1 The Submillimeter Range for Planetary Science
The submillimeter-wavelength range is the so-called Terahertz (THz) frequency range and it is usually associated with the electromagnetic spectrum of radiations from 1 mm of wavelength (300 GHz of frequency) to 0.03 mm of wavelength (10 THz of frequency). This frequency range has given rise to an intense scientific interest since at least the 1920s [Nich25], but the historical study of the millimeter and the submillimeter range starts in 1890s [Wilt84]. The research on THz science is far from being completed but a detailed summary of the development of this science can be found in [Sieg02]. The THz range remains an important technological challenge today. It has two important facets since the THz frequency range is difficult to reach by both electronic-based sources, when moving up from lower frequencies, and optics based sources, when moving down from upper frequencies. The present work is focused on the electronic-based devices for submillimeter applications. Very important advancements have been carried out in electronic-based devices and the THz range has spread out through diverse technological applications in which the absence of a compact high-efficiency and high-power THz sources is the common ground. The THz range is currently applied in the medical field [Tada04], [Pick06], in spectroscopy [Gopa98], [Encr04], [Shi04], [Hans07], [Luci10] and in THz imaging for multiple applications [Chan07], [Coop08], [Bryl09], [Coop14]. However, it is in atmospheric, planetary and astrophysics sciences as well where the development of THz technologies has been especially enhanced. The interest comes from the large amount of electromagnetic radiation in the THz and far Infra-Red (IR) ranges associated with observable universe that has attracted the attention of astronomers since the early stages of this science [Sieg02]. For example, the kinetic analysis of galaxy dust using the detection of high rotational transitions of the CO molecule lines which are redshifted to millimeter wavelengths [Cox02].
The spectroscopy has been especially important in the development of THz science which has also required the development of THz receivers. The detection of a frequency signal in electronic-based circuits usually consists of a preliminary stage conformed by a Radio Frequency (RF) antenna, which is able to capture a certain frequency range, an electronic device sensitive to those frequency signals and a Low Noise Amplifier (LNA). The first challenge found in THz science to develop THz receivers is the LNA since all solid-state transistors have a cutoff frequency from which the amplification capabilities of the device are dramatically degraded. The commercial THz amplification solid-state technology that has spread out during the last decades is able to rise to W-band (75 – 110 GHz) [IEEE std. 521-2002] using High Electron Mobility Transistors (HEMTS), metamorphic HEMTs (mHEMTs) and Heterojunction Bipolar Transistors (HBTs) mostly based on InP and GaAs semiconductor (SC) structures [Samo11]. Additionally, these devices allow THz amplifiers for broadband applications (bandwidth ~20 % of the center frequency) that can operate at room temperature. Commercial Monolithic Microwave Integrated Circuits (MMIC) amplifiers up to 100 GHz are commonly found up to W-band [Wang01], but higher frequency broadband amplifiers around 200 GHz [Wein99], [Deal07a], [Chio16], around 300 GHz [Deal07b], [Tess08] and even up to 600 GHz [Seo13], [Tess14], [Deal16] have already been demonstrated. A lot of narrowband amplifiers can be found even at higher frequencies using diverse technologies [Truce14]. However, most of these amplifiers are not commercialized and the technology is carefully controlled by the manufacturers or even restricted for military applications. This leads to lower frequency amplifiers up to W-band (75 – 100 GHz) that have been commonly commercialized. The application of this technique based on GaN amplifiers is providing higher output power at W-band during the last years [Mish08], [Sile08b], [Sche16], [Marti16] and it could eventually replace the actual GaAs and InP-based W-band amplifiers in a near future. The availability of a commercial W-band LNA has also been a prior objective that has encouraged significant research [Mei08], [Brye09], [Yang13], [Pepe15], [Zhan16]. This absence of commercial availability of higher frequency amplifiers has hampered THz detection, especially in room temperature direct detection applications at THz frequencies which is nowadays limited up to W-band [Hoef14], [Deco16].
The alternative technology that has placed itself in the THz detection field due to its sensitivity and high spectral resolution (1-100 MHz) at THz frequencies is the heterodyne detection technique. The development of a 1.2 THz heterodyne receiver is the subject of this work and the different motivations that have enhanced the accomplishment of this work are pointed out in this section.
The Heterodyne Detection
The heterodyne reception is a widespread technique in most of the THz applications developed for THz science since these kinds of receivers allow moving up on frequency beyond the 1 THz RF detection [Sieg02]. A general scheme of a heterodyne receiver for THz applications is shown in Fig. 1.1. A THz heterodyne electronic-based receiver requires a Radio Frequency (RF) mixer module sensitive to certain THz frequency ranges coming from a RF source. This mixer module needs to be pumped by a THz power source, so-called Local Oscillator (LO), able to yield a very precise frequency signal under W-band (with only some few tens of MHz wideband) that is then multiplied N times to increase the frequency up to the mixer requirement. A telescope and optical bench are usually used to increase the effective detection surface and focus the signal into the RF horn of the mixer. The RF signal at frequency fRF is mixed with the LO signal at frequency fLO by the mixing stage resulting in an Intermediate Frequency (IF) signal at frequency fIF= fRF – n•fLO, where n refers to the n-th harmonic of the LO frequency signal [Hayk08]. It is possible to differentiate between fundamental heterodyne receivers if n = 1 and sub-harmonic receivers if n ≥ 2. The heterodyne receiver developed in this work corresponds with the second kind with n = 2. These receivers are usually designed to generate an IF signal low enough to be amplified by commercial LNAs, i.e., the considered frequency of the LO signal (or its n-th harmonic) is similar to the detected RF frequency signal. Usually, no LNA is used in THz heterodyne receivers to amplify the captured RF signal by the RF antenna since there isn’t frequency LNAs able to efficiently work beyond W-band. This result in a LO input power which is usually much higher than the RF input power in these receivers. The principles of frequency mixing using non-linear devices can be found in [Man56], [Pant58], [Ebst67], [Sea69].
In conclusion, the heterodyne reception avoids the necessity of LNA beyond W-band but obtaining a THz source for the LO and an adequate device for the mixing stage are the challenges presented by these receivers.
THz Frequency Mixers
The most critical element of the heterodyne reception is the frequency mixer and especially the non-linear devise used to efficiently mix the LO and the RF signal. Several technologies have been studied in THz science for mixing applications but the most widespread devices are the Superconductor-Isolator-Superconductor (SIS), the Hot-Electron Bolometers (HEB) and the Planar Schottky Diodes (PSBDs). The SIS technology has proven to be the best device for mixing applications since frequency heterodyne receivers, with Single Side Band (SSB) noise temperature as low as 2 times the quantum noise limit (hf/k), have been demonstrated [Kerr99], [Koll02], [Chat08], [Maes10b], [Bill13]. Additionally, these mixers can perform a broadband IF band [Laur01], [Pan04] with low LO power requirements. However, the SIS mixers require cryogenic temperatures to work and they are limited up to 1.5 THz, mainly due to the disappearance of the superconductor properties [Karp07], [West13], [Zmui15]. The HEB has been placed itself as the alternative device that is able to replace the SIS beyond its frequency limit. The HEB are also required to be cooled down to 2 – 4 K but frequency receivers up to 4.7 THz have been demonstrated with DSB noise temperatures 10-20 times the quantum noise limit [Haje04], [Chat07], [Cher08], [Tret11], [Buch15]. HEB are able to provide the best performances using very low LO input power, hundreds of nW, which is possible to obtain with both electronic and optical based THz sources. The challenge of HEBs is the short IF bandwidth that it is possible to provide and the cooling system requirement. Regarding the PSBDs technology, it can provide frequency receivers with DSB noise temperature much higher than SIS and HEB, above 50 times the quantum noise limit, but it is able to work at room temperature. This results in extremely compact frequency mixers that do not require any cryogenic system [Chat07], [Maes10b]. Additionally, the performance of PSBD mixers is always improved down to 30-40 times the quantum noise limit when cooling down below 150 K [Pred84], [Schl14], [Treu16]. Very important efforts are focused now to develop frequency mixers based on Schottky technology that will overcome the SIS frequency limit in a near future [Treu16b]. The challenge of PSBD mixers is the high LO power requirements compared with HEBs and the higher noise provided as the frequency increases. This work is focused on the development of a 1.2 THz mixer based on PSBD technology. The PSBD technology has positioned itself as the most suitable option for space-borne applications that do not require the most sensitive receivers, like planetology or remote sensing of the Earth’s atmosphere, due to the compact low-weight receivers that can be achieved.
THz Local Oscillator
One of the most difficult challenges faced by the THz science community, which has not been addressed yet, is obtaining a powerful, efficient and compact THz source. The main challenge when reaching the THz range in electronic-based applications is associated to an electron transit time in the same order of the THz frequencies, hampering the proper functioning of electronic-based devices when moving up in frequency. Similarly, a THz laser based on an energy transition requires transition energy of the same order as the room temperature lattice vibrations in the material, hampering the proper functioning of optic-based devices when moving down in frequency. The development of electronic-based THz sources using solid-state oscillators has been pursued for several decades. One of the most promising THz oscillators above 100 GHz is based on InP Gunn devices, thoroughly studied by Dr. Carlstrom [Carl85] and Dr. Eisele in [Eise95], [Eise97], [Eise00], [Eise04], [Eise06], [Eise10] and by Dr. Khalid in [Khal07], [Khal13]. The Gunn diodes were widely used in the (sub)-millimeter heterodyne receivers in the past. However, Gunn diodes require phase locking, are difficult to tune and to use in broadband frequency applications. Gunn oscillators have been currently replaced by amplify-multiply chains. Interesting advancements have been carried out during the last years by Tokyo Institute of Technology in fundamental THz source based on Resonant Tunneling Diodes (RTDs) up to 1.46 THz [Feig14] and 1.92 THz [Maek16]. Regarding the optical based THz source, the Quantum Cascade Laser (QCL) technology has placed itself as the leading mid-infrared (mid-IR) source since it was proposed twenty years ago and it has shown itself to be the most suitable way to substantially reduce the frequencies even down to 1 THz [Will07]. However, the challenge faced by QCL sources is avoiding the cryogenic temperatures requirement that notably hampers this technology to spread out in the THz applications [Belk15]. Other techniques use crystals as coherent THz source [Shi02]. An intermediate technique that combines optical and electronic base devices is based on photo-mixers and it allows obtaining LO power sources suitable to pump SISs or HEBs based mixers. The photo-mixers can mix two laser beams at IR range to generate an electrical signal at required LO frequencies [Font07], but this technique usually requires cooling systems due to the low conversion efficiency of the photomixers and the high amount of power dissipated in the device. Low LO power sources can be provided by this technique.
The alternative electronic-based technology that has been able to place itself in the commercial applications at THz frequencies is based on frequency multiplication. The technique used in this work, which is able to provide the highest LO power electronic-based sources know nowadays, consists of a preliminary commercial Ultra Stable Oscillator (USO), at a single frequency of few tens or hundreds of MHz [Cand03], followed by a commercial Frequency Synthesizer [Tobi01] that is able to generate a broadband frequency window up to some tens of GHz [Jain09] using the reference given by the USO. The generated millimeter wavelength source can be amplified using available commercial amplifiers obtaining some hundreds of mW at E-band (60-90 GHz) and multiplied to move beyond 100 GHz. The multiplication chains usually consist of an assembly of frequency doublers and triplers in accordance with the desired output frequency signal and the available amplifier source [Sieg02]. It is based on the excitation, at a specific frequency f1, of a passive non-linear solid-state device whose response will contain the fundamental frequency f1 and its harmonics n f1 with n= 2, 3,…, ∞. Electronic-based frequency doublers and triplers usually consist of a waveguide system to lead the input and output frequency signals and a passive non-linear solid-state device mounted in a microelectronic chip. The microelectronic chip is designed to couple the input signal in a certain frequency and power range and it also couples the desired frequency harmonic generated by the non-linear device with the output waveguide system. The most important electronics-based device technologies used currently in the THz multiplication field are the Heterostructure Barrier Varactors (HBVs) and the PSBDs previously mentioned.
The LO source used in this work is also based on the second type of devices, i.e., the 1.2 THz receiver developed in this work is an all-solid state PSBD heterodyne receiver. The theoretical foundation of the so-called metal-semiconductor rectifiers started in 1931 when the band model of SCs was formulated [Wils31]. The fundaments of Schottky barrier diodes were proposed by W. Schottky in 1938 [Scho38] and H. A. Bethe formulated in 1942 [Beth42] the well-known thermionic emission theory that describes the electrons transport through the Schottky barriers. The HBVs were first introduced by Kollberg and Rydberg in [Koll89], [Rybd90] and they can only be used in odd harmonic multiplication due to the internal symmetry of the device. HBVs are suitable for frequency triplers [Salg03], [Vuku12] and especially useful in quintupler applications providing efficiencies higher than PSBDs [Bryl12], [Malk15]. Nevertheless, it is the Schottky technology which is widespread in millimeter and submillimeter applications, especially for ground-base and space-borne heterodyne applications since it is able to provide the highest efficiencies, output power and instantaneous bandwidth [Maes05a]. It is mainly because PSBDs can be used in different configurations to design both doublers and triplers as varactor mode, and it can also be used for mixing applications in varistor mode. It can work at room temperature, allowing PSBD based modules to be very compact and robust. They can easily pump frequency mixers based on HEB or SIS technologies. It has motivated the development of a wide range of PSBDs frequency broadband multipliers since the 1990s from 100 GHz to up to 2.7 THz [Eric93], [Schl01a], [Maes05a], [Maes06], [Maes08], [Maes10a], [Sile11a], [Maes12], [Treu14], [Sile15]. Different combinations of doublers and triplers in a multiplication chain based on PSBDs also allow the development of THz sources at different frequency ranges using the same millimeter source. A summary of the actual THz electronic and optic-based sources availability can be found in [Maes10b (Fig.1)] and [Will07], respectively.
Space missions and THz science
Some of the most important space missions where a millimeter or submillimeter instrument was proposed are summarized in this section. The common ground between these mission is the implementation of some of the different technological approaches used to define the technical specification of the SWI instrument. The development and implementation of these space missions summarized the background that has motivated the basis of JUICE-SWI project and this work.
Ground-based millimeter and Sub-Millimeter Telescope
The ALMA project [ALMA 2016] is the largest ground-based radio-telescope currently existing for millimeter and submillimeter astronomy interferometry and it is placed in the Atacama Desert, Chile. It consists of 54 antennas of 12 m diameter equipped with frequency receivers based on SIS mixers covering the 30-950 GHz [Brow04]. ALMA has operated since 2009 but it was inaugurated in 2013. ALMAs receivers are based on SIS mixers and the LO chain is mainly based on MMIC Schottky diodes multipliers [Brye05], [Morg05] that were developed by VDI. The W-band power amplifiers drivers are based on MMIC InP HEMTs technology [Samo05]. ALMA was develop to improve the performances and complement the already existing submillimeter telescopes such as the James Clerk Maxwell Telescope (JCMT) in Hawaii and interferometers as the Institut de Radio Astronomie Millimétrique (IRAM) in France. The JCMT [Murd00] is placed in Hawaii island and it is a single 15 m diameter dish dedicated to the detection of submillimeter radiation between 1.4 to 0.4 mm wavelengths. The IRAM [Maar87] was established in 1984 in Pico Veleta, Spain, and it is a 30-m millimeter radiotelescope. It had a complementary array of three 15-m antennas placed on Plateau de Bure Observatory, France, which has eventually been extended to six antennas and finally replaced by the current NOEMA project since 2014 with six additional 15-m antennas. The six array NOEMA project is expected to be completed in 2019, becoming the most powerful millimiter radiotelescope in the North Hemisphere [IRAM-2016].
Earth Observing System: AURA Satellite
The AURA satellite was launched on July 2004 as part of the NASA’s Earth Observing System (EOS) program. This science program is dedicated to monitoring the complex interactions that affect the globe. The AURA satellite is in a sun-synchronous orbit specifically dedicated to analyzing climate change by monitoring the interactions of the ozone and other chemical compounds with the radiation. AURA consists of four instruments dedicated to the analysis of the ozone and other greenhouse compounds from different aspects. The High Resolution Dynamics Limb Sounder (HRDLS) is dedicated to analyzing infrared emission, the Microwave Limb Sounder (MLS) to analyzing microwave emission, the Tropospheric Emission Spectrometer (TES) and the Ozone Monitoring Instrument (OMI).
The MLS instrument [Wate06] was one of the first important microwave space-borne instruments that enhanced the development of this technic. It is an on-board submillimeter instrument based in heterodyne reception in broad spectral regions centered at 118, 190, 240 and 640 GHz and 2.5 THz. All solid state technology was used for the local oscillators, excepting in the 2.5 THz, and MMIC amplifiers for the 118 GHz channel. All the frequency mixers were based in planar technology [Sieg93].
Microwave Instrument for Rosetta Orbiter (MIRO)
The Rosetta mission is a comet mission accepted by the ESA in November 1993 in the framework of the long-term program “Horizon 2000” [Glas07]. It consists of two mission elements, the ROSETTA orbiter and the PHILAE lander. ROSETTA was launched in March 2004 featuring the unprecedented assembly of 25 payload-experiments. The main scientific objectives of ROSETTA were the study of the origin of the solar system by studying comets. ROSETTA took contact with the comet 67P/Churyumov-Gerasimenko in summer 2014 and PHILAE lander was successfully deployed at the end of 2014. The mission was a complete success and it finished in September 2016 [ROSETTA-Mission-2016]. One of the ROSETTA instruments is the so-called Microwave Instrument for Rosetta Orbiter (MIRO). A precise description of MIRO and the molecules studied by it can be found in [Gulk07]. This is a heterodyne spectrometer working at two frequency ranges around 190 GHz and 562 GHz. The frequency mixer modules of MIRO are based on PSBD technology while the LO multiplication chain is a combination of different technologies. The LO source is based on InP Gunn devices to generate a LO signal around 95 GHz. This Gunn source is enough to directly pump the PSBD-based sub-harmonic mixer for the 190 GHz channel of MIRO, while a HBV-based frequency tripler is used to increase the frequency of the signal to pump the PSBD-based sub-harmonic mixer of the 562 GHz channel. This heterodyne receiver didn’t require any cooling system and it has successfully demonstrated the PSBD technology capabilities as mixer stage for space-borne front-end heterodyne detection.
HESHEL Space Observatory
The development of THz science has been especially encouraged by the atmospheric, planetary and astrophysics sciences which have enhanced and positioned the different available THz technologies as actual state-of-the-art. One of the first space missions that has enhanced the development of the heterodyne reception is the Far Infra-Red and Submillimeter Telescope (FIRST) mission led by the European Space Agency (ESA) in the “Horizon 2000” science plan. This telescope was expected to be put in a geostationary Earth orbit in late 2005 and featuring a payload of two instruments and at least 3 m telescope mirror [Pilb97]. The launch was delayed and the project was renamed as “HERSHEL”, in honor of Sir William Hershel who discovered the Infra-Red spectrum. The HERSHEL space observatory was finally launch in 2009 and it was the largest infra-red telescope launched until the mission ended in 2013. HERSHEL finally featured a 3.5 m telescope mirror and three instruments, the Photodectective Array Camera and Spectrometer (PACS), the Spectral and Photometric Imaging REceiver (SPIRE) and the Heterodyne Instrument of Far Infra-red (HIFI) [Pilb10]. It was the HIFI instrument which motivated an intense activity in THz heterodyne technology based on multiplication and mixing stages up to the desired frequency channels. HIFI is a set of 7 heterodyne receivers that are electronically tunable, covering a 0.48-1.25 THz range in five bands using SIS mixers and a 1.41-1.91 THz range in two bands using HEB mixers. A cryostat was included in the payload module of HERSHEL due to the cryogenic requirements of these technologies. Laboratoire d’Etudes du Rayonnement et de la Matière en Astrophysique et Atmosphères (LERMA) was involved in the development of the SIS mixer used in the first band between 480-640 GHz. The main interest of these frequency ranges was focused on water lines analysis, surveying the molecular complexity of the universe and observations of ionized carbon for redshift analysis. Each receiver had two LO multiplier chains completely based on planar Schottky diodes technology with their corresponding W-band amplifiers [Samo00]. The LO chain for each receiver development was strongly enhanced by JPL and some details can be found in [Pear00], [Pear03], [Maes06a]. The complete development of the LO chain based on Schottky diodes technology not only allowed high-power handling capability and an improved reliability and stability of the receivers, but it dramatically reduced the cryogenic system requirements of HIFI. Nevertheless, HERSHEL was operational during 3.5 years due to the limited cryogenic gas stock.
The success of HERSHEL-HIFI, together with ROSETTA-MIRO, has laid the basis of the space-borne THz science, where Schottky technology has demonstrated an important role. The progress made during the last two decades in the completion of such important scientific goals and the developed techniques have motivated a wide range of future space mission proposals. JUICE mission was elected in the framework of the ESAs “Cosmic Vision 2015-2025” program [ESA-Cosmic-Vision (2016)]. This work is focused on the development of a part of the sub-millimeter instrument of JUICE as described below.
The JUICE Project Baseline
The Jupiter Icy Moons Explorer (JUICE) is the first large class mission chosen in the framework of the Cosmic Vision 2015-2025 program of the Science and Robotic Exploration Directorate of the ESA. The mission has been chosen in May 2012 out of three possible L-class missions in the Cosmic Vision 2015-2025 program [Doug12]. A detailed description of the JUICE mission science goals and perspectives can be found in [Gass13]. The JUICE spacecraft is expected to be launched in 2022 and reach the Jovian system eight years later, where it will perform a three year minimum-tour investigating the atmosphere and magnetosphere of the giant. The JUICE mission will survey the Jovian system with a special focus on the three Galilean Moons; Europa, Ganymede and Callisto, It will be the first spacecraft ever to orbit a Moon (Ganymede) of a giant planet. The main science goal of JUICE is “the emergence of habitable worlds around gas giants” by studying the presence of necessary conditions to sustain life in current habitats of the Solar System [Doug12]. The devoted science payload of JUICE consists of 10 state-of-the-art instruments and one experiment which uses the spacecraft telecommunication system with ground-based radio telescopes. The different instruments proposed in the JUICE mission baseline are listed below and a detailed analysis of the science goals pursued by each one can be found in [Gass14] and [Plau14]:
• Gravity & Geophysics of Jupiter and Galilean Moons (3GM) to study the moon gravity fields.
• Ganymede Laser Altimeter (GALA) to study moon surface topography.
• Jovis, Amorum ac Natorum Unique Scrutator (JANUS) to study the geology and surface processes in the visible range.
• Magnetometer for JUICE (J-MAG) to study moon and Jupiter magnetic fields.
• Moons And Jupiter Imaging Spectrometer (MAJIS) to study the composition of moon surfaces and Jupiter atmosphere in the visible-IR range.
• Particle Environment Package (PEP) to study plasma particles.
• Planetary Radio Interferometer & Doppler Experiment (PRIDE) is the experiment that uses a ground-based Very-Long-Baseline Interferometry (VLBI) to provide precise determination of the moons ephemerides.
• Radar for Icy Moon Exploration (RIME) to study moon-subsurface ice shells and shallow liquid water.
• Radio & Plasma Wave Investigation (RPWI) to study the radio emission and plasma of Jupiter and moons.
• Sub-millimeter Wave Instrument (SWI) to do spectrometry of Jupiter atmosphere and the moon surfaces and exospheres.
• UV imaging Spectrograph (UVS) to study moon exospheres and Jupiter auroras.
The ensemble of these instruments comprises the science payload of the JUICE spacecraft. The instrument that has motivated the present work is the SWI instrument and it will be the most important submillimeter science space-borne instrument after HIFI.
The Submillimeter Wave Instrument (SWI)
JUICE-SWI is a submillimeter heterodyne spectrometer with a frequency resolution of ~107 covering two spectral ranges around 600 GHz and 1200 GHz [JUICE-SWI-Payload (2016)] proposed by an European international consortium headed by the Max Planck Institute für Sonnensystemforschung (MPS) of Gottingen, Germany [Hart13]. The science objectives of SWI are the investigation of the middle atmosphere of Jupiter and the atmospheres and exospheres of the Galilean satellites [Hart13]. SWI consist of three primary units: the Telescope and Receiver Unit (TRU), the Radiator Unit (RAU) and the Electronic Unit (EU).
Fig. 1.2. JUICE-SWI RTU block diagram of the different parts of the instrument, as 21th June 2016. The Receiver Unit (RU) in light blue is the only cooled part of the SWI instrument.
• The RTU consists of the Telescope Unit (TU) with a 30 cm antenna effective diameter with tracking mechanisms and a receiver optical bench, and the microelectronic modules of the sub-millimeter Receiver Unit (RU). Two independent double sideband receivers were initially proposed in the SWI baseline to obtain simultaneous observing capability for two different frequencies within the 530-625 GHz frequency range [Jaco15]. However, an alternative configuration was open in the baseline of the submillimeter instrument shown in Fig. 1.2. It consists of two independent double sideband receivers, around 600 GHz and 1200 GHz. This optional second channel was never developed based on PSBDs in Europe and the only precedent was developed by JPL in the USA [Schl14]. The 1.2 THz channel was motivated by methane transition at 1256 GHz but a wide richness of compounds can be found in this frequency range [Encr04]. However, this channel had to be demonstrated and developed by the European SWI consortium in order to be included in the fly RTU module. LERMA got involved in the development and demonstration of the 1.2 THz channel for SWI which is the motivation of this PhD. work.
• The RAU consists of a radiator used to passively cool down the Sub-millimeter detectors between 120-150 K to improve their sensitivity and their signal-to-noise ratio.
• The EU consists of several electronic devices dedicated to determine the spectral line shapes and the lines surveys. A detailed study of the SWI structural and tracking systems of the RTU is carried out in [Jaya14].
The different contributions of European institutions in the consortium were distributed as follows. LERMA represents the French contribution distributed between the RU and the EU. MPS represents the German contribution to the structure design and manufacture of the RTU and RAU as well as the project management led by Dr. P. Hartogh. Omnisys represents Sweden’s contribution divided between the RU and the EU. The Laboratorium Satelitarnych Aplikacli Ukladow FPGA (CBK) represents Polish’s contribution to the EU. RPG represents the second German’s contribution focused on the amplification stage of the RU. The National Institute of Information and Communications Technology (NICT) represents Japan’s contribution focused on the primary mirror manufacture of the TU and some components of the EU. The Institut of Applied Physics (IAP) University of Bern represents Switzerland’s contribution focused on the optical test bench design to match the signal obtained by the TU into the RU.
All-solid-state Planar Schottky Diode technology has been chosen for the microelectronic MMIC chips that conform the LO multiplication chains and the frequency mixers for SWI. A passive cooling system is proposed to control temperature conditions of the SWI instrumentation between 120-150 K. It requires a bandwidth of a 20 % around the center frequency of each receiver with 100 MHz spectral resolution. The sensibility specifications proposed less than 1500 K of DSB noise temperature at 120-150 K for the 600 GHz channel and less than 4000 K for the 1200 GHz channel. LERMA got involved in SWI project in summer 2013 and it was in charge of the industrial delivery of two USO at 100 MHz and two K-band synthesizer of the SWI-EU to generate the initial LO chain bandwidth for each receiver between 22-26 GHz. LERMA was also in charge of the development and delivery of a frequency multiplier of the SWI-RU between 270-320 GHz to complete the LO multiplication chain of each receiver. Omnisys was in charge of the frequency mixers development and delivery for the 600 GHz channels. A contract between LERMA and Omnisys was closed within the framework of the SWI project. In summer 2014, LERMA got involved in the development of a 1.2 THz frequency mixer in the framework of the European SWI consortium. LERMA of the Observatoire de Paris in close collaboration with the Laboratoire de Photonique et de Nanostructures (LPN) of the Centre National de la Recherche Scientifique (CNRS) represent the French contribution to the SWI project. LERMA was in charge of the design, optimization and test of the developed modules while LPN-CNRS was in charge of manufacturing the modules. LERMA’s contribution in the development of a 1.2 THz mixer prototype was initially supported by a contract between Centre National d’Études Spatiales CNES and LERMA-LPN, and this was complemented by this PhD work, fully supported by Labex Exploration Spatiale des Invironnements Planétaires (Labex-ESEP) and granted to this author, Diego Moro Melgar. The CNES financial support and LERMAs commitment culminated with an ESA official contract that allowed LERMA to get involved in the 1.2 THz channel for SWI. LERMAs commitment in the CNES and ESA contracts was the demonstration of a PSBD-based 1.2 THz sub-harmonic frequency mixer able to fulfill the SWI specifications (less than 4000 K DSB noise temperature at 120 K). The challenge was proposed to both LERMA and Omnisys. The demonstration of the 1.2 THz channel feasibility was successfully accomplished by both groups at the beginning of 2016. The satisfactory results obtained by LERMA in the development of a 1.2 THz receiver culminated in summer 2016 with a full financial support by CNES of LERMA’s contribution to the 1.2 THz channel delivery. However, the French contribution was not completely finished when this doctoral work was accomplish and further work on the final fly version of the 1.2 THz channel was required.
The author’s work has been focused on the development of a PSBD-based 1.2 THz mixer design, a PSBD-based 600 GHz doubler design and the improvement of the PSBDs model for harmonic balance (HB) simulators. The chapter 2 of this work is dedicated to the physical model of PSBDs which is especially important at these high frequencies. It is due to the influence of additional phenomena associated to the reduced geometry and the saturation phenomena in the semiconductors. Regarding the author’s role in LERMA’s contribution, this author has been in charge of providing regular design reports of LERMA’s progress to Omnisys in the framework of the LERMA-Omnisys contract. This author has presented LERMA’s progress on the 1.2 THz receiver in the meetings taking place in Gothenburg (Sweden) between LERMA, Omnisys and ESA members. In addition, a consortium meeting was held every six months at MPS in Gottingen (Germany) between the different members of the consortium.
Design and Optimization of PSBD-based MMIC modules
The methodology followed in the design and optimization of PSBD-based MMIC modules for multiplication and mixing applications is briefly described in this section. Further details are referenced to previous works and documentation where it is detailed and discussed. The flowchart of the design and optimization process is illustrated in Fig. 1.3. It mainly consist on a three dimensional Computer-Aided Design (CAD) software with an Electromagnetic Field Simulator implemented to solve the Maxwell equations in a defined mesh of the 3D structure. These kinds of software are focused on the solution of the S-parameters of a structure that allow characterizing the losses associated to a specific geometry, materials and impedance matching. However, these kinds of software are not sufficient to carry out the optimizations of MMIC modules since the electrical behavior of active (oscillators) or passive (PSBDs, HBVs, SIS, HEB, etc) devices are not accounted for in HFSS simulations. The complementary software used in this work to develop the virtual design of each MMIC module is the so-called Advance Design System (ADS) software. It is an electronic design automation software system developed by Keysight Techonologies which is dedicated to RF, microwave and high speed digital applications. ADS software allows us to carrying out harmonic balance simulations of non-linear electronic circuits to obtain their frequency and time domain response. The CAD software that has been mainly used in this work is the High Frequency Simulation/Structure Software (Ansys-HFSS) [Ansys-HFSS 2016].
Fig. 1.3. Design flow chart based on Ansoft HFSS and ADS suite. Courtesy of Miss Hui Wang, March 2009.
HFSS has been especially important in this work to characterize the transmission/reflection coefficients (S-parameter) between waveguide transitions, the coupling efficiency of an antenna or probe with the input or output signal and the impedance matching network of the PSBDs devices. ADS software features the electrical model of a wide range of electronic devices (Schottky diodes, lumped elements, transistors, operational amplifiers, etc) and circuit elements (waveguides, transmission lines, attenuators, current/voltage/power sources, etc).
ADS software uses the S-parameters calculated by HFSS together with the electric models of the circuit elements and the analytical Schottky diode model to predict the performance of the structure. The analytical PSBD model main parameters are the saturation current IS, the junction capacitance Cj0 and the series resistance RS. The analytical PSBD model used in this work will be thoroughly discussed in chapter two as well as the improvements included in terms of a two-dimensional Monte Carlo (2D-MC) simulator. The complementary utilization of both simulators allows the design and optimization of the MMIC modules developed in this work. The process is divided in two main stages. First, individual HFSS simulations of each impedance transition in the global structure are carried out to obtain the S-parameters of each transition of the circuit. Second, the ADS simulator is then used to simulate the transmission losses of a defined frequency signal in accordance with the calculated S-parameters, which contains each impedance transition, the materials properties of the simulated structure and the electrical path. These simulations allow the optimization of the structure in an iterative way to efficiently perform the required properties.
PSBD Frequency Multipliers
ADS-HFSS simulations of frequency multipliers for heterodyne applications are used to define the frequency and power range of the input and output LO frequency signals in the non-linear device. The design of frequency PSBD-based frequency multipliers using this methodology can be found in [Tuov95], [Maes05b], [Maes06b], [Sile09b], [Maes10a], [Maes10b], [Sile11a], [Sile11b], [Maes12], [Chen13], [Treu14]. The PSBDs’ properties have to be carefully defined aiming for the maximization of the conversion efficiency of the input LO power into one of its harmonics. The input stage of the MMIC is fully optimized in ADS-HFSS simulations to maximize the coupling efficiency of the input LO power with the antenna or probe that matches the PSBDs of the chip. The output stage of the MMIC is fully optimized to enhance the n-th harmonic generation in the PSBDs and the coupling efficiency of the generated signal with the output antenna or probe that extracts the signal from the chip. Additional structures like filters, DC circuit and specific configurations of the PSBDs, are usually required in the design of frequency multipliers to correctly filter the undesired harmonics of the input LO signal. The bias of the PSBDs is usually required to efficiently manage the input power. HFSS simulations of the PSBDs in frequency multipliers are usually affected by non-linear electromagnetic fields and the near-field phenomena that arise when optimizing the input LO coupling efficiency. This makes the final optimization stage of frequency multipliers longer. These concepts are implemented on the design of a 600 GHz frequency doubler which has been developed by this author and detailed in Chapter 4.
PSBD Frequency Mixers
ADS-HFSS simulations of frequency mixers for heterodyne reception are used to define the correct interaction between the LO and RF input signals in the non-linear device to provide the IF signal. It is possible to find this methodology in the design of PSBD-based mixers [Thom10a], [Chen12], [Hanq12], [Treu16a], [Treu16b]. The PSBDs’ properties have to be carefully defined in accordance with the available LO power and frequency range. ADS-HFSS optimization of frequency mixers is very complex since the objective of these modules is coupling both the LO and the RF input signals in a specific frequency range. The PSBDs’ sensitivity to detect the RF signal depends of a specific amount of LO power coupled with the diodes while the maximization of coupling efficiency of the RF signal and its conversion efficiency into the IF signal are especially important to reduce the noise temperature of the mixer. However, the LO and RF coupling efficiency are intimately related and the final conversion efficiency of RF signal into the IF signal by the PSBDs is a tradeoff between the LO and RF signals interaction. Additional structures like filters, DC circuits, IF adapter circuits and specific configurations of the PSBDs, are usually required in the design of frequency mixers to correctly filter the undesired frequencies and define the IF signal output path. The IF circuit adapter and DC circuit can be included in these mixer modules to improve the global performances. These concepts are implemented in the design of a 1200 GHz frequency doubler which has been developed by this author under the supervision of Dr. A. Maestrini and detailed in Chapter 6.
Structure and Objectives of this PhD Work
The THz science has achieved very important advancements during the last thirty years thanks to considerable technical developments that were used in recent space missions. The experiences, results and successes achieved have placed the multiplication technique as the most suitable way to reach this important frequency range with electronic-based devices. The high potential featured in Schottky technology in multiplications and mixing stages for space-borne applications has enhanced the development of these electronic devices. Schottky technology has usually been combined with superconductor technologies (SIS and HEB) for high sensitivity heterodyne reception, but all-solid-state PSBDs heterodyne receivers have proven to be the most suitable option when lower sensitivities are required. LERMA has had the opportunity to get involved in the development of the JUICE-SWI instrument and the success achieved within the framework of this project is now presented in this PhD. work.
The structure of this dissertation is divided into two different parts. The first part consists of a single chapter dedicated to the PSBDs devices. The improvement and systematization of the electrical PSBDs’ modeling in this work has been carried out in terms of a Two-dimensional Monte Carlo (2D-MC) simulator. The improvements and their implementation in a simple analytical model are discussed in this part as well as the implementation of the analytical model in the Harmonic Balance ADS (HB-ADS) simulator. The second part consists of four chapters dedicated to each module developed by LERMA-LPN in the framework of SWI and the culmination of LERMAs work with a functional 1.2 THz receiver. The first chapter of second part is dedicated to the LERMA-LPN 300 GHz doublers, presenting the single and power-combined approach of this multiplier. The 300 GHz MMIC chips were designed before this work, but the experimental results provided by this module have been invaluable for validating the improved PSBD analytical model developed by this author. The second chapter of the second part is dedicated to the LERMA-LPN 600 GHz doubler. Two different versions are presented, discussed and compared. The design of the first version was developed by LERMA-LPN rather than by this author while the second version of this doubler has been fully designed by this author. Experimental results of this first version will discussed in this work. However, experimental results of the second version were not available yet. The design of a second version of the 600 GHz doubler by this author was motivated by the additional LO power provided by the previous stage at 300 GHz during the development of the project. This second version is expected to provide additional LO power to pump the 1.2 THz mixer and ensure the availability of LO power in the full frequency band of the receiver. The third chapter of second part is dedicated to a LERMA-LPN 600 GHz mixer prototype that was designed before this work and has played a key role in the development of the 1.2 THz mixer. This module has provided the best performances reported at these frequencies using a PSBD-based mixer, and these results have been invaluable for the development of LERMAs contributions to SWI. The last chapter of second part culminates with the demonstration of the LERMA-LPN 1.2 THz receiver. The development of the 1.2 THz mixer is fully detailed and discussed in this section. The experimental results are analyzed, explained and validated by simulations. A novel study of the interaction between the LO multiplication chain and the mixing stage has been carried out by this author. This interaction has demonstrated to be critical in the prediction of the experimental performances of the receiver. Further improvements of the LO multiplication chain and the 1.2 THz mixer are finally proposed and discussed to enhance the LERMA’s contribution to SWI.
Table of contents :
1 THE SUBMILLIMETER RANGE FOR PLANETARY SCIENCE
1.1 THE HETERODYNE DETECTION
1.1.1 THZ FREQUENCY MIXERS
1.1.2 THZ LOCAL OSCILLATOR
1.1.3 SPACE MISSIONS AND THZ SCIENCE
1.2 THE JUICE PROJECT BASELINE
1.2.1 THE SUBMILLIMETER WAVE INSTRUMENT (SWI)
1.3 DESIGN AND OPTIMIZATION OF PSBD-BASED MMIC MODULES
1.4 STRUCTURE AND OBJECTIVES OF THIS PHD WORK
PART 1: PLANAR SHOTTKY BARRIER DIODES MODELING
2 ANALYTICAL MODEL OF PLANAR SCHOTTKY BARRIER DIODES
2.1 THE TWO-DIMENSIONAL MONTE CARLO SIMULATOR
2.1.1 PSBD STRUCTURE IN MONTE CARLO SIMULATIONS
2.2 THE CURRENT TRANSPORT AND CAPACITANCE MODEL IN PSBDS
2.2.1 BUILT-IN VOLTAGE AND BARRIER HEIGHT RELATIONSHIP
2.2.2 ANALYTICAL MODEL OF CURRENT TRANSPORT IN PSBDS
2.2.3 RESISTANCE MODEL OF PSBDS
2.2.4 CAPACITANCE ANALYTICAL MODEL IN PSBDS
2.3 THE ANALYTICAL MODEL IN HARMONIC BALANCE SIMULATIONS
2.3.1 SINGLE VARACTOR PSBD SIMULATIONS FOR DOUBLERS
2.3.2 ANTIPARALLEL VARISTOR PSBDS SIMULATIONS FOR MIXERS
PART 2: THE JUICE-SWI PROJECT
3 A POWER-COMBINED 300 GHZ FREQUENCY DOUBLER
3.1 THE 300 GHZ FREQUENCY DOUBLER CHIP
3.1.1 DESCRIPTION OF THE VIRTUAL DEVICE
3.1.2 DESCRIPTION OF THE MECHANICAL BLOCK
3.1.3 EXPERIMENTAL RESULTS
3.1.4 EXPERIMENTAL COMPARISON BETWEEN PSBD PHYSICAL MODELS
3.2 A POWER-COMBINED 300 GHZ FREQUENCY DOUBLER
3.2.1 QUADRATURE HYBRID COUPLER
3.2.2 DESCRIPTION OF THE MATCHING NETWORK DESIGN
3.2.3 MECHANICAL BLOCK DESIGN: DC CIRCUIT
3.2.4 EXPERIMENTAL RF PERFORMANCE
3.2.5 COMPARISON WITH ADS-HFSS INDIVIDUAL SIMULATIONS
4 A 600 GHZ FREQUENCY DOUBLER
4.1 TWO-ANODES 600 GHZ TWO ANODES FREQUENCY DOUBLER
4.1.1 VIRTUAL DEVICE
4.1.2 MECHANICAL BLOCK
4.1.3 EXPERIMENTAL DEVICE
4.1.4 SIMULATIONS OF THE TWO-ANODES 600 GHZ FREQUENCY DOUBLER
4.1.5 EXPERIMENTAL COMPARISON WITH COMBINED SIMULATIONS OF THE GHZPOWER COMBINED AND 600 GHZ DOUBLERS
4.2 A 600 GHZ FOUR ANODES FREQUENCY DOUBLER
4.2.1 OPTIMIZATION OF THE PSBDS PROPERTIES
4.2.2 VIRTUAL DESIGN IN ADS-HFSS
4.2.3 VIRTUAL ADS-HFSS COMPARISON BETWEEN THE 600 GHZ TWO AND FOUR ANODES DOUBLERS
4.2.4 COMPARISON WITH ADS-HFSS INDIVIDUAL SIMULATIONS
4.2.5 EXPERIMENTAL COMPARISON WITH COMBINED SIMULATIONS OF THE 300 GHZ AND THE 600 GHZ FOUR ANODES DOUBLER
5 A 600 GHZ FREQUENCY RECEIVER
5.1 QUALITATIVE DESCRIPTION OF THE DEVICE
5.1.1 IF ADAPTER CIRCUIT
5.2 ANALYSIS OF THE PSBDS IMPEDANCE MATCHING IN THE 600 GHZ MIXER
5.3 ANALYSIS OF THE PSBDS PERFORMANCES
5.3.1 ANALYSIS OF THE 3·1017 CM EPILAYER DOPING
5.3.2 ANALYSIS OF THE 5·1017 CM EPILAYER DOPING
5.3.3 IMPROVEMENT OF THE 600 GHZ SUBHARMONIC MIXER PERFORMANCE
6 A 1.2 THZ SUB-HARMONIC BIASABLE FREQUENCY MIXER
6.1 OPTIMIZATION OF THE PSBDS PROPERTIES
6.1.1 ANALYSIS OF THE 3·1017 CM EPILAYER DOPING
6.1.2 ANALYSIS OF THE 5·1017 CM EPILAYER DOPING
6.2 DESCRIPTION OF TWO DIFFERENT 1.2 THZ MIXER CHIP DESIGNS
6.2.1 IN-CHANNEL AND OUT-CHANNEL DESIGNS OF THE CHIP
6.2.2 LO ANTENNA AND HAMMER FILTER
6.2.3 DC GROUND STRUCTURE
6.2.4 RF FILTER AND ANTENNA
6.3 MECHANICAL BLOCK: IF AND DC CIRCUITS
6.3.1 IF CIRCUIT
6.3.2 DC CIRCUIT
6.4 THEORETICAL COMPARISON IN ADS-HFSS SIMULATIONS
6.4.1 CONVERSION LOSS AND NOISE TEMPERATURE OF THE RECEIVER
6.4.2 NOISE IN SCHOTTKY MIXERS
6.4.3 NOISE TEMPERATURE IN ADS SIMULATIONS
6.4.4 SIMULATED CONVERSION LOSS IN THE 1.2 THZ MIXER CHIPS
6.4.5 SIMULATED NOISE TEMPERATURE IN THE 1.2 THZ MIXER CHIPS
6.4.6 ANALYSIS OF THE BIAS PERFORMANCES AND THE RLC EQUIVALENT CIRCUIT
6.5 EXPERIMENTAL DEVICE
6.5.1 I-V CHARACTERISTICS OF THE DIODES
6.5.2 THE Y-FACTOR FOR EXPERIMENTAL MEASUREMENT
6.5.3 RF RESULTS OF THE 1.2 THZ RECEIVER AT 300 K
6.5.4 RF RESULTS OF THE 1.2 THZ RECEIVER AT 160 K
6.6 COMPARISON WITH ADS-HFSS INDIVIDUAL SIMULATIONS
6.7 SIMULTANEOUS SIMULATION OF THE 600 GHZ DOUBLER AND THE 1.2 THZ MIXER
6.7.1 SIMULTANEOUS SIMULATIONS WITH THE 600 GHZ TWO ANODES DOUBLER
6.7.2 SIMULTANEOUS SIMULATIONS WITH THE 600 GHZ FOUR ANODES DOUBLER
6.7.3 SIMULTANEOUS SIMULATIONS OF THE 1.2 THZ MIXER WITH HIGHLY DOPED PSBDS
6.8 UPDATED EXPERIMENTAL STATUS OF THE 1.2 THZ RECEIVER
6.8.1 LOCAL OSCILLATOR CHAIN
6.8.2 I-V CHARACTERISTICS OF THE NEW SET OF PSBDS 1.2 THZ CHIPS
6.8.3 UPDATED RESULTS OF THE 1.2 THZ RECEIVER AT 150 K
7 CONCLUSIONS AND PERSPECTIVES