as stellar out°ows and their impact onthe interstellar medium, and extragalactic studies,such as the strength and shape of important diag-nostic emission lines.
The instrument's grating mode spectroscopy isa powerful tool for the study of water ice emission ina wide range of objects, an important capabilitylargely absent for more than 20 years since theKuiper Airborne Observatory (KAO) and InfraredSpace Observatory (ISO).
1.1. Science drivers
The past decade has seen considerable advance-ments in our understanding of exoplanets: their di-verse properties (e.g. masses, size, orbits) andfundamental trends (e.g. with metallicity or hostmass). These observed properties place importantboundary conditions on the key processes thatgovern the formation and evolution of planetary
S. N. Richards et al.
1840015-2
J. A
stro
n. In
stru
m. 2
018.
07. D
ownl
oade
d fr
om w
ww
.wor
ldsc
ient
ific.
com
by 7
6.21
8.10
7.31
on
01/1
0/19
. Re-
use
and
dist
ribut
ion
is st
rictly
not
per
mitt
ed, e
xcep
t for
Ope
n A
cces
s arti
cles
.
systems. Analogous studies of the proto-planetarydisks orbiting young stars provide the vital initialconditions for making planetary systems, as well asthe early coevolution and interaction of planets andtheir birth environments. The observed propertiesof both disks and exoplanets are needed to inform,test, and re¯ne models of the planet formationprocess (Andrews, 2015, and references therein).Studying the dynamics and chemistry of molecularand atomic gas in the inner regions of the disk(Radius < 10AU) provides key information on thereservoir available for the formation of gas giants,and the generation and eventual delivery of suchchemicals to terrestrial planets (Henning &Semenov, 2013; van Dishoeck, 2006). ISO, Spitzer,and Herschel pioneered the studies of proto-planetary disks at mid-infrared to sub-millimeterwavelengths. SOFIA, in synergy with other groundand space based observatories, will revolutionizethis ¯eld over the next decade, and thus not onlyprovide the astronomical community with uniquedata, but also drive the science requirements anddesign of future missions like the Origins SpaceTelescope (OST; Battersby et al., 2018).
Development of ideas about the role of water indisk surfaces is currently based on limited data,taken at low spectral resolution, or based uponupper limits. The critical transition region from thewet inner disk to the dry outer disk at a few to10AU is beyond our current observational reach.
Wavelengths shorter than ! 30!m (accessible withJWST) trace the innermost disk at "1AU, whileHerschel traced the cooler outer disk at distances>1AU. However, neither of these observatoriesspectrally resolve the molecular lines nor do theyprovide information about the location of theemission. While HIRMES cannot observe the waterlines tracing the dilute water vapor beyond thesnow line at temperatures <150K, observing fromthe stratosphere on-board SOFIA does allowHIRMES to observe water lines tracing gas at 200–300K. By spectrally resolving the lines, HIRMEScan map the surface abundance of water inside ofthe transition region. This is achieved by targetingthree key spectral lines, speci¯cally: H2O : 34.9!m,[OI] : 63.1!m, and HD : 112!m (see Fig. 1).However, there are hundreds of potentially brightwater lines accessible to HIRMES over its band-width. These three key lines are selected as theyprobe di®erent regions and properties of the proto-planetary disk, and when combined, they break thedegeneracy between mass and abundance that ispresent in other observational techniques, enablingdetermination of the timescales over which water isimplanted into icy bodies to seed habitable worlds.With this, the proto-planetary disk gas mass, seenas the most fundamental quantity that determineswhether planets can form, is now within the grasp ofobservational astrophysics. Furthermore, HD hasbeen proposed as the best tracer of the total H2
Fig. 1. An artistic representation of a proto-planetary disk, sliced edge-on, with a size scale similar to that of our Solar System[Mercury ’ 0:4AU, Earth ¼ 1AU, Saturn ’ 9:5AU]. The temperature decreases with distance from the star and with thermalshielding, so any water present transits through its various states accordingly. Analyzing the three key spectral features shown herein high-resolution; H20, [OI] and HD, allows HIRMES to probe di®erent regions of disks and their properties.
SOFIA-HIRMES: Looking Forward to the HIgh-Resolution Mid-infrarEd Spectrometer
1840015-3
J. A
stro
n. In
stru
m. 2
018.
07. D
ownl
oade
d fr
om w
ww
.wor
ldsc
ient
ific.
com
by 7
6.21
8.10
7.31
on
01/1
0/19
. Re-
use
and
dist
ribut
ion
is st
rictly
not
per
mitt
ed, e
xcep
t for
Ope
n A
cces
s arti
cles
.
gas mass in proto-planetary disks (Bergin et al.,2013; McClure et al., 2016). HIRMES providesunique access to this mass tracer, which can in turnbe used to derive precise molecular abundances indisks, including observations obtained with ALMA(Bergin & Williams, 2017; Cleeves et al., 2018).
The ice/rock ratio in the solar nebula is thoughtto have been signi¯cantly larger than unity beyondthe snow line. That is, the solid mass reservoir wasdominated by ice by factors of two or more. Con-sequently, most core-accretion models form giantplanets beyond the snow line (e.g. Ida & Lin, 2008;Dodson-Robinson et al., 2008). Compared to sili-cates and other refractory materials, water ice alsosubstantially increases the sticking probability incollisions between dust grains, catalyzing the ¯rststage of planet formation by growing micron-sizeddust grains to centimeter and meter-sized icy bodies(Blum & Wurm, 2008). Solar system comets arelikely primordial tracers of the original ice/rockratios, and indeed suggest that ice was a dominantsolid mass reservoir in the Solar System. While icesare often observed via their mid-infrared bands inthe 3–20!m range, these features cannot generallybe used to measure bulk ice in disks, as dusthot enough to emit at these wavelengths no longerretain ice. The longer-wavelength phonon modes(43–63!m), on the other hand, are expected to beseen in emission in typical proto-planetary disks.The strongest feature of crystalline ice at 43 micronhas not been accessible since ISO (see Fig. 2; Malfaitet al., 1998).
Beyond spectroscopy of proto-planetary disks,HIRMES will be able to access a diverse set of ¯nestructure lines, including [FeII] 25.99!m, 35.35!m,[SI] 25.25!m, [SIII] 33.48!m, [SiII] 34.81!m,[NeIII] 36.0!m, [NIII] 57.30!m, [NII] 121.90!m,[OI] 63.18!m, and [OIII] 51.81!m, 88.35!m.These, in particular, yield line ratios that probethe abundances, ionization state, and density inshocked gas. This will permit shock models tobe tested, yielding quantitative mass °ow ratemeasurements, as well as generating velocity-re-solved line pro¯les that will elucidate the kinematicsof shocked gas. Spitzer and Herschel have observedand mapped several such transitions in proto-stellarout°ows and supernova remnants, allowing thespatial distributions of the various shock tracers tobe compared. However, because their line pro¯leswere unresolved, these observations did not supplykinematic information. HIRMES has the ability to
provide the ¯rst complete data cubes in this ¯eld,allowing supersonic motions of the shock-heated gasto be measured as a key test of shock models.
Designed to be as versatile as possible, HIRMESwill be a major enhancement to SOFIA's suite ofinstruments, supplying the astronomical communi-ty with data not yet seen. We describe the instru-ment design in Sec. 2, and the various observingmodes and techniques and data reduction in Sec. 3.The authors must stress that the design and sensi-tivities are still preliminary (correct at the timeof writing). As the instrument undergoes I&Tthrough Commissioning and Acceptance, it is likelythat some parameters presented here will beupdated. As such, we recommend the reader to keepvisiting the SOFIA-HIRMES webpage(a) for themost up-to-date information.
2. HIRMES Instrument Design
2.1. Overview
The HIRMES instrument is a vacuum cryostatusing a TransMIT pulse-tube cryocooler (PTC)with a Commercial O®-The-Shelf (COTS) 4He re-frigerator and an Adiabatic DemagnetizationRefrigerator (ADR) to cool the Transition EdgeSensor (TES) bolometers to 70mK. The instrumentblock diagram (see Fig. 4; top) identi¯es the sub-systems required to achieve the aforementionedscience, including the cryostat, Fabry–Perot
Fig. 2. An adaptation of McClure et al. (2015)'s ¯gure,showing the emission/absorption coe±cients of the 43, 47, and63!m water–ice features that observationally infer the thermalhistory of grain mantles.
ahttps://www.so¯a.usra.edu/science/instruments/hirmes.
S. N. Richards et al.
1840015-4
J. A
stro
n. In
stru
m. 2
018.
07. D
ownl
oade
d fr
om w
ww
.wor
ldsc
ient
ific.
com
by 7
6.21
8.10
7.31
on
01/1
0/19
. Re-
use
and
dist
ribut
ion
is st
rictly
not
per
mitt
ed, e
xcep
t for
Ope
n A
cces
s arti
cles
.
Interferometers (FPIs), optics, mechanisms, detec-tors, mechanical structures, and instrument controland data handling electronics. The following sectionsets out to detail the primary subsystems. A cut-section of the latest HIRMES CAD model is given inFig. 3. HIRMES is !1m in diameter and !2m inlength.
2.2. Optical design
2.2.1. Window
HIRMES uses a 101.6 mm (4-inch) diameter (3-inchclear aperture) Topas(b) (COC, Cyclic Ole¯n Co-polymer) window, as the transmissive vacuumboundary between the instrument vacuum spaceand the Telescope Assembly (TA). Topas has lowabsorption over mid- to far-infrared wavebands andis also transparent in the visible, facilitating align-ment and inspection. The average transmission is93%, with a re°ection of 5%, and an absorption of2%. A series of tests have been performed on the
Topas window, including vacuum, stress, thermal,and environmentally testing a range of windowthicknesses and curvatures. The result was to use awindow with thickness of 400!m and a radius ofcurvature of 66.5mm. This radius of curvature is aformed curvature, concave to the instrument, whichdoes not appreciably change under vacuum load.
2.2.2. Optics
The optical system provides e±cient coupling to thetelescope, controls stray light, provides ¯ltering,and images the spectrum/image onto the detectorarrays (see Fig. 4; middle). The ¯xed optical com-ponents are manufactured with standard diamondturning process, and all of the components are fab-ricated from aluminum, thus the design of the op-tical bench is isothermal and contracts conformally.This allows the alignment to be completely warmwith assurance that it will still be within tolerancewhen the instrument is cooled. The instrument'sopto-mechanical implementation is given in Fig. 4(bottom).bhttps://topas.com/products/topas-coc-polymers.
Fig. 3. A cut-section of the HIRMES CAD model. Some components have been removed or phantomized to aid clarity. The lightfrom the Telescope Assembly (TA) enters through the window on the right.
SOFIA-HIRMES: Looking Forward to the HIgh-Resolution Mid-infrarEd Spectrometer
1840015-5
J. A
stro
n. In
stru
m. 2
018.
07. D
ownl
oade
d fr
om w
ww
.wor
ldsc
ient
ific.
com
by 7
6.21
8.10
7.31
on
01/1
0/19
. Re-
use
and
dist
ribut
ion
is st
rictly
not
per
mitt
ed, e
xcep
t for
Ope
n A
cces
s arti
cles
.
Fig. 4. The HIRMES sub-system block diagram shows the instruments internal and external components, the optical path withelements (middle, not to scale), and a labeled CAD model of the optical bench with opposing vantage points (bottom).
S. N. Richards et al.
1840015-6
J. A
stro
n. In
stru
m. 2
018.
07. D
ownl
oade
d fr
om w
ww
.wor
ldsc
ient
ific.
com
by 7
6.21
8.10
7.31
on
01/1
0/19
. Re-
use
and
dist
ribut
ion
is st
rictly
not
per
mitt
ed, e
xcep
t for
Ope
n A
cces
s arti
cles
.
Stage 1:
The SOFIA telescope delivers a di®raction-limitedf/19.5 beam over the HIRMES bandpass. HIRMES'slits are 113 arcsec long, with a selectable slit widthfrom 3.0 to 8.7 arcsec to match the di®raction imagesize at the selected wavelength range. The slitlength and the image ¯eld of view are well withinSOFIA's 8 arcmin diameter ¯eld. The telescope'simage rotation, generated by a moving observatoryand the telescope's 3-axis spherical bearing mount,is handled operationally by adjusting the telescope's¯ne-drive to produce a ¯xed sky rotation for a pe-riod of time, typically 10–30mins depending onobservational geometry (for more details see Temiet al., 2014). As such there is no hardware image de-rotator. A system of two folding mirrors redirectsthe beam onto a collimator that produces a parallelbeam at Stage 1. This O®-Axis-Paraboloid (OAP)collimator also forms an image of the telescope pupilof 20mm diameter onto a cold stop. The ¯lterwheel, accommodating 12 positions for the instru-ment ¯lters, and the low-resolution FPI wheel, areplaced at the approximate location of this pupilimage. Following the low-resolution FPI wheel,there is the slit wheel with four slits and the openaperture for the Spectral Imaging mode. After theslit wheel, two folding mirrors redirect the beamonto the Stage 2 collimator.
Stage 2:
An OAP collimator produces parallel beams to gothrough the mid-resolution and high-resolutionFPIs ¯lter wheels. It also produces an 80mmdiameter pupil image onto the mid-resolution FPI.The mid-resolution FPIs wheel is tilted by ! 0:1degrees, and the high-resolution FPIs are placed at50mm from the mid-resolution FPIs to eliminateparasitic ringing between the mid-resolution andhigh-resolution FPI mirrors. Another OAP relaysthe telescope image onto another intermediate focalplane with an aperture to prevent scattered lightgoing through the system. Due to the walk-o® of thebeam through the high-resolution FPIs (primarilyat the longer wavelengths) the outgoing beam isexpanded.
Stage 3:
Another OAP collimates the beam after the inter-mediate focus and forms a 40mm pupil for thegrating wheel, and the camera OAP relays thespectrally-dispersed light to the detectors. The °at
mirrors in Stage 3 were added to package the opticsinto the available cold space. Final order-sorting isperformed by a set of three re°ective di®ractiongratings. The linear dispersion at the detector arrayis set to spread the FSR of the mid-resolution FPIsonto at least 1 pixel at short wavelengths, and tospread the FSR wider than the point spread func-tion (PSF) at long wavelengths. A rotary stagemechanism is used to select the grating and to setits angle over a range of $8:1 degrees to choose thespectral line of interest. Each grating is blazed tooptimize the e±ciency over three sub-bands of thewavelength range. A mirror used for the spectralimaging mode occupies the fourth position on thestage. The spectral image (R ! 2000) is projectedonto one side of the low-resolution detector array.
The optical imaging system is di®raction limited(% "=14 at 24!m) over the whole spectral range ofthe instrument. In fact, the quality of the re°ectiveoptical elements are such that they could be used foroptical wavelengths. The instrument's internalstray light ba®les are blackened with a 500 micronthick layer of an absorptive mixture comprised of65% Epotek 377, 30% fumed (pyrogenic) silicapowder, and 5% graphene by volume (alternatively,50.7% Epotek 377, 42.8% fumed silica, and6.5% graphene by weight) (Chuss et al., 2017). Amonolayer of K1 borosilicate glass microspheressieved to diameter <100!m can be incorporated onthis lossy dielectric surface and overpainted with!50!m of Aeroglaze Z306,(c) where dilutionthrough scattering is desirable to control the opticalresponse. The resulting coating is CTE-matched(Coe±cient of Thermal Expansion) for use on me-tallic substrates, non-magnetic, and robust underthermal cycling. Any rejected light is expected toexit the optical system or be absorbed by the par-titioning blackened ba®les in the front end of theinstrument.
2.2.3. Filters
The ¯lters use a combination of technologies tocover the full 25–122!m spectral range: a dielectricmultilayer ¯lter to provide the 25!m cut-on atshort wavelength, and crystal ¯lters at the longerwavelengths. Table 1 details the speci¯cations ofeach available ¯lter mounted on the Stage 1 ¯lter
cLord Corporation Chemical Products, \Aeroglaze Z306 FlatBlack Absorptive Polyurethane Low-Outgassing Paint", 2000West Grandview Blvd., P.O. Box 10038, Erie, PA 16514-0038.
SOFIA-HIRMES: Looking Forward to the HIgh-Resolution Mid-infrarEd Spectrometer
1840015-7
J. A
stro
n. In
stru
m. 2
018.
07. D
ownl
oade
d fr
om w
ww
.wor
ldsc
ient
ific.
com
by 7
6.21
8.10
7.31
on
01/1
0/19
. Re-
use
and
dist
ribut
ion
is st
rictly
not
per
mitt
ed, e
xcep
t for
Ope
n A
cces
s arti
cles
.
wheel (see Fig. 4). The ¯ve \First-order FPIþlongpass" ¯lters are used in the Spectral Imaging modeas order-sorting ¯lters that each consist of a coupled¯xed-width FPI and long-pass ¯lter. Either plastic¯lm or vapor deposited parylene are used for theanti-re°ection coatings.
2.2.4. Fabry–Perot interferometers
HIRMES uses a °eet of FPIs for its various ob-serving modes. The high spectral resolution mode ofR ¼ 100; 000 is achieved with FPIs. Also, FPIs withlow spectral resolution (R ! 2000) will be used forSpectral Imaging.
An FPI consists of two highly re°ective plane-parallel mirrors that form a resonant cavity. Theresonance condition is 2 ' d ¼ n ' ", where d is thecavity spacing, " is the wavelength, and n is aninteger order. The spectral resolution of an FPI isthe product of the ¯nesse, F (the FPI e±ciencydetermined by the absorption in the re°ector andthe mean number of re°ections in the cavity), andthe order, n. To create a spectrum, the cavityspacing d is adjusted, changing the resonant wave-length. Since any wavelength that ful¯lls the reso-nance condition will be transmitted by the FPI, it isnecessary to employ additional ¯lters to sort out theunwanted wavelengths. HIRMES uses additionalFPIs with a resolution of about 12,000 (mid-reso-lution FPIs) and the re°ective grating to select thedesired wavelength (see Fig. 5).
Rays within either the axial beam or o®-axisbeams that pass the FPI at an angle with respect tothe normal axis will resonate at a slightly shorterwavelength. The angle is minimized when the FPIsare located at pupil positions and have a large ap-erture, as is the case for the HIRMES FPIs. Inaddition, a high spectral resolution requires oper-ating the FPI at a high order, hence a large cavity
spacing. These two conditions determine the phys-ical size of the FPIs. Although o®-axis beams thatpass the FPI at an angle with respect to the normalaxis of the mirrors resonate at shorter wavelength,we will use this feature as an advantage. By slewingthe telescope during the observation, and thusmoving the science target from axial to o®-axispixels, we also spectrally sweep over a part of thenearby spectrum. To obtain the full desired spec-trum thus requires less steps of changing the cavityspacing of the FPIs (see Fig. 6).
The FPIs in HIRMES use free-standing nickelmetal meshes with a gold-°ash as their highly re-°ective mirrors (Ulrich et al., 1963). The ¯nesseobtained with the metal meshes is a function ofwavelength. Optimal ¯nesses of FPIs made withthese meshes ranges between 30–60 (a ¯nesse that istoo high would reduce the overall transmission).Thus, for each of the mid- and high-resolutionmodes, HIRMES has three separate FPIs to coverthe full wavelength range between 25 and 122!m.The low-, mid-, and high-resolution FPIs have ascanning mechanism using piezo elements to selectthe desired cavity spacing, and step over a wave-length range to create spectra. In addition to thetunable FPIs, HIRMES will employ FPIs with a¯xed cavity spacing that are tuned to speci¯c ¯ne-structure lines. The overall transmission that eachof the FPIs will achieve is over 70%. A full write-upof the FPIs used in HIRMES can be found atDouthit et al. (2018) and Cothard et al. (2018).
2.2.5. Slits
The slit wheel holds four slits of length ¼ 113 arcsec,and widths ¼ 8:7, 6.1, 4.2, and 3.0 arcsec, respec-tively. These are selected based on the desired cen-tral wavelength and resolution. Additional to theslits, there is also a 2D image-stop of dimensions
Table 1. Properties of the ¯lters used in the Stage 1 ¯lter wheel.
Filter Design Details Mean transmission (%)
23–32!m; Bandpass Multi-layer dielectric On CdTe substrate 7026!m; Long Pass Al2O3 with near IR blockers Parylene AR (P.AR) coating 7540!m; Long Pass Al2O3 þCaF2 stack P.AR coating on outer layers 7565!m; Long Pass Al2O3 þCaF2 þBaF2 stack P.AR coating on outer layers 7551.8!m; [OIII] First order FPI þ long pass Metal mesh 5057.3!m; [NIII] First order FPI þ long pass Metal mesh 5063.2!m; [OI] First order FPI þ long pass Metal mesh 5088.4!m; [OIII] First order FPI þ long pass Metal mesh 50121.9!m; [NII] First order FPI þ long pass Metal mesh 50
S. N. Richards et al.
1840015-8
J. A
stro
n. In
stru
m. 2
018.
07. D
ownl
oade
d fr
om w
ww
.wor
ldsc
ient
ific.
com
by 7
6.21
8.10
7.31
on
01/1
0/19
. Re-
use
and
dist
ribut
ion
is st
rictly
not
per
mitt
ed, e
xcep
t for
Ope
n A
cces
s arti
cles
.
Fig. 5. Top to bottom: transmission of high-resolution FPI,mid-res-FPI, grating, and their product. The product is veryspectrally pure. These ¯gures are purely graphical, and are notbased on measurements.
Fig. 6. Top to bottom: transmission pro¯le at 112!m and63!m (high-resolution FPI) and 122!m and 52!m (imaging,low-resolution FPI) for beams on and o® the optical axis.1 beam ¼ 1 di®raction limited PSF at the detector.
SOFIA-HIRMES: Looking Forward to the HIgh-Resolution Mid-infrarEd Spectrometer
1840015-9
J. A
stro
n. In
stru
m. 2
018.
07. D
ownl
oade
d fr
om w
ww
.wor
ldsc
ient
ific.
com
by 7
6.21
8.10
7.31
on
01/1
0/19
. Re-
use
and
dist
ribut
ion
is st
rictly
not
per
mitt
ed, e
xcep
t for
Ope
n A
cces
s arti
cles
.
113:0 ( 106:8 arcsec used in the Spectral Imagingmode, which is then projected onto 16 ( 16 pixels ofthe low-resolution detector array.
2.2.6. Gratings
In the HIRMES wavelength range, echelette(blazed) gratings have near-ideal performance, sotheir e±ciency can be calculated accurately usingdi®raction models and groove geometry. Thesegratings, with an average e±ciency of ! 0:9 (seeFig. 7), were chosen to optimize performance at themost important spectral lines.
2.3. Detectors
The HIRMES detector layout consists of two TESbolometric detector arrays (see Fig. 8) that equipHIRMES with eight subarrays of 1 ( 16 pixels forthe high-resolution mode (low saturation power),and a 64 ( 16 array for the mid- and low-resolutionand Spectral Imaging modes (high saturationpower). A full description of the HIRMES detectorscan be found in Brown et al. (2018) and Barrentineet al. (2018).
The \low-resolution" 64 ( 16 array is com-prised of 1mm ( 1mm square pixels with a 50%absorptive frequency-independent coating. Thethermal isolation design consists of eight single-crystal silicon legs that are 1.4!m thick, 50!mwide, and 30!m long, and will provide an expectedNoise Equivalent Power (NEP) !2 ( 10 ) 17W=ffiffiffiffiffiffiHz
pand saturation power !25 pW, inclusive of the
50% absorption e±ciency.
The \high-resolution" array is comprised ofeight, 1 ( 16 pixel subarrays, and is operated in amanner in which only one detector subarray isused at a given time. Each subarray is separatelyoptimized for operation at a speci¯c wavelength,by matching to the Full-Width Half-Maximum(FWHM) of the beam at the wavelength, and pro-viding e±cient absorption using a quarter-waveback short (Miller et al., 2018). The central wave-length for each subarray is: 30, 36, 43, 51, 61, 73, 88and 105!m, and their respective physical pixel-widths range from 0.4–1.4mm. This system, withbackground-limited detectors, is near optimal forthe high-resolution spectroscopy mode throughoutthe entire spectral range. The detectors are Mo/AuTES bilayers deposited on photolithographically-de¯ned leg-isolated 0.45!m thick, 5!m wide,and 30!m long single-crystal silicon membranes.These detectors have an expected NEP !3 (10 ) 18W=
ffiffiffiffiffiffiHz
pand saturation power !0:13 pW.
The detectors are read out using NIST 1 ( 11multiplexers operating at the detector temperatureof 70mK. The signal is ampli¯ed by a SQUID seriesarray operating at !4K, and controlled by aroom temperature Multi-Channel Electronics(MCE) controller (Hassel¯eld, 2013; Hendersonet al., 2016). The low-resolution and high-resolutiondetectors are packaged in a single, common FocalPlane Assembly (FPA), and kinematically mountedusing a Kevlar and magic-Titanium [15-3-3-3]support system to provide thermal and mechanicalisolation.
2.4. Cryostat and ADR
HIRMES implements a tri-layer all aluminum(6061-T6) cryostat design (see Figs. 3 and 4), withthe load path extending from the °ange ring (wherethe instrument mounts to the TA), through thecentral `bellyband' support region of the cryostat.The external hemispherical end caps, as well as the65K and 4K heat shields, are supported from thisbellyband, along with the monolithic 4K opticalbench (from a three-point mount). This benchsupports all working instrument components andoptics. The heat shields and optical bench are sup-ported by aluminum rings suspended using 12 tita-nium struts with titanium end caps from the bellyband. The optical bench, bulkheads, and structuralcomponents are made of the same forged 6061-T6aluminum as the cryostat, providing high strengthand thermal expansion coe±cients that match the
Fig. 7. The three selectable di®raction gratings provide >90%average e±ciency of both s- and p-polarizations over theHIRMES spectral region.
S. N. Richards et al.
1840015-10
J. A
stro
n. In
stru
m. 2
018.
07. D
ownl
oade
d fr
om w
ww
.wor
ldsc
ient
ific.
com
by 7
6.21
8.10
7.31
on
01/1
0/19
. Re-
use
and
dist
ribut
ion
is st
rictly
not
per
mitt
ed, e
xcep
t for
Ope
n A
cces
s arti
cles
.
mirrors and mirror mounts. All aluminum mirrorsand the optical bench are machined, annealed, andassembled to ensure dimensional stability at cryo-genic operating temperatures, preventing hysteresisfrom repeated cool-downs.
The HIRMES thermal design uses a two-stagePTC to cool the instrument and optical bench.A 4He sorption fridge coupled to an ADR; salt-pillis subsequently used to cool the detectors locatedin the FPA to 70mK, with an expected operationalhold-time of >12 h. A TransMIT PTD406C
pulse-tube head is mated to one of the two SOFIAon-board Cryomech CP2870 He compressors. ThePTC ¯rst stage cools the outer radiation shieldbelow 65K. This lowers the radiation heat load fromthe cabin-temperature cryostat shell to a level thatallows the second stage to cool the optical benchand inner radiation shield to 4K.
A 4He sorption fridge then lowers the ADR heatsink temperature to 1.3K, allowing a quicker recycletime and enhanced cooling power. From this lowstarting temperature, the ADR is able to cool the
Fig. 8. (Top) One half of the 64 ( 16 pixel low-resolution detector. (Middle-Left) A zoom in of the four most lower-right pixels ofthe Top image. (Middle-right) The eight, 1 ( 18 pixel subarrays of the high-resolution detector, in which the pixels located at theedges of each subarray are not read out, resulting in 1 ( 16 active pixel subarrays. (Bottom) A schematic of the layout of bothdetectors on the Focal Plane.
SOFIA-HIRMES: Looking Forward to the HIgh-Resolution Mid-infrarEd Spectrometer
1840015-11
J. A
stro
n. In
stru
m. 2
018.
07. D
ownl
oade
d fr
om w
ww
.wor
ldsc
ient
ific.
com
by 7
6.21
8.10
7.31
on
01/1
0/19
. Re-
use
and
dist
ribut
ion
is st
rictly
not
per
mitt
ed, e
xcep
t for
Ope
n A
cces
s arti
cles
.
detector to 70mK. Additionally, a separate 3Hesorption fridge acts as a 0.3K thermal intercept tothe ADR, reducing its thermal load stress.
2.5. Instrument calibration
HIRMES calibration captures the wavelength set-tings for observations, instrumental spectral pro¯le,and radiometric calibration to allow the line in-tensity to be determined. The spectral calibrationuses gas cells for absolute wavelength calibration onthe ground, and two Quantum Cascade Lasers(QCLs) for in-°ight veri¯cation (! 63 and 83!m,R ! 106). The QCLs are mounted to the 65Kstage, and are injected into the instrument by re-°ection o® a mirror on the back of the slit wheel.The QCLs provide good short-term wavelengthstability; absolute knowledge is provided by a ¯xedetalon used with the QCL. With knowledge of thespectral setting and instrumental pro¯le, radio-metric calibration is achieved by observing knowncontinuum sources, including rocky moons andasteroids. During operation, the FPI spacing ismeasured and monitored with FPI capacitive sen-sors. Parallelism is established at room temperatureand only requires minor adjustment when cold by aknown, ¯xed amount, through the usage of low-voltage, tilt PICMA piezo actuators from PhysikInstrumente.
3. Observing with HIRMES
3.1. Observing modes
By combining the direct-detection arrays (TESbolometers), grating-dispersive spectroscopy, and ahost of Fabry–Perot tunable narrow-band ¯lters,HIRMES can provide four primary observingmodes: High resolution (R ! 100; 000), mid-resolu-tion (R ! 10; 000) and low-resolution (R ! 600)spectroscopy, and Spectral Imaging (R ! 2000).Figure 9 shows how the various optical and spectralelements can combine to support each primary ob-serving mode. Figures 5 and 6 are useful to refer towhen reading through the following mode descrip-tions. HIRMES is a complex instrument with manycon¯gurable elements, however, the combinationselection for various modes will be operationallyautomatic, based on the desired science. One onlyneeds to select the central wavelength, delta wave-length, number of steps, and one of the four modes.If instrumental resolution is desired, then one only
needs to specify the wavelength range and themode. Having said that, any combination of opticaland spectral elements is technically feasible tocreate additional modes. A summary table of theobserving modes and their respective properties isgiven in Table 2.
3.1.1. Low-resolution spectroscopy
The light ¯rst passes through a bandpass ¯lter, thena slit, then a re°ective grating. The spectrum isunder-sampled by the 64 ( 16 pixel detector, pro-ducing a resolution of R ! 320–635 (see Fig. 9,lower-right), and instantaneous bandwidths of5–15!m, depending on the wavelength. It wouldtake nine di®erent wavelength settings to obtain thefull 25–122!m spectral range in this mode.
3.1.2. Mid-resolution spectroscopy
This mode keeps the con¯guration of the low-resolution grating mode, however, inserts amid-resolution FPI into the optical path. The e®ectof this is a narrow, sharp wavelength peak that isgreatly under-sampled in one of the pixels of the64 ( 16 pixel array. The FPI is stepped through oneFSR (in roughly 10–50 steps depending on wave-length and desired spectral sampling) to produce aspectrum over the full instantaneous spectral cov-erage of the low-resolution grating mode.
3.1.3. High-resolution spectroscopy
Going one step further to the mid-resolution mode,an additional high-resolution FPI is inserted intothe optical path, and the central wavelength iscentered onto the appropriate column of the eight1 ( 16 pixel linear subarrays of the high-resolutiondetector (which linear subarray depends on wave-length; subarray pixel size is proportional to wave-length). That means there is only a single pixel inthe spectral dimension, which is closely matched tothe PSF. However, due to the radially dispersivenature of FPIs, stepping spatially up and downthe slit also results in a slight wavelength shift(! 1–5 km/s depending on wavelength, see Fig. 6).Combining this feature with stepping the FPIs bydiscrete steps, results in the spectral sampling of adesired wavelength range (typically a single narrowspectral line). Figure 5 visualizes the product of theFPIs working together with the grating to produce ahigh-resolution line spectrum.
S. N. Richards et al.
1840015-12
J. A
stro
n. In
stru
m. 2
018.
07. D
ownl
oade
d fr
om w
ww
.wor
ldsc
ient
ific.
com
by 7
6.21
8.10
7.31
on
01/1
0/19
. Re-
use
and
dist
ribut
ion
is st
rictly
not
per
mitt
ed, e
xcep
t for
Ope
n A
cces
s arti
cles
.
Fig. 9. The four primary observing modes and their combination of optical and spectral elements. The lower-left plot shows alloptical and spectral elements. The lower-right plot shows the resolution as a function of wavelength for the low-resolution gratingmode. Note that it would take nine di®erent wavelength settings to obtain the full 25–122!m spectral range in this mode (three foreach grating).
SOFIA-HIRMES: Looking Forward to the HIgh-Resolution Mid-infrarEd Spectrometer
1840015-13
J. A
stro
n. In
stru
m. 2
018.
07. D
ownl
oade
d fr
om w
ww
.wor
ldsc
ient
ific.
com
by 7
6.21
8.10
7.31
on
01/1
0/19
. Re-
use
and
dist
ribut
ion
is st
rictly
not
per
mitt
ed, e
xcep
t for
Ope
n A
cces
s arti
cles
.
3.1.4. Spectral imaging
This mode changes the con¯guration completely byswitching out the initial bandpass ¯lter with anarrow-band ¯lter on the same ¯lter wheel. Thisnarrow-band ¯lter is actually a combination of a¯xed-width FPI, and its own bandpass ¯lter, bothof which are tailored for speci¯c spectral lines (seeTable 1). The other con¯gurable elements used inoptical path are a low-resolution FPI, a squareimage-stop instead of a slit, and a mirror instead ofthe grating. The 2D image is then placed on one sideof the 64 ( 16 pixel array, to produce a 16 ( 16 pixelspectral image (113:0 ( 106:8 arcsec), whose wave-length is also variable over the image, due to theradially dispersive nature of the FPI (see Fig. 6).
3.2. Sensitivity limits
As the instrument is still being built, full charac-terization and calibration of the various compo-nents is yet to be completed. The instrumentsensitivities or capabilities presented here are thebest current estimates based on analysis of thedesign. In particular, Fig. 10 visualizes the esti-mates of Minimum Detectable Line Fluxes(MDLFs) for the various observing modes. One canextrapolate an estimate of the total time on-sourcefor a given °ux from Fig. 10, and later apply at-mospheric transmission factors. Observationaloverheads are not taken into account, and will befully realized once cold function checks andcommissioning are underway.
3.3. Observing techniques
Each detector read-out has associated astrometricand timing data, enabling HIRMES to be able tosupport any TA observing mode (e.g. Lissajous,raster-scan, slit-scan, mapping, chop and nod,etc.). The typical TA Observing mode when ineither of the Low-, Mid-, or high-resolution modeswould be slit-scan (scanning up and down thelength of the slit), and Lissajous for Spectral Im-aging. Both of these modes would be performedwithout chopping.
Fig. 10. Visual representation of the Minimum DetectableLine Fluxes (MDLFs) for the di®erent modes, assuming noatmosphere. Note that the Spectral Imaging mode cannot coverthe entire wavelength range, only the discrete wavelengths,shown by square line-markers. The wavelengths of key spectralline features are marked by vertical blue dotted lines.
Table 2. Properties of the ¯lters used in the Stage 1 ¯lter wheel.
Parameters High-res. Mid-res. Low-res. Imaging
Sensitivity (5#, 1 h) . 1 ( 10 ) 17 W/m2 ! 1 ( 10 ) 16 W/m2
Resolving power (R ¼ "=$") 50,000–100,000 ! 12; 000 325–635 ! 2; 000
Angular resolution Di®raction limited
Slit size/FOV (arcsec) Length: 113"; Width: 8.7", 6.1", 4.2" and 3.0" 113.0" ( 106.8"Spectral range 25–122!m Selected linesa
Simultaneous spectral coverage ($"=") "/R 0.1" 0.001"Detector format 8 ( 16 pixb 64 ( 16 arrayc
Detector type Transition edge sensor (TES)
Notes: aSingle wavelength setting for selected ¯lters (63.2!m [OI]; 51.8!m, 88.4!m [OIII]; 57.3!m [NIII]; 121.9!m[NII]).bHigh resolution detector consists of eight, 1 ( 16 pixel linear subarrays; whose pixel size increases per subarray. Shorterwavelength light is positioned onto the smaller pixel subarrays, longer wavelength light onto the larger pixel subarrays.cSpectral Imaging uses only a 16 ( 16 pixel section of the 2D array.
S. N. Richards et al.
1840015-14
J. A
stro
n. In
stru
m. 2
018.
07. D
ownl
oade
d fr
om w
ww
.wor
ldsc
ient
ific.
com
by 7
6.21
8.10
7.31
on
01/1
0/19
. Re-
use
and
dist
ribut
ion
is st
rictly
not
per
mitt
ed, e
xcep
t for
Ope
n A
cces
s arti
cles
.
The intent of selecting slit-scan and Lissajousfor typical TA observing modes is two-fold: (1) tomaximize the spectral bandwidth by moving spa-tially across the FPIs, and (2) to break the degen-eracy of having the same sky and/or source °ux onthe same pixels, thereby increasing the ability tocharacterize the detector for data reduction. Thisis also achieved by a new TA Observing modein development that allows the sky to rotatewhilst tracking. Atmosphere-less moons and aster-oids will be regularly observed, in addition to blank-sky/sky-dips, for °ux calibration and telluriccorrection.
3.4. Control and data reduction software
The HIRMES software system design includes: (1)Graphical User Interface (GUI) for instrumentcontrol and monitoring; (2) detector data acquisi-tion; (3) detector and readout tuning and control;(4) calibrate mechanisms and spectroscopicelements; (5) interface with ancillary devices; (6)interfacing with telescope and executing observa-tions via the SOFIA Command Language (SCL);(7) power and thermal management; (8) health andsafety monitoring, providing timely feedback to theinstrument operator; (9) data reduction tools andpipeline Levels 1–4 science product generation;(10) data archival following all SOFIA Data CycleSystem (DCS) requirements; (11) user documenta-tion. The HIRMES software architecture consists oftwo sub-systems: the HIRMES Command and DataHandling (CDH) software system, and the DataAnalysis and Products (DAP) software.
The HIRMES CDH software is based on theAurora application framework (Steve Maher, pri-vate communication), which provides °exible,platform-independent, Java-based utilities for sci-enti¯c data systems and devices. Science data isarchived by converting the detector data, telescopesettings and astrometry, and instrument con¯gu-ration into Flexible Image Transport System(FITS) format ¯les. The detector data and telescopeastrometry are tightly synchronized using theSOFIA Inter-Range Instrument Group (IRIG)timing framework. Indexes generated from the FITSheader information permit fast retrieval andreporting of science/engineering data. The time-se-ries FITS output of Aurora constitutes Level-1 data,and is archived in its entirety through the DCS.
The Level-1 FITS data is then fed into a mod-i¯ed version of Comprehensive Reduction Utility
for SHARC-2 (CRUSH(d); Kov!acs, 2008), whichproduces Level-2 data after performing the follow-ing steps: evenly spectrally and spatially resampled(FITS data cube: RA, DEC, wavelength); re-orien-tated to North-up, East-left; instrument calibrated(removal of correlated noise, pixel masking, biasing,etc); and WCS and DCS compliance. The DAPpipeline can then take this Level-2 data and produceLevel-3 data after °ux calibration and telluric cor-rection. After this has been achieved, DAP canproduce Level-4 data by combining data frommultiple observations, stitching map-pointingstogether, extracting 2D or 1D spectra, etc.
3.5. Data formats and access
All of the Level-1 to Level-4 data will be provided inmulti-extension FITS data cubes of °ux, variance,and instrument and observation parameters. Alldata are fully FITS and DCS compliant. Addition-ally, the data will be compliant with the NASA/IPAC InfraRed Science Archive (IRSA(e)), to sup-port plans for the DCS archive to be injestedtherein, within the next year or so. Viewing andexploration of these FITS data cubes will be possiblevia a future modi¯ed version of SpexTool,(f) in ad-dition to your local, friendly data cube viewers, suchas QFitsView.(g)
Acknowledgments
The development of HIRMES is funded by theNASA/SOFIA 3rd Generation Instrument solicita-tion to the NASA GSFC and partnering institu-tions. We thank the USRA/SOFIA ScienceOperation Center sta® and NASA Armstrong B703sta® for their ongoing support during the develop-ment of HIRMES and its upcoming commissioning.
This research was conducted [in part] at theSOFIA Science Center, which is operated by theUniversities Space Research Association undercontract NNA17BF53C with the National Aero-nautics and Space Administration.
We would like to extend our thanks to DavidFranz, Kevin Denis, Manuel Balvin, George Manos,and Elissa Williams for their contribution towards
dCRUSH: https://github.com/attipaci/crush.e IRSA: https://irsa.ipac.caltech.edu/.fSpexTool: http://irtfweb.ifa.hawaii.edu/!spex/observer/.gQFitsView: http://www.mpe.mpg.de/!ott/QFitsView/.
SOFIA-HIRMES: Looking Forward to the HIgh-Resolution Mid-infrarEd Spectrometer
1840015-15
J. A
stro
n. In
stru
m. 2
018.
07. D
ownl
oade
d fr
om w
ww
.wor
ldsc
ient
ific.
com
by 7
6.21
8.10
7.31
on
01/1
0/19
. Re-
use
and
dist
ribut
ion
is st
rictly
not
per
mitt
ed, e
xcep
t for
Ope
n A
cces
s arti
cles
.
useful discussions and fabrication support on thedetector arrays.
References
Andrews, S. M. [2015] Publ. Astron. Soci. Pac. 127, 961.Barrentine, E. M., Rostem, K., Brekosky, R. P. et al. [2018]
J. Low Temp. Phys. 193, 241.Battersby, C., Armus, L., Bergin, E. et al. [2018] Nat. Astron.
2, 596.Bergin, E. A., Cleeves, L. I., Gorti, U. et al. [2013]Nature 493, 644.Bergin, E. A. & Williams, J. P. [2017] Astrophys. Space Sci.
Libr. 445, 1.Blum, J. & Wurm, G. [2008] Annu. Rev. Astron. Astrophys.
46, 21.Brown, A.-D., Brekosky, R., Franz, D. et al. [2018] J. Low
Temp. Phys, doi: 10.1007/s10909-018-1914-3.Chuss, D. T., Rostem, K., Wollack, E. J. et al. [2017] Rev. Sci.
Instrum. 88, 104501.Cleeves, L. I., Öberg, K. I., Wilner, D. J. et al. [2018] ApJ 865,
155.Cothard, N. F., Abe, M., Nikola, T. et al. [2018] Proc. SPIE
10706, 107065B.
Dodson-Robinson, S. E., Bodenheimer, P., Laughlin, G. et al.[2008] ApJ 688, L99.
Douthit, G., Stacey, G., Nikola, T. et al. [2018] Proc. SPIE10708, 107081P.
Hassel¯eld, M. [2013] Galaxy Cluster Cosmology with the Ata-cama Cosmology Telescope, PhD Thesis, The University ofBritish Columbia, Vancouver.
Henderson, S. W., Stevens, J. R., Amiri, M. et al. [2016] Proc.SPIE 9914, 99141G.
Henning, T. & Semenov, D. [2013] Chem. Rev. 113, 9016.Ida, S. & Lin, D. N. C. [2008] ApJ 685, 584.Kov!acs, A. [2008] Proc. SPIE 7020, 70201S.Malfait, K., Waelkens, C., Waters, L. B. F. M. et al. [1998]
A&A 332, L25.McClure, M. K., Espaillat, C., Calvet, N. et al. [2015] ApJ 799,
162.McClure, M. K., Bergin, E. A., Cleeves, L. I. et al. [2016] ApJ
831, 167.Miller, T. M., Brown, A.-D., Costen, N. et al. [2018] J. Low
Temp. Phys, doi: 10.1007/s10909-018-1939-7.Temi, P., Marcum, P. M., Young, E. et al. [2014] ApJS 212, 24.Ulrich, R., Renk, K. F. & Genzel, L. [1963] IEEE Trans.
Microw. Theory Tech. 11, 363.van Dishoeck, E. F. [2006] Proc. Nat. Acad. Sci. 103, 12249.
S. N. Richards et al.
1840015-16
J. A
stro
n. In
stru
m. 2
018.
07. D
ownl
oade
d fr
om w
ww
.wor
ldsc
ient
ific.
com
by 7
6.21
8.10
7.31
on
01/1
0/19
. Re-
use
and
dist
ribut
ion
is st
rictly
not
per
mitt
ed, e
xcep
t for
Ope
n A
cces
s arti
cles
.
