Prospects of Optical Recording in Tera Era Han-Ping D. Shieh Inst. of Electro-Optical Engineering Nat l Chiao Tung University Hsinchu, Taiwan 30010 e-mail: hpshieh@cc.nctu.edu.tw
Disk Storage Roadmap
$/MB
Schematic Diagram of HDD Spindle motor Magnetic W/R head Magnetic disk medium Head suspension??? Base Head actuator VCM (voice coil motor)
Magnetic Recording Process
Write (thin film) & Read (MR & GMR) heads Structures
Transitions of magnetic head technology Bulk Head Thin Film Head MR/GMR Head 1979 ~ 1991~ 1998~ Coil Core Write/Read Gap Magnetized Bit Write Head GMR Read Head Write/Read Gap Semiconductor-like/ Micro-fabrication process
Super-paramagnetic effect Stability of small- size marks Ferromagnetic state Super-paramagnetic state Energy 40k b T Energy 40k b T E G 0 p 0 k b T < < E G E G = KuV: Magnetic anisotropy energy of crystal K u : Magnetic anisotropy energy constant V: Crystal size (volume) k b : Boltzmann s constant T: temperature E G k b T ~ E G p
Optical Disk Storage Status: Density ~ Gb/in 2, Access time ~ 100 msec, Data rate ~ MB/sec, form factor ~ ½ inch, single disk. Optical Diffraction Limit (NA, l)??? Optics, laser, mechanics, electronics, material, etc. integration. Objective: Density ~ Tb/cm 2, access time ~ msec
Relationship Between NA and
Minimum Recorded Mark Size MO Wmin = σw/mshc or 2 Ddomain wall Wmin 40 nm 20 nm (theory) 45 nm (Exp.) [by Rugar] Wmin 75 nm (MAMMOS using LP - MFM) [by Awano] PC Grain size 5 nm [by Hosaka] 60 nm (SNOM) λ = 785 nm, Pw =7.6 mw/ 5 ms, Pr = 0.2 mw [by Hosaka]
Ultra-high Density Storage Roadmap - Reading MR (%) 200 150 M rt = 1 m emu/cm 2 SNR = 23 db 55 Gb/in 2 : 38 nm 308 nm 350 Gb/in 2 : 18 nm 102 nm 1000 Gb/in 2 : 14 nm 46 nm 1000 500 100 100 50 50 0 2001 02 03 04 05 06 07 08 09 10 Year 10
Issues in optical detection High-density optical recording by short-wavelength light source Signal Number of photons, Kerr/Faraday rotation angle Number of photons = P RD / (hν) λ Shorter wavelength => Weaker interaction conversion efficiency (A/W) 0.4 0.2 0.0 theoretical limit down 400 800 1200 wavelength (nm) Kerr rotation angle (min.) 30 20 10 0 weaker interaction amorphous GdCo 2000 1000 600 400 wavelength (nm) Hamamatsu Photonics; Si PIN photo diode, (S5973-02). K. Sato, and Y. Togami; J. Magn. Magn. Mater., 35, (1983), 181.
Issues in optical detection Narrow read-power margin by shrinkage optical spot size Upper limit => Read-spot destroys recorded marks. Lower limit => Shot-noise and amp.-noise become dominant. Signal i sign I PD Shot-noise i shot (2qk B I PD ) 1/2 Amp.-noise i amp = const. Other difficulties... Readout signal is strongly affected by quality of optical spot. Developing light source devices of shorter wavelength Many difficulties in optical detection!!! Alternative method: Magnetic Flux detection
2D ultra-high Density Recording Objective Lens Spot Size NA 0.6~0.7 Objective Lens Solid Immerse Lens (SIL) High NA >1 Below-Diffraction-Limit Recording Near-field Recording Probe Recording 3D Storage Multi-function Volumetric optical disks Holographic Memory
Hybrid Recording Thermal Assist Magnetic Recording Optical head Magnetic coil high readout sensitivity. limitation: High coercive media super-paramagnetic effect perpendicular recording, amorphous materials limitation: Optical diffraction limit low readout sensitivity Hybrid Recording: Optical Writing / Magnetic Reading Recording
NIST Awards Thermal Assist Magnetic Recording (HAMR) Seagate, Carnegie Mellon Univ., U. of AZ, MEMS Optical Inc., et al. 5yrs, +20 Millions. Enabling recording density ~ 1000 Gb/in 2
Writing Process in Hybrid Recording LS-MFM head Applied field Carbon protective layer Protective layer Conductive layer Protective layer Readout layer Memory layer Protective layer Substrate Requirements of memory layer : 1. Higher Tc 2. High Ku 3. High switching rate Disk rotating direction
Reading Process in Hybrid Recording GMR head Output signal Carbon protective layer Protective layer Conductive layer Protective layer Readout layer Memory layer Protective layer Substrate Fringing field Requirements of readout layer : 1. Higher Mr 2. High switching rate
Remnant Magnetization (Mr) Requirement in Hybrid Recording 1. CNR log ( V signal /V noise ) V signal M r x mark size x thickness x Dr Dr : the percentage change of resistivity 2. Mr requirement in optical writing/magnetic reading recording 3. CNR of 45 db required for system. Mr (emu/cc) Mark size (um) Thickness (nm) Head sensitivity (compared with MR head) CNR (db) Recording density ( GB/in 2, track width=0.1um) *(1) 90 1.07 50 1 45 *(2) 90 0.274 50 1 37 389 0.1 25 5 45 60 243 0.1 25 8 45 60 195 0.1 25 10 45 60 405 0.06 25 8 45 Over 100 (1) and (2): IEEE trans. Magn., 36(1), Jan. 2000
TbFeCo Dual-layer Media Structure magnetization (emu/cm3) 600 400 200 0-200 -400-600 TM-rich 20nm -4-2 0 2 4 applied field (koe) TM-rich TbFeCo, Ms = 370 emu/cm 3, Mr= 355 emu/cm 3, Hc= 0.95 koe RE-rich TbFeCo, Mr= 113 emu/cm 3, Hc= 4.5 koe SiN 20 nm Readout layer ( h 1 ) Memory layer ( h 2 ) SiN 20 nm Si substrate M t (memu/cm2) memory layer 20 nm/readout layer 20 nm 0.8 0.4 0.0-0.4-0.8 Magnetization reversal of readout layer 6 koe Magnetization reversal of memory layer -10-5 0 5 10 applied field (koe) Readout layer Hc = 6 koe ( = Hc of single layer + s w /2M 1 h 1 ) NCTU/OSDLAB C.C. Lin
Optical Disk Drive
Optical Pick-Up by Micro-Optical Bench Technologies
Flying Optical Head Slider objective lens, SIL Flying height (air-bearing)< near-field recording Objective Lens slider Solid Immerse Lens (SIL) air bearing < λ disk motion
Thermal Writing Apparatus by use of Fiberlens Spinning Medium Laser Diode SMF Coupling objective NA=0.2~0.3 Microlens Driver Circuit System throughput efficiency > 50% Laser diode output power (20mW) Fiberlens output power (>10mW) Marks ~1.6 um
Procedure of the Parabolic Lens Making r»60 m NA=0.6
Fiberlens-SIL-aperture combination S i O 2 S i substrate Photo resist AZP4620 Step 1: deposit SiO2 as sacrificial layer Step 4: oxide etching Step 7: photolithography mask #2 Photo resist FH6400 Cu Step 2: photolithography mask #1 Step 5: sputtering Cu as seed layer Step 8: reflow Ni SU8 Step 3: reflow Step 6*: electro-plating Ni to reduce aperture Step 9: photolithography mask #3 * Trans Lane and Wensyang Hsu: ISOM01 p252
Integrated fiberlens-sil-aperture combination Step 10: release support fiberlens SIL aperture
Far-field emission light thru sub-um aperture S ~ 600 nm s @ 600 nm AFM image of aperture Far-field emission light thru A sub-um aperture
Planar Pick-Up Fiber+Mirror Light Source
Planar Pick-up Converge Light Planar Microlens
Planar Near-field Aperture Minimize Spot Size Aperture Part
Integrated with Thin-Film Magnetic Coils Apply to Thermo-Magnetic Recording Top View Micro Tip Planar Copper Coil Perspective (upside down) A A A-A section
Integrated Planar Pick-Up Thermo-Magnetic Recording with GMR Head Planar Thermal-Magnetic Recording GMR Head Recording Layer GMR Head
Planar Integrated Pickup for Hybrid Recording focusing mirror formed by isotropic etching or flat mirror formed by anisotropic etching V-groove for fiber alignment fiber lens fiber suspension bonding SIL GMR head for magnetic flux reading planar coil for bias magnetic field Semiconductor-Like Fabrication Processes
Disk Storage Perspective Magnetic recording density and data rate are superior and continuously extend beyond forecast. Super paramagnetic limit keeps pushing further. ROM/Write-Once optical storage are very unique for applications. Portable is the niche for erasable optical disks. > 300 Gb/in 2 and beyond, fast access time require more innovation in write/read elements, media, and others. Thermally assisted magnetic/optical (Hybrid) recording using perpendicular media is regarded > Tb/in 2.