07_Introduction (part 1) - Yiwei666/08_computional-chemistry-learning-materials- GitHub Wiki

1. An Ab Initio Molecular Dynamics Simulation of Liquid FeO–SiO2 Silicate System with Sulfur Dissolving

https://doi.org/10.1007/s11663-021-02263-x

Sulfur enters molten steel from raw materials such as pig iron and direct reduced iron. The sulfur content in steel should be as low as possible. Otherwise, due to the formation of eutectic iron sulfide, hot embrittlement of the steel may occur. Even ultra-low levels of sulfur can limit an alloy’s useful operating life in critical applications like aerospace engine components. Thus, ultra-low sulfur production is always an important issue in the metallurgical process.[1–6]

Generally, sulfur can be removed by the silicate system slag during the steelmaking or refining process. In the 1950s, Richardson et al.[7] explained that S exists in terms of S2 in the molten silicate when the oxygen potential is lower than 10^7 atm from a thermodynamic viewpoint. Then, several parameters have been proposed to evaluate the desulfurization capacity of silicate system slags, such as sulfur capacity[7–10] or optical basicity.[11–14] Meanwhile, the gas/slag equilibrium technique was widely used to determine the effects of temperature, oxygen partial pressure as well as slag chemical composition on the sulfur removal or participation ratio.[15–17] Nevertheless, the behavior of sulfur in the silicate structure remains unclear.

Due to the limitations of high-temperature experiments, it is difficult to explain the mechanism of macroscopic property changes from the perspective of microstructure. With the development of computational capacity, the simulation of the silicate structure has been carried out in the past decades.[19–22] Classical molecular dynamics (MD) has been used to provide useful information on simple binary silicate systems,[23–28] it can provide the structural unit changes of silicate systems with different compositions and different temperatures, but it cannot conduct further research on the electric charge. Ab Initio molecular dynamics (AIMD) based on first principles can calculate the electronic structure and energy of the simulated system at a certain time step and the force of each nucleus. This method overcomes the limitations of classical molecular dynamics simulation and can combine first principles or semi-empirical electronic structure calculation methods with molecular dynamics simulation to realize the essential understanding of the melting structure information of silicate under high-temperature conditions. This method is often used to study catalysis,[29,30] proton transfer,[31] etc. Meantime, the ab initio molecular dynamics simulations have been successfully applied in the melting process of liquid metal[32–35] and the silicate oxide system.[20,36–40] In the silicate oxide system, Johannes[41] studied the structure and electronic properties of liquid and amorphous SiO2. De Koler[37] and Georg Spiekermann[42] study the thermodynamic properties and structural vibration properties of the binary system MgO–SiO2 by AIMD. For the FeO–SiO2 system, Huang[43] investigated Fe–Si–O ternaries from 3800 K to 4800 K at CMB pressure (136 GPa). However, there is little research on charge information, element valence information, and desulfurization mechanism for FeO–SiO2, and further research is still needed.

In the current work, systematic investigations of FeO–SiO2 silicate were performed to obtain its structural feature and reveal the influence of sulfur corporation on the internal bonding and diffusion features using AIMD simulations.

2. Comparison of desulfurization mechanism in liquid CaO-SiO2 and MnO-SiO2: An ab initio molecular dynamics simulation

https://doi.org/10.1016/j.jallcom.2021.163008

1. Introduction

Consequently, this study focuses on the comparative analysis of the different desulfurization mechanisms in the two liquid CaO-SiO2 and MnO-SiO2 slag.

3. Effect of alkaline oxides on aluminate slag structure by first principles calculation

https://doi.org/10.1016/j.molliq.2023.123088

1. Introduction

In recent years, in order to comply with the application of high-alumina based iron ore 1(https://www.sciencedirect.com/science/article/pii/S0167732223018949?via%3Dihub#b0005), the improvement of refining effect 2(https://www.sciencedirect.com/science/article/pii/S0167732223018949?via%3Dihub#b0010), and the production of high alloy steel (high manganese and high aluminum steel) 3(https://www.sciencedirect.com/science/article/pii/S0167732223018949?via%3Dihub#b0015), the research of CaO-Al2O3-based slag has been paid more attention. Particularly, aluminate-based mold flux has become the research focus because of the alleviation or avoidance of steel-slag reactivity. Among the properties of mold flux, the melting temperature and viscosity directly affect the lubrication effect, and the breaking temperature of viscosity-temperature curve (Tbr) and crystallization properties influence the heat transfer performance. Moreover, the properties of the slag are closely related to its structure, so the variation of properties is usually explained by studying the structure.

In order to realize the effective regulation of alkaline oxide on the macroscopic properties of aluminate mold flux, it is necessary to master the mechanism from the microstructure level. Due to limited computational conditions, only relatively simple systems are studied, thus providing theoretical guidance for complex systems. In the aluminate slag, alkaline oxides partly stabilize the Al-O network structure by playing a charge compensation role of the excess negative charge. However, it is not known whether the alkaline cation is bonded to Al or O in the [AlO4]5- tetrahedron. Although it is common to use molecular dynamics simulation to study the microstructure of slag, there are some limitations for multicomponent slag simulation because the potential parameters and potential functions are obtained based on the diatomic systems. Therefore, in this paper, the first-principles molecular dynamics simulation is used to study slag structure, and the partial density of states (PDOS) and electron localization function (ELF) of atom are analyzed to realize the multi-scale characterization of slag microstructure at different levels of electron-atom-molecule, providing theoretical guidance for the stability control of slag structure and properties.

4. First-principles study on microstructure of CaO-Al2O3-B2O3 slag

https://doi.org/10.1016/j.molliq.2022.120738

1. Introduction

Since the slag viscosity and conductivity are closely related to its composition and temperature, it is essentially dependent on the microstructure. Therefore, only by studying the microstructure can control the performance of slag thoroughly. However, the structure research is limited in obtaining structural parameters from atomic and molecular scales to analyze the structure evolution by spectroscopy experiments and molecular dynamics simulation at present, while there is rare research on explaining atomic bonding from the electronic level. In general, the slag viscosity is reduced 16(https://www.sciencedirect.com/science/article/pii/S0167732222022772?via%3Dihub#b0080), 17(https://www.sciencedirect.com/science/article/pii/S0167732222022772?via%3Dihub#b0085), 18(https://www.sciencedirect.com/science/article/pii/S0167732222022772?via%3Dihub#b0090), 19(https://www.sciencedirect.com/science/article/pii/S0167732222022772?via%3Dihub#b0095), 20(https://www.sciencedirect.com/science/article/pii/S0167732222022772?via%3Dihub#b0100), 21(https://www.sciencedirect.com/science/article/pii/S0167732222022772?via%3Dihub#b0105), 22(https://www.sciencedirect.com/science/article/pii/S0167732222022772?via%3Dihub#b0110) due to the structure depolymerization with C/A addition 12(https://www.sciencedirect.com/science/article/pii/S0167732222022772?via%3Dihub#b0060), unless solid phase particles precipitated 2(https://www.sciencedirect.com/science/article/pii/S0167732222022772?via%3Dihub#b0010), 3(https://www.sciencedirect.com/science/article/pii/S0167732222022772?via%3Dihub#b0015), 4(https://www.sciencedirect.com/science/article/pii/S0167732222022772?via%3Dihub#b0020) when C/A beyond a certain range. Nevertheless, the structure will be more complex alongside the increase of C/A 23(https://www.sciencedirect.com/science/article/pii/S0167732222022772?via%3Dihub#b0115), due to the excessive charge compensation, which makes the free oxygen and non-bridge oxygen transform into bridge oxygen and tricluster oxygen, meanwhile, QAl1 and QAl2 polymerize into QAl4 and QAl5. Since the B or Al in CaO-Al2O3-B2O3-based slag can only lose three electrons in the outermost layer to form B3+ or Al3+, it presents excess negative charge when more than three O atoms connected with B or Al, which requires metal cations to play a charge compensation role in stabilizing the network structure. However, the details about bonding mode between atoms and how charges compensate for [AlO4]5- and [BO4]5- tetrahedra are unknown.

The classical molecular dynamics simulation mostly uses the Born-Mayer-Huggins (BMH) potential, in which the Coulomb potential is an important part. The qi and qj represent the charges of particles, which are usually set as constant in previous studies 20(https://www.sciencedirect.com/science/article/pii/S0167732222022772?via%3Dihub#b0100), 24(https://www.sciencedirect.com/science/article/pii/S0167732222022772?via%3Dihub#b0120), 25(https://www.sciencedirect.com/science/article/pii/S0167732222022772?via%3Dihub#b0125), 26(https://www.sciencedirect.com/science/article/pii/S0167732222022772?via%3Dihub#b0130). In the study of CaO-Al2O3-B2O3 system, the existence of multiple coordination conditions of Al and B inevitably followed with significant changes of charge. Therefore, a train of deviations need to be settled in the classical molecular dynamics simulation, whereas the first-principles molecular dynamics simulation 27(https://www.sciencedirect.com/science/article/pii/S0167732222022772?via%3Dihub#b0135) can analyze the charge and electron localization of each atom to improve the accuracy of results. Moreover, the multi-scale characterization of slag microstructure at different levels of electron-atom-molecule can provide theoretical guidance for the stability control of slag structure and performance.

5. Stabilization mechanism of arsenic-sulfide slag by density functional theory calculation of arsenic-sulfide clusters

https://doi.org/10.1016/j.jhazmat.2020.124567

1. Introduction

Arsenic (As) is a typical deadly pollutant in the wastewater of smelting industries, which is normally detoxified by chemical precipitation (Liu and Yu, 2016, Luukkonen et al., 2016, Zhu et al., 2017, Wang et al., 2018, Wei et al., 2018, Zhang et al., 2019, Li et al., 2020). As such, more than 10 tons of As sulfide slag (As‒S slag) could be produced annually (Drahota and Filippi, 2009, Morales et al., 2010). However, the relatively weak stability of As‒S slag in ambient conditions would allow the As release again into the environment, causing severely secondary pollution (Qu et al., 2013, Yang et al., 2013, Roessler et al., 2014, Tao et al., 2020, Liu et al., 2018, Seidl et al., 2019, Wang et al., 2015).

To stabilize the As‒S slag, solidification methods have been developed, but they suffer from relatively high enlargement ratio and cost, and complicated process (Kumar et al., 2016, Li et al., 2018, Li et al., 2018, Lu et al., 2017, Min et al., 2019). Most recently, hydrothermal treatment was proposed to dispose the As‒S slag by creating necessary conditions to re-build the structure of As‒S slag (Qiu et al., 2016, Yao et al., 2018). As a consequence, the structure of As‒S slag was modified with improved stability towards the harsh environment. However, the understanding of relationship between As‒S structure and stability of As‒S slag is missing, which in turn hampers a further development of As stabilization.

In recent years, the rapid development of clusters science boosts the fundamental research of materials since clusters structure can be regarded as a basis to understand physicochemical properties (Bouveyron, 2016, Gagne et al., 2010, Trippl et al., 2015, Mardirossian and Head-Gordon, 2017, Chamroukhi and Nguyen, 2019). Currently, limited examples are emerging on the cluster structure of As‒S compounds. For instance, based on Gaussian-03 scheme calculation, Yang et al. investigated the structure of As-rich As‒S clusters and predicted that the clusters are stable (Yang et al., 2013, Hou et al., 2014). Guillermo et al. revealed the geometrical structure of S-rich As‒S clusters with combination of results and analysis of mass spectroscopy (Ramírez-Galicia et al., 2010). Although great progress has been made, these researches are irrelevant with stabilization of As‒S slag (Splat et al., 2005, Pangavhane et al., 2010). As revealed, the As‒S structure is an important factor to the stability of As‒S slag and the clusters can be regarded as basic molecular skeleton in As‒S slag (Liang et al., 2011). Therefore, the structure of As‒S clusters and further the correlation of cluster structure with its properties, typically the bonding behavior and electronic information, are very important for a rational design of stabilization strategy (Yao et al., 2019, Xu et al., 2020). However, seldom effort has been paid to figuring out how the S-to-As molar ratio influences the clusters structure and why specific structure could increase the slag stability.

Here we investigated the geometric structure of various As‒S clusters based on density functional theory to simulate their structures. Typically, the cluster energy, and electron and orbital characteristics were analyzed, and the As‒S interaction behavior was revealed. The bonding characteristics and energy analysis demonstrate the structure of S multimers-covering-(As2S3)n is of the highest stability. Furthermore, the orbital composition indicates that the stability of this structure is stemmed from the 4p-orbital in As atom binding with 3p-orbital of S atom, which brings down the HOMO energy level. Motivated by this theoretical result, rational design was conducted on hydrothermal treatment of As‒S slag by adding pure S powders. Such a hydrothermally treated slag possesses a very high stability as revealed by leaching experiment that the As concentration in the leaching solution is only 0.8 mg/L.

6. Ab initio molecular dynamics assessment of thermodynamic and transport properties in (K,Li)Cl and (K, Na)Cl molten salt mixtures

https://doi.org/10.1016/j.molliq.2020.115262

1. Introduction

Due to their highly desirable physicochemical properties, molten salts (MSs) have numerous important technological applications [1(https://www.sciencedirect.com/science/article/pii/S0167732220375048?via%3Dihub#bb0005), 2(https://www.sciencedirect.com/science/article/pii/S0167732220375048?via%3Dihub#bb0010), 3(https://www.sciencedirect.com/science/article/pii/S0167732220375048?via%3Dihub#bb0015), 4(https://www.sciencedirect.com/science/article/pii/S0167732220375048?via%3Dihub#bb0020), 5(https://www.sciencedirect.com/science/article/pii/S0167732220375048?via%3Dihub#bb0025)]. Used as electrolytes, MSs [5] offer high energy density and high power density that makes them ideal for these high temperature electrochemical energy storage devices. Their excellent thermal, chemical, and radiochemical stabilities make MSs appealing choices as coolants [6], and/or fuels in nuclear reactors for energy production [3]. At a nuclear reactor core, where large amounts of heat as well as different actinides are generated, quick heat transport is crucial. MSs show numerous advantages over gases (H2O, CO2, etc.), as they have much higher heat capacities, higher boiling points, lower vapor pressure, and higher chemical stability against radiation [4]. In separations, particularly in spent nuclear fuel reprocessing [4,7], pyroprocess using MSs, is an attractive approach, largely due to their low chemical sensitivity to high levels of radiation, and proliferation resistance. Electro-separation of actinides from rare earths [7] in MSs is also a widely used separation application of actinide in MSs systems. These typical examples, amongst many others, are at the center of renewed interest in fundamental studies of many molten salt (MS) properties.

Molecular understanding of MSs has mainly relied on experimental techniques such as neutron scattering or X-ray diffraction 8]. However, measuring properties of MSs at very high temperatures under hazardous or corrosive conditions is difficult and costly. In addition, MSs are liquid-phase multi-component mixtures whose local structure and speciation are very difficult to determine. Luckily, computer simulations using theoretical modeling techniques such as molecular dynamics or Monte Carlo provide a low-cost alternative approach [[8(https://www.sciencedirect.com/science/article/pii/S0167732220375048?via%3Dihub#bb0040), 9(https://www.sciencedirect.com/science/article/pii/S0167732220375048?via%3Dihub#bb0045), 10(https://www.sciencedirect.com/science/article/pii/S0167732220375048?via%3Dihub#bb0050)]. For atomistic simulations of MS systems, the treatment of inter-atomic interactions is crucial. Since the early days of computer-based molecular simulations, MS modeling has been largely done using force field potentials 8,9,11]. Most commonly, the inter-atomic forces are described with additive pairwise potentials consisting of terms like Coulombic attraction/repulsion, van der Waals dipole-dipole/dipole-quadrupole interactions [12,13]. These so-called rigid ion models omit inter-atomic forces arising from polarization. Rahman et al. showed that on top of these two-body potentials, the polarization, treated in a shell-model type fashion [14], leads to a significant increase in ionic diffusion coefficient, and in a reduction in the characteristic frequencies [15]. The parameterization of these potentials largely depends on reproducing experimental data. More recently, sophisticated procedures based on first principles for potential parameterization assessing polarization have been applied and showed certain success in reproducing static and dynamic properties of MSs system [9,[16(https://www.sciencedirect.com/science/article/pii/S0167732220375048?via%3Dihub#bb0080), 17(https://www.sciencedirect.com/science/article/pii/S0167732220375048?via%3Dihub#bb0085), 18(https://www.sciencedirect.com/science/article/pii/S0167732220375048?via%3Dihub#bb0090)]. However, such a procedure based on the fitting with first principles potentials becomes challenging as the time scale and the complexity of multicomponent systems increase. One can always question the transferability of these potentials as physical and chemical properties of an ion can change significantly from on coordination environment to another. Furthermore, classical force fields implicitly ignore the electronic degrees of freedom when ions change their coordination environments. As such, processes including charge transfer are not captured, and the most obvious consequence is that important chemical processes such as redox cannot be described.

Recently quantum mechanics-based molecular dynamics has been applied to study MSs [19(https://www.sciencedirect.com/science/article/pii/S0167732220375048?via%3Dihub#bb0095), 20(https://www.sciencedirect.com/science/article/pii/S0167732220375048?via%3Dihub#bb0100), 21(https://www.sciencedirect.com/science/article/pii/S0167732220375048?via%3Dihub#bb0105), 22(https://www.sciencedirect.com/science/article/pii/S0167732220375048?via%3Dihub#bb0110), 23(https://www.sciencedirect.com/science/article/pii/S0167732220375048?via%3Dihub#bb0115), 24(https://www.sciencedirect.com/science/article/pii/S0167732220375048?via%3Dihub#bb0120), 25(https://www.sciencedirect.com/science/article/pii/S0167732220375048?via%3Dihub#bb0125)]. Galamba et al. [19,20]., by employing density functional theory (DFT) – based molecular dynamics (MD) simulations to study molten NaCl and NaI, found that the main differences in structural and transport properties between results from the rigid ion and ab initio approaches are related to polarization effects. Bengtson et al. [22], performing ab initio molecular dynamics (AIMD) simulations on the eutectic (K, Li)Cl mixture, showed that AIMD is a useful predictive tool for properties such as bulk modulus of which experimental data are not available, or complex thermodynamics properties like the free energy of mixing which help understand phase stability.

In practical applications, MS mixtures are more often used than pure MSs. In principle, a mixture of MSs would lead to tunable properties compared to either of the pure components. This is particularly true in the case of thermodynamic properties of the salts. For example, the melting point of pure LiF, NaF, and KF is 848, 993, and 858 °C, respectively, while that of their eutectic mixture is 454 °C 26]. Using mixtures, instead of single-component MSs, reduces harmful effects in electrochemical cell such as the corrosion of the cell and high vapor pressure of the salt [27]. Moreover, according to Bloom “mixtures of molten salts often behave as if they contain new ionic species, i.e., complex ions, formed by interaction between simple ions” [28]. Thus, applications aside, MS mixtures represent an interesting research subject. In the study of MS mixtures, physicochemical properties are usually discussed as a function of composition (molar fraction) [[29(https://www.sciencedirect.com/science/article/pii/S0167732220375048?via%3Dihub#bb0145), 30(https://www.sciencedirect.com/science/article/pii/S0167732220375048?via%3Dihub#bb0150), 31(https://www.sciencedirect.com/science/article/pii/S0167732220375048?via%3Dihub#bb0155), 32(https://www.sciencedirect.com/science/article/pii/S0167732220375048?via%3Dihub#bb0160), 33(https://www.sciencedirect.com/science/article/pii/S0167732220375048?via%3Dihub#bb0165), 34(https://www.sciencedirect.com/science/article/pii/S0167732220375048?via%3Dihub#bb0170), 35(https://www.sciencedirect.com/science/article/pii/S0167732220375048?via%3Dihub#bb0175)]. Combined with Job's method [36], this allows us to know whether complex ions are formed in a mixture [32] and how to determine their composition [32,37]. In most cases, properties of MSs are usually not strictly additive, i.e., the values of a property do not depend linearly on the composition. This includes physical properties like electrical conductivity [29] and surface tension [33,38], and chemical properties like activity [34]. The nonlinear correlation is indicative of strong interactions between MS components in the mixtures. On the other hand, other properties like molar volume show varying behaviors in different mixtures. The molar volume of systems like (Ag, K)Br, (Ag, K)Cl, and Ag(Cl, Br) mixtures is strictly additive [39], i.e., it obeys a linear correlation with composition, as in mixtures of non-interacting systems, whereas, (Pb, K)Cl, and (Cd, K)Cl both show positive deviations from additivity [39]. The behavior in (Pb, K)Cl or (Cd, K)Cl indicates a higher degree of covalence in the mixture than in each components [32]. To understand this in detail, it is necessary to track down the partial coordination numbers (CN) as well as the atomic charges of the ions in the systems. Experimentally, this is not a trivial task for MS mixtures, but AIMD can be used to study such fluctuations. To date, very little has been done on MS mixtures using AIMD.

In this work, we report on an AIMD study of structural, thermal, electronic and transport properties of MS mixtures at different compositions and temperatures. We considered two mixtures of three common metal halides KCl, NaCl, and LiCl: (K, Li)Cl and (K, Na)Cl. Most of the properties of these mixtures were found to depend on composition and temperature. While atomic charge distributions showed an additive behavior, the ionic conductivity showed negative deviation from additivity. Our structural and transport results appeared to be more consistent with experimental data compared with classical force field results. We will also show that the mixing of the salts is mainly driven by entropy and that, KCl and LiCl mix better than KCl and NaCl. This study provides a much-needed look into MS systems and also sets a new benchmark for atomistic simulations of these systems.

7. Influence of Glass Composition on the Luminescence Mechanisms of CdSe Quantum-Dot-Doped Glasses

https://doi.org/10.1021/acs.jpcc.1c04665

Quantum dots (QDs) exhibit remarkable electronic and optical properties induced by quantum confinement. Colloidal quantum dots can be produced at a low cost with a high quantum efficiency and narrow size distribution, which makes them promising for photovoltaics, (1) lighting, (2) and labeling (3) applications. Incorporation of QDs into glasses can prevent the agglomeration and enhance the chemical and thermal abilities, making them appealing for nonlinear optical devices and LEDs. (4,5)

Due to the high surface-to-volume ratio of quantum dots, the main obstacle for the success of their applications is to achieve a precise control of the emission properties, which mainly depends on the surface chemistry of the nanocrystals, and the interface with the surrounding medium. Surface defects are ubiquitous to QDs, and an in-depth understanding of their atomic origin is necessary to tune the emission and design defect-free QDs. In spite of a wealth of experimental data reported for CdSe QDs, the question of exactly how surface defects influence the photophysical properties of QDs is still open. It is experimentally challenging to resolve spectral features originating from QD–ligand interactions. Density functional theory (DFT) is able to address these issues, gaining comparable gaps between the highest occupied (HOMO) and the lowest unoccupied molecular orbitals (LUMO), IR, and Raman spectra to experimental findings. (6)

In the literature, Cd33Se33 QDs are the most commonly used representative models of quantum dots for computational studies, as they are believed to have the highest degree of stability, exhibiting basic optical features similar to larger QDs. Kilina et al. (7) found that the electron trap states were very sensitive to ligand position and the solvent polarity. Sargent et al. (8) demonstrated that surface vacancies can improve the fluorescence yield compared to vacancy-free surfaces. Beratan et al. (9) performed comparative studies of the structural and electronic properties of pristine and NH3–, SCH3–, and OPH2– capped Cd33Se33 QDs, showing that an increase in capping ligand concentration causes noticeable changes in the capped QD electronic structure, leading to different surface states and HOMO–LUMO gaps. Pudzder et al. (10) reported significant reconstruction of the nanoparticle surface, while the wurtzite core was maintained, leading to the opening of an optical gap without the aid of passivating ligands, thus “self-healing” the surface electronic structure.

All of those theoretical studies helped understand the structural origin of defect emission and effective surface passivation of colloidal QDs. However, relatively little research effort has been devoted to the computational study of CdSe quantum dots embedded in glasses compared to their colloidal counterparts. The quantum efficiency of the quantum dots embedded in glass matrices is extremely low, which is usually attributed to their surface passivation and broad size distribution. Recent work showed that the size distribution can be controlled by tuning the heat treatment and the concentration of dopants. (11) However, surface control techniques used in colloidal QDs, such as post-synthetic passivation by surfactants or growth of an iso-structural shell, are difficult to perform when the QDs are surrounded by dense and amorphous glass matrices. Hence, defect emission, rather than intrinsic emission, is most frequently observed in the CdSe quantum-dot-doped glasses. It is of predominant importance to fully comprehend the structural origin of defect emission and the impact of QD/glass interface on the electronic structure. The atomic structure of the CdSe quantum-dot-doped glasses is exceedingly difficult to characterize due to the typical low density of QDs in the glass and the instability of glass matrices under electron beam. In a previous study, (12) we used network modifiers as well as nonbridging oxygen atoms to be capped with Cd33Se33 clusters, to explore the impact of these additional surface atoms on the morphology and electronic structure on the CdSe QD. However, this simplistic model was far from being a realistic environment of the quantum dot in glass matrices. In later work, we created a more detailed atomic structure using a combination of classical and ab initio molecular dynamics, demonstrating the complex interfacial chemical environment between the CdSe quantum dot and a surrounding glass matrix. (13)

In the present work, we used DFT-based methods to determine the structure of CdSe quantum-dot-doped glasses with varying composition and to determine their electronic structure to probe the impact of the glass matrices on the optical properties of the CdSe quantum dot. Besides, the sodium ion was found to sharply decrease the HOMO–LUMO gap in previous DFT calculations. (12) In their role of glass modifiers, sodium ions can alter the structure of silicate glasses, and it was experimentally observed that there is a disappearance of the visible absorption of Se–Se color centers upon increasing concentration of Na2O in a silicate glass doped with ZnSe. (14) The Se–Se color centers are considered to be the nucleation sites of the CdSe QD. Therefore, in this study, we changed the amount of Na2O in the glass (Na2O)x(SiO2)1–x (x = 0, 0.25, 0.33, and 0.5 in molar fraction) to explore the compositional dependence of the atomic and electronic structures of CdSe quantum dot in glass matrices. The results demonstrate that an increase in the amount of Na2O contributes to the formation of Cd–O and Se–Na bonds, with breaking of Si–O, O–Na, and Cd–Se bonds. The density of states (DOS) and projected density of states (PDOS) were also calculated. Although with the same composition, the dominant luminescence mechanisms in different configurations are also different. The top of the valence band and the bottom of the conduction band are decided by the hybrid QD in most compositions. However, in most configurations of CdSe, quantum-dot-doped glass with composition 0.33 Na2O–0.67 SiO2 exhibits totally different luminescence mechanisms that the top of the valence band is determined by the hybrid glasses.

Most importantly, based on the analysis of atomic structure, HOMO–LUMO gap distribution, and DOS, the intrinsic emission from pristine QD is found to be negligible due to the little possibility of QD without any interaction with glass matrix. The photoluminescence of CdSe quantum dots doped glass is originated from the intrinsic emission of these complicated hybrid system rather than the intrinsic or defect emission of pristine QD, which is a common model proposed by experimentalists. These results provide a better understanding of the electronic structure and luminescence mechanisms of these complex systems, giving guidance for future compositional design of highly luminescent glass containing quantum dots.

8. Ab Initio Molecular Dynamics of CdSe Quantum-Dot-Doped Glasses

https://doi.org/10.1021/jacs.9b12073

The tunable optical and electronic properties of quantum dots induced by quantum confinement have stimulated enormous research interest. These properties were extensively explored for applications in the fields of lasers, (1) light-emitting diodes (LEDs), (2) and biolabeling. (3) Colloidal quantum dots exhibit excellent luminescence properties; however, the agglomeration of these QDs in solution severely restricts their practical applications. As an alternative, quantum-dot-doped glasses combine the good thermal and chemical stability of glass with easy access to device fabrication, hence offering potential applications in nonlinear optical devices (4) and LEDs. (5)

Quantum dots were first fabricated in a glass matrix by Ekimov et al. in 1981. (6) Brus (7) et al. first synthesized quantum dots in colloidal solutions in 1983. The past few decades have witnessed the blossoming of research in quantum dots, and chemically fabricated quantum dots can exhibit a high photoluminescence quantum yield (PLQY) after surface passivation. Compared to their colloidal counterparts, quantum dots embedded in a glass matrix display poor PLQY, and the highest PLQYs for CdSe embedded in glass have been measured at 3% (8) until now. The enormous difference in the PLQY between QDs in colloidal solutions and in glass matrixes naturally leads to the following question: why is the quantum efficiency in the glass matrix so low? How can this situation be improved to promote the practical applications of quantum dots? Experimentalists have worked for decades to shed light on these issues, and some mechanisms have been established to explain this phenomenon. It has been assumed that the presence of defects at the interface between QD and the glass matrix can quench the excitonic emission and produce unfavorable defect emission, which is detrimental to applications. (9) However, it is experimentally challenging to probe the origins of the interfacial defects and to characterize the structure and chemical environment of CdSe quantum-dot-doped glass because of the inadequate resolution of the existing techniques, the instability of the glass matrix under an electronic beam, and the relatively low concentration of QDs in the glass.

Theoretical modeling has been applied as an alternative to bring insight into these issues, and a comprehensive exploration of the local atomic structure can give guidance on the design of highly luminescent glasses. Computational materials studies have become very popular due to increasing computational power and the development of efficient numerical algorithms and can investigate a system at the atomistic level, which is not directly possible in experiments, gaining insight into both physical and chemical properties of materials. The two main computational methods for materials are classical approaches and ab initio simulations. Both classical molecular dynamics (MD) and ab initio molecular dynamics (AIMD) have been successfully used in modeling various multicomponent glass materials (10−16) and CdSe nanocrystals. (17−23) However, in spite of these extensive theoretical studies for these two separate systems, no atomistic calculations have been conducted so far on the composite QD/glass systems and their interfacial properties.

In this work, we have employed ab initio molecular dynamics in order to model CdSe QD embedded in a glass matrix, with a particular focus on the reconstruction occurring at the QD/glass interface. We started from classical MD calculations of the glass matrix and quantum chemical models of the quantum dots. Because the interaction between the CdSe QD and glass matrix cannot be described accurately in classical MD due to the lack of the appropriate interatomic potentials, we used ab initio methods. Our ab initio MD methodology is based on quantum mechanics at the density functional theory (DFT) level, which means it does not rely on a fixed functional form for interatomic interactions, but the electronic degrees of freedom of each atom are fully modeled. Due to the limitation of computational power cost, we chose a model of the composite system with the composition of 60Na2O-120SiO2-33CdSe to conduct the AIMD simulations. The radial distribution functions for Si–O, Na–O, Cd–Se, Cd–O, Se–O, Se–Na, Se–Se, and Cd–Cd were calculated to compare with experimental counterparts. The coordination environment and the ring structures were analyzed. The results demonstrate that enormous structural reconstruction happens simultaneously in QD and the glass matrix, with the creation of Cd–O bonds and Se–Na bonds at the interface. The incorporation of the CdSe QD disrupts Na–O bonds, while stronger SiO4 tetrahedra are reformed. The glass matrix contributes to great structural reconstruction at the external surface of the quantum dot, making it hard to maintain the bulk structure even at its core. The results reported here can give a better fundamental understanding of this complex system and give insight into the fabrication of highly luminescent glasses containing quantum dots.

9. Thermodynamic criteria of the end-of-life silicon wafers refining for closing the recycling loop of photovoltaic panels

https://doi.org/10.1080/14686996.2019.1641429

1 Introduction

As a clean and renewable energy source, solar energy shows great environmental advantages compared with fossil-fuel energy, such as reducing greenhouse gas (GHG) emissions, air and water pollution, and saving natural energy resources [Citation1,Citation2]. The usage of solar energy based on photovoltaic (PV) technology has skyrocketed during the past decade, as shown in Figure 1 [Citation3]. The total capacity of solar PV installed reached 404.5 GW by the end of 2017, which increased over 300 times from the beginning of the century (2000: 1.3 GW) [Citation3]. Moreover, this value is expected to exceed 500 GW in 2018, and reach terawatt (TW) level (1270.5 GW) by the end of 2022 [Citation3]. The PV panels installed generally have a long service life of 25–30 years, and the first batch of PV panels installed are now being retired. Given the growth of the PV market, the cumulative amount of end-of-life (EoL) PV panels was estimated to be over 70 million tonnes in 2050 [Citation4].

The coming boom of the EoL PV panels presents another environmental issue because they are classified as hazardous and toxic waste [Citation2,Citation5]. The EoL PV panels contain heavy metals, such as lead, tin, and cadmium, which cause environmental pollution and pose threats to human health [Citation5]. The European Union recently revised the European Waste Electrical and Electronic Equipment (WEEE) Directive, where EoL PV panels were added under the category of discharged electrical/electronic waste [Citation6]. Henceforth, it is mandatory to define alternative strategies to landfilling, and recycling the EoL PV panels has become an obligation. On the other hand, recycling of EoL PV panels shows great environmental advantages, such as energy conservation, CO2 emissions reduction, natural resources conservation, reduction of landfilling, and reduction of heavy metals pollution [Citation2,Citation4,Citation5,Citation7].

Developing an efficient recycling system is becoming crucial to face the coming boom of the EoL PV panels [Citation4,Citation5,Citation8]. By now, over 90% of the global PV market is dominated by the crystalline silicon (c-Si)-based PV panels, where the silicon wafer is the most important core part [Citation9]. The silicon wafer, as shown in Figure 2(a), produced from solar-grade silicon (SoG-Si) (>99.9999% purity), is typically a moderately doped p-type c-Si semiconductor with a heavy doped n+-type layer on the top side and heavy doped p+-type layer on the back side. Figure 2(b) schematically shows the recycling loop of silicon wafers from EoL c-Si PV panels for new SoG-Si. The recycling process of the EoL c-Si PV panels starts from the disassembly of the sandwich layer-like structure of the EoL silicon wafers. The silicon wafers can be separated by methods such as mechanical crushing, pyrolysis, organic solvent, or acid etching to remove the encapsulant and electrode materials [Citation4,Citation5,Citation8,Citation10]. The wafers obtained without any damages can be reused directly, while most of them with edge chippings, micro-cracks, or electronic damages would be subject to a refining process to recycle the silicon metal [Citation11,Citation12].

Refining the EoL silicon wafers becomes the key to close the recycling loop of the PV panels [Citation13–Citation15]. Figure 3 compares the concentrations of typical impurity elements in EoL silicon wafers and metallurgical-grade silicon (MG-Si), the raw materials with purity of approximately 98% produced by reducing quartz from natural ore [Citation16,Citation17]. The typical impurity elements in MG-Si are aluminum, iron, titanium, and calcium with relatively fixed concentration. On the other hand, the impurity elements in the EoL silicon wafer are mainly from the doping elements used to form the semiconductors. The typical doping elements include the group IIIA atoms as acceptors for p-type semiconductors, such as boron, and group VA atoms as donors for n-type semiconductors, such as phosphorus, arsenic, and antimony [Citation17,Citation18]. The mixture of different types of wafers and the incomplete separation of the attached materials, such as aluminum alloy frame, silver grid line, and tin or copper wire, make the composition of the collected EoL c-Si products more complicated and variable than that of MG-Si. The contamination by these impurity elements dramatically downgrades the recycled silicon. Thus, the impurity elements must be eliminated before the use of silicon in new PV panels [Citation13–Citation15,Citation19].

Chemical etching using solutions such as HNO3+ HF + CH3COOH + Br2 or HF + HNO3 + H2SO4 + CH3COOH + CMP-MO-2 (surfactant) has been attempted to remove the metal impurities such as metal electrodes and n-p junction [Citation10,Citation20,Citation21]. However, long treatment duration and a complicated hazardous acid mixture discharge make it difficult to be employed in practical industrial applications. On the other hand, the metallurgical refining process, which has been attempted for the MG-Si purification (Supporting Information) [Citation16,Citation22–Citation25], is simple and well-established. The metallurgical refining process is also a promising method for purification of the EoL silicon wafers as the new resource for SoG-Si in new PV panels.

This study is meant to systemically examine the thermodynamic criteria of the metallurgical refining process of the EoL silicon wafers for closing the recycling loop of EoL c-Si PV panels. The elimination limitation of impurity elements by three most typical metallurgical refining processes: oxidation refining, evaporation refining, and solvent refining, was quantitatively evaluated using a developed thermodynamic method [Citation26–Citation33]. A total of 42 impurity elements that are likely to be present in the collected EoL PV panels were considered. All the thermodynamic parameters were used in the evaluation, and the influence of the physico-chemical factors including temperature, oxygen partial pressure, and slag composition was extensively examined.

10. The effect of CaO(MgO) on the structure and properties of aluminosilicate system by molecular dynamics simulation

https://doi.org/10.1016/j.molliq.2018.07.123

1. Introduction

Aluminosilicates are the most common and important slag systems in the metallurgical process [1(https://www.sciencedirect.com/science/article/pii/S016773221833157X#bb0005), 2(https://www.sciencedirect.com/science/article/pii/S016773221833157X#bb0010), 3(https://www.sciencedirect.com/science/article/pii/S016773221833157X#bb0015), 4(https://www.sciencedirect.com/science/article/pii/S016773221833157X#bb0020), 5(https://www.sciencedirect.com/science/article/pii/S016773221833157X#bb0025), 6(https://www.sciencedirect.com/science/article/pii/S016773221833157X#bb0030), 7(https://www.sciencedirect.com/science/article/pii/S016773221833157X#bb0035)], especially playing a significant role in the blast furnace ironmaking. The structure and properties of slag have a great dependence on its composition. Changes in each component will affect the structural complexity, conductivity and viscosity of the system [3,8,9]. Therefore, the change of slag composition has a vital influence on its properties which finally affects the operation of the blast furnace. For example, high viscosity slag may cause difficulty in tapping; high melting point of slag may impede blast furnace operation and cause accident [10]. The relationship between blast furnace slag performance and composition has always been the focus in the field of metallurgy.

Up to now, the structure of aluminosilicate has been systematically studied through experiments and simulations. And there are also a few studies reported by molecular dynamics simulations [11(https://www.sciencedirect.com/science/article/pii/S016773221833157X#bb0055), 12(https://www.sciencedirect.com/science/article/pii/S016773221833157X#bb0060), 13(https://www.sciencedirect.com/science/article/pii/S016773221833157X#bb0065), 14(https://www.sciencedirect.com/science/article/pii/S016773221833157X#bb0070), 15(https://www.sciencedirect.com/science/article/pii/S016773221833157X#bb0075)]. Wu et al. [16] have successfully assessed the structure and properties for the CaO-SiO2 and CaO-Al2O3 systems. Mongalo et al. [8] studied the structural properties and electrical conductivities of CaO-MgO-Al2O3-SiO2 melts. The structure of aluminosilicates is mainly composed of network connected by SiO4 and AlO4 tetrahedron through bridge oxygen and tricluster oxygen [17,18].

The CaO/SiO2 mass ratio is usually defined as the basicity to judge the blast furnace slag condition. When the basicity is greater than 1, it is helpful for desulfurization and dephosphorization [19(https://www.sciencedirect.com/science/article/pii/S016773221833157X#bb0095), 20(https://www.sciencedirect.com/science/article/pii/S016773221833157X#bb0100), 21(https://www.sciencedirect.com/science/article/pii/S016773221833157X#bb0105), 22(https://www.sciencedirect.com/science/article/pii/S016773221833157X#bb0110)]. And in the field of silicates, many scholars have studied the effect of CaO on aluminosilicate. Ca2+ ions act as alkaline cations, which play a different role from Si4+ and Al3+ ions in aluminosilicate systems [1,9]. Ca2+will destroy the network structure of the system and reduce the degree of polymerization by reducing the content of the polymer. And the increase of CaO content in the slag will also cause the viscosity to decrease.

With the increasing consumption of high-aluminum ore in ironmaking process, the content of Al2O3 in blast furnace slag gradually increases, which leads to the increase of slag viscosity. Although the viscosity reduction could be achieved by increasing CaO content, it will increase the melting temperature of the slag, which will cause difficulties in smelting. Alternatively, MgO was used to decrease the slag viscosity considering that role of MgO in the silicate system is somewhat similar to that of CaO [23(https://www.sciencedirect.com/science/article/pii/S016773221833157X#bb0115), 24(https://www.sciencedirect.com/science/article/pii/S016773221833157X#bb0120), 25(https://www.sciencedirect.com/science/article/pii/S016773221833157X#bb0125), 26(https://www.sciencedirect.com/science/article/pii/S016773221833157X#bb0130), 27(https://www.sciencedirect.com/science/article/pii/S016773221833157X#bb0135)].

The influence of CaO or MgO on aluminosilicates has been studied a lot, but the similarities and differences of their effects on slag have not been systematically compared by scholars. Based on the difficulties and instability of the high-temperature experiment, the molecular dynamics simulation was used to simulate the effect of CaO and MgO on the aluminosilicate system from atomic scale at the same mass fraction. Results showed that both CaO and MgO have a destructive effect on the network structure of the aluminosilicate system and the damage degree of MgO to aluminosilicate is obviously higher than that of CaO at the same mass content. These results will provide a significant reference for the property control of aluminosilicate slags.

11. Taking the core temperature

https://doi.org/10.1038/46696

The chemical and physical properties of materials are described by quantum mechanics with great accuracy. However, for anything more complicated than the hydrogen atom, there are no exact solutions to Schrödinger’s equation. The principal difficulty is the many-body nature of the problem, which is where density functional theory (DFT) comes in. This is based on the observation that the ground-state properties of materials are rigorously and uniquely determined by the one-electron density. The many-body Schrödinger’s equation is replaced by an equation with effective interactions that depend only on the electron density. Although the exact form of these effective interactions is not known, several approximations exist that yield very accurate structural and physical properties of molecules and condensed materials.

Ab initio DFT methods refer to calculations that rely on DFT but use no empirical information about the material. Given the atomic numbers of the constituent atoms and known physical constants such as Planck’s constant and electron charge, these calculations yield equilibrium structures, total energies, equations of state, vibrational frequencies, and so on. Density functional theory has long appealed to geophysicists interested in understanding the physics of materials subjected to the extreme conditions in deep planetary interiors. Not surprisingly, the Earth’s core was the early subject of such investigations. Today, DFT methods are being used successfully to examine the effects of pressure on the structure, elasticity, and thermodynamics of complex silicates thought to make up the deep Earth.

Developments such as that described by Alfè et al. promise that it won’t be long before ab initio methods will be used almost routinely to study such geochemical processes as partitioning of elements among coexisting phases, high-temperature and high-pressure phase transformations, and even the chemical reactions between core and mantle materials.