Heterodyne Receiver for Origins

Abstract. The Heterodyne Receiver for Origins (HERO) is the first detailed study of a heterodyne focal plane array receiver for space applications. HERO gives the Origins Space Telescope the capability to observe at very high spectral resolution (R  =  107) over an unprecedentedly large far-infrared (FIR) wavelengths range (111 to 617  μm) with high sensitivity, with simultaneous dual polarization and dual-frequency band operation. The design is based on prior successful heterodyne receivers, such as Heterodyne Instrument for the Far-Infrared /Herschel, but surpasses it by one to two orders of magnitude by exploiting the latest technological developments. Innovative components are used to keep the required satellite resources low and thus allowing for the first time a convincing design of a large format heterodyne array receiver for space. HERO on Origins is a unique tool to explore the FIR universe and extends the enormous potential of submillimeter astronomical spectroscopy into new areas of astronomical research.

an ultrastable mid-infrared (mid-IR) spectrometer. To fully exploit the capabilities of Origins, two upscope instruments have been designed: the Mid-Infrared Spectrometer and Camera: Wide Field Imager (MISC-WFI) 7

and the Heterodyne Receiver for Origins (HERO).
HERO's principal goal is to trace the trail of water 8-10 from the interstellar medium (ISM), through the different stages of star formation to (habitable) planets. HERO will measure the amount, location, and temperature of water and its isotopologues in the ISM, protostars, and protoplanatary disks. This science goal requires HERO to have extremely high spectral resolution of up to 10 7 , to cover the major water lines and their isotopologues between 500 and 2700 GHz, and to have moderate instantaneous fields of view of a few arc minutes, sufficient to observe water in starless cores and protostars. This led the HERO team to design the first heterodyne focal plane array (FPA) receiver for a space mission. Table 1 summarizes HERO's capabilities. Figure 1 shows a block diagram of HERO. HERO has two sets of four frequency bands, one set for each linear polarization. Each frequency band has a small 3 × 3 pixel FPA.
Section 2 describes the instruments and its components in some detail. Section 3 puts HERO in the context of the Origins mission: it describes the resources HERO requires, the risks it carries, and shows the performance the receiver achieves, including comparison with prior missions. Section 4 discusses the technology roadmap. Section 5 briefly mentions an extension of HERO to be used as an element of the Event Horizon Telescope. The paper concludes with a summary and an outlook of the future.

Instrument Description
HERO uses the standard principle of a heterodyne receiver, in which the signal from the sky is mixed with a much stronger internally produced monochromatic reference signal called the local oscillator (LO). The beat signal is at the difference frequency, called the intermediate frequency (IF), which is amplified and spectrally analyzed. This beating of the two signals downconverts the sky signal to a lower frequency for amplification and makes very high spectral resolution measurements straightforward (spectral resolution up to 10 11 have been demonstrated 11 ). Extremely low noise mixers and first stage IF amplifiers are required for high sensitivity of such receivers and the tuning range of the mixers along with that of the LO determines the bandwidth of the receiver.
The design builds, in particular, on the experience from the receivers of the Atacama Large Millimeter Array (ALMA), 12 the German REceiver at THz (GREAT)/ 4GREAT/ up-GREAT 13,14 on the Stratospheric Observatory For Infrared Astronomy (SOFIA), and the Heterodyne Table 1 Summary of the performance of HERO. HERO can carry out dual polarization and dualfrequency observations over a large frequency range, i.e., effectively using 2 × 2 × 9 pixels at the same time. The receiver is ideally suited for very high spectral resolution observations of many molecular and atomic lines that are critical for answering the key science questions discussed above. Instrument for the Far-Infrared (HIFI) 15 on the Herschel satellite. The HERO design also takes the experience from the first ground and airplane-based FPAs 14,16-20 into account. HERO goes well beyond these earlier instruments by combining the largest frequency coverage of any heterodyne receiver and incorporates the first FPA design for space. The difficulty of cooling in space and the limited electrical power required us to look for innovative solutions for most of the receiver components. An earlier paper 21 describes the available technologies and justifies the choices we made for a preliminary version of HERO, here we describe the design submitted to the Decadal Review and the remaining challenges that need to be addressed with a technology roadmap.   has a 5.9-m on-axis primary mirror. A secondary mirror (not visible) reflects the light, to a tertiary mirror, and finally to a small field steering mirror. Around the focal plane, each of the four instruments picks up the radiation from their corresponding field-of-view. The HERO pickoff mirror is slightly in front of the Origins field of view. It directs the light into an Offner relay (1 in Figs. 3 and 4), which has a movable mirror that can send the light to one of the four frequency bands. The cold receiver optics (2a) divides the sky signal into its two linear polarizations and reimages each onto a 3 × 3 focal plane mixer array. There are eight LO chains 2 for each of the four frequency bands. To minimize coupling of the thermal radiation from the warm spacecraft bus to the cold Cryo-payload module (CPM), the eight LOs are combined into two beams in the warm LO optics (3): dichroics superimpose bands 1 on band 3 and band 2 on band 4, and polarization grids superimpose the vertical polarization of band 1 plus 3 with the horizontal polarization of band 1 plus 3 and the same for bands 2 and 4. (A design with four dichroics would be unable to provide continuous frequency coverage or have unacceptable losses.) The combined beams transfer the LO signals to the CPM, where the cold LO optics (2b) separates the LO beams into the four frequency bands and the two polarizations. A wire grid superimposes each LO signal onto a sky signal.

Instrument Overview
The LO and sky signals have orthogonal polarizations with respect to each other. Two mirrors and a lenslet array reimage the sky and the LO signals onto the focal plane mixer array. A set of 3 × 3 mixer feed horns captures the superimposed sky and the LO signal. A waveguide orthomode-transducer (OMT) at 45 degrees to both input signals divides the signals in sky plus LO and sky minus LO. These two signals are mixed and their outputs recombined. This method of superposition requires much less LO power than beam splitters, has no limitation on the IF bandwidth unlike Martin-Puplett interferometers, but there are some losses in the OMT. 22 The HERO mixers consist of superconducting tunnel junction [superconductor-isolator-superconductor (SIS)] mixers or hot-electron bolometer (HEB) mixing junctions, depending on the target frequency band. A paper giving details of the HERO optics by Jellema is in preparation. Fig. 2 HERO is the first heterodyne array receiver designed for space. The schematic diagram shows a relatively simple design that provides the widest frequency coverage and the highest sensitivity of any heterodyne receiver. HERO uses sensitive state-of-the-art components with low power consumption and weight.

Optical losses
HERO optics has a throughput of 95%, due to not perfect reflectivity and losses in the polarization grid, see Table 2. This calculation does not include the matching of the beams and truncation. We calculated the throughput of the HERO optics at its mid frequency of 1.6 tetrahertz (THz), see Table 2. The HERO mirrors are gold plated and at 4.5 K. Using the surface impedance Z at 80 K for gold, the emissivity is 0.146% at 1.6 THz (epsilon ¼ 4Z∕Z 0 , with Z 0 ¼ 376.7 Ω). To account for slight surface roughness and misalignments, we doubled that number and used a mirror throughput of 99.7%. For a wire grid using 10 μm wire at 35-μm spacing, Defrance 23 calculates at 1.4 THz a transmission of 99.8% at 0°and 99.7% at 45°, for the reflections the numbers are slightly lower at 99.1% and 98.3%, respectively. Jacob et al. 24 obtained a good agreement between their calculations with measurements at 640 GHz, where transmission and reflections were above 99% for nearly all grids tested. For the HERO calculations, we used the most pessimistic number of 98.3%. The lenslet array is an array of individual mirrors to refocus each beam onto the mixer array. We assumed a reflectivity of 99%, worse than that of a single mirror. The Offner relay with the pick-off mirror consists of five mirrors (see Fig. 4, insert 1) and has thus a throughput of 98.5%. The mixer optics (Fig 4, insert 2a) contains one wire grid, three mirrors, and the lenslet array and hence has a throughput of 96.4%. Together, the HERO optics has 95% throughput.
The radiation is collected by corrugated feed horns. Munoz 25 estimates the loss of the feed horn to be −0.22 dB mainly due to losses in the walls of the waveguide, i.e., 95% throughput. Behind the feed horn is an OMT, which is expected to have about −0.3 dB loss, i.e., 6.7%. Altogether, the optical losses and the losses in the feed horn and the OMT are expected to be 16%.
Estimating the final aperture efficiency requires a diffraction calculation to determine the coupling efficiency of the corrugated feed horn to the primary mirror, via the whole optics system.

Challenges
The greatest challenges of optics design are the wide bandwidth coverage of nearly three octaves, which requires a nearly frequency-independent design. A lenslet array in the cold receiver optics is used to map the sky, and the LO to fit the spacing of the mixers in the FPA. This component needs careful alignment, and optimization in the design phase.
Another challenge is the LO transmission from the warm spacecraft bus to the mixers, that are over 2 m away. As mechanical stability between the warm spacecraft bus and the cold instrument cannot be guaranteed, the warm LO optics has a tip-tilt mirror (see Fig. 4, insert 3) that can be adjusted to maximize the LO power on the mixer.
In addition, the IR heat transfer needs to be kept minimal. In addition to superimposing the eight LOs onto two combined LO beams, these combined LO beams are kept as narrow as possible and the beam passes through 5 thermal blocking filters. These bandpass filters need to provide nearly lossless transmission of the LO signal, but block any out of band IR radiation efficiently. A single heat sunk filter is able to block out of band radiation to −23 dB and thus the 5 filters reduce the heat load to 2.4 mW. An alternative solution would be low loss fiber-like waveguides for 500 to 2700 GHz, with a loss of <2 dB per m. Currently, these fibers do not exist yet; for an overview see, e.g., Humbert. 26

Local Oscillators
Eight LO chains consisting of a combination of amplifier and multiplier stages multiply the frequency of a low noise W-band synthesizer signal to the required target LO frequency. 27,28 The LO signal is divided by waveguide power divider into nine outputs to match the mixer arrays, similar to the 16 beam example shown in Fig. 5.  The HERO LOs have bandwidth of nearly 50% each to reduce the number of LOs and hence weight and volume in the spacecraft. Broadband amplifiers and multipliers are difficult to design, but multipliers for submillimeter LOs with 40% bandwidth have been already demonstrated, e.g., Virginia Diodes (VDI). 29 HERO requires extremely efficient multipliers as we have only allocated 2 W of DC power to each amplifier chain. The design of amplifier multiplier chains up to 2.7 THz with sufficient power to pump an array of mixers is difficult, but the prototype chains already exist for single-pixel applications, e.g., Mehdi et al. 27 for an overview, VDI 28 with 5 μW at 2.7 THz. An alternative approach for the LO source is the Quantum Cascade Laser (QCL). [30][31][32][33] These components have sufficient output power, but require cooling and generally have very limited tunability. Hübers 34 , and Hübers et al. 35 have packaged a QCL with 0.6 mW output power in a compact cooler requiring 85 W input power. For continuous tuning, a multimode QCL with a frequency selecting Fabry-Pérot interferometer has been developed. 36

Mixer
HERO uses the most sensitive mixers available. SIS mixers [37][38][39] are utilized in the FPAs at the two lower frequency bands 486 to 756 GHz, and 756 to 1188 GHz. HERO's SIS mixers will have noise temperatures of 2 hf∕k (h ¼ Planck constant, f ¼ frequency, and k = Boltzmann constant), which has been already achieved for HIFI/Herschel 15 for band 1. HEB 40-43 mixers make up the FPAs for the bands 1188 to 1782 GHz and 1782 to 2700 GHz. The HEBs will have noise temperatures below 4 hf∕k. Currently, the upGREAT 14 Instrument on SOFIA demonstrated the best noise temperatures of 700K double sideband (DSB) at 1.9 THz, and HEBs are continuously improving, with a 780 K receiver noise temperature (all components included) measured at 2.7 THz. 44 An overview of mixer noise temperatures is given in Fig. 6.
HERO has eight FPA (four frequency bands with two polarizations each). Each FPA consists of 3 × 3 mixers in a square configuration. The SIS mixers that require a magnetic field have a footprint of 10 × 10 mm 2 , and the HEBs 5 × 5 mm 2 . Projected onto the sky, the arrays have 2 × full width half maximum beam spacing, which is the closest spacing that can be achieved without considerable cross-talk between pixels and low antenna efficiency. A similar FPA with 4 × 4 pixels is shown in Fig. 7.
Eight mixers of each array are DSB balanced mixers. The balanced mixer design is adopted to reduce the LO power required and to enhance the stability by suppressing LO amplitude modulation noise. 42 One mixer in each FPA is a sideband separating (2SB) mixer to allow sideband calibration of the DSB mixers. Each junction can be individually DC biased for optimal operation.
The baseline IF range is 0.5 to 6.5 GHz (goal of 0.5 to 8.5 GHz) so that the IF covers at least 670 km∕s at the highest frequency, sufficient for observations of the inner Milky Way and of nearby galaxies. This IF bandwidth is typically not a problem for SIS mixers but is challenging for HEBs. A priori, NbN HEBs will be used, because NbTiN HEB mixers have narrower IF bandwidth (3 GHz), NbN HEBs need more LO power but have wider IF bandwidth. Recently, NbN HEB mixers on a GaN buffer-layer operating at ∼1.3 THz with 7.5 GHz IF bandwidth have been demonstrated, 46 and efforts to increase the IF bandwidth are ongoing. An alternative are MgB 2 mixers 47 that have IF bandwidth of more than 8 GHz, however, they do not have the same sensitivity, yet (1400 K at 1600 THz), but progress is being made. The common Origins closed-cycle cryocoolers cool the mixers to about 4.5 K.

Intermediate Frequency and Backend
The IF signal at the output of the mixer is immediately amplified by ∼20 dB by an LNA at 4.5 K physical temperature, then again by ∼20 dB at a physical temperature of 35 K, and finally in the warm spacecraft bus. The first stage low-noise (<6 K), low-power SiGe cryogenic amplifiers [48][49][50] [ Fig. 8(a)] are directly behind the mixer arrays and connected to the mixers by a short transmission line on a flexible Mylar bus. 52,53 Low power dissipation of the LNAs is essential to prevent the mixers from heating up and to minimize the cooling requirements. Bardin et al. 48 have recently demonstrated a state of the art SiGe amplifier integrated with a SIS mixer having an instantaneous bandwidth of 4 GHz and with power dissipation as low as 0.3 mW. An alternative are the mature InP High-Electron-Mobility Transistor (HEMT) amplifiers, 54 which have already been used on Herschel. Normally, InP amplifiers would require around 5-mW power per pixel, but tests operating at only 0.5 mW have shown only slightly higher noise temperatures. 55 If InP amplifiers operating with less than the nominal power are stable, they could serve as a backup alternative to SiGe.
Passive power combiners multiplex the IFs of different bands, such that the 2 × 9 pixels of any two frequency bands can be operated simultaneously. The multiplexing minimizes IF cables and back-ends, thus economizing on weight and thermal conduction. Coaxial cables carry 40 IF channels to the 35 K LNAs for further amplification, and then to the spacecraft bus. Miniaturized  THz mixer array with smooth-walled spline feed horns in a square grid configuration. The block, including circular-to-rectangular waveguide transitions and pockets to hold the hot-electron-bolometer mixers, was directly machined from copper. 45 integrated IF circuits 51 [Figs. 8(b) and 8(c)] further amplify, filter, and condition the signal for the backends.
HERO requires 40 low-power (∼1 W) spectrometers, each covering a bandwidth of 6 to 8 GHz. Complementary Metal Oxide Semiconductor (CMOS) Application-Specific Integrated Circuit (ASIC) backends as developed by Tang et al. 56,57 and others have very low power consumption, thousands of spectral channels and are rapidly advancing in bandwidth. 58,59 The spectrometers contain analog-to-digital converters, FFT processors, and data accumulators. They digitize to 3-or 4-bit resolution, sufficient for our application. Alternatively, HERO could use backends based on field-programmable gate array (FPGA) technology. 60,61 They show excellent performance, but their power consumption has to be lowered from currently around 70 W to about 2 W per spectrometer to be usable in HERO.
To increase the frequency resolution and minimize spectral leakage, HERO will use window functions in the spectrometer such as Blackman-Harris windows that reduce the side lobes to below −61 dB. The HERO spectrometer will take full advantage of prior experiments and knowledge, described e.g., by Klein et al. 62

Control Electronics
There are three different control units, one for each subsystem to make testing before integration easier. The LO control unit monitors and controls the synthesizer and the LO Unit with the amplifier multiplier chains. The focal plane control unit monitors and controls all the components mounted at 4.5 K, in particular, the mixer bias, the amplifier bias, and the Offner relay mirror that is used to select between the different frequency bands. It also monitors and regulates the temperature of the calibration loads. The instrument control unit (ICU) monitors and controls the IF and the spectrometers. This unit also reads the data from the spectrometers and compresses it before passing it to the Origins bus computer for downlinking. The ICU is the interface between the Origins bus computer and the other control units of the HERO instrument.
The control units use state of the art onboard processors such as the GR740 rad-hard quadcore LEON4FT. They use spacewire as an interface between the units and MIL1553STDB as an interface to the spacecraft. All control units are mounted on the spacecraft bus in pairs, one operating unit and one redundant spare that is normally not operating.

Thermal Design
The thermal design contains on one hand requirements for the instrument to operate correctly and on the other hand the requirement that the instrument's heat load on the satellite should be minimized. The HERO front-end is located in the CPM at 4.5 K, the LOs, back-end, and the control units in the warm (∼300 K) spacecraft bus. The CPM is cooled by four cryocoolers, which can each lift 50 mW of heat, i.e., there is only 200 mW cooling power available in the CPM for all instruments, mechanical structures, and thermal radiation.
The only components that require cryogenic temperatures for operation are the superconducting mixers. They only work below ∼6 K and have slightly better performance at lower temperatures. The mixer mounts have brass straps to two of the cold fingers of the ISM and hence are cooled to ∼4.5 K. A Mylar cable bus separates the mixers from the LNAs, which have significant heat dissipation of around 0.5 mW∕amplifier. The second components that need to be cooled are the LNAs, which add less noise to the signal when cold. There is no strict temperature limit, so we have compromised between instrument performance and cooling demands on the satellite. The outcome is that the first stage LNAs with 20 dB gain are on the CPM at slightly above 4.5 K, the second stage with another 20 dB gain at 35 K, and the third stage with 60 dB gain in the warm spacecraft bus. The emissivity of reflective optics elements at submillimeter wavelengths is small (5%, see Table 2), and since the noise temperatures are 50 to 300 K, the temperature of the optics has only a small impact on the system noise temperatures.
The second design criterion is to keep the cooling demands of the satellite low. On the 4.5 K stage, the LNAs are the most important heat source. We have selected SiGe amplifiers, 48 which can operate at around 0.3 mW for 20 dB gain, as opposed to InP LNAs 53 as used in HIFI/ Herschel, which had 5 mW power dissipation. All amplifiers not in use are turned off.
The passive heat load on the 4.5 K stage comes from three major sources: the warm calibration load, the two holes through which the LO beams pass, and the cable harness. We carefully designed a thermally suspended calibration load with a shutter that can be closed when HERO is not in operation. We minimized the IR radiation through the LO beam by adding five heat sunk bandpass filters that block all IR radiation except that in the observing range of HERO. The filters reduce the heat load to 2.4 mW per beam, i.e., 4.8 mW in total. In addition, the LO beams have a shutter that is closed when HERO is not in operation to minimize stray light to the other instruments. In the cable harness, the coaxial IF cables conduct the most heat. We have designed a passive multiplexing system on the 4.5 K stage so that only 40 IF signals for the 2 × 2 × ð9 þ 1Þ beams are carried back to the spacecraft bus (þ1 because one mixer in each FPA is sideband separating and has a second IF signal). The cable harness is heat sunk on the 35 K stage.
The thermal design on the spacecraft bus is less critical, because the operating temperature range is larger between ∼200 K and 320 K, and heat can be removed by radiators. Nevertheless, we strive to use components with low power consumption, in particular spectrometers requiring only 2 W of power each, and efficient LO frequency multiplier chains. Alltogether HERO requires 205 W of power.

Mechanisms
HERO has only four mechanisms ( Table 3). The Offner Relay 63-66 directs the sky beam to the respective frequency array with the help of a band selection mirror (see Fig. 4, insert 1) that has two axis of rotation. The tip-tilt mirror (Fig. 4, insert 3) on the warm side of the LO relay compensates for any mechanical displacement between the warm spacecraft bus and the focal plane. Finally, the calibration load shutter and the LO door [location of LO door see Fig. 3, for the design see light tight shutter design for SPace Infrared telescope for Cosmology and Astrophysics (SPICA) 67 ] can be closed when HERO is not in use to minimize heat transfer and to ensure a dark environment for the other instruments. All mechanisms are based on prior designs and are hence relatively low-risk.

HERO Tests, Integration, Alignment, and Calibration
The Herschel/HIFI testing, integration, alignment, and calibration procedures (see PhD thesis by Jellema 68 ) will be used for HERO. The HERO instrument team tests, calibrates, and  internally aligns the instrument in existing cryostats prior to delivery. Origins is designed such that each instrument can be removed separately for maintenance and upgrades. Consequently, the National Aeronautics and Space Administration (NASA) team can integrate the instruments in any order. The usual procedures are used to calibrate HERO in flight: HERO has two internal calibration loads to calibrate the receiver response, one of the mixers in each array is sideband separating to allow a calibration of the sideband ratio of the DSB mixers, and calibration on astronomical continuum sources (such as planets) are used to obtain a bandpass and allow the determination of the optical coupling/ aperture efficiency. Observations of point sources are used to determine the main beam efficiency.
HERO has two calibration loads, a hot load at a few tens of K and a cold load at 4.5 K. The loads are made of stycast or a similar material that looks like a black body at THz frequencies.
The loads are cone shaped to have many internal reflections and to avoid any backscattering into the receiver. To minimize any heating of the CPM, the loads have infrared filters. The filters are mounted at an angle to avoid reflection and standing waves. The loads also have a shutter that will be closed when HERO is not in operation.

Performance
The HERO has near quantum-limited performance and has unrivaled sensitivity over a very large frequency range (see Fig. 11). It consists of small mixer arrays, observes in both polarizations simultaneously and allows dual-frequency observations. It uses very little power per pixel and is the first heterodyne array receiver designed for space. The size of the HERO FPAs are well matched to prestellar cores thus requiring generally only a single pointing. For point source observations, HERO is 5 to 20 times more sensitive than Herschel/HIFI, i.e., can observe a factor of 25 to 400 times faster. This is partly due to the large collecting area of the Origins 5.9-m primary mirror (versus 3.5 m for Herschel) and partly due to improved sensitivity especially of the HEB mixers (e.g., at 1.9 THz HERO has 300 K receiver temperature versus 1100 K for HIFI). We used an aperture efficiency of 0.8 for Origins, versus 0.65 for Herschel. When mapping extended sources convolved to the same beam size, HERO is 10 to 30 times more sensitive, i.e., 100 to 900 times faster than HIFI/Herschel. HERO can map so much faster because it has 9-pixel arrays, can do dual-frequency observations and has more sensitive mixers, especially at the high frequencies. Being on a satellite, the full FIR spectrum will be available to HERO, whereas ground based telescopes and even SOFIA operate behind a more or less opaque curtain produced by atmospheric absorption.

Technology Roadmap
HERO builds on the current state-of-the-art receivers but surpasses them in different aspects of performance. For the current design, the latest and most promising Research and Development (R&D) results have been taken into account. Table 5 gives an overview of prior missions, the HERO design and the technical readiness level (TRL) of the HERO components. Concerning the LO, multiplier amplifier chains have been used for HIFI/Herschel, however, the HERO LOs need to have wider relative bandwidth, higher efficiency to pump nine mixers and need to reach 2.7 THz versus 1.9 THz for HIFI; we hence consider that this proven technology has TRL 5. SIS, and HEB mixers have flown on HIFI/ Herschel. HERO will have arrays of mixers, however these are essentially closely packed individual mixers, so no difficulty is expected. HERO will have more sensitive mixers, in particular more sensitive HEB mixers than HIFI/Herschel. More sensitive mixers are already used in up-GREAT on SOFIA, though some improvement is still needed for HERO. Up-GREAT has already demonstrated mixers in waveguide up to 4.7 THz, so this technology can be used for HERO up to 2.7 THz. Herschel has used InP LNAs, but Stratospheric TeraHertz Observatory (STO) 69 and GREAT have used SiGe; both worked well. For HERO we need the SiGe LNAs but with lower power consumption and space qualified (TRL 4). All missions have so far flown digital FFTs. CMOS-based system on chip (SOC) have been characterized and tested in the laboratory and have the required bandwidth and power, but they still need to be space qualified (TRL 5). Table 6 lists the risks and mitigations that could be taken if these goals are not all met. In summary, HERO is a very low risk mission. In the design, we pushed current technologies further and used come innovative components. However, if any of these components should not be ready, there is always the HIFI/Herschel technology to fall back on. This would allow us to either built HERO with less pixels, or to build the HERO described here but with higher cooling power, electrical power, and mass budgets.
It is essential to carry out R&D work on components to ensure the HERO mission, we envisage. Figure 12 summarizes the technology roadmap for HERO.

EHT Extension
The concept of using the Origins Space Telescope (Origins) to study black hole physics on event horizon scales has recently been raised as a potential extension of the HERO science case. 70 Fig. 11 HERO is more sensitive by an order of magnitude than any prior heterodyne instrument and has the widest frequency coverage of any instrument. It is worth noting that the sensitivity calculations for SOFIA assume a perfectly transparent atmosphere, which is not true close to the water lines.   Preliminary science concept exploration shows much promise and merits further study. The unprecedented angular resolution resulting from the combination of Origins with existing ground-based submillimeter/millimeter telescope arrays would increase the number of spatially resolvable black holes by a factor of 10 6 , permit the study of these black holes across all cosmic history, and enable new tests of general relativity by unveiling the photon ring substructure in the nearest black holes. Expanding the HERO instrument to be an interferometric station will require adding broadband single pixel receivers at 86, 230, 345, and 690 GHz, a very accurate time reference, fast on-board data processing, and a fast downlink. A top-level examination of these requirements has not revealed any insurmountable obstacles to achieving this goal.

Conclusion
The HERO is the first heterodyne FPA designed for space. It builds on the solid foundations of prior heterodyne receivers and in particular on the successful Herschel/HIFI instrument. HERO incorporates the latest mixer developments to reach unprecedented sensitivity. Using innovative components and a compact design, HERO demands only relatively modest resources of the satellite. HERO also surpasses any prior instrument in terms of wavelength coverage and observing efficiency; it allows dual polarization and simultaneous dual-frequency observations with four 9-pixel arrays. HERO will help to map the trail of water from the ISM to habitable worlds. With its unique capacities HERO opens a new window onto the physics and chemistry of the ISM and invites observers to explore the still largely unknown FIR Universe.
have taken place if NASA had not invited international members to participate in the Origins Science and Technology Definitions Team (STDT). The help and support of the NASA engineering team was very much appreciated. We would like to thank everybody on the Origins and HERO team for a fantastic study, interesting science, great engineering ideas, hard work, and the friendly and cooperative atmosphere. in mechanical engineering from the University of Maryland, a master of engineering in mechanical engineering, and a master of science in technical management from Johns Hopkins University.
Victor Belitsky (M'95-SM'07) received his MSc degree in electrical engineering from Moscow Telecommunication Institute, 1977, and a PhD in experimental physics from the Institute of Radio Engineering and Electronics, U.S.S.R. Academy of Sciences, Moscow, in 1990. He is a professor and leads the Group for Advanced Receiver Development, Department of Space, Earth, and Environment, Chalmers University, Gothenburg, Sweden. His research interests include terahertz and superconducting electronics and components, instrumentation for radio astronomy, and environmental science. Brian Ellison leads the Millimetre-wave Technology and Chilbolton Observatory Group at the Rutherford Appleton Laboratory Space Department, UK. He oversaw the UK construction contribution to the Atacama Large Millimeter/submillimeter Array and at the California Institute of Technology, USA, developed receivers for its Sub-millimeter-wave Observatory. He has led millimetre-wave instrument development programs for astronomy and Earth observation and is visiting professor at The Open University, UK, and a Fellow of the Royal Academy of Engineering.
Martin Eggens received his BSc degrees in mechanical and electrical engineering from Hogeschool Drenthe, Emmen, The Netherlands, in 1995 and 1997. Since then, he has been senior mechanical design engineer at SRON, Groningen, The Netherlands. He is mainly involved in engineering of optomechanical designs, analysis (vibration, thermal), and mechanisms of cryogenic space instrumentation like ESA/Herschel/HIFI and ground equipment. As mechanical lead engineer he was involved in the development of the SAFARI instrument for the ESA/ JAXA/SPICA mission.
Juan-Daniel Gallego is the head of the Instrumentation and Technological Development Department of the Yebes Observatory (Spain). He received his doctoral degree in physics from the Universidad Complutense (Madrid). Among other projects, he has been in charge of the development and construction of the cryogenic amplifiers of the of Herschel ESA mission and of the amplifiers for the European contribution to the ALMA. He is member of IEEE, URSI, and IAU. Paul F. Goldsmith is a senior research scientist at the Jet Propulsion Laboratory. He was a professor of astronomy at the University of Massachusetts, Amherst, and Cornell University. He has developed millimeter and submillimeter radiometric systems. He has carried out millimeter and submillimeter observations of molecular, atomic, and ionic fine structure lines, particularly CO, [ Fabrice Herpin is an astronomer at the Laboratoire d'Astrophysique de Bordeaux, France. He was instrument co-I of HIFI on-board Herschel Space Telescope, active member of the HIFI-Instrument Calibration Center, and Co-I of the WISH HSO Large Program. He has carried out millimeter, submillimeter, and far-IR observations of molecular lines, particularly water to understand star formation, but also various other molecules, specially SiO maser emission with the goal to constrain the chemical content and evolution of circumstellar envelopes of evolved stars.
Richard E. Hills, FRS, is emeritus professor of radio astronomy at the University of Cambridge, UK, having gained a BA degree there in 1967 and a PhD from UC, Berkeley, in 1973, and then held a post-doctoral position at the MPIfR, Bonn, before returning to Cambridge. He works on observational and instrumental aspects of astronomy, focusing on sub-millimetre wavelengths. He served as project scientist on JCMT and ALMA and as telescope scientist for the Planck satellite.
Michiel R. Hogerheijde is an associate professor at Leiden University and adjunct professor at the University of Amsterdam. He received his PhD from Leiden Univeristy, and held a Miller Research Fellowship at the University of California (UC), Berkeley, and a Bart J. Bok Fellowship at the University of Arizona. He has made extensive use of the Herschel Space Observatory and leads the Netherlands ALMA Regional Center node. His research focuses on the formation of stars and planets.
Leslie K. Hunt is a research director at the INAF-Arcetri Astrophysical Observatory in Firenze. After her undergraduate degree from the UC, Berkeley (USA), a master's at the University of Firenze (Italy), and her PhD from the University of Groningen (NL), she has worked at Arcetri, with a brief period at the European Space Agency for the Hubble Space Telescope in Germany. She studies mainly star-formation processes and the ISM in galaxies, and in particular the evolution of galaxies in extreme conditions. André Laurens is a space systems expert in CNES Early Mission Studies team, and referent for Space Sciences. He received his master degree in computer science from Toulouse Science University, France. He joined CNES in the early 1980s and contributed to many space missions, working in satellite image processing, spacecraft control and command, and balloon-borne science (astrophysics and atmospheric chemistry) missions. He is conducting future mission studies in collaboration with French and international science communities.
David T. Leisawitz is a NASA study scientist for the Origins Space Telescope and is most interested scientifically in the development of habitable conditions during planet formation. He was PI on the Space Infrared Interferometric Telescope mission concept study and served as NASA Goddard study lead for the Submillimeter Probe of the Evolution of Cosmic Structure. He was a mission scientist for the Wide-field Infrared Survey Explorer and deputy project scientist for the Cosmic Background Explorer.
Dariusz C. Lis received his PhD from the University of Massachusetts, Amherst in 1989. He was a senior research associate at the California Institute of Technology and deputy director of the Caltech Submillimeter Observatory, professor and director of Laboratoire d'Etudes du Rayonnement et de la Matière en Astrophysique et Atmosphères at the Sorbonne University/ Paris Observatory, and is currently scientist at the Jet Propulsion Laboratory, California Institute of Technology. He is an expert in high-resolution molecular spectroscopy of the ISM and solar system small bodies.

Margaret
Meixner is a director of the SOFIA Science Mission Operations center. She received her bachelor's degree in electrical engineering and math (1987) from the University of Maryland, College Park, and PhD in astronomy (1993), UC, Berkeley. Her research includes infrared instrumentation and dust evolution in galaxies. She is a member of the JWST/MIRI Science Team. She is the community co-chair of NASA's STDT for the Origins Space Telescope mission concept. Formerly, she is a JWST project scientist and distinguished astronomer at STScI. She is a fellow of the AAAS.
Gary Melnick is a senior astronomer at the Harvard-Smithsonian Center for Astrophysics specializing in infrared/submillimeter spectroscopy and astrochemistry. For more than 45 years, he has conducted airborne, balloon-borne, and space-based observations focused on the main coolants of interstellar clouds and understanding the formation, abundance, and distribution of interstellar water. He was PI of the Submillimeter Wave Astronomy Satellite (SWAS) mission, Co-I on the Herschel mission, and is the science lead for the Ices Investigation on the upcoming SPHEREx mission.
David A. Neufeld is a professor at the Johns Hopkins University in Baltimore, Maryland. He received his undergraduate degree at Clare College, Cambridge, and his PhD from Harvard University, and was a postdoctoral fellow at the UC Berkeley. His scientific interests include molecular astrophysics, astrochemistry, astrophysical masers, dark matter, and submillimeter and infrared astronomy. Serena Viti is a professor of astrophysics at Leiden Observatory, in the Netherlands. She received her undergraduate degree from Queen Mary University of London and her PhD from the University College London (UCL). She was a professor at UCL (where she is still affiliated to) for many years and head of the Astronomy Group until June 2020. Her research interests span a wide range of topics but are all centred around the role of molecules in space, especially in the dense gas of the ISM and star forming regions. Her recent work concentrates on the interpretation of molecular observations in extragalactic regions, and on devising novel techniques for astrochemistry involving machine learning.
Biographies of the other authors are not available.