Effect of dead space on the excess noise factor and time responseof avalanche photodiodes

The effect of dead space on the statistics of the gain process in continuous-multiplication avalanche photodiodes (APDs) is determined using the theory of age-dependent branching processes. The dead space is the minimum distance that a newly generated carrier must travel in order to acquire sufficient energy to cause an impact ionization. Analytical expressions are derived for the mean gain, the excess noise factor, and the mean and standard deviation of the impulse response function, for the dead-space-modified avalanche photodiode (DAPD), under conditions of single carrier multiplication. The results differ considerably from the well-known formulas derived by R.J. McIntyre and S.D. Personick in the absence of dead space. Relatively simple asymptotic expressions for the mean gain and excess noise factor are obtained for devices with long multiplication regions. In terms of the signal-to-noise ratio (SNR) of an optical receiver in the presence of circuit noise, it is established that there is a salutory effect of using a properly designed DAPD in place of a conventional APD. The relative merits of using DAPD versus a multilayer (superlattice) avalanche photodiode (SAPD) are examined in the context of receiver SNR; the best choice turns out to depend on which device parameters are used for the comparison     

Excess noise in GaAs avalanche photodiodes with thin multiplicationregions

It is well known that the gain-bandwidth product of an avalanche photodiode can be increased by utilizing a thin multiplication region. Previously, measurements of the excess noise factor of InP-InGaAsP-InGaAs avalanche photodiodes with separate absorption and multiplication regions indicated that this approach could also be employed to reduce the multiplication noise. This paper presents a systematic study of the noise characteristics of GaAs homojunction avalanche photodiodes with different multiplication layer thicknesses. It is demonstrated that there is a definite “size effect” for multiplication regions less than approximately 0.5 μm. A good fit to the experimental data has been achieved using a discrete, nonlocalized model for the impact ionization process     

低过剩噪声级联倍增雪崩探测器设计

基于弛豫空间倍增理论数值模型和修正的弛豫空间倍增理论模型,分析了不同倍增级数和不同载流子初始能量时级联倍增雪崩探测器的过剩噪声.研究了不同碰撞离化倍增层厚度、不同电子预加热层厚度、不同电场控制层掺杂浓度对过剩噪声因子的影响.同时,比较了DSMT模型、Van Vilet模型和McIntyre模型得到的结果.通过调整碰撞离化倍增层厚度、电子预加热层厚度和电场控制层掺杂浓度,DSMT数值模拟获得了一个相对优化的结构,其过剩噪声与Van Vliet模型k_s=0. 057时相当.     

Parametric manufacturing yield modeling of GaAs/AlGaAs multiplequantum well avalanche photodiodes

GaAs/AlGaAs multiple quantum well (MQW) avalanche photodiodes (APD's) are of interest as an ultra-low noise image capture mechanism for high-definition systems. Since literally millions of these devices must be fabricated for imaging arrays, it is critical to evaluate potential performance variations of individual devices in light of the realities of semiconductor manufacturing. Specifically, even in a defect-free manufacturing environment, random variations in the fabrication process will lead to varying levels of device performance, Accurate device performance prediction requires precise characterization of these variations. This paper presents a systematic methodology for modeling the parametric performance of GaAs MQW APD's. The approach described requires a model of the probability distribution of each of the relevant process variables, as well as a second model to account for the correlation between this measured process data and device performance metrics. The availability of these models enables the computation of the joint probability density function required for predicting performance using the Jacobian transformation method. The resulting density function can then be numerically integrated to determine parametric yield. Since they have demonstrated the capability of highly accurate function approximation and mapping of complex, nonlinear data sets, neural networks are proposed as the preferred tool for generating the models described above. In applying this methodology to MQW APD's, it is shown that using a small number of test devices with varying active diameters, barrier and well widths, and doping concentrations enables prediction of the expected performance variation of APD gain and noise in larger populations of devices. This approach compares favorably with Monte Carlo techniques and allows device yield prediction prior to high volume manufacturing in order to evaluate the impact of both design decisions and process capability     

Multiplication noise in multi-heterostructure avalanche photodiodes

Using a multi-feedback network representation, multiplication noise in two-stage and three-stage heterostructure avalanche photodiodes has been calculated for different degrees of carrier feedback over heterobarriers. Results show appreciable dependence of noise on carrier feedback in III–V compound APD's, the effect being less with larger number of stages.     

Absolute noise characterisation of avalanche photodiodes

Excess noise in four types of commercially obtained avalanche photodiodes (a.p.d.s) has been measured absolutely, by comparing avalanche noise from the a.p.d. with shot noise from an illuminated p-i-n diode. The method used yields directly the noise-current spectral density, simplifies the deduction of the quantum efficiency keff and hence the true value of the multiplication factor, and ultimately yields a measured value of the noise parameter x.     

Effect of impact ionization in the InGaAs absorber on excess noise of avalanche photodiodes

The effects of impact ionization in the InGaAs absorption layer on the multiplication, excess noise and breakdown voltage are modeled for avalanche photodiodes (APDs), both with InP and with InAlAs multiplication regions. The calculations allow for dead space effects and for the low field electron ionization observed in InGaAs. The results confirm that impact ionization in the InGaAs absorption layer increases the excess noise in InP APDs and that the effect imposes tight constraints on the doping of the charge control layer if avalanche noise is to be minimized. However, the excess noise of InAlAs APDs is predicted to be reduced by impact ionization in the InGaAs layer. Furthermore the breakdown voltage of InAlAs APDs is less sensitive to ionization in the InGaAs layer and these results increase tolerance to doping variations in the field control layer.     

High-gain GaAs avalanche photodiodes with proton-implanted guard ring

A GaAs avalanche photodiode with a multiplication factor as high as 8000 was prepared by Zn diffusion and proton double implantation. The proton-implanted guard ring completely prevented edge breakdown, and multiplication occurred uniformly over the junction area. Dark current was proved to be due to a leakage current at the periphery between junction and implanted layer.     

Temperature dependence of multiplication noise in silicon avalanche photodiodes

The temperature dependence of multiplication noise in silicon avalanche photodiodes with a low-high-low impurity density profile is calculated. The variation of multiplication noise by temperature change can be neglected in practical use at a constant multiplication factor, which is in agreement with experimental results.     

Improved figure of merit for noise in avalanche photodiodes

The authors propose a new noise figure for avalanche photodiodes (APDs). This new noise figure overcomes the difficulty of estimating the internal multiplication and quantum efficiency in complex APD structures, such as III-V SAM APDs. Measurements of the new noise figure are presented for two commercial SAM APDs and the authors show theoretically that it represents a more complete figure of merit for comparing the performance of one APD with another, and with the ideal     

Wavelength dependence of multiplication noise in silicon avalanche photodiodes

Theoretical and experimental results on wavelength dependence of multiplication noise in silicon avalanche photodiodes are described. When the photodiode has a p-n⁺-junction and is illuminated from the n⁺-side, multiplication noise increases by decreasing optical wavelength. Effective ionization coefficient ratio keffis equal tokexp (2Kw_{a}) for a uniform junction electric field, wherekis the ratio of ionization coefficients of electrons α and holes β. The multiplication noise depends on the product of optical absorption coefficientKand the avalanche-region width wa. Calculations show that there exists an optimum wafor minimizing multiplication noise at a given wavelength. Theoretical results are shown to agree with results of experiments on diodes with a low-high-low impurity profile. Measured ionization coefficient ratiokvalues are 0.04 and 0.08 at 0.811- and 0.633-µm wavelength, respectively.     

Silicon avalanche photodiodes with low multiplication noise and high-speed response

Low-noise and high-speed silicon avalanche photodiodes with low breakdown voltage are reported. The diode structure with a low-high-low impurity density profile is proposed to have low-noise characteristics. Multiplication noise and depletion layer width of several structures are compared theoretically, and effects of impurity density profile of the avalanche region are discussed. Built-in field is also provided to realize high-speed response without increasing operating voltage. Silicon avalanche photodiodes with the above mentioned structure have been fabricated with long time substrate annealing, ion implantation, and epitaxial growth. Attained performances are as follows: noise parameter k = 0.027 - 0.040, output pulse half width τ = 260 ps for a mode-locked Nd:YAG laser pulse, gain-bandwidth product up to 300 GHz at M = 400, quantum efficiency 0.55 - 0.66 at the 0.81- to 0.83-µm wavelength, and breakdown voltage about 100 V.     

Signal and noise response of high speed germanium avalanche photodiodes

Germanium avalanche photodiodes, providing gain at microwave frequencies, have been fabricated and tested. The diodes employ a guard ring structure to achieve a uniform, microplasma-free, multiplying region with an active diameter of 40 microns. Low-frequency chopped light current gains of greater than 200, and small-signal 6 GHz current gains of greater than 10 have been obtained at room temperature for a carrier wavelength of 1.15 microns. In the normal operating range, the signal output power is found to vary as the square of the multiplication, and the noise is found to vary as the cube of the multiplication. This limits the maximum useful multiplication of the diode to that level which gives a diode noise equal to the receiver noise. A small-signal equivalent circuit with lumped elements corresponding to the physical processes occurring within the diode, is introduced to describe the small signal behavior. The model is valid over the entire multiplication range, up to frequencies of about 10 GHz.     

A resonant-cavity, separate-absorption-and-multiplication,avalanche photodiode with low excess noise factor

We report on the design, fabrication, and performance of a photodiode that combines the advantages of a resonant cavity with a separate-absorption-and-multiplication avalanche photodiode. The device is grown on GaAs using molecular beam epitaxy and is designed to detect light near 900 nm. This photodetector has exhibited the following characteristics: an external quantum efficiency of 70%, a spectral linewidth of less than 7 nm, an avalanche gain in excess of 30, and low dark current. In addition, a low excess noise factor corresponding to 0.2⩽k⩽0.3 has been achieved     

An analytical approximation for the excess noise factor ofavalanche photodiodes with dead space

Approximate analytical expressions are derived for the mean gain and the excess noise factor of avalanche photodiodes including the effect of dead space. The analysis is based on undertaking a characteristic-equation approach to obtain an approximate analytical solution to the existing system of recurrence equations which characterize the statistics of the random multiplication gain. The analytical expressions for the excess noise factor and the mean gain are shown to be in good agreement with the exact results obtained from numerical solutions of the recurrence equations for values of the dead space reaching up to 20% of the width of the multiplication region     

Photo-induced excess low frequency noise in HgCdTe photodiodes

We have investigated the properties of excess low frequency noise in illuminated mid wavelength infrared and long wavelength infrared HgCdTe photodiodes at zero bias. The current power spectrum (S_i) dependence is usually close to inverse frequency (f), but substantial variations have been observed. The magnitude of l/f spectra is voltage independent for small bias voltages, but is proportional to the square of the photocurrent (I). Consequently, the l/f knee increases, with photocurrent. Variable area device studies indicate that the noise sources are more closely associated with the device area (A_j) than perimeter, indicating bulk limitations. The power spectrum can be represented by an empirical relationship of the form S_i=α_phI ²/fA_j. This defines a figure of merit, α_ph which takes into the account the relationship between current dependence and device geometry. α_ph is device dependent, suggesting that randomly distributed defects play a role in the difference. This is also supported by noting that devices fabricated in material grown on lattice matched substrates have lower α_ph (10⁻¹⁶ cm²) than those fabricated in material grown on nonlattice matched substrates (10⁻¹⁴ cm²), which usually have two orders of magnitude larger dislocation density. We conclude that photo-induced l/f noise can be reduced via defect reduction and is not fundamental. Data on our best devices indicates that α_ph is somewhat lower for smaller band gap material. The temperature dependence of photo-induced excess low frequency noise is much weaker than that of bias induced excess low frequency noise, indicating unrelated generation mechanisms. In addition, photo-induced l/f adds in quadrature with bias induced l/f noise and is not well correlated in magnitude with either bias induced l/f noise or detector dark currents.     

Effect of dead space on gain and noise ofdouble-carrier-multiplication avalanche photodiodes

The effect of dead space on the statistics of the gain in a double-carrier-multiplication avalanche photodiode (APD) is determined using a recurrence method. The dead space is the minimum distance that a newly generated carrier must travel in order to acquire sufficient energy to become capable of causing an impact ionization. Recurrence equations are derived for the first moment, the second moment, and the probability distribution function of two random variables that are related, in a deterministic way, to the random gain of the APD. These equations are solved numerically to produce the mean gain and the excess noise factor. The presence of dead space reduces both the mean gain and the excess noise factor of the device. This may have a beneficial effect on the performance of the detector when used in optical receivers with photon noise and circuit noise     

Effect of stochastic dead space on noise in avalanche photodiodes

A stochastic dead-space model for impact ionization is developed and used to study the effect of the soft nature of the ionization capability of carriers on the excess noise factor of avalanche photodiodes. The proposed model is based on the rationale that the gradual, or soft, transition in the probability density function (PDF) for the distance from birth to impact ionization can be viewed as that resulting from uncertainty in the dead space itself. The resulting soft PDF, which is parameterized by a tunable softness parameter, is used to establish the limitations of the existing hard-threshold ionization models in ultrathin multiplication layers. Calculations show that for a fixed operational gain and fixed average dead space, the excess noise factor tends to increase as a result of the softness in the PDF in very thin multiplication layers (viz, <70 nm), or equivalently, under high applied electric fields (viz., >800 kV/cm). A method is proposed for extracting the softness parameter from noise versus multiplication measurements.