correlated electronic variations

Transcription

1 Surface and bulk defects of diamond and correlated electronic variations Outline: Christoph E. Nebel Fraunhofer Institute of Applied Solid State Physics (IAF) Freiburg, Germany I. Surface Properties Morphology Termination Surface Band Bending Schottky Contact Properties II. Bulk properties Mobilities P1 and H1 Centers µτ-products Deep Defects

2 Acknowledgement Dr. Zeisel (ODLTS) WSI, Germany Dr. Rezek (AFM) AIST, Japan Dr. Yang (electrochemistry) AIST, Japan Dr. Takeuchi (TPYS) AIST, Japan Dr. Ri (wet chemical etching) NIMS, Japan Dr. Tokuda (homo-epi growth) AIST, Japan Dr. Watanabe (homo-epi. growth) AIST, Japan Dr. Misouchi (EPR) Univ. Tsukuba, Japan Dr. Gräff (EPR) Uni. Sao Paulo, Brasil

3 I.a) Surface Morphologies Atomically flat surface with step-etch growth. Terraces run parallel to the (110) direction H.Okushi, Diam. Rel. Mat. 10, 281 (2001)

4 Homo-Epitaxial Growth of CVD Diamond (AIST) H. Watanabe et al., New Diamond and Frontier Carbon Technology, 12, 369 (2002) but: extremly low growth rate (10 nm/h)

5 Surface Properties of Homo-Epi. Diamond Atomically Smooth Surface RMS = 0.8 Å

6 Truly Atomically Smooth CVD Diamond tilt angle initial substrate surface atomic steps N. Tokuda et al. sub. (111) basal plane 300 [nm] B B Homoepitaxial growth diamond sub. step-flow growth direction step-free surface epitaxial film diamond sub. step-free surface 0.5 [nm] [µm] B B cross-section 500 [nm] epitaxial film diamond sub.

7 Polycrystalline CVD Diamond: Growth side Substrate side fast growth, large area M. Nesladek

8 Diamond Surface: Poly CVD Diamond Crystallite roughness about: µm Roughness on planes: nm

9 Polished Diamond Polishing Roughness: 2-15 nm Fine structure roughness: 5 A µm

10 I.b) Diamond Surface Terminations C(100)(2x1) Hydrogen Terminated Diamond Wet-chemically or electrochemically oxidized Reconstructed Diamond Plasma-oxidized Diamond Ketone or Top Site Model Ether or Bridge Site Model

11 H-Termination after 6 h H-plasma H-termination parameter sample date d stage temperature [ C] 800 microwave power [W ] 750 total gas pressure [Torr] 25 total H2 flow [sccm] 400 total CH4 flow [sccm] 0 exposure time [min] 5 plasma off at [ C] 700

12 Defect Free: C(100)-(2x1):H H-termination: ex-situ at 800 o C, for 1h Microwave Plasma CVD Hot Filament Technique Atomically flat surface with mono-atomic steps A. Maynev, G. Dujardin, New Diamond and Frontier Carbon Technology 15 (2005) p. 265.

13 Summary H-termination If H-termination is optimized: Unpinned Surface Fermi level High quality surface with minimized defect density Schottky-Mott Law ok? Schottky Barrier Hight: qφ Βn = q(φ m - χ) But: Data in literature does not show NEA properties 1.2 ev shift

14 Electronic Surface Properties Total Photoelectron Yield (rel. units) Surface(VBM) 4.4±0.1 ev E g 800 C 10 min Photon Energy (ev) TPY=A(hν-6.784) 3.5 H-termination: negative electron affinity: -1.1 ev Maier, Ley, Ristein et al. Takeuchi et al. UHV annealed sample (800 o C): positive electron affinity: +1.2 ev

15 Band Bending of H-Terminated Diamond A A 2 Built-in Potential V bi Acceptor Density N D Doping Profiling as function of x: A 2 C=ε o ε r A/w A 2 Gives Depth Gives Acceptor Density

16 H-term. Boron doped Diamond in 1 M H 2 SO 4 Capacitance Voltage Spectroscopy 8.0x10 13 N A = 3 x10 20 cm -3 1M H 2 SO 4 Freq. 1 khz 6.0x V Built-In Potential: V [1/C 2 ] / F x x HSBD UK HSBD 8 HSBD 9 HSBD7 0.8 V 0.84 V 1.1 V Energy Barrier: ev Capacitance of Depletion Layer: 5 µf/cm 2 E / V (Ag / Ag Cl)

17 Diamond Interface To Buffer Remember Metals: Charge separation at interface: 1 to 2 Å ( Helmholtz Capacitance ): Large Capacitance! Diamond Capacitive Layer Width: 10 Å Small Capacitance! Electrolyte Background Current: I = v C

18 Energy Levels of Buffers

19 Local Oxidation (STM): C(100)-(2x1):H H-termination: ex-situ at 800 o C, for 1h a): Atomically flat surface with mono-atomic steps b) After desorption of individual H by STM tip. Circles are defects K. Bobrov, G. Dujardin, et al. Surface Science 528, p. 138 (2003)

20 Surface Degradation DNA-FET in buffer Drain-Source Current (A) 2.5x x x x x10-7 Phosphate Buffer, ph 7 U G = -0.4 V U DS = -0.6 V Significant degradation FET in Air Time (h) Amplitude decrease: 70 % (700 h) H-term. Surface: Decrease: 40 % (700 h) H-terminated diamond in air is chemically not stable

21 Wet-Chemical/ Electrochemical Oxidation Boiling in: HNO 3 :H 2 SO 4 = 1:3 at 230 o C for 1 h Termination with OH but: degree of termination (100 %)?

22 H-termination evaluated chemically Decreased to 10 % Still H-terminated!

23 Plasma Oxidation: Ketone or Top Site Model O 2 -Plasma: Ether or Bridge Site Model 300 W, 25 Torr, 5 min.

24 Reconstruction (ann o C) H-free surface, clean diamond (2x1) reconstruction with π-bondet dimers K. Bobrov, G. Dujardin, et al., Nature 413, 616 (2001)

25 Capacitance/Voltage Evaluation 2,5x10 14 N A = 5x10 20 cm -3 2,0x10 14 electrochemical Oxygen Plasma 1/C -2 (F -2 ) 1,5x ,0x10 14 reconstructed 5,0x10 13 H-term 1.65 V 2.5 V 3.7V 0, Potential (V) Frequency: 100 Hz Buffer : Phosphate (no FeCN)

26 Diamond Interface after Surface Modifications (data are not convincing?)

27 Surface Fermi Level (UPS/XPS): E = ev (111)-(2x1) reconstructed: 0.88 above E VBM J.B. Cui et al. Phys. Rev. Lett. 81 (1998) p (100)-(1x1):O 1.2 ev downward bending Murret et al, Semicond. Sci. Technol. 19 (2004) p. 1. (100)-(2x1):H Murret et al, Semicond. Sci. Technol. 19 (2004) p ev downward Bending (100)-(2x1):H 0.5 ev above E VBM Kono et al. Surf. Science 529 (2003) p.180. Surface defects

28 Real Schottky Contact Surface Fermi-level pinning by surface defects Schottky barrier is not dependent on metal.

29 Schottky Properties: Clean and Oxygen Terminated Diamond from: M. Werner, Semicond. Sci. Technol. 18 (2003) S41

30 Cleanness: SEM image of Bar-Contact structures partially wiped in ethanol

31 Cleaning by contact mode AFM Tapping Mode AFM Surface Morphology AFM Image Contact Mode AFM Cleanning Surface is covered with a thin (1-10) nm thick adhesive layer. SEM Image B. Rezek, C.E. Nebel, M. Stutzmann Diam. and Rel. Mat. 13, (2004)

32 AFM carbon deposit in air and liquids in air in water in water in isopropanol C-O C-H Au contact c-afm region ~110nN Surface layer is also attached to diamond in: Water Isopropanol not removable by ultrasonic cleaning

33 Electrostatic Attraction of Ionized Nano-Particles Accumulation of charged particles on the surface, in the water adsorbate? Due to electrostatic force, only small particles will be attached CH Dipole CO Dipole

34 Summary Surface: H-terminated Surface Technique: H-plasma, H-hot filament Unpinned Surface Fermi Level Best interface for Schottky Metal Contacts In air not stable Diamond surface oxidation: Techniques: wet-chemically (OH), electrochemically (OH), plasma (O) Not fully characterized with respect to: Degree of termination, Surface defect density and distribution Clean diamond surface Realized only in UHV at T > 800 C Missing Data about: Stability in air, Defect density

35 Roadmap Surface Optimization Polishing causes surface damage (not yet characterized!) Optimization: 1) Polishing 2) Plasma Etching 3) Homo-Epitaxial Overgrowth 4) H-Termination

36 Bulk Properties

37 Hole Mobilities T -3/2 : acoustic phonon scattering T -2.8 : optical phonon scattering 10 4 ~T -3/2 Theory Reggiani et al. Dean et al. Konorova et al. Isberg (intrinsic) AIST Boron doping Isberg et al. : time-of-flight on undoped CVD diamond (Science 297, p (2002): 3800 cm 2 /Vs) Reggiani: Time-of-flight on undoped natural diamond Dean and Konorova: Hall Mobilities Dr. Okushi et al. AIST: Hall effect on boron doped CVD diamond Hole Mobility (cm 2 /Vs) ~T Temperature (K)

38 Electron Mobilities in natural and P-doped Diamond: T -3/2 : acoustic phonon scattering Isberg et al.,: Time-of-flight in undoped CVD diamond ( Science 297, p (2002): 4500 cm 2 /Vs) Nava: Time-of-flight on natural undoped diamond Konorova: Hall effect Redfield: Hall effect Electron Mobility (cm 2 /Vs) -3/2 ~T Phosphorus doped diamond Isberg et al. Theory Nava et al. Redfield Konorova et al. Koizumi Koizumi et al.: Hall effect on Phosphorus doped diamond Temperature (K)

39 Defects: H1 Center (g = ) In homoepitaxially grown single crystalline diamond: Typical Density: 5x10 18 cm -3 Zhou et al. PRB 54 (1996) p N. Mizuochi et al. DRM. Effect on transport and recombination not clear!

40 Defect Model: X. Zhou, G.D. Watkins et al. PRB 54 (1996) p. 7881

41 Defects: P1 Center (N-Dopant, g = ) EPR of substitutional nitrogen (g=2.0024). Satellite positiondepends on magnetic field with respect to (111)-orientation

42 Dislocations lattice distortion grow from substrate into epi-layer killer of high voltage devices Minimization: soft etching of substrate befor growth, to remove distortions

43 µτ Product as Function of Morphology N-content < cm -3 µτ e Single Crystalline Diamond µτ h µτ (cm 2 /Vs) Single Crystalline CVD Factor 1000 better 10-6 Polycrystalline CVD Sample Polycrystalline Diamond J. Isberg et al., Diam. Rel. Mat. 13 (2004) 872.

44 µτ-product as Function of Nitrogen Content Nitrogen not important 10-5 N-content < cm -3 µτ h µτ (cm 2 /V) µτ e N-content: ( ) cm Polycrystalline CVD Diamond Sample J. Isberg et al., Diam. Rel. Mat. 13 (2004) 872.

45 Comparision of structural properties: P. Bergonzo et al Grain boundary effect: best SCD

46 Polycrystalline CVD Diamond

47 Polycrystalline CVD Diamond

48 Deep trapping of carriers (electrons and holes) Memory Effect 6,0x10-3 4,0x10-3 A F = 1.2x10 3 V/cm PD Current (A) 2,0x10-3 0,0 B F = 0 V/cm -2,0x10-3 0,0 5,0x10-9 1,0x10-8 1,5x10-8 2,0x10-8 Time (s)

49 Model:

50 After laser exposure the internal field is reversed:

51 Traps or defect, can be occupied by electrons and holes!

52 Deep Defect Characterization (ODLTS)

53 Carbon Implantation

54 Energy resolved defect variation:

55 Defect Annealing

56 Annealing Dynamics: Carbon Vacancy! T > 600 K

57 Carbon Implantation induced Defects:

58 Summay Bulk Minor problems from: Defects (H1) Nitrogen (low density) Major problems from: Grain boundaries (dislocations?) Defect characterization not established: V NV NiN...