Transcript AlN
Progress report on 'Smart' Die Coatings for Al Pressure Die Casting J.J Moore, M. Hasheminiasari, J. Lin Advanced Coatings and Surface Engineering Laboratory, Colorado School of Mines, Golden, Colorado P.Ried Ried and Associates, LLC, Portage, Michigan Outline Research background. Progress in the AlN piezoelectric thin film deposition using P-CFUBMS Progress in the AlN piezoelectric thin film piezoelectric property measurement Why a ‘smart’ die coating? Typical failure process: increased strain followed by release of strain on cracking Need for real time detection/monitoring of strain in coating system prior to catastrophic failure Need to detect incipient failure at elevated temp & harsh environment, e.g., oxidation, corrosion, etc. In-service detection of incipient failure, e.g., cracking of coating before catastrophic failure of die coating occurs Optimized ‘smart’ tribological coating system working layer 3 2 1 Optimized Tribological coating Compositionally graded Superlattice ‘Smart’ layer Thin film electrodes /adhesion layer Piezoelectric layer Die substrate (Ferritic Nitrocarburized) Graded CrAlN CrN/AlN Graded CrN Graded CrN Cr Cr Technical examples of the optimized tribological layers How a ‘smart’ die coating works Piezoelectricty: Electronic Response to Mechanical Stress or Mechanical Response to Electrical stimulation The voltage generate in piezoelectric materials is proportional to magnitude of applied mechanical stress. The proportionality constant is given by the piezoelectric constant, ε 31 , as : V =E.d = ε3,1.σ1,2 (E= electric field, d= film thickness, σ1,2= stress in x, y direction, 3 is in z direction) F - + ++ + The actual sensor element might be very small in size. For simplicity, we have assumed a continuous thin film, but we can also use a more complex thin film sensor network using photolithography. The selection of the piezoelectric materials - High piezoelectric constant/response: Materials with permanent polarisation (non-centro-symmetric crystal): ferroelectrics (e.g. PbTiO3, Bi4Ti3012, LiNbO3), & non-ferroelectric wurtzite nitrides (e.g. AlN, GaN) P Requires a c-axis oriented film P -High curie temperature TC Stable at high temperature - Low dielectric constant to avoid electric charge breakdown T Why ‘AlN’ candidate? Properties Current response (31,f) [C/m2] Voltage response (31,f /o33) [GV/m1] Coupling Coefficient (Kp,f)2 on Si Curie Temperature (Tc) [C] CTE () [10-6 K-1] Pb(ZrxTi1-x)O3 [PZT] -14.7 -1.2 0.2 ~300 7.2 LiNbO3 -5.8 N/A 0.02 1210 11 AlN -1.0 -10.3 0.11 ~1100 4 ZnO -0.7 -7.2 0.06 N/A 5 • AlN and LiNbO3 are more suitable than traditional PZT for ‘smart’ die coatings because of their high electrically stable Curie temperatures; • AlN exhibits low dielectric constant (dielectric constant K33 of 12) and high resistivity (1016 cm) which lead to its high resistance to the electrical breakdown; • AlN films exhibit good structural and bonding compatibility to the overall tribological coating system which is based on nitrides. Deposition systems (P-CFUBMS) • Deposition system: Pulsed closed field unbalanced magnetron sputtering (P-CFUBMS); The Cr and Al targets were used for sputtering Cr electrode/adhesion layer and the AlN piezoelectric layer; The Al target was pulsed at 100-300 kHz; The Al target was powered at 1000 W; the working pressure is 3 mTorr , 100% N2 flow. How to obtain the maximum piezoelectric effects in AlN thin films Piezoelectric materials have two crystalline configurations: Generate voltage or electricity under stressinduced structural distortion Changing the shape of the material under voltage or electricity AlN in the bulk form exhibits no piezoelectric activity. However, when properly oriented on a compatible substrate, AIN thin films exhibit piezoelectric properties. Theoretically, a strong c-axis growth in the AlN, which is related to a (002) preferred orientation, is recognized to exhibit potential good piezoelectric response. Can be optimized by controlling the ion energy and ion flux in the plasma dcMS: effect of TiN layer on ε3,1 Sample no. Architecture 7 AlN/TiN/Ti/Si Thickness Processing parameter Piezoelectric coeff. (C/m2) 900 DC power: on TiN/Ti/Si substrate -0.90 5 AlN/Ti/Si 700 nm DC power: on Ti/Si substrate -0.74 6 AlN/Pt/Ti/Si Not measured DC power: on Pt/Ti/Si substrate -0.49 TiN interlayer (diffusion barrier) produce best piezoelectric effect: probably due to its best lattice matching with AlN, leading to reduced leakage current. 3500 3000 AlN/Ti/SS AlN/SS Ti/SS AlN <101> AlN/TiN/Ti/Si, dcMS unbiased substrate AlN <002> 4000 Si 1000 SS Ti <002> TiNx ? 1500 SS TiN Ti 2000 TiNx ? AlN Intensity 2500 500 0 30 40 50 2 60 70 80 dcMS: effect of TiN & bias on ε3,1 when using Ti as electrode/adhesion layer Sample no. Architecture 1 AlN/TiN/Ti/Si (TiN layer prevents Ti diffusion) 900 nm 2 AlN/TiN/Ti/Si 500 nm 3 4 AlN/Ti/Si AlN/Ti/Si Thickness 700 nm 470 nm Processing parameter Piezoelectric coefficient (C/m2) -0.90 -50 V bias (introduces inplane residual stress)/150 (Reported value: W power 1.05 C/m2) Same sample, after annealing No substrate bias (less residual stress) No substrate bias (less residual stress) Insertion of TiN layer enhanced piezoelectric constant Effect of increased ion energy (bias) Effect of annealing -0.87 -0.74 -0.14 Effect of thickness: (decrease in conduction leads to higher piezoelectric signal in thick samples) I-V characteristics: thickness dependence, dcMS 0.01 + + + + + + + + + + + 2 Current density (A/cm ) Films thinner than 0.4μm show considerably lower resistance. The resistance of a film with 0.9μm is 105 times higher. Implies that the charge leakage could be surface roughness induced. The other reason for the high leakage could be related to porosity & defects in the film. 1E-3 1E-4 Thick film (~0.90 micron) Thinner film (~0.35 mcron) 1E-5 1E-6 Bre 1E-7 1E-8 High field region - - - - - - - - ------cavities Rough surface ++ + + + + + + + + + + ++ + High field region - - - - - - - - - - - 0.1 1 V (Volts) 10 o akd wn Effects of the pulsing frequency on the preferred (002) orientation of AlN/Cr films Measured by XRD: Rocking curve: 25000 20000 0.98 0.96 Intensity (002) Orientation Percentage 1.00 0.94 15000 200 kHz 150 kHz 350 kHz 250 kHz 300 kHz 0 kHz 10000 0.92 3mTorr/floating bias 3mTorr/-50V 0.90 5000 0.88 0 0 50 100 150 200 250 300 350 400 6 8 10 12 14 16 18 20 22 Theta Pulse Frequency [kHz] Optimized pulsing regimes of 100-200kHz with 1μs reverse time were identified to achieve high degree of (002) orientation in the AlN/Cr films. 24 AlN films: P-CFUBMS – effect of frequency pulsed biased substrate: 100 kHz, 2 S, - 50V AlN TiN Ti Si substrate Sample Architecture Thickness no. Processing parameter Piezoelectric coeff. (C/m2) 5 AlN/TiN/Ti/Si 800 nm 100 Hz/1microSec/1kW/closed field/3mT/N2//0 bias/3hr -0.69 6 AlN/TiN/Ti/Si 800 nm 200 kHz/1microSec/1kW/closed field/3mT/N2/0 bias/3hr -0.82 800 nm 300 kHz/1microSec/1kW/close field/3mT/N2/0bias/3hr -0.92 7 AlN/TiN/Ti/Si Increased ion energy Why use pulsed magnetron sputtering (PMS)? Pulsing the Al target can effectively reduce the arcing problems generated during the AlN film deposition. It was found that the pulsed plasma in the middle frequency range exhibits much higher ion energy and ion flux than in the dc discharged plasma. Technical example: CrAlN coating deposition. Ion energy distribution C B 29 + N2 Intensity [Counts/Sec] 350/1.4 (51%) 350/1.0 (65%) 350/0.4 (86%) 100/5.0 (50%) 100/2.5 (75%) 2.4x10 7 2.0x10 7 1.6x10 7 1.2x10 7 8.0x10 6 4.0x10 6 29 Ion Flux [Counts/Sec] A Ion flux change 36 N2 + + Ar 27 + Al 52 + Cr 0.0 100/1.0 (90%) 0 20 40 60 80 100 120 Ion Energy [eV] 140 100/1.0 100/2.5 100/5.0 350/0.4 350/1.0 350/1.4 160 Pulsing Parameters Microstructure of AlN coatings deposited under various pulsing frequencies Working pressure=3 mTorr, substrate bias=floating bias 0kHz-dc Non-columnar structure 150kHz 200kHz 250kHz 300kHz Thermal Stability of AlN: P-CFUBMS – 100kHz, 1μs reverse time, 5mTorr 0.40 0.25 0.4 0.4 DSC First derivative Heat flow [mW/mg] 0.3 0.3 0.20 0.30 0.15 o 0.2 1322 C 0.10 0.1 0.2 0.05 0.0 0.1 0.0 -0.1 0.00 -0.2 -0.05 1300 1310 1320 0.35 1330 1340 0.25 0.20 0.15 0.10 1350 0.05 -0.1 0.00 -0.2 exo 0 200 First derivative [mW/mg/min] 0.5 -0.05 400 600 800 1000 1200 1400 Temperature [ C] o DSC: no apparent phase change or reaction below about 1318°C. Thermal Stability of AlN: P-CFUBMS – 100kHz, 1μs reverse time, 5mTorr After DSC test Intensity AlN (002) XRD analysis confirmed: No phase changes of AlN coatings after 1400 oC DSC test The diffraction peaks of Cr adhesion layer and Si substrate disappeared after 1400 oC DSC test due to the formation of CrSi. As-deposited Si Si Cr 30 35 40 45 50 55 60 Diffraction angle [2-Theta] 65 Thermal Stability of AlN: P-CFUBMS – 100kHz, 1μs reverse time, 5mTorr As-deposited Top-view After 1400 oC Cross-sectional SEM analysis confirmed: No obvious grain growth of AlN coatings after the annealing test at 1400 oC in Ar. The columnar structure of AlN coatings was retained well after the annealing test at 1400 oC in Ar. Piezoelectric characterization using Capacitance measurements A-ΔA A No voltage applied C0 d+Δd d voltage applied Cv A ) A Cv C0 d (1 ) d (1 0 A d 0 ( A A) d d Cr 1 A A ( ) C x Cv d r , x 33 , Cr C0 d31 d d ( 1 Cr ) Cr x 1 Cr Electrodes (before voltage applied) (after voltage applied) d33= Piezo strain coeff. AlN Substrate Impedance Analyzer Piezoelectric Measurements Results (5mtorr, 60%Nitrogen) 9.0E-07 7 10 V 8.0E-07 Charge Constant d33 (pm/v) 6 Capacitance (F) 7.0E-07 6.0E-07 5.0E-07 5V 4.0E-07 3.0E-07 0V 2.0E-07 5 4 3 2 1 1.0E-07 0.0E+00 1000 0 0.E+0 10000 100000 Frequency (Hz) 1000000 2.E+5 4.E+5 6.E+5 Frequency (Hz) 8.E+5 1.E+6 New Setup for Direct Piezoelectric Measurements Function Generator Actuato r AlN Si Substrate Cr Electrodes Lock-in Amplifier Results (Direct Piezoelectric Measurements) 80% Nitrogen (3.0 mtorr) 20% Nitrogen (5.0 mtorr) 3.5 12 CFAlN 6 (5mtorr, 20%Nitrogen) CFAlN 4( 5mtorr, 80%Nitrogen) 3 10 R² = 0.9996 Piezo-response (mV) Piezo-response (mV) R² = 0.9994 2.5 2 1.5 1 8 6 4 2 0.5 0 0 0 5 10 15 20 25 Applied Force (N) 30 35 0 5 10 15 20 25 Applied Force (N) 30 35 CHALLENGES • Measurement of high temperature thin film piezoelectric properties • Incorporating complex piezoelectric thin film network into overall tribological coating system using photolithography • In-situ monitoring of piezoelectric response under service conditions Conclusions • Films of single phase aluminum nitride with (002) orientation were successfully deposited by DC and pulsed DC reactive magnetron sputtering. • Pulsed DC sputtered films yielded the phase formation at much lower sputtering power and contained in-plane stress. • There was a charge breakdown problem associated with the films, which was minimized at higher thickness ≥ 900 nm. • Optimum AlN/Cr film was identified by degree of (002) orientation and rocking curve. At present, the optimum parameters are a pressure of 5mTorr, and pulsing regime of 100-200kHz with 1μs reverse time • High (002) orientation growth was obtained from P-CFUBMS system. Proper Cr or TiN/Ti underlayer favored high (002) hexagonal AlN orientation growth Future Work • Optimize P-CFUBMS deposition process correlated with plasma properties of ion energy & ion flux using Hiden EQP plasma diagnostics that achieve high temperature piezoelectric properties, minimum residual stress • Improve the piezoelectric measurement techniques • Incorporate piezoelectric thin film sensors into one of ACSEL’s optimized multi-functional tribological coatings and evaluate ‘smart’ capabilities. • Prepare the Li-Nb composite target and reactively sputter this target to produce LiNbO3 piezoelectric thin films & compare with AlN with respect to deposition & piezoelectric property/performance.