GaN/AlGaN DEVICES

SOLAR BLIND ALGaN AVALANCHE PHOTODIODES

  For many applications, where very low levels of light are to be detected, it is desirable to use a detector with a large sensitivity. This means that there is   a need for large optical gain. Large gains can be obtained in an avalanche photodiode (APD). The device is essentially a reverse biased photodiode   operated at voltages close to the breakdown voltage. The recent developments in GaN/AlGaN material growth technology have led to the fabrication of   high performance photodetectors operating in the UV spectral region. AlGaN based Schottky [1,2], p-i-n [3,4] and MSM [5] photodetectors with   excellent detectivity performances have been reported. However, very few GaN based avalanche photodiodes (APDs) were reported in the literature   [6], and there are not any publications reporting AlGaN based APDs. In this work, I made the design, fabrication and characterization of AlGaN based   APDs operating in the solar-blind spectral region.

  The samples were fabricated by using a five-step microwave-compatible fabrication process in a class-100 clean room environment. First, the ohmic   contact regions were defined via reactive ion etching (RIE) under CCl2F2 plasma, a 20 sccm gas flow rate, and 100W RF power. After an ohmic etch   of ~1.2 μm, Ti/Al (100 Å /1000Å) contacts were deposited via thermal evaporation and left in an acetone solution for the lift-off process. The contacts   were annealed at 650oC for 60 s in a rapid thermal annealing system. A ~100 Å thick Au film was evaporated in order to form Au/AlGaN Schottky   contacts. Mesa structures of the devices were formed via the same RIE process, by etching all of the layers (>2.1μm) down to the sapphire substrate   layer for better mesa isolation. Then, a 200 nm thick Si3N4 was deposited via plasma enhanced chemical vapor deposition (PECVD) for passivation.   Finally, a ~ 0.6 μm thick Ti/Au interconnect metal was deposited and lifted off to connect the Schottky layers to the coplanar waveguide transmission   line pads.

Fig.1 The microscope image and SEM picture of a 30 micron AlGan APD device

   The resulting devices had breakdown voltages higher than 50V. Figure.2-a shows the dark current of a 60 μm diameter device. Dark current density   at a 5 V reverse bias was 5.3x10-11 A/cm2. Up to 10V, the dark current was less than 10fA. At 50V bias, the device had a dark current ~2nA. The   low dark current values proved the high growth quality of AlGaN wafer with low dislocation densities. The differential resistance of our detectors at zero   bias was in excess of 1.88x1015 Ω in the 0-20 V range.

Fig.2: a) Dark current of a 60 μm diameter photodetector. b) Spectral noise measurement of a high-leakage 100 μm diameter photodetector with a varying applied reverse bias

     Figure 3-a shows the current-voltage characteristics of a device with a 100 μm diameter.

   The device showed almost unity-gain behavior for voltages between 10 and 50 V, and the photocurrent was approximately 10 nA. After 50 volts,   avalanche gain has been observed from the device. As shown in Figure 3-b, the maximum reproducible avalanche gain was 25 at 72V reverse bias. The   onset avalanche gain field was estimated to be approximately 1MV/cm. Although we had observed avalanche gains as high as 200 (at ~85V reverse   bias) from some of the devices, the performance of these devices degraded rapidly after the measurements.

   Figure 3-c shows the quantum efficiency measurements of a 100 μm diameter device for different bias voltages. Under a 25 V reverse bias voltage,   the device had a maximum quantum efficiency of 55% at 254 nm. The cut-off wavelength was ~270 nm for all measurements. Figure 3-d shows the   responsivity measurements of the same photodetector. The peak responsivity was 0.11 A/W at a wavelength of 254nm. The device had a rejection ratio   of more than four orders of magnitude with wavelengths larger than 362 nm at a 25 V reverse bias. Using the thermal-noise limited detectivity (D*)   formula D*=Rλ(R0A/4kT)1/2, where Rλ is the device responsivity at zero bias, R0 is the zero volt dark impedance and A is the detector area, D* is   found as 4.68x1013 cmHz1/2/W which is higher than our earlier results(2). The corresponding noise equivalent power (NEP) is 1.89x10-16 W/Hz1/2.

     Fig.3: a) Dark current and photocurrent measurement of a 100 micron diameter photodetector b) Corresponding avalanche gain of the same device. c)   Quantum efficiency measurements of a 100 μm diameter photodetector. d) Responsivity measurement of the same device in semilog scale.

  Finally, the noise characterization of the solar-blind detectors was carried out using a fast Fourier transform spectrum analyzer, current amplifier, DC   voltage source, and a microwave probe station. In the 1Hz-10 kHz range, our low-leakage, high breakdown voltage solar-blind photodetectors had   noise power densities below the resolution level of the instrument. Even under 30 V, the detector noise did not exceed the measurement setup noise floor   of 3x10-29 A2/Hz at 10 kHz. Therefore, we measured devices with higher leakage currents in order to observe the bias dependence of the spectral   noise density. We took a 100 μm diameter device which has higher dark current (5x10-7 A at 35 V) and a lower breakdown voltage near 50V. As can   be seen in Fig.2-b, 1/f (flicker) noise is the dominant noise mechanism in our detectors. 1/f noise is known to result from contamination and crystal   imperfection. Up to 30V reverse bias voltage, the noise power density is nearly the same. At a 0V bias voltage and 10Hz, Sn(f) is ~9.8x10-21 A2/Hz,    at 30V bias it is only ~3x10-20 A2/Hz  and at 50 V it is ~ 1.4x10-16 A2/Hz . The noise curves obey the Sn=S0/fb relation. S0 depends on current,   which is why the noise increases with an applied bias voltage in low frequencies.

  In summary, we fabricated the first solar blind AlGaN based avalanche photodiodes with low dark current, low noise, and high detectivity. The   photocurrent-voltage characteristics indicated a reproducible gain of > 25 at 72 V. The devices exhibited a maximum quantum efficiency of 55% and a   peak responsivity of 0.11 A/W at a wavelength of 254 nm.

     References

  [1]     A. Osinsky, S. Gangopadhyay, B. W. Lim, M. Z. Anwar, M. A. Khan, D. V.

     Kuksenkov and H. Temkin, Appl. Phys. Lett. 72, 742 (1998).

  [2]     T. Tut, N. Biyikli, I. Kimukin, T. Kartaloglu, O. Aytur, M. S. Unlu and E. Ozbay, Solid State Electronics. 49, 117 (2005).

  [3]     U. Chowdhury, M. M. Wong, C. J. Collins, B. Yang, J. C. Denyszyn, J. C.  Campbell and D. Dupuis, J. Cryst. Growth. 248, 552 (2003).

  [4]     N. Biyikli, I. Kimukin, O. Aytur, E. Ozbay, IEEE Photon. Technol. Lett. 16, 1718 (2004).

  [5]     T. Li, D. J. H. Lambert, A. L. Beck, C. J. Collins, B. Yang, J. M. M. Wong, U. Chowdhury, R.D. Dupuis and J. C. Campbell, Electron. Lett.   36, 1581 (2000).

  [6]     J. C. Carrano, D. J. H. Lambert, C. J. Eiting, C. J. Collins, T. Li, S. Wang, B. Yang, A. L. Beck, R. D. Dupuis, and J. C. Campbell, Appl. Phys.   Lett. 76, 924 (2000).

   Nanophotonic Light Sources at UV Wavelengths

 

   MOCVD Growth of High-Al Content P-Type AlGaN on AlN/Sapphire Template

 

   Solar-Blind AlGaN Avalanche Photodiodes

 

   Low Dark-Current AlGaN MSM Photodiotes

 

   GaN Based Nanotransistors for High Power Applications

 

   Fabrication and Characterization of GaN Based LEDs

Metamaterials   -    Photonic Crystals  -  GaN/AlGaN Devices  -  Other Semiconductor Devices and Fabrication Techniques

© Copyright - com.web.tr. This page is best viewed with IE v.5.0 or higher, at minimum 1024x768 resolution.