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

82

3 authors, including: Some of the authors of this publication are also working on these related projects: Diamon electronics View project String and Field theory View project Jan Isberg Uppsala University

163 PUBLICATIONS 3,160 CITATIONS

SEE PROFILE

Markus Gabrysch Uppsala University

40 PUBLICATIONS 323 CITATIONS

SEE PROFILE

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

SBDD XIII February 25 – 27, 2008 cultuurcentrum Hasselt, Hasselt, Belgium

Uppsala University: Markus Gabrysch Saman Majdi Jan Isberg KTH Stockholm: Anders Hallén Margareta Linnarsson Acreo AB Stockholm: Adolf Schöner Element Six Ltd: Daniel Twitchen

Contents

• Introduction
• Hall-effect measurements
• SIMS measurements
• C-V measurements on Schottky

diodes

• Summary & Conclusion

Contents Introduction Hall-effect SIMS C-V of Schottky Summary

Introduction

Contents Introduction Hall-effect SIMS C-V of Schottky Summary

• SC-CVD diamond is a very promising material for

electronic devices due to its outstanding physical and electronical properties.

• BUT: For successful diamond devices the growth
• f doped diamond films with low concentrations
• f defects and residual impurities is essential.
• Deep level impurities causes compensation

effects: electrons / holes are trapped instead of emitted to the conduction / valence band. → leads to reduced free carrier concentration ⇒ Compensation ratio (i.e. ratio between the dopant and the compensating defect concentration) should be kept to a minimum for many device applications

• Hole concentration p in valence band of non-degenerate

p-type semiconductor is the solution to the equation if both acceptors NA and compensation donors ND are present, and if neglecting intrinsic carriers (T < 1000K).

) / exp( ) ( T k E g N p N N N p p

B A a V D A D

− = − − +

p-type doping in diamond

• B is shallowest known p-type dopant for diamond
• Activation energy: EA = 0.37 eV

Contents Introduction Hall-effect SIMS C-V of Schottky Summary

(1)

with ( )

2 2

2 D V D A V D V

N N N N N N N p + ′ − − ′ +       + ′ = ) / exp( 2 2

2 / 3 2

T k E h T k m g N

B A B h a V

−         = ′

∗

π (2)

valence band effective DOS spin degeneracy factor

• The solution to Eq. (1) is

p-type doping in diamond

Example: Diamond with [B] = 1018 cm-3

• without compensation

→ hole conc. at RT: 2x1015 cm-3

• with ND/NA = 1%

→ hole conc. drops by almost a decade

Contents Introduction Hall-effect SIMS C-V of Schottky Summary

• Nitrogen impurities cause compensation in boron-

doped diamond.

(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 [B] concentration

• comp. ratio

Fox et al. [1] 5.9×1016 cm-3 4.9 % Hatta et al. [2] 7×1016 cm-3 6 % Teraji et al. [3] 6.4×1016 cm-3 1.1 % 1.3×1018 cm-3 0.45 % Yamanaka et al. [4] 1.0×1017 cm-3 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)

Contents Introduction Hall-effect SIMS C-V of Schottky Summary

Samples

Contents Introduction Hall-effect SIMS C-V of Schottky Summary

• two with very low compensation ratio (Sample 1 & 3)
• one with average compensation ratio for comparison

Three SC-CVD samples made by ElementSix selected: Sample growth

• grown on specially prepared HPHT substrates
• B2H6 addition to H2/CH4/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

(more samples were investigated but some of them showed only hopping conduction at all investigated temperatures)

Sample 2 and 3:

Samples

Contents Introduction Hall-effect SIMS C-V of Schottky Summary

p i

intrinsic CVD layer of high purity grown on the HPHT structure

p

Sample 1 is different:

• n top: ≈40µm thick boron-

doped layer was deposited.

Freestanding B-doped homoepitaxial (100) CVD layers

≈ 5×5×0.5 mm3

Hall-effect measurements

Sample preparation

• Electrodes were formed at the four corners by ion

implantation of boron (2.5×1016 cm-2 at 37keV) → followed by annealing (15min at 1100°C)

• deposition of Ti/Al contacts (100/200 nm)

→ followed by final contact anneal (15min at 500°C) to obtain good Ohmic contact

• Hall-effect measurements for 80 < T < 450K at 0.5 T

using the van der Pauw configuration

Contents Introduction Hall-effect SIMS C-V of Schottky Summary

Hall-effect measurements

hole concentration data → for low temperatures hopping conduction

Contents Introduction Hall-effect SIMS C-V of Schottky Summary

1000/T (K-1)

1 2 3 4 5 6 7 8

Hole Concentration (cm-3)

1010 1011 1012 1013 1014 1015 1016 1017 1018 Sample 1 Sample 2 Sample 3

Hall-effect measurements

valence band conduction dominates for higher T → corresponding data is shown with solid symbols

Contents Introduction Hall-effect SIMS C-V of Schottky Summary

1000/T (K-1)

1 2 3 4 5 6 7 8

Hole Concentration (cm-3)

1010 1011 1012 1013 1014 1015 1016 1017 1018 Sample 1 Sample 2 Sample 3

Hall-effect measurements

Contents Introduction Hall-effect SIMS C-V of Schottky Summary

T in °C

700 200 100 30

• 50
• 100
• 150

Hole Concentration (cm-3)

1010 1011 1012 1013 1014 1015 1016 1017 1018

1000/T (K-1)

1 2 3 4 5 6 7 8 Sample 1 Sample 2 Sample 3

from least-square fits to data the relevant parameters could be obtained, such as EA, NA, and ND

Fit results

Contents Introduction Hall-effect SIMS C-V of Schottky Summary

Hall-effect measurement Sample EA (eV) NA (cm-3) ND (cm-3) 1 0.37 ± 0.02 4.8 ± 3.7×1018 7.2 ± 3.8×1013 2 0.34 ± 0.01 1.5 ± 0.5×1018 6.0 ± 2.7×1016 3 0.36 ± 0.01 2.9 ± 0.9×1018 < 2.7×1014

Fit results

Contents Introduction Hall-effect SIMS C-V of Schottky Summary

Hall-effect measurement Sample EA (eV) NA (cm-3) ND (cm-3) Comp. ratio 1 0.37 ± 0.02 4.8 ± 3.7×1018 7.2 ± 3.8×1013 < 10-4 2 0.34 ± 0.01 1.5 ± 0.5×1018 6.0 ± 2.7×1016 0.04 3 0.36 ± 0.01 2.9 ± 0.9×1018 < 2.7×1014 < 10-4

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(O2)+ ions, secondary 11B+ ions)

Contents Introduction Hall-effect SIMS C-V of Schottky Summary

depth (µm)

10 20 30 40

boron concentration (cm-3)

1016 1017 1018 1019 Sample 1 Sample 3

average: 3.9 x 1019 cm-3

surface bulk

average: 1.5 x 1018 cm-3

Interface between intrinsic and boron-doped layer

C-V measurements

Capacitance-Voltage measurements for analysis of the net acceptor concentration NA−ND

• Samples treated in a graphite etch

(HNO3:HClO4:H2SO4 = 1:1:1, 180°C, 40min) and a mild

• xygen plasma to terminate the surface
• Circular Schottky gold contacts formed on one side

and Ti/Al contacts were deposited on the other side of the samples.

Contents Introduction Hall-effect SIMS C-V of Schottky Summary

• First: I-V analysis and only contacts with leakage

current < 5×10-4 A cm-2 (at 5 V reverse bias) were 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 characteristics for a one- sided abrupt junction with Cj being the depletion layer capacitance per unit area and Vbi the built-in potential

C-V measurements

bias (V)

• 4
• 2

2 4

current density (mA/cm2)

• 2

2 4 6 8 10 12 14

bias (V)

• 4
• 2

2 4

current density (A/cm2)

1e-9 1e-8 1e-7 1e-6 1e-5 1e-4 1e-3 1e-2 1e-1

Example: Diode on Sample 3

Contents Introduction Hall-effect SIMS C-V of Schottky Summary

bias (V)

• 5
• 4
• 3
• 2
• 1

1

1/C2 (cm4/F2)

3.8e+14 4.0e+14 4.2e+14 4.4e+14 4.6e+14 4.8e+14 5.0e+14 5.2e+14

( ) ( )

D A S bi j

N N q V V C − − = ε 2 1

2

⇒ Obtain NA−ND from fit to data and taking the average

Results

• SIMS result by a factor of 2 higher than C-V result

→ 50% might be passivated by some mechanism

Contents Introduction Hall-effect SIMS C-V of Schottky Summary

Hall-effect measurement C-V SIMS Sample EA (eV) NA (cm-3) ND (cm-3) Comp. ratio NA–ND (cm-3) [B] (cm-3) 1 0.37±0.02 4.8±3.7×1018 7.2±3.8×1013 < 10-4 7±1×1017 1.5±0.3×1018 2 0.34±0.01 1.5±0.5×1018 6.0±2.7×1016 0.04 2.4±0.3×1018 — 3 0.36±0.01 2.9±0.9×1018 < 2.7×1014 < 10-4 1.8±0.2×1018 3.9±0.8×1018

• EA < 0.37eV and depends on boron incorporation
• NA values from C-V and Hall seem to be consistent

→ for Sample 1 less good agreement, most likely due to underestimation of thickness of active volume (only B-layer was taken into account)

Summary

• Compensation ratio in three samples of

boron doped Single-Crystal CVD diamond determined by fits to Hall-effect data.

• Results found to be in agreement with

SIMS and C-V measurements on Schottky diodes.

• Very low compensation ratios below 10-4

have been observed in two samples.

Contents Introduction Hall-effect SIMS C-V of Schottky Summary

Thank you!

Contents Introduction Hall-effect SIMS C-V of Schottky Summary

View publication stats View publication stats