1 High Sensitivity Magnetic Field Sensor Technology overview David P. Pappas National Institute of Standards & Technology Boulder, CO

2 Outline High sensitivity applications & signal measurements Description of various types of sensors used New technologies Comparison of sensor metrics

3 Market analysis - magnetic sensors 2005 Revenue Worldwide - $947M Growth rate 9.4% Type Application “World Magnetic Sensor Components and Modules/Sub-systems Markets” Frost & Sullivan, (2005)

4 Geophysical Astronomical Archeology Health Care Data storage 120 observatories world-wide Fluxgates Proton magnetometers Earth interior dynamics Mineral exploration GPS stability Satellite electronics Increased radiation “Magnetic Monitoring of Earth and Space Jeff Love, USGS Physics Today Feb Halloween magnetic storm North Caroline Department of Cultural Resources “Queen Anne’s Revenge” shipwreck site Beufort, NC Blackbeard’s last stand Magneto-encephalography Magneto- Cardiography “Biomagnetism using SQUIDs: Status and Perspectives” Sternickel, Braginski, Supercond. Sci. Technol. 19 S160–S171 (2006). Mars Global Surveyor Magnetic anomalies Mars Global Explorer (1998) Applications SI units

5 SI - Le Système International d’Unitès I (A) r (m) H-field “Magnetic field intensity”  A/m  B = flux density = “Magnetic induction” field =  H includes medium Frequency What do we measure? Use  0 = permeability of free space = 4  x Wb/Am B-field  tesla (T) kg/(As 2 ) e.g. 1 1 m: A  = B  A B = 2  T

6 B-field Ranges & Frequencies Adapted from “Magnetic Sensors and Magnetometers”, P. Ripka, Artech, (2001) 1 fT 1 nT  , Magnetic field Range 1 pT Geophysical Industrial Magnetic Anomaly Magneto- cardiography Magneto- encephalography 1 fT ( ) , Frequency (Hz) B-field 1 pT ( ) Geophysical Industrial Magnetic Anomaly Magneto-cardiography Magneto- encephalography 1 nT (10 -9 ) 1 gauss  T ~ Earth’s B-field 1 mT (10 -3 ) Technologies 1  T (10 -6 ) 1 1 m 1 aT ( ) Non-destructive evaluation & hdd’s 

7 B-field effects Induction (Faraday’s Law)  Search Coil  Fluxgate  Giant magneto-impedance ( + skin effect) Torque -  Magnetic resonance (optical pumping) Proton – f ~ 4 kHz/gauss Electron – f ~ 3 MHz/gauss  Magneto-striction Scattering  Magneto-resistance (AMR)  Spintronics Giant MR, Tunneling MR, Spin Xtor…  Hall Effect (Lorentz force)  Magneto-optical Wave function interference  Superconducting Quantum Interference Device (SQUID) State B * FM NM FM * *

8 State measurement Low noise excitation source -  Voltage, current, light, … Detector volume -  = A  h Sense state  e.g. sensitivity = V/T Flux feedback is typical  Linearize  Dynamic Range  Complicated  Limits slew rate & bandwidth B ext  BfBf B S Noise V A h

9 Noise metrology Low noise preamplifiers Magnetically shielded container (or room) Spectrum analyzer  Noise power vs. frequency = V 2 /Hz  field noise =  power / sensitivity State Measurement SQUID magnetometer (w/gradiometer) Hz /f White Units Preamp pT/  Hz

10 e.g. Magneto-cardiography pT/  Hz frequency 1pT/  Hz Magnetometer 1 Magnetometer 2 ~60 pT => Use noise at lowest frequency in bandwidth as benchmark extra noise

11 Benchmark properties: SQUID State variableV, f, etc B-field measurementVector/scalar B noise - 1 Hz pT/  Hz Detector volume (  ) cm 3 -mm 3 Operating temperature, TCryogenic/RT/heated Power – form factorLine/Battery

12 Superconducting Quantum Interference Devices ILIL IRIR I Tunnel Junctions B Left-Right phase shifted by B  Signal (V) PU loop Superconductor

13 Pickup loops for SQUIDS One loop: measure B Z Two opposing loops:  dB Z /dz (1 st order gradiometer)  Good noise rejection Opposing gradiometers:  dB Z 2 /dz 2 (2 nd order gradiometer)  High noise rejection Universal magnetometer configurations NSNS SQUID specs Z

14 SQUID Magnetometer “The SQUID Handbook,” Clarke & Braginski, Wiley-VCH 2004 Commercial: 10 – 100’s k$ MRI Discrete components Integrated Systems State variable Voltage (10’s  V) B-fieldVector, gradients B 1 Hz sources pT/  Hz Shunt resistors Shielding eddy currents  - Volume ~ 10 cm 3 coil Operating Tcryogenic PowerLine

15 SQUID detected Magnetic Resonance Image Image of human forearm Low polarizing B-field  60 mT Low precession field  132  T Low resonance f  ~ 6 kHz Don’t need superconducting magnets B Clarke, et. al, ANRV 317-BE09-02 (2007)

16 Resonance magnetometers Proton Nuclear spin resonance  f = 43 MHz/T Water, methanol, kerosene Overhauser effect  Transfer e - spin to protons  He 3, Tempone B ext f  B ext Proton

17 Proton magnetometer Commercial: 5 k$ Kerosene cell Toroidal excitation & pickup Olsen, et. al (1976) From Ripka, (2001) e - spin State variableFrequency ~ kHz B-fieldScalar B 1 Hz sources ~10 pT/  Hz depolarization  - Volume 1 cm 3 cell Operating T-20 => 50 o C PowerBattery

18 Resonance magnetometers (2) Electron spin resonance  f = 28 GHz/T  He 4 : 2 3 S 1 - optical pumping  Alkali metals (Na, K, Rb) f  B ext Proton Vapor cell Photodetector B ext RF Coils /4 Filter Laser

19 He 4 e - -spin magnetometer JPL - SAC-C mission Nov. (2000) Smith, et al (1991) from Ripka (2001) CSAM State variableFrequency B-fieldscalar B 1 Hz sources 1 pT/  Hz Depolarization e.g., precession & collisions w/walls  - Volume ~10 cm 3 cell Operating Tambient Powerbattery

20 e - -spin magnetometer Spin-exchange relaxation free Kominis, et. al, Nature 422, 596 (2003) K metal vapor Low field, high density of atoms Line narrowing effect All-optical excitation & pickup =>Optimized for high sensitivity Solid state State variableFrequency B-fieldVector B 1 Hz pT/  Hz  - Volume ~ 3 cm 3 Operating T180 o C Powerline

21 Solid state magnetometers Semi-conductors  Hall effect Ferromagnetic based:  Magneto-resistive  AMR – Anisotropic MR  Spintronic GMR – Giant MR, TMR – Tunneling M  Fluxgate  Giant magneto-impedance Disruptive technologies  Hybrid superconductor/solid state  Colossal magneto-resistance  Magneto-electric  Spin Transistors Hall

22 V I Hall Effect State variableVoltage B-fieldVector B 1 Hz problems 300,000 pT/  Hz Resistive noise + Small signal – need high electron mobility  - Volume mm 3 Operating TRT Powerbattery Commercial: ~$ 0.1  1 In-line with CMOS Applications Keyboard switches Brushless DC motors Tachometers Flowmeters etc. noise B Fruounchi, Demiere, Radjelovic, Popovic, ISSC (2001)

23 Spectral noise measurements MR Stutzke, Russek, Pappas, and Tondra, J. Appl. Phys. 97, 10Q107 (2005) Yuan, Halloran, da Silva, Pappas, J. Appl. Phys. submitted (2007) 1 p 1 n 1  TMR GMR GMI Hall Fluxgate AMR Noise floor (T/Hz 1/2 ) Frequency (Hz)

24 “Thin Film Magneto-resistive Sensors S. Tumanski, IOP (2001). Magneto-resistive (MR) sensors AMR - Anisotropic MR  Single ferromagnetic film NiFe  2% change in resistance Spintronic:  GMR trilayer w/NM spacer 60%  R/R min Co/Cu/Co “Spin Valve”  TMR – Insulator spacer 500%  R/R min at R.T. CoFeB/MgO/CoFeB Hayakawa, APL (2006) I M FM NM FM * * * * Low field

25 MR as low field sensors AMR Large area films GMR Small sensors Commercial: ~ 1$ 2 Unshielded sensors 2 shielded Flux concentrators TMR “Low frequency picotesla field detection…” Chavez, et. al, APL 91, (2007). F.C. State variableResistance B-fieldVector B 1 Hz sources ~200 pT/  Hz 1/f mag noise Temp fluct. M Johnson/Shot Perming  - Volume mm 3 film Operating TRT Powerbattery

26 Application of flux concentrators B ext a Need: Soft ferromagnet High M =  No hysterisis  Gain up to ~50  No increase in noise  Increase in volume 2 cm Noise TMR with F.C.

27 Spectral noise measurements fluxgate without flux concentrator with external flux concentrator Stutzke, Russek, Pappas, and Tondra, J. Appl. Phys. 97, 10Q107 (2005) Yuan, Halloran, da Silva, Pappas, J. Appl. Phys. submitted (2007) 1 p 1 n 1 

28 Fluxgate M H B ext H mod (f) M State variableInductive 2f B-fieldVector B 1 Hz Sources 10 pT/  Hz Thermal magnetic Johnson Perming  - Volume ~1 cm 3 Operating TRT Powerbattery “Magnetic Sensors and Magnetometers” P. Ripka, Artech, 2001 Commercial: ~1 k$ Pickup(2f) Drive(f) GMI B ext

29 Giant Magneto-impedance (GMI) “Giant magneto-impedance and its applications” Tannous C., Gieraltowski, Jour Mat. Sci: Mater. in Electronics, V15(3) pp (2004) Magnetic amorphous wire I ac M CoFeSiB frequency external field   Enhanced skin effect in magnetic wire GMI spec.

30 GMI specifications Commercial: ~100 $ Disruptive State MHz B-fieldVector B 1 Hz sources ~3000 pT/  Hz 1/f mag noise Temp fluct. M Johnson Perming  - Volume 0.01 mm 3 (wire) Operating TRT PowerBatter

31 Disruptive technologies? Superconducting flux concentrator State variableGMR voltage B-fieldVector B 1 Hz sources.03 pT/  Hz Sensor noise  - Volume 0.1 cm 3 Operating Tcryogenic Powerline Field Gain Nb ~ 500 YBCO ~2000 Hybrid S.C./GMR “…An Alternative to SQUIDs” Pannetier, et. al, IEEE Trans SuperCond 15(2), 892 (2005) Colossal Sensor

32 Colossal Magneto-resistance Manganite materials  La 1-x M x MnO 3  Phase transition – Jahn-Teller distortion Low T – Ferromagnetic High T – Paramagnetic semiconductor ~ 500% change of resistance Barriers to commercialization  Optimal  R/R at ~260 K  High fields  Single crystal materials  High growth temperatures Can integrate with superconducting flux concentrators Haghiri-Gosnet, Renard, J. Phys. D: Appl. Phys. 36 (2003) R127–R150 ME

33 Magneto-electric Disruptive No power required Two terminal device High impedance output State variablePiezo voltage B-fieldVector B 1 Hz sources 100 pT/  Hz pyro/static  - Volume 1 mm 3 Operating T -40 to 150  C Power0 Zhai, Li, Viehland, Bichuin, JAP (2007) Dong, et. al APL V86, (2005). Magnetostrictive + piezo-electric multilayer Terfenol-D H PMN-PT V ME EMR

34 Extraordinary MR (EMR)  Hall effect with metal impurity Based on 10 6 MR in van der Pauw disks  Non-magnetic materials   Mesoscopic devices   R/R ~ 35% in field Replacement for GMR in hdds?  Can’t compete with TMR in MgO  May scale better at small sizes Au InSb B Semiconductor Au impurity “Magnetic Field Nanosensors, Solin, Scientific American V291, 71 (2004) V I Compilation

35 SensorBB n ( pT/  1 Hz) VolumePower SQUID v cm 3 Line e - - SERF v cm 3 Battery-line Hybrid GMR/SCv cm 3 line Proton s 110 cm 3 battery e - - He 4 s 11 cm 3 Battery-line Magneto-opticv1.41 cm 3 line Fluxgate v 101 cm 3 battery MEv1001 mm 3 0 MR v mm 3 battery GMI v mm 3 battery Hallv30, mm 3 battery Compilation Trend: Noise increases as Volume decreases Bn vs. V

36 B noise vs. Volume Volumetric energy SERF He4 SQUID Proton F.G MO Hybrid CSAM ME GMI Hall MR 1.0E E E E E E E-071.0E-051.0E-031.0E-011.0E+01 Volume (cm^3) Noise (pT/rtHz)

37 Conclusions High sensitivity magnetometers research very active Many advances to be made in conventional devices  Potentially disruptive technologies  Move to smaller, lower power, nano-fabrication Noise floor decreases with volume Can look at intrinsic energy resolution of sensor Also need to evaluate high sensitivity against many other parameters:  Spatial resolution  bandwidth  dynamic range  cost, … Acknowledgement. Pick the right tool for the job!

38 Acknowledgements Steve Russek Bill Egelhoff John Unguris Mike Donahue John Kitching Fabio da Silva Sean Halloran Lu Yuan Jeff Kline

39 Noise vs. volume in magnetic sensors  Fluxgate magnetometers Increase volume (  ) & decrease loss (  )  AMR – make up for low  R/R by: Large arrays of elements (volume) good magnetic properties (reduce   Flux concentrators Increase volume Softer, low hysterisis to reduce  E resolution “Fundamental limits of fluxgate magnetometers…” Koch et al, APL V75, 3862 (1999)

40 MR Sensors – spatial resolution 256 element AMR linear array Thermally balanced bridges High speed magnetic tape imaging – forensics, archival NDE imaging Cassette Tape – forensic analysis 4 mm 45 mm erase head stop event write head stop event da Silva, et al., subm. RSI (2007) Current flow in VLSI RAM w/short 2 cm -I y +I x -I x +I y 4 mm V+ V- I- 16  m x 256 I+ BARC.

41 Innovations in Fluxgate technology Apply current in core Single domain rotation 1 pT/  1 Hz Circumferential Magnetization I M Planar fabrication 80 pT/  1 Hz Micro-fluxgates Koch, Rosen, APL 78(13) 1897 (2001) Kawahito S., IEEE J. Solid State Circuits 34(12), 1843 (1999) GMI

42 e - -spin magnetometer Chip scale atomic magnetometer P. D. D. Schwindt, et al. APL 90, (2007). Rb metal vapor Optimized for low power Very small form factor 4.5 mm SERF State variableFrequency B-fieldScalar B 1 Hz 5 pT/  Hz  - Volume 20 mm 3 Operating T110 o C PowerSmall battery

43 Magneto-optic State variableLight intensity B-fieldVector B 1 Hz 1.4 pT/  Hz  - Volume 1 cm 3 Operating Tambient Powerline Mirror-coated iron garnet Magnetometer head Ferrite Flux Concentrators Light polarization changes in garnet Rotation  B-field (Faraday effect) Sensed with interferometer Fiber- optic Deeter, et. al Electronics Letters, V29(11), p 993 (1993). Youber, Pinassaud, Sensors and Actuators A129, 126 (2006). Disruptive Light not affected by B Remote sensors High speed Imaging capability (light) NDE Spintronic B

44 Spin transistors  Tunnel junction based devices Spin dependent hot e- transmission  Cu  Tunnel Barrier  Spin Valve  Schottky barrier 3400% magneto-conductance at 77 K Relatively low currents (10  A) van Dijken, et. al, APL V83(5) 951 (2003) Compilation