Avalanche gain and breakdown voltage in most wide bandgap semiconductor materials are dependent on temperature and most instruments utilizing APDs rely on temperature stabilization or voltage compensation circuitry to maintain a constant avalanche gain. The complexity in operation circuitry can be reduced by incorporating material with inherently superior temperature stability in its avalanche gain and breakdown voltage. In state of the art APDs, the temperature dependence of avalanche breakdown voltage is quantified by the temperature coefficient of avalanche breakdown, Cbd. We report on the temporal and temperature stability of avalanche gain and breakdown voltage of 100 nm thick avalanche layers of Al0.85Ga0.15As0.56Sb0.44 (AlGaAsSb). The Cbd (1.60 mV/K) is smaller compared to state of art InP and InAlAs APDs for similar avalanche layer thickness. The temporal stability of avalanche gain for the AlGaAsSb APD was also evaluated in temperature ranges of 294 K to 353 K. The APD was biased at room temperature gain of 10 and maximum fluctuation of ±0.7% was recorded at 294 K which increases to ±1.33% when the temperature was increased to 353K. The promising temperature stability of gain indicates the potential of AlGaAsSb lattice matched to InP in achieving higher tolerance to temperature fluctuations and reduction of the operational complexity of circuitry. The dark currents are robust and do not show significant thermal degradation after gain measurements at elevated temperatures.
The usefulness of avalanche photodiodes (APDs) resides in their ability to produce internal gain via impact ionization without generating excessive noise. This process is stochastic and the gain values fluctuate around a mean value, giving rise to the so-called excess noise. In this work, we evaluate the gain fluctuations in APDs using a multi-channel analyzer (MCA). Two Al0.85Ga0.15As0.56Sb 0.44 APDs, one p-i-n and one n-i-p were used. Illuminated with a pulsed light source, the APDs were connected to a charge-sensitive amplifier, counting the number of charges created by each avalanche event initiated by the light pulse. The signal was subsequently sent to an MCA, recording the gain values and outputting a gain spectrum. Both APDs were investigated for mean gains up to ~ 9. For a given mean gain, the gain distribution for the n-i-p diode was found to be significantly broader than for the p-i-n diode, as expected from the excess noise values previously measured in those devices. The coefficient of variance (CoV), defined as the ratio of standard deviation to mean value of the gain peaks, was found to be low for the p-i-n APD, consistent with the low excess noise values in this material. For higher mean gain values, the CoV of the n-i-p APD gave higher values than for the p-i-n APD, again corroborating the conventional excess noise measurements.
InAs avalanche photodiodes (APDs) can be designed such that only electrons are allowed to initiate impact ionization, leading to the lowest possible excess noise factor. Optimization of wet chemical etching and surface passivation produced mesa APDs with bulk dominated dark current and responsivity that are comparable and higher, respectively, than a commercial InAs detector. Our InAs electron-APDs also show high stability with fluctuation of ~0.1% when operated at a gain of 11.2 over 60 s. These InAs APDs can detect very weak signal down to ~35 photons per pulse. Fabrication of planar InAs by Be implantation produced planar APDs with bulk dominated dark current. Annealing at 550 °C was necessary to remove implantation damage and to activate Be dopants. Due to minimal diffusion of Be, thick depletion of 8 μm was achieved. Since the avalanche gain increases exponentially with the thickness of avalanche region, our planar APD achieved high gain > 300 at 200 K. Our work suggest that both mesa and planar InAs APDs can exhibit high gain. When combined with a suitable preamplifier, single photon detection using InAs electron-APDs could be achieved.
Sensitive detection of mid-infrared light (2 to 5 μm wavelengths) is crucial to a wide range of applications. Many of the applications require high-sensitivity photodiodes, or even avalanche photodiodes (APDs), with the latter generally accepted as more desirable to provide higher sensitivity when the optical signal is very weak. Using the semiconductor InAs, whose bandgap is 0.35 eV at room temperature (corresponding to a cut-off wavelength of 3.5 μm), Sheffield has developed high-sensitivity APDs for mid-infrared detection for one such application, satellite-based greenhouse gases monitoring at 2.0 μm wavelength. With responsivity of 1.36 A/W at unity gain at 2.0 μm wavelength (84 % quantum efficiency), increasing to 13.6 A/W (avalanche gain of 10) at -10V, our InAs APDs meet most of the key requirements from the greenhouse gas monitoring application, when cooled to 180 K. In the past few years, efforts were also made to develop planar InAs APDs, which are expected to offer greater robustness and manufacturability than mesa APDs previously employed. Planar InAs photodiodes are reported with reasonable responsivity (0.45 A/W for 1550 nm wavelength) and planar InAs APDs exhibited avalanche gain as high as 330 at 200 K. These developments indicate that InAs photodiodes and APDs are maturing, gradually realising their potential indicated by early demonstrations which were first reported nearly a decade ago.
We demonstrate an AlInP detector grown on lattice-matched GaAs substrate for underwater communication applications.
This detector has a narrow inherent spectral response of 22 nm with central wavelength at ~ 480 nm and is capable of
having avalanche gain of ~ 20 which gives peak responsivity of ~ 2 A/W. A much higher multiplication of ~167 was
shown in the previous work. The full-width-half-maximum (FWHM) and responsivity of this detector is fairly
insensitive to the angle of the incident light. These properties enable it to detect an optical signal at 480 nm even in the
presence of high background illumination.
We report on the evaluation of InAs photodiodes and their potential for low temperature sensing. InAs n-i-p photodiodes were grown and analyzed in this work. Radiation thermometry measurements were performed at reference blackbody temperatures of 37 to 80°C to determine photocurrent and temperature error. The uncooled InAs photodiodes, with a cutoff wavelength of 3.55 μm, detect a target temperature above 37°C with a temperature error of less than 0.46°C. When the photodiode was cooled to 200 K, the temperature error at 37°C improves by 10 times from 0.46 to 0.048°C, suggesting the potential of using InAs for human temperature sensing.
Al0.52In0.48P is the largest bandgap material in III-V non-nitride semiconductors that is lattice matched to a readily available substrate (GaAs). Having a bandgap narrower than that of GaN enables it to detect wavelengths around 480 nm. Such wavelengths have the best transmittance underwater and may be used as a carrier in underwater communication systems. We present an Al0.52In0.48P homo-junction Separate-Absorption-Multiplication-Avalanche-Photodiode (SAMAPD) as a high sensitivity detector for such an application. By increasing the neutral and space-charge region thicknesses, the peak response wavelength can be tuned to longer wavelengths with a narrower full-width-half-maximum (FWHM). The quantum efficiency of the detector reduces with FWHM and this is compensated by having an avalanche gain. At room temperature, the SAM-APD has a dark current of <20 pA for a 210 μm radius device up to 99.9% of breakdown voltage. The structure gives a narrow spectral FWHM of 22 nm with centre wavelength of 482 nm. An external quantum efficiency of 33% and 6410% at 482 nm is obtained at bias voltage of -19 V and -92.6 V respectively.
The dilute nitride GaInNAs(Sb) alloy system is challenging to grow and defects can cause short diffusion lengths and
high background doping densities. Despite these difficulties, extremely high cell efficiencies have recently been
achieved in multi-junction solar cells utilising 1 eV GaInNAs absorber layers. This study aims to highlight the tradeoffs
between the electrical and optical characteristics related to the performance of GaInNAs(Sb) diode structures
grown by molecular beam epitaxy , with band gaps ranging from 0.90 to 1.04 eV. Post-growth annealing was necessary
in some instances to reduce the background doping and dark current densities. The incorporation of Sb into GaInNAs
has enabled the possibility of producing a dilute nitride cell with a band gap lower than 0.80 eV, although with an
increased dark current.
InGaAsN is a promising material system to enable low-cost GaAs-based detectors to operate in the telecommunication
spectrum, despite the problems posed by the low growth temperature required for nitrogen incorporation. We
demonstrate that InGaAsN p+-i-n+ structures with nominal In and N fraction of 10% and 3.8%, grown by molecular
beam epitaxy (MBE) under non-optimal growth conditions, can be optimized by post growth thermal annealing to match
the performance of optimally grown structures. We report the findings of an annealing study by comparing the
photoluminescence spectra, dark current and background concentration of the as-grown and annealed samples. The dark
current of the optimally annealed sample is approximately 2 μA/cm2 at an electric field of 100 kV/cm, and is the lowest
reported to date for InGaAsN photodetectors with a cut-off wavelength of 1.3 μm. Evidence of lower unintentional
background concentration after annealing at a sufficiently high temperature, is also presented.
Important avalanche breakdown statistics for Single Photon Avalanche Diodes (SPADs), such as avalanche breakdown
probability, dark count rate, and the distribution of time taken to reach breakdown (providing mean time to breakdown
and jitter), were simulated. These simulations enable unambiguous studies on effects of avalanche region width,
ionization coefficient ratio and carrier dead space on the avalanche statistics, which are the fundamental limits of the
SPADs. The effects of quenching resistor/circuit have been ignored. Due to competing effects between dead spaces,
which are significant in modern SPADs with narrow avalanche regions, and converging ionization coefficients, the
breakdown probability versus overbias characteristics from different avalanche region widths are fairly close to each
other. Concerning avalanche breakdown timing at given value of breakdown probability, using avalanche material with
similar ionization coefficients yields fast avalanche breakdowns with small timing jitter (albeit higher operating field),
compared to material with dissimilar ionization coefficients. This is the opposite requirement for abrupt breakdown
probability versus overbias characteristics. In addition, by taking band-to-band tunneling current (dark carriers) into
account, minimum avalanche region width for practical SPADs was found to be 0.3 and 0.2 μm, for InP and InAlAs,
In this work, we present the study on In0.53Ga0.47As/GaAs0.51Sb0.49 type-II heterojunction PIN diodes and
Separate Absorption, Charge and Multiplication (SACM) APDs utilising In0.52Al0.48As as the multiplication layer and
In0.53Ga0.47As/GaAs0.51Sb0.49 type-II heterostructures as the absorption layer. In0.52Al0.48As lattice matched to InP has been
shown to have superior excess noise characteristics and multiplication with relatively low temperature dependence
compared to InP. Furthermore, the type-II staggered band line-up leads to a narrower effective bandgap of approximately
0.49 eV corresponding to the APD cut off wavelength of 2.4 μm. The device exhibited low dark current densities near
breakdown. The device also exhibited multiplication in excess of 100 at 200 K.
There are many applications where the ability to detect optical signals in the 1.65 - 3 μm wavelength range
would be of considerable interest. In this paper we discuss two technologies that offer considerable promise for high
speed, high sensitivity detection in this region utilising avalanche gain. InGaAs/GaAsSb Type II superlattices as the
absorption region and InAlAs as the multiplication region can be combined to form a separate absorption and
multiplication (SAM) avalanche photodiode (APD), all grown lattice matched on InP substrates. Detection at room
temperature up to 2.4 μm can be readily achieved as can gains in excess of 40. InAs homojunction p-i-n diodes are
capable of detecting light with wavelengths > 3 μm, even at 77 K. Although controlling the surface leakage current is a
major challenge in mesa devices of InAs, gains in excess of 40 have also been obtained in these devices at room
temperature. InAs is also the only III-V semiconductor material that appears to show excess noise-free avalanche gain
when electrons are used to initiate the avalanche multiplication. We will discuss recent developments in these two
material systems to date and the current state of the technology.
Electroluminescence (EL) and its temperature dependence of InAs quantum dots embedded in In0.15Ga0.85As quantum
well [dots in a well (DWELL)] have been investigated as functions of the growth temperature of the GaAs spacer layer.
The EL intensity at room temperature increases as the spacer growth temperature increases. The integrated EL intensity
as a function of injection current at room temperature for all samples shows that at low currents, the gradients are
superlinear but this superlinearity decreases as the spacer growth temperature is increased. From a simple analysis of the
generation-recombination rate equations, it can be shown that the superlinearity stems from the nonradiative
recombination being the dominant recombination process. As the spacer growth temperature is increased, this
nonradiative recombination become less dominant. An Arrhenius plot of the temperature dependence of the EL intensity
gives an activation energy of ~300 ± 15 meV at high temperature. The dominant loss mechanism is therefore concluded
to be the electron escape from the quantum dot ground state to the GaAs barrier.
We report the findings of work undertaken to develop InAs photodiodes with low reverse leakage current, for detection
of mid-wave infrared wavelengths up to 3.5μm. Good quality epitaxial growth of InAs and the lattice matched ternary
AlAs0.16Sb0.84 was developed using molecular beam epitaxy. A photodiode structure was designed, grown and
characterized using an AlAs0.16Sb0.84 layer to block the diffusion of minority electrons. Further reductions in the reverse
leakage current were achieved through studies of wet etching using a range of etchants. A sulphuric acid based etchant
provided the lowest surface leakage current for a single etch step, however the surface leakage current was further
reduces when a two steps etching process was employed, starting with a phosphoric acid based etchant and finishing off
with a sulphuric acid based etchant. Surface profile analysis showed that higher etching rates were obtained in the
direction parallel to the <100> direction. The atomic composition of the etched surface was investigated using Auger
analysis. By etching a test pixel array, the potential for fabricating small pitch focal plane arrays by wet etching was
Realization of high-speed avalanche photodiodes (APDs) requires the use of thin avalanche regions to reduce carrier transit time. A systematic investigation on the effect of dead space on the current impulse response and bandwidth of short APDs was carried out using a random path length model assuming a constant carrier velocity. The results indicate that, although dead space suppresses large multiplication values in a short device to give low excess noise, the number of impact ionization a carrier can undergo in a single transit is reduced. Consequently, multiple carrier feedback processes are necessary to achieve a given multiplication value. This results in an increase in the response time and reduces the bandwidth of short APDs. Conventional local models that take no account of the dead space effect will tend to overestimate the operating speed of these devices.
Conventional models of the time response of avalanche photodiodes (APDs) assume that carriers travel uniformly at their saturated drift velocity, vsat. To test the validity of this drift velocity assumption (DVA) the model was used to compute the distribution of exit times of electrons generated in an avalanche pulse and the results were compared with those of Monte-Carlo (MC) simulations. The comparison demonstrates that, while the DVA is valid for thick (1um) avalanching regions, it does not take account of non-equilibrium effects which occur in thin avalanching regions, nor of the effects of diffusion. As a consequence, the DVA model may increasingly underestimate the speed of APDs as the width of the avalanche region is reduced.