Quenched-in centers in silicon p+n junctions

Silicon p⁺n junctions heat treated at high temperatures (1200°C) for a long time (2–20 hr) and then quenched to room temperatures or below shows two deep donor levels (E_C?264meV andE_C?542meV) in the n-type side of the junction depletion layer which appear to originate from the same imperfection center. The concentration of these levels ranges from 10¹³ to 10¹⁴ cm^?3. The junction leakage current comes from carrier generation at the deeper level in the depletion region. Phosphorus gettering was found ineffective in reducing the concentration of these quenched-in levels, but they are annealed out by very slow cooling (25°C/hr to 650°C then quench to room temperatures). The thermal emission and capture rates of electrons and holes at these levels are measured as a function of temperature and electric field by the junction high frequency capacitance and d.c. leakage current transient techniques. It is demonstrated that the detailed balance relationship does not hold. The origin of this double donor center is yet to be identified.     

A difficulty in the theory of p+-n junction diode noise

It is shown that the usual theory of p⁺-n junction diode noise leads to the following difficulty. The spectrum of the open-circuit noise voltage of the diode does not go to zero fast enough at sufficiently high frequencies, so that the spectrum cannot be integrated from 0 to ∞. The difficulty is removed when the effect of the capacitance C_j of the junction space charge region is taken into account. This is a particular example of a more general result.     

Heat flow resistance measurements of silicon p+-v-n+ diodes under forward and reverse biased conditions

A detailed study of the heat flow resistance measurements in a p⁺-v-n⁺ diode is studied in both forward and reverse biased conditions. Measurements are made by continuously switching the diode from the power dissipation state into the temperature measuring state. Safe operating power limits are identified for the diodes depending upon their mode of operation either as a microwave switch or as an IMPATT oscillator.     

High injection phenomena in p+in+ silicon solar cells

A very simple analytic analysis of p⁺in⁺ solar cell structures is undertaken based in the quasi-neutrality condition (n = p), which validity is verified. The only differential equation to be solved in the model is the classical continuity one. Non linearities, appea ronly in the boundary conditions. This model can be applied to p⁺nn⁺ or n⁺pp⁺ cells under very high injection conditions. High carrier concentrations are required to support the current flow even in short-circuit conditions, so enhancing the recombination which, on the other hand, is reduced by high injection lifetime increase. Under-linearity between the short-circuit current and the photon flux is also deduced at very high irradiance. The short-circuit current under bifacial illumination is higher than the sum of those currents under front and back illumination alternatively, so leading to inherently better bifacial cells. In the V_oc vs J_sc curve values of m in the logarithmic slope mkT/q range from values slightly below one to two. The value m = 1 is not to be expected even if the base recombination is negligible due to Dember potential effects.     

Mobility fluctuation1/f noise in silicon p+-n-p transistors

A study of the current dependence of base 1/f noise and of collector 1/f noise in p⁺-n-p transistors shows that the former is most likely of the mobility fluctuation 1/f noise type and that the latter is most probably not of that type. The current dependence of 1/f noise in transistors is a powerful tool in the interpretation of the noise.     

Post-breakdown bulk oscillations in gold-doped silicon p+−i−n+ double-injection diodes

Experimental work on post-breakdown bulk oscillations in n-type gold-doped phosphorus-compensated p⁺?i?n⁺ double-injection diodes is presented. An empirical relationship for the frequency of oscillation in this region is derived for the first time and discussed.     

Junction potentials of strongly illuminated n+ - p - p+ solar cells

The low-junction (LHJ) model is applied to an n⁺ - p - p⁺ solar cell having finite dimensions, in order to investigate its performance under intense illumination. Ambipolar transport equations are solved in the three sections of the cell using appropriate boundary conditions. Expressions for junction currents are derived, and the junction potentials under open-circuit conditions are computed by the Newton-Raphson method.The theory presented here includes the effects of high level injection. The generalized current density equations which are derived here for an n⁺ - p - p⁺ device are shown to reduce to the ideal Shockley diode equation with appropriate modifications. The effects of p - p⁺ low-high junction on the open-circuit voltage of the cell are explained. The theoretical results of this paper are consistent with the experimental results of others.     

Electron bombardment effect on the reverse I–V characteristics and tensosensibility of p+-nGaAs diodes

The effect of ≈2 MeV electron bombardment on the reverse characteristics and tensoelectric properties in p⁺-nGaAs diodes is investigated. The reverse breakdown voltage showed a weak increase due to irradiation and an anomalous temperature dependence in the range 77–300 K. The I–V characteristics in electron-irradiated diodes revealed a high pressure sensitivity (tensosensibility) to the external hydrostatic pressure (up to 6 · 10⁸ Pa). The peculiarities in the reverse I–V characteristics of diodes investigated point to the presence of a radiation-induced deep trap (acceptor-type level at about E₀ ? 0.3 + 0.4 eV), which is attached to the Γ₁₅^V-maximum of the valence band.     

Characteristics of three-terminal metal-tunnel if oxide-n/p+ devices

The device described here comprises a p⁺ substrate containing an epitaxial n-layer, on the surface of which is grown a thin (～50 Å) tunnel oxide. A metal cathode is deposited on the oxide surface, and a metal anode on the back side of the p⁺ substrate. A third terminal, the gate electrode, is connected to the n epilayer to provide for biasing the n-p⁺ junction.The I-V characteristic exhibit two stable states: a high-impedance state and a low-impedance state which are separated by a negative-resistance region. The high-impedance state is stable for applied voltages up to the intrinsic threshold voltage, V_s. When the switching voltage is exceeded, the device switches rapidly to the low-impendance state, which is characterized by a current that increases with little increase in the voltage across the device.The switching voltage may be reduced below V_s by current or voltage biasing of the n-p⁺ junction by means of the gate electrode. Gate efficiencies, the ratio of the change in switching voltage with d.c. gate voltage or current, of 10 V/V and 1.0 V/μA have been observed. Pulsed gate measurements are also presented, and it found that for pulse widths down to 0.1 μs the gate switching characteristics follow the d.c. characteristics. For pulse widths less than 0.1 μs the gate efficiencies are degraded. Suggestions for improving the device characteristics and the turn-on and turn-off time of the device and device reliability are discussed.     

Low field study of P+NN+ or P+PN+ structures in the forward direction

It is shown, on the basis of P⁺NN⁺ and P⁺PN⁺ models, how injection processes modify the field distribution in two-electrode systems, and thereby the effective resistivity.     

Some properties of ion-implanted p−n junctions in silicon

Diodes have been made by implantation of boron or gallium ions in n-type, and phosphorus ions in p-type silicon. The doses range from 5 × 10¹² to 10¹⁵ ions/cm², and the energies from 20 to 70 keV. In all diodes the reverse current shows a sharp recovery step upon annealing at 500–600°C. The reverse current after this annealing is typically of the order of 1 nA/cm² at 1 V reverse bias. To overcome the problem of low breakdown voltages usually found for implanted junctions, methods have been developed to enlarge the effective radius of curvature at the edge of the implanted junction. In a planar process with oxide masking, breakdown voltages of 150 V for 3 Ωcm or 1500 V for 300 Ωcm silicon are obtained. This is done by implanting the ions through a tapered oxide, where the oxide walls make an angle of only 3–5° with the silicon surface. The junction depth in this case is 0.4 μm.Another method uses a mask, placed free in front of the slice. Slice and mask rotate during implantation. In this way, a breakdown voltage of 2700 V is obtained with 300 Ωcm silicon.     

A discrete element small-signal equivalent circuit for forward-biased p+n diodes containing deep levels

A discrete element small-signal equivalent circuit model for p-n diodes containing deep defect levels is developed, by extending an existing model for undamaged devices. With the aid of a simple analytical expression which accurately describes the forward bias d.c. current, the enhanced small-signal conductance due to carrier recombination in the depletion region is included in the model. The influence of trapped charge on the space charge capacitance is incorporated using a simplified version of the analysis of Beguwala and Crowell. The predictions of the model are verified by experimental data from silicon p⁺n diodes, in which deep levels have been induced by electron irradiation. It is shown that the deep level activation energies may be estimated from the forward bias capacitance-voltage characteristics, yielding values which agree well with those obtained by established techniques.     

A theoretical and experimental study of recombination in silicon p−n junctions

This paper describes the results of a numerical solution of the one-dimensional transport equations as applied to p?n junctions. Generation-recombination is included in the model, and the method of solution is based on that of De Mari. The modifications of Choo are included to allow the solution of problems involving heavy recombination. Theoretical current-voltage characteristics are presented for both uniform and non-uniform distributions of recombination centres, and a comparison is made with experimental results on carbon implanted, silicon, p?n diodes. The dependence of the $\text{I}\text{V}$ characteristics on activation energy, capture cross sections, and density of recombination centres is described. Good agreement between theory and experiment is obtained, and the results suggest that the $\text{I}\text{V}$ characteristics can be described by the sum of two independant components of current. The first component is the generation-recombination current with an exp (eV/mkT) dependence, and the second is the diffusion current with an exp (eV/kT) dependance.     

Transition-capacitance calculations for double-diffused p-n junctions

A study of the depletion-layer characteristics of double-diffused p-n junctions, formed by successive diffusion of opposite type impurities into a semiconductor, is presented. The general form of the impurity profile is taken to be N(r) = N_SA e^?ar(ar+2k) where N_S, A, a and k are constants and r is the distance. This general form reduces to Gaussian, erfc and two-step diffusion profiles as particular cases. Results are shown graphically for typical values of these constants.It is shown that the double-diffused junctions can be well approximated by equivalent double-exponential profiles in the impurity range of practical interest and C-V relations are obtained analytically. The results obtained are convenient for ready engineering calculations. The accuracy and the range of validity of the approximation are discussed.A simple and accurate analytical method is outlined for the calculation of reverse-biased sidewall capacitance of double-diffused structures, with special reference to the emitter junction of a planar transistor. It is shown that by approximating the impurity profiles by Gaussian distributions, the metallurgical junction formed is given by an ellipse and this leads to a simple evaluation of the sidewall capacitance. An expression obtained for the impurity gradient along the sidewall enables the computation of forward-biased depletion-layer capacitance when the mobile carriers in the depletion region are to be considered.     

Crystal damage and the properties of implanted p-n junctions in silicon

The influence of crystal damage on the properties of implanted p-n junctions has been studied by variation of the amount of initial damage, variation of the recovery process, and variation of the residual damage. This was done by carrying out implantations at - 196, 25 and 700°C with 10¹⁵ B⁺/cm² at an energy of 50 keV, and at 25°C with 10¹⁵ BF₂⁺ at an energy of 250 keV and 10¹⁵ Ga⁺/cm² at an energy of 70 keV. Substrate orientations of both 〈111〉 and 〈100〉 were used, and annealing was done in a temperature range between 400 and 1100°C. Gettered as well as non-gettered slices were used for 〈111〉 oriented substrates. The diode properties were analyzed with the aid of Shockley-Read-Hall recombination statistics. Depending upon crystal history and processing, different traps are found to dominate the reverse current. Traps caused by the gettering of contamination as well as those caused by the damage itself play a role. The number of traps is found to be smaller than 10¹²/cm³ for well annealed diodes, resulting in a reverse current density of 0.2 nA/cm² at 1 V reverse bias.     

Noise in near-ballistic n+nn+ and n+pn+ gallium arsenide submicron diodes

Direct-current (d.c.) characteristics and noise measurements in the range 1 Hz-25 kHz are reported for n⁺nn⁺ and n⁺pn⁺ near-ballistic devices, with n regions (p regions) of 0.4 μm (0.45 μm), fabricated by molecular beam epitaxy at Cornell. The n⁺nn⁺ mesa structures show very low 1/? noise. indicating a Hooge parameter α_H = 6.0 × 10^?. This very low noise is attributed to the near absence of phonon collisions. The thermal (? like) noise above 1 kHz is equal to Nyquist noise at the lowest currents, rising to slightly above Nyquist noise for high currents, indicating the presence of carrier drag effects. The n⁺pn⁺ noise, on the contrary, is quite high. It seems to be associated with the ambipolar effects occurring for low injection of electrons in the p region. The importance of noise measurements for confirming ballistic or near-ballistic behavior is discussed.     

Boundary conditions at p-n junctions

A discussion is presented of the relationships between carrier densities at a p-n junction and the junction potential. The major purpose of the work is to discuss the relationship between the Fletcher and Misawa boundary conditions. In addition, a discussion is presented of an extension of the normal boundary conditions to account for current flow within the junction space charge region.     

The physical behaviour of an n+p silicon solar cell in concentrated sunlight

The performance of a simple n⁺p silicon solar cell at various illumination levels is analysed by a modified form of the Gummel and De Mari numerical algorithms. Effects of high doping, such as bandgap narrowing together with the correction to density of states are included. The effective recombination life time of the charge carriers due to both Shockley-Read-Hall recombination via traps and Auger recombination is taken into account. The base acceptor doping concentration is 10¹⁶ cm^?3. The light concentration is varied from 1 to 200 AM1. The physical mechanisms of the device at various levels of illumination are discussed by determining the cell parameters, namely, saturation current density, short circuit current density, ideality factor and fill factor. The ideality factor which is close to 1 at low illumination suggests that the cell is controlled by diffusion-recombination processes. The high value of the ideality factor, which is very much greater than 1 but less than 2, at high-illumination is attributed to high-injection effect. The efficiency reaches a maximum around 100 AM1 and starts falling beyond that. This fall is due to the high-injection and the voltage drop in the base layer. The fill factor starts falling at high-illumination.     

Drift of the breakdown voltage in p-n junctions in silicon (walk-out)

A quantitative model for the time behaviour of the walk-out phenomenon in planar p-n junctions is given. The injection of hot carriers into SiO₂ and subsequent trapping of part of them is assumed to be the origin of the walk-out. The model is found to be in reasonable agreement with the experimental results on both p⁺?n and n⁺?p junctions. The parameters in the model are discussed in relation with the experiments.     

Impurity distributions and p−n junctions in semiconductors

We have obtained a closed-form solution of the Poisson-Boltzmann equation for a class of quasi-linearly graded p?n junctions. The solution reveals that asymmetric distributions can produce electrostatic potential barriers on the low impurity density side of the junction. For majority carriers these barriers become potential traps producing a secondary space charge next to the primary dipole at the junctions center. The depth of the potential wells depend on the difference in bulk acceptor and donor densities as well as the impurity density slope at the junctions center. Numerical values are given for silicon and germanium.