Search

This paper presents experimental studies on the formation of ring defects during high-temperature annealing in both electronic-grade and upgraded metallurgical-grade (UMG) Czochralski-grown silicon wafers. Generally, a faster onset of ring defects, or shorter incubation time, was observed in the UMG samples ([Oi] = 6.3 × 1017 cm−3) in comparison to the electronic-grade samples ([Oi] = 3.9 × 1017 cm−3) used in this work. By applying a tabula rasa (TR) treatment prior to annealing, the incubation time can be increased for both types of wafers. We show that TR temperatures above 1000 °C are necessary to effectively dissolve grown-in oxygen precipitate nuclei and limit the subsequent formation of ring defects. A 30 min TR treatment at 1000 °C resulted in the longest incubation time for both types of samples used in this work, as it achieved the best balance between precipitate nuclei dissolution and precipitate re-growth/ripening. Finally, a nitrogen ambient TR step showed a short incubation time for the formation of ring defects in comparison to an oxygen ambient TR step. [ABSTRACT FROM AUTHOR]

Significant reductions in interstitial iron (Fei) concentrations occur during annealing Fe-containing silicon wafers with silicon nitride films in the temperature range of 250 °C-700 °C. The silicon nitride films are known to release hydrogen during the annealing step. However, in co-annealed samples with silicon oxide films, which are hydrogen-lean, changes in the Fei concentrations were much less significant. The precipitation of Fei is ruled out as a possible explanation for the significant reductions. The hydrogen passivation of Fei, which is the complexing of monatomic H and isolated Fei forming a recombination-inactive hydride, is proposed as the most probable model to explain the reductions. Under the assumption that the reduction is caused by the hydrogenation of Fei, the reactants' charge states in the hydrogenation reaction are determined by two independent approaches. In the first approach, illumination is found to have a small but detectible impact on the reaction kinetics in the lower temperature range. The dominating reactants' charge states are concluded to be Fe0 + H+ as revealed by modelling the injection-dependent charge states of isolated Fei and monatomic H. In the second approach, the reaction kinetics are fitted with the Arrhenius equation over a large temperature range of 250 °C-700 °C. A reasonable fit is only obtained when assuming the reacting charge states are Fe0 + H+. This supports the conclusion on the reacting charge states and also gives a value of the activation energy of hydrogenation in the 0.7-0.8 eV range. [ABSTRACT FROM AUTHOR]

There are a number of existing models for estimating the charge states of defects in silicon. In order of increasing complexity, these are (a) the Fermi-Dirac distribution, (b) the Shockley-Last model, (c) the Shockley-Read-Hall model, and (d) the Sah-Shockley model. In this work, we demonstrate their consistency with the general occupancy ratio α, and show that this parameter can be universally applied to predict the charge states of both monovalent and multivalent deep levels, under either thermal equilibrium or steady-state conditions with carrier injection. The capture cross section ratio is shown to play an important role in determining the charge state under nonequilibrium conditions. The application of the general occupancy ratio is compared with the quasi-Fermi levels, which are sometimes used to predict the charge states in the literature, and the conditions where the latter can be a good approximation are identified. The general approach is then applied to the prediction of the temperature- and injection level-dependent charge states for the technologically important case of multivalent monatomic hydrogen, and several other key monovalent deep levels including Fe, Cr, and the boron-oxygen complex in silicon solar cells. For the case of hydrogen, we adapt the model of Herring et al., which describes the charge states of hydrogen in thermal equilibrium, and generalize it for non-equilibrium conditions via the inclusion of the general occupancy ratio, while retaining the pre-factors which make the model more complete. Based on these results, the impact of temperature and injection on the hydrogenation of the key monovalent defects, and other pairing reactions, are discussed, demonstrating that the presented model provides a rigorous methodology for understanding the impact of charge states. [ABSTRACT FROM AUTHOR]

Injection-dependent lifetime spectroscopy of both n- and p-type, Cr-doped silicon wafers with different doping levels is used to determine the defect parameters of Cri and CrB pairs, by simultaneously fitting the measured lifetimes with the Shockley-Read-Hall model. A combined analysis of the two defects with the lifetime data measured on both n- and p-type samples enables a significant tightening of the uncertainty ranges of the parameters. The capture cross section ratios k=σn/σp of Cri and CrB are determined as 3.2 (-0.6, +0) and 5.8 (-3.4, +0.6), respectively. Courtesy of a direct experimental comparison of the recombination activity of chromium in n- and p-type silicon, and as also suggested by modelling results, we conclude that chromium has a greater negative impact on carrier lifetimes in p-type silicon than n-type silicon with similar doping levels. [ABSTRACT FROM AUTHOR]

The band-band absorption coefficient in crystalline silicon has been determined using spectral photoluminescence measurements across the wavelength range of 990-1300 nm, and a parameterization of the temperature dependence has been established to allow interpolation of accurate values of the absorption coefficient for any temperature between 170 and 363K. Band-band absorption coefficient measurements across a temperature range of 78-363K are found to match well with previous results from MacFarlane et al. [Phys. Rev. 111, 1245 (1958)], and are extended to significantly longer wavelengths. In addition, we report the band-band absorption coefficient across the temperature range from 270-350K with 10K intervals, a range in which most practical silicon based devices operate, and for which there are only sparse data available at present. Moreover, the absorption coefficient is shown to vary by up to 50% for every 10K increment around room temperature. Furthermore, the likely origins of the differences among the absorption coefficient of several commonly referenced works by Green [Sol. Energy Mater. Sol. Cells 92, 1305 (2008)], Daub and Würfel [Phys. Rev. Lett. 74, 1020 (1995)], and MacFarlane et al. [Phys. Rev. 111, 1245 (1958)] are discussed. [ABSTRACT FROM AUTHOR]

This article presents experimental and simulation studies on the formation of recombination‐active ring‐like defects during thermal donor (TD) formation at 450 °C in n‐type Czochralski‐grown silicon wafers. With increasing anneal duration from 1 to 24 h, the resistivity and interstitial oxygen concentration of samples decrease, consistent with the formation of TDs. However, after a subsequent TD annihilation treatment, the resistivity recovers completely, while the interstitial oxygen concentration recovers only partially. Furthermore, the wafers exhibit ring‐like defects in photoluminescence images after the TD generation and remain persistent even after annihilation treatment. By applying a tabula rasa (TR) treatment prior to TD generation anneals, the net loss of interstitial oxygen is reduced, and the incubation time of the ring‐like defects is extended in comparison to as‐grown wafers. Finally, to investigate the possibility of low‐temperature precipitation, simulation of oxygen precipitation (OP) kinetics during the 450 °C annealing steps is performed using an enhanced effective diffusivity of oxygen. These simulations demonstrate that OP can occur simultaneously during TD formation, resulting in recombination‐active ring‐like defects. [ABSTRACT FROM AUTHOR]

The defect parameters of isolated Cri and chromium-boron (CrB) pairs are reassessed by conducting lifetime spectroscopy on both n- and p-type, Cr-doped silicon samples with different doping levels, and fitting the lifetimes with the Shockley-Read-Hall (SRH) model. The uncertainty ranges of the parameters are significantly tightened through a combined analysis of the two defects with the lifetime data measured on both n- and p-type samples. Both the experimental data and the modelling results suggest that Cri has a greater negative impact on lifetimes in p-Si than n-Si. The possible hydrogen passivation of Cri in n- and p- Si is then investigated both through charge state modelling and experimentally. A general occupancy factor is applied to predict the charge states of both monovalent Cri and multivalent H as a function of temperature and injection level. The modelling results suggest that above 800K, both the injection and the doping become unimportant as silicon becomes intrinsic. However, despite some positive hints, the experimental evidence for hydrogenation of Cri is inconclusive at this stage. [ABSTRACT FROM AUTHOR]