HTS Materials and Devices for RF Applications

Transcription

1 slide 1 HTS Materials and Devices for RF Applications John Talvacchio Northrop Grumman Corporation, Baltimore, MD Potential markets / motivation for HTS RF electronics Preparation of materials Measurement of RF properties Northrop Grumman s interest in HTS Analog subsystems for Cryoradar Specific devices and applications Microwave filters Wireless communications Oscillators Delay lines Tunable devices Navy HTSSE program Comments on refrigeration Predictions

2 slide 2 HTS Materials and Devices for RF Applications: Bibliography The proceedings of Applied Superconductivity Conferences from 1992 to 1998 are the best general sources for relevant papers. These are published in issue No. 2 of Vols. 3, 5, 7, and 9, respectively, of the IEEE Trans. on Applied Superconductivity. M. J. Lancaster, Passive Microwave Device Applications of High-Temperature Superconductors (Cambridge University Press, Cambridge, 1997). M. M. Fitelson, Cryogenic Electronics in Advanced Sensor Systems, IEEE Trans. on Applied Superconductivity 5(2), 3208 (1995). S. H. Talisa, M. A. Janocko, D. L. Meier, J. Talvacchio, C. Moskowitz, D. C. Buck, R. S. Nye, S. J. Pieseski, and G. R. Wagner, High-Temperature Superconducting Space-Qualified Multiplexers and Delay Lines, IEEE Trans. Microwave Theory and Techniques 44(7), 1229 (1996). M. M. Driscoll and R. W. Weinert, "Low-Noise, Microwave Signal Generation Using Cryogenic, Sapphire Dielectric Resonators: An Update," Proc. IEEE Symposium on Frequency Control, 157 (1992).

3 slide 3 Viable Electronic Applications of HTS: Grouped by Markets Radar / Military RF Front-end preselection Low-phase-noise waveform generator Antenna matching networks High dynamic range A/D conversion Communications Low-loss, small-size filters Channelizers / multiplexers Spread spectrum comm High data-rate switching Magnetic Anomaly Sensors Mine detection Submarine detection / ASW Geophysics Medical Systems Magneto-encephalography Magneto-cardiography NMR and MRI pick-up coils Computing Crossbar switches Cryo-CMOS MCM interconnects Infrared Imaging On-FPA preprocessors VLWIR detection Instrumentation Voltage and current standards Spectrum analyzer Sampling oscilloscope/ timedomain reflectometer

4 slide 4 Passive RF Applications of HTS are Critical to the Cryogenic Electronics Industry SQUID Sensors: Market is too small to develop industrial infrastructure Instrumentation (e.g. voltage standards): Market is too small to develop industrial infrastructure LTS Digital: High-speed signal processing capabilities demonstrated Integrated circuit fabrication well developed No one wants the size, cost, power consumption, and reliability risk of coolers HTS Digital: Integrated circuit fabrication capability is relatively primitive HTS Microwave Devices: Pay the bills at large and small companies specializing in superconducting electronics LTS 2 x 2 Network Switch HTS 39-Jct Digital Circuit

5 Orders of Magnitude Performance Advantage From Superconductivity and Cryogenics 1. Low Surface Resistance: Improved Performance of Microwave Devices R s (ohms) 2. Reduced Power Dissipation and Delay: High-Speed Logic Delay Colorado 7-00 RF slide K COPPER f 1/ f 2 YBCO Frequency (GHz) 3. Unique Quantum Accuracy: Voltage Standard, DAC, ADC 1 ns 100 ps 10 ps 1 ps 0.1 ps 0.1 mw JJs: Flux Quantum Logic CMOS JJs: Latching Logic CMOS 77K Silicon Bipolar GaAs MESFET MODFET 77K 10 mw 1 mw Power Dissipation / Gate 4. Low Noise from Cryogenic Operation 100 mw

6 Two-Fluid Model of a Superconductor Relates dc and Microwave Properties Energy Normal Electrons R N s N n N e-d / kt Colorado 7-00 RF slide 6 D (2 30 mv) SC Pairs L S For a Normal Metal: 1 Skin depth, d 2 N m o s N w >> l 2 For a Superconductor: 1 m o s S w Magnetic penetration depth is independent of frequency R surface (w/s N ) 1/2 R surface w 2 s N

7 slide 7 What did we have to learn to do? Low RF Surface Resistance of YBCO Epitaxial films grown on single-crystal substrates C-axis orientation (Cu-O planes parallel to substrate) R s (W) R s (W) 1 10 GHz 77K COPPER YBCO Temperature (K) COPPER f 1/2 f 2 YBCO Frequency (GHz) Having obtained low R s, other factors will determine whether passive HTS devices are ultimately successful: manufacturing costs - CAIV power handling dynamic range (linearity) weight and volume

8 slide 8 Large-Area, Double-Sided YBCO Films: Materials Base for a First Generation of Devices STALO Resonator Single Film Layer UHF Antenna YBCO Substrate Filterbank Channel Microstrip Transmission Lines YBCO Substrate Stripline Transmission Lines Delay Line YBCO Substrate Substrate Northrop Grumman Proprietary

9 slide 9 High-T c Superconductors Reduce Refrigeration Requirements Transition Temperature, T c (K) 140 Hg-Ba-Ca-Cu-O Tl-Ba-Ca-Cu-O Bi-Sr-Ca-Cu-O Y-Ba-Cu-O Liquid Nitrogen La-Sr-Cu-O La-Ba-Cu-O NbN Nb 3 Sn Nb 3 Ge 20 Pb Nb Hg Liquid Helium Year of Discovery Transition Temperature, T c (ºC)

10 slide 10 Film Deposition Techniques for Epitaxial Oxide Superconductors Multi-Source Evaporation / Sputtering Heated Substrates Single-Source Off-Axis Sputtering Heated Substrates RF Power Y Ba Cu YBCO Target Excimer Laser Beam Laser Ablation Heated Substrates O 2 MOCVD YBCO Target Ar Y Ba Cu

11 slide 11 Oxygen Phase Diagram for YBCO: Oxygen Order and Stoichiometry are Keys to Performance O 2 Pressure (Torr) C 800 C 700 C 600 C 500 C 400 C Tetragonal Y 1 Ba 2 Cu 3 O x MOCVD Orthorhombic 60K LASER ABLATION SPUTTERING Orthorhombic 90K Shaded areas indicate typical growth conditions for high-quality YBCO films YBCO decomposed to: Y 2 BaCuO 5 + BaCuO 4 + CuO EVAPORATION / Temperature (10-3 K -1 )

12 Film Growth for Microwave Applications: The Garching Process #1 Colorado 7-00 RF slide 12 All of the sophistication of this process is in the heater design High vacuum permits high deposition rates and good rate control Oxygen gas pocket permits YBCO phase formation

13 slide 13 Measurement of RF Surface Resistance Most measurements of R s use a resonant cavity and infer R s from the measured Q where, 1 / Q = 1 / Q s + 1 / Q dielectric + 1 / Q radiation + 1 / Q normal-metal Measured If the cavity is loaded with a dielectric (= tan d) e.g., cavity with Cu walls R s 1 / Q s Energy may be lost by radiation For a sensitive measurement, Q s should be the smallest of all of these Q s, i.e., the low-loss superconductor should be the lossiest part of the device A well-designed measurement apparatus will have low Q s even when R s is small In contrast, a well-designed device will have a high Q

14 slide 14 Measurement Techniques for R s CYLINDRICAL COPPER CAVITY (End-Wall Replacement) HTS / DIELECTRIC RESONATOR 1/2" dia. high sapphire puck PARALLEL-PLATE RESONATOR (Taber Technique) 2-inch dia. HTS film Two 2-inch dia. HTS films HTS film spacer HTS film Gray indicates area measured Large area but poor Small area; excellent sensitivity (~5 mw) m sensitivity (> 10x reference film) Good sensitivity but wafer must be diced

15 slide 15 Measurement of Non-Linear Response Insertion Loss, S 21 2f 1 - f 2 f 1 f 2 2f 2 - f 1 Filter Response Frequency fi 1. Apply two high-power tones within filter passband 2. Measure output power as a function of input power at f 1, f 2, 2f 2 - f 1, and 2f 1 - f 2 (Third-order products would not exist for a perfectly linear response) System Dynamic Range is a function of materials and design (keep current density low) LNAs usually limit dynamic range

16 slide 16 Summary of RF Properties of HTS Films Only epitaxial c-axis films have the low rf loss, R s (77K, 10 GHz) < 1 mw, needed for applications TBCCO or YBCO but l(t) for YBCO is still changing at 77K High-quality films and clever device designs that minimize current density permit up to 100s W devices - no problem for receive applications Low signal attenuation is only one benefit of HTS Cryogenic operation fi Low noise Elimination of amplification stages fi High dynamic range Yield is longer a critical factor but overall film production costs are still high

17 slide 17 Transmission-Line Dimensions for Microstrip and Stripline The primary requirement is for a 50 W characteristic impedance, Z = (L/C) 1/2 = function of (h/w) For a 50 W line on LaAlO 3 (e = 24), the conductor width must be: Wafer Thickness 10 mils 20 mils Microstrip 88 mm 176 mm Stripline 22 mm 44 mm YBCO YBCO Microstrip Substrate Stripline Substrate w h Substrate Reducing wafer thickness reduces the overall device size proportionally For thin-film dielectrics (e.g. 1 mm) linewidths must be < 1/2 mm

18 Optimum Dielectric Thickness for Compact HTS Microwave Components For HTS Microstrip Transmission Lines, 2 Configurations Now Available (also applies to stripline): substrate 250 µm thick Substrate Dielectric: Low HTS Conductor Loss Relatively Large Size YBCO 0.5 µm thick Microstrip Q Ideally, use an intermediate dielectric thickness Materials Parameters: R s (77K, 10 GHz) = 0.5 mω Single Crystal: LaAlO 3 Dielectric Films: Sr 2 AlTaO 6 tan δ= 3 x 10-5 tan δ= 1 x 10-4 Dielectric Thickness 250 mm Colorado 7-00 RF slide 18 substrate µm thick Thin Film Insulator ~0.5 µm thick Thin Film Dielectric: Extremely Compact Relatively High Conductor Loss mm 25 mm Frequency (GHz)

19 slide 19 Substrates for HTS Microwave Devices LaAlO 3 was the most widely used substrate for development programs Problems with LaAlO 3 : anisotropic dielectric constant movement of twin boundaries for mm-wave applications, the high e results in structures that are too small Problems with alternate substrates: Thermal expansion mismatch of Si and sapphire to YBCO limits films thicknesses Loss tangent is much too high in YSZ, somewhat too high in NdGaO 3 30% LaAlO % Sr(Ta,Al)O 3 (LSAT) is untwinned but e is not sufficiently uniform MgO is not readily available in large wafers; cleaves easily Nevertheless, best alternative available today

20 slide 20 HTS Technology Enables CRYORADAR to Find Targets in Clutter Cryoelectronic Radar Subsystems Provide: Pure Transmit Signal 100x increase in microwave resonator Q 50x increase in dynamic range 50x reduction in size Low Noise / High Dynamic Range Reception 10x increase in speed 10x reduction in power of logic circuits ~20 db improvement in target detectability in clutter Large Background Signals Establish Full Scale... but Small Signals Can Be Important

21 Superconducting Filters Uniquely Provide Low Loss and Small Volume Insertion Loss (db) 20 Element» Filter» 7 R Lumped- Combline Filter Interdigital Superconducting Filter Dual Mode Dielectric Resonator Waveguide Filter Waveguide Filter Volume (cubic in.) 1 cm - X-Band Filters - Refrigeration volume amor- tized over a bank of filters Colorado 7-00 RF slide 21

22 slide 22 Filter Loss Calculation 4-Pole Chebychev Filter, 1.25% Bandwidth Center Frequency at 4 GHz 1.0 Midband Loss (db) Au Gnd. Plane HTS Gnd. Plane: R s = 0.5 mw R s = 5 mw Strip R s (mw at 10 GHz)

23 slide 23 Conductus is Betting Its Existence on HTS Filters for Cellular and PCS Base Stations Compressor Receiver YBCO Filters combined with cryogenic LNA reduce noise Low insertion loss permits higher number of poles, sharper skirts Compact, lightweight systems can be mounted on towers Cellular systems (800 MHz) in the field; PCS in development Similar military systems fielded

24 slide 24 STI is Betting Its Existence on HTS Filters for Cellular and PCS Base Stations Cellular A-Band TBCCO Filters combined with cryogenic LNA reduce noise Low insertion loss permits higher number of poles, sharper skirts Compact, lightweight systems can be mounted on towers Cellular systems (800 MHz) in the field; PCS in development Similar military systems fielded Recently added spectrum for A and B is, The FCC s gift to HTS

25 Cryocooler Technology is Making Significant Advances in Affordability, Reliability, and Size Most users want integrated, closed-cycle cryocoolers - their existence transparent to the operator Only volume sales can bring down cooler costs / Cost $20k $15k $3k $1.5k Reliability 5,000 hrs. 15,000 40, ,000 (MTBF) Size 1x 1/3x Based on 4W Heat Lift at 77K Colorado 7-00 RF slide 25

26 slide 26 HTS Four-Channel Filterbank: Example of HTSSE II Device Insertion Loss (db) 0 77K Frequency (GHz) Centered at 4 GHz 50 MHz-Wide Channels 4 YBCO Films on 2 Wafers Integrated 50 W Terminations Integrated Branchline Couplers Integrated Channel Interconnections Northrop Grumman Proprietary

27 slide 27 Motivation for Switched Filterbanks The Problem Radar Agile Frequency Slots Interfering Signals RADAR SIGNAL System RF Bandwidth The Solution: A Switched Filterbank All Filters Turned Off Except That Encompassing Desired Signal

28 Preselector Switched Filterbank: YBCO Films Packaged with GaAs FET Switches Colorado 7-00 RF slide 28 Input Through-Port (All Switches Open) Selected Channel Switched Filter Channel Channel 1 Channel 2 Channel N 0 Insertion Loss On State Off State GaAs Etch-Back FET Switch ETCHBFT2.JPG Frequency Northrop Grumman Proprietary

29 slide 29 Cryo Resonator for STALO: Example of Cryocooler Integration Key Fabrication Issues Vibration Isolation Grounding HTS Films Frequency Trimming Temperature Stability System Delivered to NRL in 1997 Demonstrated in Navy Radar Testbed Significant Improvement in Radar Sensitivity (Limited by ADC Used) PWR. Supplies Array Driver STALO Temp. Control Cryogenic Resonator Assembly Advanced cryo package will reduce volume by 10x

30 slide 30 Introduction / Motivation for Tunable Microwave Devices Interest is in materials where dielectric constant, e (E), is a function of applied electric field Used to produce tunable capacitors: capacitance, C e (E) For a length of transmission line: tunable resonators: wavelength, f (CL) 1/2 (e m) 1/2 tunable delay: phase velocity, v p (1 / CL) 1/2 (e m) -1/2 However, characteristic impedance, Z (L / C) 1/2 (m / e) 1/2 Ideally, impedance would be independent of tuning Analogous magnetic field tuning is easier, already in use, but potential for dm / m is smaller than for de / e DARPA s program, Frequency Agile Materials for Electronics, (FAME) started in goal is for factor of 2 shift in frequency

31 slide 31 Motivation for Tunable Filters The Problem Radar Agile Frequency Slots Amplitude Frequency Interfering Signals RADAR SIGNAL Potential Solution #1: A Switched Filterbank All Filters Turned Off Except That Encompassing Desired Signal Potential Solution #2: A Single Tunable Filter Filter s Passband Shifted to Match Signal Frequency

32 slide 32 Integrated HTS / Ferroelectric Band Reject Filter Insertion Loss (db) Increasing dc voltage Superconducting Core Technology, Inc. Signal In Frequency (GHz) Signal Out SrTiO 3 film covers interdigitated capacitor Four notch filters, each with 5% tuning range, Q»200, K»10 Thin-film SrTiO 3 ferroelectrics and YBCO conductors (60 K operation) Note higher Q with applied field

33 slide 33 Other RF Devices Based on HTS China Lake NWC: Electrically short UHF antenna matching networks; HTS improves antenna efficiency Brucker Instruments: NMR or MRI pick-up coils improve sensitivity for small samples and low magnetic fields Neocera, Inc.: RF circulators for antenna manifolds, etc.; compact and low loss but HTS films are exposed to H = 0.2 tesla Lincoln Labs: Variable phase shifters for beam steering combine HTS and ferrites

34 slide 34 Navy High-T c Superconductor Space Experiments (HTSSE I and II) HTSSE I Simple passive devices, mostly filters and resonators Failed to achieve orbit HTSSE II More complex subsystems but still just for testing Originally scheduled for August, 1996 launch - delayed Launched Feb 23, 1999

35 slide 35 Cryogenic Packaging Fundamentals Control Lines (DC) Vacuum Housing Conductive Heat Load RF Line HTS Device RF Line Radiation Shield Cold Head Thermal Connector Radiated Heat Load Cryocooler Engine

36 slide 36 Conclusions Analog HTS electronics are based on low RF surface resistance Materials technology is relatively mature No trade-off between LTS and HTS Device performance can be accurately modeled Microwave filters for the commercial wireless market Best (only?) hope for a substantial market -- big enough for volume to reduce costs and pay for special tools Assist defense electronics development with experience in scaling up production of films, packaging, and cryocooler integration Performance has been demonstrated for a range of devices -- but few applications are based solely on performance Cost and reliability of cryocoolers is a major barrier to wider application