Skip to content

Excess noise in high-current diamond diodes

MetadataDetails
Publication Date2022-02-07
JournalApplied Physics Letters
AuthorsSubhajit Ghosh, Harshad Surdi, Fariborz Kargar, Franz A. Koeck, Sergey Rumyantsev
InstitutionsPolish Academy of Sciences, Arizona State University
Citations21
AnalysisFull AI Review Included

This study investigates low-frequency excess noise in high-current diamond p++-i-n diodes to establish a baseline for noise spectroscopy-based reliability assessment in high-power diamond electronics.

  • Noise Mechanism Identification: The electronic excess noise is dominated by generation-recombination (G-R) noise, appearing as Lorentzian features, or 1/f noise, depending on the device quality (turn-on voltage).
  • Defect Correlation: G-R noise is characteristic of diodes with lower turn-on voltages (fewer defects), while 1/f noise dominates in high turn-on voltage diodes (higher trap concentration).
  • Unique Current Dependence: The noise spectral density (SI) exhibits three distinct regions: scaling as I2 at low (<10 µA) and high (>10 mA) currents, but remaining nearly constant in the intermediate current density range (0.1 to 100 A/cm2).
  • Trap Time Constants: Characteristic trap time constants, extracted from G-R noise data, show a uniquely strong dependence on current (fc ~ Jβ), attributed to the deep, partially ionized donor/acceptor states and hole-dominated transport.
  • Thermal Performance: Diode performance improves with increasing temperature; the ideality factor (n) decreases, and the noise level remains almost constant, which is highly beneficial for high-power switching applications.
  • Reliability Tool Development: The large variation in noise levels (spanning three orders of magnitude at low currents) between different diodes confirms the potential of noise spectroscopy as a sensitive, non-destructive predictor for diamond diode lifetime (Mean-Time-To-Failure, MTTF).
ParameterValueUnitContext
Substrate Materialp-type <111> HPHT diamondN/AHighly B-doped (p++)
Substrate Doping (B)~2 x 1020cm-3p++ layer
i-layer Thickness~0.2µmIntrinsic layer
n-layer Thickness~0.15µmModerately P-doped layer
n-layer Doping (P)~1018cm-3P-doped layer
NanoC Layer Thickness~0.1µmN-doped, near-metallic cathode contact
Low Turn-on Voltage (VT)~5VDevices showing G-R noise
High Turn-on Voltage (VT)~10 or higherVDevices showing 1/f noise
Typical Trap Energy Range0.2 to 1.7eVDefects in diamond bandgap
P Dopant Activation Energy0.43 to 0.63eVIn the n-layer
Low Current Noise ScalingSI ~ I2N/ACurrent I < 10 µA
High Current Noise ScalingSI ~ I2N/ACurrent I > 10 mA
Intermediate Current Density0.1 to 100A/cm2SI is nearly constant
Corner Frequency Exponent (β)0.31, 1.15, 1.39, 1.35N/Afc ~ Jβ for devices 1, 2, 3
Measurement Temperature Range296 to 400KElevated temperature testing
Fixed Noise Frequency (f)10HzUsed for SI vs J plots

The diamond diodes utilized a p++-i-n structure grown on a highly B-doped <111> single crystal diamond plate.

  1. Material Growth (PECVD):

    • i-layer Growth: Plasma Enhanced Chemical Vapor Deposition (PECVD) using H2:CH4:O2 mixture.
      • Chamber Pressure: 63 Torr.
      • Microwave Power: 1000 W.
    • n-layer Growth: PECVD using H2:CH4:TMP mixture.
      • Chamber Pressure: 60 Torr.
      • Microwave Power: 2000 W.
    • Cathode Contact Layer: A near-metallic highly conductive N-doped nano-carbon (nanoC) layer was grown on the n-layer to reduce contact resistance.
  2. Device Fabrication:

    • Active Area Definition: Defined by partially mesa etching the diamond into the i-layer.
    • Hard Mask: SiO2 hard mask.
    • Etching Chemistry: O2/SF6 chemistry in a Reactive Ion Etcher (RIE).
    • Contact Definition: UV photolithography and e-beam deposition.
    • Metal Stack (Cathode/Anode): Ti-Ni-Au metal stack with thicknesses of 50 nm - 50 nm - 300 nm.
  3. Electrical and Noise Characterization:

    • I-V and Noise Measurement: Conducted in vacuum using an Agilent/Lake Shore setup.
    • Noise Spectra Acquisition: Acquired using a dynamic signal analyzer (Stanford Research).
    • Noise Analysis: Low-frequency excess noise (SI) was analyzed as a superposition of 1/f noise and G-R noise (Lorentzian features).
    • Trap Time Constant Extraction: G-R noise was fitted using the Lorentzian expression SI(f) = S0/[1 + (2Ļ€fĻ„)2] to determine the characteristic corner frequency (fc = 1/(2πτ)).

The research focuses on developing reliable, high-performance diamond devices, leveraging diamond’s superior material properties for demanding electrical applications.

  • High-Power Electronics and Switches: Diamond diodes are targeted specifically for high-current switching applications where high critical electric field and thermal conductivity are essential.
  • Electricity Grid Infrastructure: Applications in power converters and inverters requiring highly reliable, ultra-wide bandgap (UWBG) semiconductors to meet increasing efficiency demands.
  • Device Reliability Assessment: Implementation of noise spectroscopy as a non-destructive, short-time measurement technique to predict device Mean-Time-To-Failure (MTTF) and assess material quality.
  • High-Temperature Operation: Utilization of diamond’s weak dependence of current and noise on temperature, allowing for stable and improved performance in high-temperature operating environments (up to 400 K tested).
  • Advanced Semiconductor Manufacturing: Providing a noise-level baseline necessary for optimizing diamond Chemical Vapor Deposition (CVD) growth and processing techniques by correlating noise signatures (G-R vs. 1/f) with defect concentrations.
View Original Abstract

We report the results of an investigation of low-frequency excess noise in high-current diamond diodes. It was found that the electronic excess noise of the diamond diodes is dominated by the 1/f and generation-recombination noise, which reveals itself as Lorentzian spectral features (f is the frequency). The generation-recombination bulges are characteristic of diamond diodes with lower turn-on voltages. The noise spectral density dependence on forward current, I, reveals three distinctive regions in all examined devices—it scales as I2 at the low (I < 10 μA) and high (I > 10 mA) currents and, rather unusually, remains nearly constant at the intermediate current range. The characteristic trap time constants, extracted from the noise data, show a uniquely strong dependence on current. Interestingly, the performance of the diamond diodes improves with the increasing temperature. The obtained results are important for the development of noise spectroscopy-based approaches for device reliability assessment for high-power diamond electronics.