Deposition Technologies for >500GB/in 2 and HAMR Write Heads PMR

Transcription

1 Deposition Technologies for >500GB/in 2 and HAMR Write Heads PMR

2 Outline Background New technologies for PMR pole deposition Optical films for HAMR write heads Summary 2

3 Background Technology Roadmap > 2015 Density < 1TB/in TB/in 2 > 5 TB/in 2 Technologies PMR DFH HAMR Shingled (SWR) DFH HAMR Shingled (SWR) 2D read-back BPM Key Areas Sigma reduction Damascene writer Reduced HOC Light delivery Near-field transducer 2D signal processing Media patterning PMR HAMR BPM 3

4 Damascene Processes for Advanced PMR Poles

5 Motivation for Damascene Processes Two types of process are used for PMR poles Subtractive process (etch) Additive process (Damascene) PR RIBE IBE IBE Alumina Pole PR PR Pole PR CMP PR IBE Alumina Alumina Alumina Damascene Processes are Preferred as Geometries Shrink 5

6 Requirements for Damascene Pole Processes Key steps in the damascene process Deposition of seed layer Plating of magnetic materials CMP Seed layer requirements 1. Conformal 2. Conductive / Low resistivity 3. Void free plating of magnetic materials with desired properties 4. Deposited at TFMH compatible temperatures (<200ºC) 5. Good thickness control 6. Adhesion CVD deposition of Ru meets the above requirements Fabrication Sequence Etch CVD Ru Deposition Plate + CMP CVD Alumina Alumina Alumina Plated pole 6

7 Ru CVD for Damascene Writer Pole Enabling Features CVD Ru is unique in being able to combine several key features 1. Appropriate properties as a seed layer for plating Conformal Conductive / Low resistivity 2. Appropriate as a seed layer for Magnetic properties of plated layer Void free plating of magnetic materials High Bs plated layer a b Step coverage % ~ 0.4 m trenches CVD- Ru 3. Provides ability to fine tune final track-width of the PMR pole Good thickness control a Compensate for trench etch variation Ru CVD also used as a plating seed for wrap around shields Enables narrower track-widths by reduced cross track interference 7

8 Ru CVD Chemistry Precursor : RuO 4 ToRuS Blend from Air Liquide Mixture of RuO 4 in two fluorinated solvents Liquid at room temperature High Vapor Pressure Rapid Reaction practical deposition rates Low Temperature Process (170 C) Small molecule that nucleates on all tested surfaces Two step reaction (heated wafer) A. RuO 4 RuO 2 + O 2 (not self limiting) B. RuO 2 + 2H 2 Ru + 2H 2 O Accomplished as cyclic CVD in 4 steps Steps 1,2 : Reaction A: + Inert Gas Purge Steps 3,4 : Reaction B: + Inert Gas Purge Repeat to get desired thickness 8

9 System Architecture Cross flow Design Inlet Outlet Heated substrate Showerhead Design Inlet Heated substrate Outlet Outlet RuO 4 decomposition reaction to RuO 2 is catalyzed by Ru but not self limiting Cross flow architecture requires self limiting reactions to achieve good WiW uniformity and precursor utilization Showerhead architecture allows for XRF thickness (A) Lower ToRuS partial pressure Higher ToRuS partial pressure localized flow optimization to achieve 180 optimum WiW uniformity and precursor utilization ToRuS pulse time (s) NEXUS Ru CVD uses a showerhead design 9

10 Deposition Conformality 10 Cross-section SEM ~400Å CVD Ru on SiO2 Deposition conformality similar at process temperatures from C Deposition conformality similar for features of various sizes Conformality of adhesion enhancing underlayers can impact stack conformality

11 CVD Ru: Thickness Control, Resistivity Rate Linearity of RuCVD on SiO2 CVD Ru on SiO2 - Rate vs Dep Temp 3.3 XRF Thickness (A) C y=2.8x C y=2.6x # of Cycles Growth Rate(A/cycle) Dep Temp(degC) Growth rate is linear with number of cycles Thickness can be accurately targeted Initial nucleation delay <7 cycles at 200ºC Growth rate varies with temperature Resistivity change slightly with temperature < 20µΩ-cm at 400Å on SiO 2 at C Resistivity (uohm-cm) Resistivity vs Dep Temp (~400A on SiO2) Dep Temperature (degc) Good thickness control, appropriate electrical properties for plating 11

12 Film Stress Engineering CVD only CVD with periodic plasma anneal Substrate Substrate Ru CVD film stress ~2600MPa Stress can be tuned to desired value by periodic plasma annealing Stress driven more compressive with: More plasma anneals Higher power plasma anneals Lower pressure plasma anneals 12

13 Marathon Test Through Ampoule Life Data from one 3.6L ampoules with single process recipe on Lab Tool 150mm wafers, 190 C deposition temp 632 wafers processed within tight specifications (black line shows End Of Life) Mean thickness at 401.6A with 3 variation of 1.45% Ampoule Utilization at 83A/cc Film Thickness (XRF) and Resistivity with # of wafers Sheet Resistance and Non-Uniformity w ith # of wafers Center XRF Thickness (A) WTW 0.34% 1σ WTW 0.36% 1σ Wafer Number Resistivity(uohm-cm) Sheet Resistance (ohm/sq) WTW 0.49% 1σ Mean 1.38% Wafer Number 3s Rs unif (%) Good WTW and WIW repeatability across ampoule life < 2% 3 Non-Uniformity up to 650 wafers 13

14 CVD Ru: Uniformity, Repeatability (200mm) 12-wafer lot on Lab Tool: WIW < 3% 3 WTW ~1% 3 Multiple ampoules on Lab Tool: WIW ~3% 3 WTW ~2.5% 3 Data through ampoule life: WIW <6% 3 WTW < 2% 3 Rs (ohm/sq) ~400A CVD Ru on 200mm SiO2-200C (49pt 6mmEE me asurement) Average WIW uniformity 0.81% 1σ WTW repeatability 0.35% 1σ Sheet resistance 0.1 WIW uniformity Wafer # Nu (%, 1s) Rs (ohm/sq) Sheet Resistance and Uniformity vs Ampoule Cycles Three Ampoules - 200mm wafer - 200C - Lab Tool WTW repeatabilty 0.83% 1σ Average WiW 1σ non-uniformity 0.88% 1σ Nos. of cycles on three 1.2l ampoules WiW 1s non-unif (%) Good uniformity and repeatability through multiple ampoules < 6% 3σ WIW, < 2% 3 WTW through ampoule life 14

15 Summary NEXUS CVD Ru meets the film properties required of a conformal seed layer for damascene processing of high areal density write poles and shields Conformality: >0.98 Deposition temperature: <200 C Resistivity: <20 cm Uniformity: <3%3 Repeatability: <2%3 High efficiency of chemical use: >80Å/ml The showerhead architecture is ideal for the ToRuS based CVD Ru process The NEXUS CVD Ru system has been proven in production Demonstrated >90% uptime Stable film performance over lifetime of ampoule 15

16 Deposition of Optical Materials for HAMR Write Heads

17 HAMR Technology Overview HAMR is the leading candidate for >1TB/in 2 areal densities for write heads Concept At high areal densities, write heads cannot produce enough field to record data in high coercivity media By heating the media locally, the coercivity can be reduced and the data recorded HAMR write heads incorporate a laser light and near field transducer to locally heat the media Challenges 1. Integration of light source and delivery mechanism 2. Thermal management 3. Low loss optical materials for light transmission Magnetic Field Field needed for smaller bit size Maximum Head Field Light transmission via low loss waveguides Head Media Source: IDEMA Presentation 12/6/07, Mark Re, Seagate HAMR is a key technology for >1Tb/in 2 17

18 HAMR: Waveguide Materials Laser with optical waveguide Core/Cladding for waveguides Requirements Low and high index films Low optical loss / defect free Good Throughput Parameter Index k Optical Loss Rate Uniformity Requirement High n: >2 (Ta 2 O 5 ) (core) Low n: <2 (Al 2 O 3 ) (cladding) 0 (below ellipsometer limit) <10 db/cm (cladding) <5 db/cm (core) Å/min (core) Å/min (cladding) <3% R/M (200mm) 18

19 PVD Deposition Technologies for Waveguide Films Property Technology Loss Uniformity (R/M 200mm) Rate Comment IBD <0.5 db/cm <3% Å/min Good optical properties Low Rate / Poor CoO PVD1 Dielectric Mode <5 db/cm <3% Ta 2 O Å/min Core PVD1 Metal Mode <10 db/cm <3.5% Al 2 O 3 >1000 Å/min Cladding PVD processing has flexibility of hardware and deposition modes Hot chuck capability (up to 400ºC) Dielectric and metal modes of deposition RF diode / DC magnetron 19

20 Ta 2 O 5 Dielectric/Poisoned Mode Films Target Voltage(V) TaOx Hysteresis Not poisoned Poisoned Oxygen flow (sccm ) No measurable change in optical properties (ellipsometry) operating at different O 2 flows in poison mode Processes show acceptable loss properties Process O 2 flow (sccm) Power (W) Deposition Rate (A/min) U% 1σ% 200 mm Index Index U% 1σ% Loss Results Db/cm Process A Flow Process B Flow Process C Flow

21 Closed-Loop Control Metal Mode Al 2 O 3 Deposition for High Deposition Rates Characteristic Hysteresis Loop of Target V vs. O 2 V (Volts) Target surface metallic O2 decr O2 incr Target surface dielectric O2 Flow (sccm) Ideal operating point High rate Good film properties Difficult to control due to target poisoning Closed-loop control with High Speed MFC partial pressure controller Target voltage continuously monitored High speed piezo MFC: oxygen flow adjusted to maintain target voltage Can now operate in forbidden range of voltages Rate ~ 50% of Al deposition rate achievable 21

22 Al 2 O 3 : Optical Loss 30 Optical Loss vs. Wafer Temperature Optical Loss vs. Deposition Temperature Optical Loss [db/cm] Temperature Strong dependence of optical loss on wafer deposition temperature No significant changes observed using ellipsometry measurement Measure extinction coefficient zero for all above films. Loss not correlated to measured refractive index Within range 22

23 Al 2 O 3 Post Deposition Annealing 8 0 db 7 Length of Light Streak [cm] 6 5 db 5 10 db db db 0 Optical Loss vs. Annealing Time and Temperature Optimized Annealing Process reference:3.3 db/cm reference: 8.0 db/cm 175C dep + 200C anneal (Ar) 175C dep + 175C anneal (Ar) 175C dep + 175C anneal (Ar) (cummulative anneal time) 175C dep C anneal (Ar/O2) 0 Reference 2 data (no 4 anneal) Annealing Time [hrs] Light Streak 3.33 db/cm (7cm light streak) 270 C deposition Light Streak wafer Prism Coupler wafer 8 db/cm (4-5 cm light streak) 250 C deposition Appropriate post-deposition annealing (<2hrs) results in low loss Al 2 O 3 Veeco has developed optimized annealing conditions that enable low temperature deposition of alumina with low-loss optical properties 23

24 Ta 2 O 5 Thickness and Index Repeatability Ta2O5 Thickness Repeatability Ta 2 O Å Data Thickness [A] Run # WTW Thickness (1 ) WIW Thickness (1 ) WTW Index (1 ) 0.75% (300 A/min) 0.75% 0.08% Ta2O5 Index Repeatability WIW Index (1 ) 0.24% Index 2.13 Index Optical Loss <2dB/cm Stress [MPa] -147 Run # 24

25 Al 2 O 2 Thickness and Index Repeatability Al2O3 Thickness Repeatability Alumina 1 m Data Thickness [A] Run # Ta2O5 Index Repeatability WTW Thickness (1 ) WIW Thickness (1 ) WTW Index (1 ) WIW Index (1 ) Index 0.3% (1300 A/min) 0.4% 0.03% 0.1% 1.67 Index Optical Loss <5dB/cm Stress [MPa] -126 Run # 25

26 Bi-layer Optical Loss Repeatability Al 2 O 3 1 m / Ta 2 O Å 8 7 Alumina/Tantala Bilayer Loss Cassette 1 Cassette 2 Cassette 3 Cassette 4 6 Loss [db/cm] Run # Repeatable optical loss <5dB/cm obtained with Al 2 O 3 1 m / Ta 2 O Å 12 wafers measured (total of 48 wafers run) Four cassettes run over two days 26

27 HAMR Data Summary Ta 2 O 5 process developed for low loss core layers Loss: <2 db/cm Rate: 334 Å/min U%: 0.7% 1 WTW% 0.7% 1 Al 2 O 3 for cladding layers Loss: <2dB/cm (with post deposition annealing max temp 200 C) Rate: >1000 Å/min U%: 0.4% 1 WTW%: 0.3% 1 Bi-layer Al 2 O 3 /Ta 2 O 5 films show <5dB/cm loss SiO 2 films in development 27