https://uvadoc.uva.es/handle/10324/28026 Electrónica - Artículos de revista 2026-04-07T02:13:32Z https://uvadoc.uva.es/handle/10324/77892 We performed classical molecular dynamics simulations to investigate, from an atomistic point of view, the formation of dislocations during the epitaxial growth of Ge on Si. We show that simulations at 900 and 1000 K with deposition rates of 10 monolayers per second provide a good compromise between computational cost and accuracy. In these conditions, the ratio between the Ge deposition rate and the ad-atom jump rate is analogous to that of out-of-equilibrium experiments. In addition, the main features of the grown film (intermixing, critical film thickness, dislocation typology, and surface morphology) are well described. Our simulations reveal that dislocations originate in low-density amorphous regions that form under valleys of the rough Ge film surface. Atoms are squeezed out of these regions to the surface, releasing the stress accumulated in the film and smoothing its roughness. Amorphous regions grow until atoms begin to rearrange in dislocation half-loops that propagate throughout the Ge film. The threading arm ends of the dislocation half-loops move along the surface following valleys and avoiding islands. The film surface morphology affects the propagation path of the dislocation half-loops and the resulting dislocation network. 2026-01-01T00:00:00Z https://uvadoc.uva.es/handle/10324/77889 A Deep Learning approach is devised to estimate the elastic energy density at the free surface of an undulated stressed film. About 190000 arbitrary surface profiles are randomly generated by Perlin noise and paired with the corresponding elastic energy density profiles , computed by a semi-analytical Green’s function approximation, suitable for small-slope morphologies. The resulting dataset and smaller subsets of it are used for the training of a Fully Convolutional Neural Network. The trained models are shown to return quantitative predictions of , not only in terms of convergence of the loss function during training, but also in validation and testing, with better results in the case of the larger dataset. Extensive tests are performed to assess the generalization capability of the Neural Network model when applied to profiles with localized features or assigned geometries not included in the original dataset. Moreover, its possible exploitation on domain sizes beyond the one used in the training is also analyzed in-depth. The conditions providing a one-to-one reproduction of the “ground-truth” profiles computed by the Green’s approximation are highlighted along with critical cases. The accuracy and robustness of the deep-learned are further demonstrated in the time-integration of surface evolution problems described by simple partial differential equations of evaporation/condensation and surface diffusion. 2025-01-01T00:00:00Z https://uvadoc.uva.es/handle/10324/58974 The surface diffusion and intermixing of Ge ad-atoms over Si (001) 2 × 1 substrates using classical molecular dynamics (CMD) simulations are characterized here. Several interatomic potentials, parametrizations, and parameter mixing rules are contemplated. A novel simulation scheme is devised to characterize the effective frequency of surface diffusion and intermixing events overcoming the inherent difficulties related to their interdependency in heteroepitaxial systems. The effective energy barriers of these events encompass different atomistic mechanisms weighted by their occurrence probabilities. The overall description of surface diffusion and intermixing based on Stillinger–Weber (SW) potential is in agreement with ab initio calculations and experimental observations, though some atomistic details differ. This study is extended to Si(001) substrates with stressed Ge monolayers grown on top. It is found that Ge ad-atom dynamics is accelerated with respect to the case of the pure Si substrate and that diffusion across dimer rows is mainly mediated by the atomic exchange of the Ge ad-atom with a Ge atom on the surface. 2023-01-01T00:00:00Z https://uvadoc.uva.es/handle/10324/55603 The improved radiation hardness of p-type Si detectors is hindered by the radiation-induced acceptor removal process, which is not fully understood yet. Through atomistic modeling of displacement damage and dopant interactions, we analyze the acceptor removal under neutron irradiation, providing physical insight into its microscopic origin. Our results show that the fast decay of the effective dopant concentration (Neff) at low irradiation fluences is due to B deactivation caused by Si self-interstitials. The intriguing increase of the acceptor removal parameter with the initial dopant concentration (Neff,0) is explained by the limited number of mobile Si self-interstitials that survive annihilation and clustering processes. The sublinear dependence of the removal parameter on Neff,0 is associated to the inhomogeneity of damage for low Neff,0 and the formation of B-interstitial clusters with several B atoms for high Neff,0. The presence of O and C modifies B deactivation mechanisms due to the key role of BiO defects and the trapping of vacancies and Si self-interstitials, but for the impurity concentrations analyzed in this work ([O] >> [C]) it has little effect on the overall amount of removed acceptors. At high irradiation fluences, the reported increase of Neff is attributed to the formation of defect-related deep acceptors. From the analysis of the defect concentrations resulting from neutron irradiation and the occupancy of small clusters with acceptor levels reported in literature, we point out the tetra-vacancy cluster as one of the main contributors to Neff with negative space charge. 2022-01-01T00:00:00Z https://uvadoc.uva.es/handle/10324/55159 Intra-die device variation due to pattern layout effects associated with the development of ultra-fast annealing processes is one of the major scaling challenges for advanced CMOS devices. In this paper, we show that an excellent and universal correlation can be established between on-die device variation and a new reflectance characterization technique with sufficient resolution. This approach has the potential to be universally applicable to virtually any structure pattern. In addition, we conducted simulations of the thermal annealing effect on 2D doping profiles by considering the effects of temperature sensitivity, reflectivity, and active dopant fraction. Our results show that the observed on-die variation was caused mainly by using a rapid thermal annealing (RTA) process rather than by flash annealing (FLA). We further concluded that pattern-induced device variation is mainly due to the redistribution of the dopants, instead of from dopant activation. To mitigate the pattern loading effect from thermal annealing, we employed a light absorbing layer to eliminate the within-die reflectivity variation. We found that we could successfully reduce electrical on-die variation by 50%. 2022-01-01T00:00:00Z https://uvadoc.uva.es/handle/10324/54389 The relentless scaling of semiconductor devices pushes the doping level far above the equilibrium solubility, yet the doped material must be sufficiently stable for subsequent device fabrication and operation. For example, in epitaxial silicon doped above the solubility of phosphorus, most phosphorus dopants are compensated by vacancies, and some of the phosphorus-vacancy clusters can become mobile around 700 °C to further cluster with isolated phosphorus ions. For efficient and stable doping, we use microwave annealing to selectively activate metastable phosphorus-vacancy clusters by interacting with their dipole moments, while keeping lattice heating below 700 °C. In a 30-nm-thick Si nanosheet doped with 3 × 1021 cm−3 phosphorus, a microwave power of 12 kW at 2.45 GHz for 6 min resulted in a free-electron concentration of 4 × 1020 cm−3 and a junction more abrupt than 4 decades/nm. The doping profile is stable with less than 4% variation upon thermal annealing around 700 °C for 5 min. Thus, microwave annealing can result in not only efficient activation and abrupt profile in epitaxial silicon but also thermal stability. In comparison, conventional rapid thermal annealing can generate a junction as abrupt as microwave annealing but 25% higher sheet resistance and six times higher instability at 700 °C. 2022-01-01T00:00:00Z https://uvadoc.uva.es/handle/10324/51137 Emergent alternative Si processes and devices have promoted applications outside the usual processing temperature window and the failure of traditional defect kinetics models. These models are based on Ostwald ripening mechanisms, assume pre-established defect configurations and neglect entropic contributions. We performed molecular dynamics simulations of self-interstitial clustering in Si with no assumptions on preferential defect configurations. Relevant identified defects were characterized by their formation enthalpy and vibrational entropy calculated from their local vibrational modes. Our calculations show that entropic terms are key to understand defect kinetics at high temperature. We also show that for each cluster size, defect configurations may appear in different crystallographic orientations and transformations among these configurations are often hampered by energy barriers. This induces the presence of non-expected small-size defect cluster configurations that could be associated to optical signals in low temperature processes. At high temperatures, defect dynamics entails mobility and ripening through a coalescence mechanism. 2022-01-01T00:00:00Z https://uvadoc.uva.es/handle/10324/51136 The effective dopant concentration in p-type Si detectors reduces with irradiation fluence at low fluences due to the acceptor removal process, which degrades detector performance and shortens its lifetime. This effect has been experimentally characterized and parametrized, but its microscopic origin is still unknown. We use atomistic simulations to gain insight into acceptor removal in neutron irradiation by modeling damage generation and defect-dopant interactions. We analyze the effect on dopant deactivation of the Si di- and tri-interstitial diffusion, the inhomogeneity of irradiation damage and the wafer temperature rise during irradiation. We characterize defect generation rates and identify the relevant defect-dopant interactions. Acceptor removal occurs mainly through the formation of Bi pairs and small boron-interstitial clusters, and it is limited by the availability of mobile Si interstitials. The presence of impurities (O, C) modifies B-complexes favoring the formation of BiO, but has a limited effect on the amount of removed acceptors. 2022-01-01T00:00:00Z https://uvadoc.uva.es/handle/10324/45124 Junction stability and donor deactivation in silicon at high doping limit has been a long-standing issue in advanced semiconductor devices. Recently, heavily doped epitaxial Si:P layer with phosphorus concentrations as high as 3 × 1021 at./cm3 has been employed in nanowire field-effect transistor (FET) devices for sub-3 nm technology node as low resistance source-drain and channel stressor. In such highly doped Si:P, the actual dopant activation is much less than nominal phosphorus concentration due to inactive phosphorus atoms arising from dopant-vacancy defects (PnV) clustering phenomenon. Even with state-of-the-art high temperature millisecond annealing, this epitaxial film is thermally unstable upon subsequent thermal treatments. To overcome this limitation, we demonstrate a selective dopant activation scheme which results from the dipole moments of inactive PnV structures within the crystal lattice and their direct energy coupling with the external electric field. It's found that superior stability in dopant activation can be achieved through microwave annealing when a specific temperature and field conditions are met using a triple-parallel-susceptor setup in the microwave cavity. Based on experimental results and ab-initio calculation, we proposed a model, whereas the microwave-PnV interactions result in a specific distribution of dopant defect dominated by thermally stable P4V clusters through elimination of unstable low order PnV, leading to the suppression of donor deactivation and achieving thermally stable junction. 2021-01-01T00:00:00Z https://uvadoc.uva.es/handle/10324/33729 By using classical molecular dynamics simulations and a novel technique to identify defects based on the calculation of atomic strain, we have elucidated the detailed mechanisms leading to the anomalous generation and growth of {001} loops found after ultra-fast laser annealing of ion-implanted Si. We show that the building block of the {001} loops is the very stable Arai tetra-interstitial [N. Arai, S. Takeda, M. Kohyama, Phys. Rev. Lett. 78, 4265 (1997)], but their growth is kinetically prevented within conventional Ostwald ripening mechanisms under standard processing conditions. However, our simulations predict that at temperatures close to the Si melting point, Arai tetra-interstitials directly nucleate at the boundaries of fast diffusing self-interstitial agglomerates, which merge by a coalescence mechanism reaching large sizes in the nanosecond timescale. We demonstrate that the crystallization of such agglomerates into {001} loops and their subsequent growth is mediated by the tensile and compressive strain fields that develop concurrently around the loops. We also show that further annealing produces the unfaulting of {001} loops into perfect dislocations. Besides, from the simulations we have fully characterized the {001} loops, determining their atomic structure, interstitial density and formation energy. 2019-01-01T00:00:00Z https://uvadoc.uva.es/handle/10324/32405 We combine focused experiments with molecular dynamics simulations to investigate in detail the formation of {001} loops in nanosecond laser-annealed silicon. We demonstrate that at temperatures close to the melting point, self-interstitial rich silicon is driven into dense liquid-like droplets that are highly mobile within the solid crystalline matrix. These liquid droplets grow by a coalescence mechanism and eventually transform into {0 0 1} loops through a liquid-to-solid phase transition in the nanosecond timescale. 2018-01-01T00:00:00Z https://uvadoc.uva.es/handle/10324/32400 The construction of realistic atomistic models for amorphous solids is complicated by the fact that they do not have a unique structure. Among the different computational procedures available for this purpose, the melting and rapid quenching process using molecular dynamics simulations is commonly employed as it is simple and physically based. Nevertheless, the cooling rate used during quenching strongly affects the reliability of generated samples, as fast cooling rates result in unrealistic atomistic models. In this study, we have determined the conditions to be fulfilled when simulating the quenching process with molecular dynamics for obtaining amorphous Si (a-Si) atomistic models structurally compatible with experimental samples. We have analyzed the structure of samples generated with cooling rates ranging from 3.3 1010 to 8.5 1014 K/s. The obtained results were compared with experimental data available in the literature, and with samples generated by other state-of-the-art and more sophisticated computational procedures. For cooling rates below 1011 K/s, a-Si samples generated had structural parameters within the range of experimental samples, and comparable to those obtained from other refined modeling procedures. These computationally slow cooling rates are of the same order of magnitude than those experimentally achieved with pulsed energy melting techniques. Samples obtained with faster cooling rates can be further relaxed with annealing simulations, resulting in structural parameters within the range of experimental samples. Nevertheless, the required annealing times are on the order of microseconds, which makes this annealing step non practical from a computational point of view. 2018-01-01T00:00:00Z https://uvadoc.uva.es/handle/10324/31966 In this paper we discuss from an atomistic point of view some of the issues involved in the modeling of electrical characteristics evolution in silicon devices as a result of ion implantation and annealing processes in silicon. In particular, evolution of electrically active dose, sheet resistance and hole mobility has been investigated for high B concentration profiles in pre‐amorphized Si. For this purpose, Hall measurements combined with atomistic kinetic Monte Carlo atomistic simulations have been performed. An apparent anomalous behavior has been observed for the evolution of the active dose and the sheet resistance, in contrast to opposite trend evolutions reported previously. Our results indicate that this anomalous behavior is due to large variations in hole mobility with active dopant concentration, much larger than that associated to the classical dependence of hole mobility with carrier concentration. Simulations suggest that hole mobility is significantly degraded by the presence of a large concentration of boron‐interstitial clusters, indicating the existence of an additional scattering mechanism. Copyright © 2009 John Wiley & Sons, Ltd. 2010-01-01T00:00:00Z https://uvadoc.uva.es/handle/10324/31965 We have used ab initio simulations to study the doping efficiency of amorphous semiconductors, in particular of B-doped amorphous Si. We have found that even in the optimum case of substitutional doping in dangling-bond free amorphous Si the holes provided by B atoms do not behave as free carriers. Instead, they are trapped into regions with locally distorted bond angles. Thus, the effective activation energy for hole conduction turns to be the hole binding energy to these traps. In the case of high B concentration, the trap states move deeper in the gap and the binding energy and spatial localization of holes increase. In addition, B atoms have lower energies for shorter bond lengths, configurations favored in the vicinity of these traps. 2010-01-01T00:00:00Z https://uvadoc.uva.es/handle/10324/31964 Ion implantation continues being the dominant technique to introduce dopants in Si devices. With the device feature size in the nanometer scale, the accurate and detailed description of as-implanted dopant and damage profiles is becoming key as advanced annealing techniques are almost diffusionless. The demanding requirements for ultrashallow junction formation are stimulating the development of improved and detailed models for molecular implants and for the kinetics of amorphous damage. Additional challenges arise in the doping of advanced architectures, such as fin field effect transistors, because the introduction of highly tilted ions is quite inefficient and, in addition, the regrowth of amorphous regions in narrow structures is hampered by the slow regrowth at free interfaces and {111} planes. Atomistic simulations play a relevant role to provide the understanding for the development of simplified physically based models computationally more efficient. 2010-01-01T00:00:00Z https://uvadoc.uva.es/handle/10324/31963 We have analyzed the atomic rearrangements underlying self-diffusion in amorphous Si during annealing using tight-binding molecular dynamics simulations. Two types of amorphous samples with different structural features were used to analyze the influence of coordination defects. We have identified several types of atomic rearrangement mechanisms, and we have obtained an effective migration energy of around 1 eV. We found similar migration energies for both types of samples, but higher diffusivities in the one with a higher initial percentage of coordination defects. 2011-01-01T00:00:00Z