Compensation in boron-doped CVD diamond, Presentation Conference - PDF document

See discussions, stats, and author profiles for this publication at: https://www.researchgate.net/publication/275350418 Compensation in boron-doped CVD diamond, Presentation Conference Paper · February 2008 CITATIONS READS 0 82 3 authors , including: Jan Isberg Markus Gabrysch Uppsala University Uppsala University 163 PUBLICATIONS 3,160 CITATIONS 40 PUBLICATIONS 323 CITATIONS SEE PROFILE SEE PROFILE Some of the authors of this publication are also working on these related projects: Diamon electronics View project String and Field theory View project All content following this page was uploaded by Jan Isberg on 24 April 2015. The user has requested enhancement of the downloaded file.

Compensation in boron-doped CVD diamond Uppsala University: Markus Gabrysch Saman Majdi Jan Isberg KTH Stockholm: Anders Hallén Margareta Linnarsson Acreo AB Stockholm: Adolf Schöner Element Six Ltd: Daniel Twitchen SBDD XIII February 25 – 27, 2008 cultuurcentrum Hasselt, Hasselt, Belgium

Contents • Introduction Contents • Hall-effect measurements Introduction Hall-effect • SIMS measurements SIMS C-V of Schottky • C-V measurements on Schottky Summary diodes • Summary & Conclusion

Introduction • SC-CVD diamond is a very promising material for electronic devices due to its outstanding physical and electronical properties. Contents • BUT: For successful diamond devices the growth Introduction of doped diamond films with low concentrations of defects and residual impurities is essential. Hall-effect • Deep level impurities causes compensation SIMS effects: electrons / holes are trapped instead of C-V of Schottky emitted to the conduction / valence band. → leads to reduced free carrier concentration Summary ⇒ Compensation ratio (i.e. ratio between the dopant and the compensating defect concentration) should be kept to a minimum for many device applications

p -type doping in diamond • B is shallowest known p -type dopant for diamond • Activation energy: E A = 0.37 eV Contents • Hole concentration p in valence band of non-degenerate Introduction p -type semiconductor is the solution to the equation valence band effective DOS Hall-effect + p ( p N ) N = − D V exp( E / k T ) (1) − − A B N N p g SIMS A D a spin degeneracy factor C-V of Schottky if both acceptors N A and compensation donors N D are present, and if neglecting intrinsic carriers (T < 1000K). Summary • The solution to Eq. (1) is ′ ′ + 2 +   N N ( ) N N ′ = + − −   (2) V D V D p N N N V A D   2 2 3 / 2   π ∗ 2 2 m k T ′ =   − h B with N exp( E / k T )   V A B 2   g h a

p -type doping in diamond Example: Diamond with [B] = 10 18 cm -3 • without compensation → hole conc. at RT: 2x10 15 cm -3 Contents Introduction • with N D /N A = 1% → hole conc. drops by almost a decade Hall-effect SIMS • Nitrogen impurities cause compensation in boron- C-V of Schottky doped diamond. Summary (N donor level is 1.7eV below conduction band) • BUT: Some nitrogen is always present. ⇒ Achieving low compensation ratios in CVD diamond can be difficult.

p -type doping in diamond Examples of low compensation ratios reported for B- doped diamond in the literature Contents [B] concentration comp. ratio Introduction 5.9 × 10 16 cm -3 Fox et al . [1] 4.9 % Hall-effect 7 × 10 16 cm -3 Hatta et al . [2] 6 % SIMS 6.4 × 10 16 cm -3 1.1 % C-V of Schottky Teraji et al. [3] 1.3 × 10 18 cm -3 0.45 % Summary 1.0 × 10 17 cm -3 Yamanaka et al. [4] 0.4 % [1] B. A. Fox et al ., Diam. Rel. Mater. 4, 622 (1995) [2] A. Hatta et al ., Diam. Rel. Mater. 8, 1470 (1999) [3] T. Teraji et al ., Diam. Rel. Mater. 15, 602 (2006) [4] S. Yamanaka et al ., Diam. Rel. Mater. 9, 956 (2000)

Samples Three SC-CVD samples made by ElementSix selected: • two with very low compensation ratio (Sample 1 & 3) • one with average compensation ratio for comparison Contents (more samples were investigated but some of them showed only Introduction hopping conduction at all investigated temperatures) Hall-effect SIMS Sample growth C-V of Schottky • grown on specially prepared HPHT substrates Summary • B 2 H 6 addition to H 2 /CH 4 /Ar source gas mixture • different gas-phase boron concentration resulted in different boron concentrations in the solid • Separation from HPHT substrate by laser cutting technique and polishing

Samples Freestanding B-doped homoepitaxial (100) CVD layers Sample 2 and 3: Contents Introduction ≈ 5 × 5 × 0.5 mm 3 Hall-effect p SIMS C-V of Schottky Sample 1 is different: Summary on top: ≈ 40µm thick boron- doped layer was deposited . p i intrinsic CVD layer of high purity grown on the HPHT structure

Hall-effect measurements Sample preparation • Electrodes were formed at the four corners by ion implantation of boron (2.5 × 10 16 cm -2 at 37keV) Contents → followed by annealing (15min at 1100 ° C) Introduction • deposition of Ti/Al contacts (100/200 nm) → followed by final contact anneal (15min at Hall-effect 500 ° C) to obtain good Ohmic contact SIMS • Hall-effect measurements for 80 < T < 450K at 0.5 T C-V of Schottky using the van der Pauw configuration Summary

Hall-effect measurements hole concentration data → for low temperatures hopping conduction Contents 10 18 Introduction Sample 1 10 17 Sample 2 Hall-effect Hole Concentration (cm -3 ) Sample 3 10 16 SIMS 10 15 C-V of Schottky 10 14 Summary 10 13 10 12 10 11 10 10 1 2 3 4 5 6 7 8 1000/T (K -1 )

Hall-effect measurements valence band conduction dominates for higher T → corresponding data is shown with solid symbols Contents 10 18 Introduction Sample 1 10 17 Sample 2 Hall-effect Sample 3 Hole Concentration (cm -3 ) 10 16 SIMS 10 15 C-V of Schottky 10 14 Summary 10 13 10 12 10 11 10 10 1 2 3 4 5 6 7 8 1000/T (K -1 )

Hall-effect measurements from least-square fits to data the relevant parameters could be obtained, such as E A , N A , and N D T in °C Contents 700 200 100 30 0 -50 -100 -150 10 18 Introduction Sample 1 10 17 Sample 2 Hall-effect Sample 3 Hole Concentration (cm -3 ) 10 16 SIMS 10 15 C-V of Schottky 10 14 Summary 10 13 10 12 10 11 10 10 1 2 3 4 5 6 7 8 1000/T (K -1 )

Fit results Hall-effect measurement Contents N A (cm -3 ) N D (cm -3 ) Sample E A (eV) Introduction 0.37 ± 0.02 4.8 ± 3.7 × 10 18 7.2 ± 3.8 × 10 13 1 Hall-effect 0.34 ± 0.01 1.5 ± 0.5 × 10 18 6.0 ± 2.7 × 10 16 2 SIMS 0.36 ± 0.01 2.9 ± 0.9 × 10 18 < 2.7 × 10 14 3 C-V of Schottky Summary

Fit results Hall-effect measurement Contents Comp. N A (cm -3 ) N D (cm -3 ) Sample E A (eV) ratio Introduction 0.37 ± 0.02 4.8 ± 3.7 × 10 18 7.2 ± 3.8 × 10 13 < 10 -4 1 Hall-effect 0.34 ± 0.01 1.5 ± 0.5 × 10 18 6.0 ± 2.7 × 10 16 2 0.04 SIMS 0.36 ± 0.01 2.9 ± 0.9 × 10 18 < 2.7 × 10 14 < 10 -4 3 C-V of Schottky Summary But are these results in agreement with SIMS and C-V measurements?

SIMS measurements • Analysis of the boron depth distribution • Cameca ims 4f micro-analyzer (sputtering beam of 8.2 keV 32(O 2 ) + ions, Contents secondary 11 B + ions) Introduction 10 19 average: 3.9 x 10 19 cm -3 Hall-effect boron concentration (cm -3 ) SIMS 10 18 C-V of Schottky average: 1.5 x 10 18 cm -3 Summary 10 17 Interface between intrinsic Sample 1 and boron-doped layer Sample 3 surface bulk 10 16 0 10 20 30 40 depth (µm)

C-V measurements Capacitance-Voltage measurements for analysis of the net acceptor concentration N A − N D • Samples treated in a graphite etch Contents (HNO 3 :HClO 4 :H 2 SO 4 = 1:1:1, 180 ° C, 40min) and a mild oxygen plasma to terminate the surface Introduction • Circular Schottky gold contacts formed on one side Hall-effect and Ti/Al contacts were deposited on the other side of SIMS the samples. C-V of Schottky First: I-V analysis and only contacts with leakage • current < 5 × 10 -4 A cm -2 (at 5 V reverse bias) were Summary considered for a C-V measurement at RT. • C-V setup: EG&G 7265 lock-in amplifier at f = 1011 Hz, bias applied by a Keithley 2400 sourcemeter

C-V measurements Example: Diode on Sample 3 14 1e-1 12 current density (mA/cm 2 ) 1e-2 current density (A/cm 2 ) 10 1e-3 Contents 8 1e-4 6 1e-5 Introduction 4 1e-6 2 1e-7 Hall-effect 0 1e-8 -2 1e-9 -4 -2 0 2 4 -4 -2 0 2 4 bias (V) SIMS bias (V) C-V characteristics for a one- C-V of Schottky 5.2e+14 sided abrupt junction ( ) − 1 2 V V = 5.0e+14 bi Summary ( ) ε − 2 C q N N 4.8e+14 j S A D 1/C 2 (cm 4 /F 2 ) with C j being the depletion layer 4.6e+14 capacitance per unit area 4.4e+14 and V bi the built-in potential 4.2e+14 4.0e+14 ⇒ Obtain N A − N D from fit to data 3.8e+14 and taking the average -5 -4 -3 -2 -1 0 1 bias (V)