Mechanism Of Glass Formation Research Articles

Evaluation of GlassNet for physics‐informed machine learning of glass stability and glass‐forming ability

AbstractGlassy materials form the basis of many modern applications, including nuclear waste immobilization, touch‐screen displays, and optical fibers, and also hold great potential for future medical and environmental applications. However, their structural complexity and large composition space make design and optimization challenging for certain applications. Of particular importance for glass processing and design is an estimate of a given composition's glass‐forming ability (GFA). However, there remain many open questions regarding the underlying physical mechanisms of glass formation, especially in oxide glasses. It is apparent that a proxy for GFA would be highly useful in glass processing and design, but identifying such a surrogate property has proven itself to be difficult. While glass stability (GS) parameters have historically been used as a GFA surrogate, recent research has demonstrated that most of these parameters are not accurate predictors of the GFA of oxide glasses. Here, we explore the application of an open‐source pre‐trained neural network model, GlassNet, that can predict the characteristic temperatures necessary to compute GS with reasonable performance and assess the feasibility of using these physics‐informed machine learning (PIML)‐predicted GS parameters to estimate GFA. In doing so, we track the uncertainties at each step of the computation—from the original ML prediction errors to the compounding of errors during GS estimation, and finally to the final estimation of GFA. While GlassNet exhibits reasonable accuracy on all individual properties, we observe a large compounding of error in the combination of these individual predictions for the PIML prediction of GS, finding that random forest models offer similar accuracy to GlassNet. We also break down the performance of GlassNet on different glass families and find that the error in GS prediction is correlated with the error in crystallization peak temperature prediction. Lastly, we utilize this finding to assess the relationship between top‐performing GS parameters and GFA for two ternary glass systems: sodium borosilicate and sodium iron phosphate glasses. We conclude that to obtain true ML predictive capability of GFA, significantly more data needs to be collected.

Organic-Inorganic Hybrid Glasses of Atomically Precise Nanoclusters.

Organic-inorganic atomically precise nanoclusters provide indispensable building blocks for establishing structure-property links in hybrid condensed matter. However, robust glasses of ligand-protected nanocluster solids have yet to be demonstrated. Herein, we show [Cu4I4(PR3)4] cubane nanoclusters coordinated by phosphine ligands (PR3) form robust melt-quenched glasses in air with reversible crystal-liquid-glass transitions. Protective phosphine ligands critically influence the glass formation mechanism, modulating the glasses' physical properties. A hybrid glass utilizing ethyldiphenylphosphine-based nanoclusters, [Cu4I4(PPh2Et)4], exhibits superb optical properties, including >90% transmission in both visible and near-infrared wavelengths, negligible self-absorption, near-unity quantum yield, and high light yield. Experimental and theoretical analyses demonstrate the structural integrity of the [Cu4I4(PPh2Et)4] nanocluster, i.e., iodine-bridged tetranuclear cubane, has been fully preserved in the glass state. The strong internanocluster CH-π interactions found in the [Cu4I4(PPh2Et)4] glass and subsequently reduced structural vibration account for its enhanced luminescence properties. Moreover, this highly transparent glass enables performant X-ray imaging and low-loss waveguiding in fibers drawn above the glass transition. The discovery of "nanocluster glass" opens avenues for unraveling glass formation mechanisms and designing novel luminescent glasses of well-defined building blocks for advanced photonics.

Rapid identification of highly glass-forming alloys in Cu–Zr binary system by laser glazing

The high-efficient experimental identification of novel metallic glasses with high glass-forming ability (GFA) is crucial for breaking through the traditional trial-and-error approach, which is time-consuming and costly. A rapid detection method for highly glass-forming alloys using laser glazing was introduced and validated in a classic binary Cu–Zr system with well-known GFA across a wide compositional range. The vitrification degree induced by laser glazing was assessed for each selected alloy and shows a strong correlation with GFA determined by traditional rapid solidification methods. The combination of experimental kinetic analysis and theoretical finite element simulation revealed possible thermodynamic mechanisms of glass formation and crystallization in the heat-affected zone (HAZ) formed by laser glazing. The results offer insights into not only the rapid identification method of highly glass-forming alloys using laser glazing but also the mechanism of glass formation under fast heating and cooling processes, providing guidance on the construction of bulk metallic glasses through laser additive manufacturing.

A novel strategy with high targeting performance in designing metallic glasses conducted on the light of melting entropy

In this study, a new design strategy of metallic glasses is proposed using melting entropy to guide the compositions, which allows for the consideration of intermetallic compounds to serve as the starting materials. The melting entropy based strategy is verified by the detailed studies on the Cu-Zr-Hf alloys. The metallic glass design triggered by intermetallic compounds challenges the traditional cognition that the intermetallic compounds should be preferentially excluded when focusing on the glass formation compositions. This reinforces the viewpoint that low melting entropy is conducive to glass formation. The new strategy is of high targeting performance in search of starting materials to design glasses, distinct from the trial-and-error mode. This study also provides a new perspective for the understanding of glass formation mechanism.

Data-driven prediction of the glass-forming ability of modeled alloys by supervised machine learning

The ability of a matter to fall into a glassy state upon cooling differs greatly among metallic alloys. It is conventionally measured by the critical cooling rate Rc, below which crystallization inevitably happens. There are a lot of factors involved in determining Rc for an alloy, including both elemental features and alloy properties. However, the underlying physical mechanism is still far from being well understood. Therefore, the design of new metallic glasses is mainly by time- and labor-consuming trial-and-error experiments. This considerably slows down the development process of metallic glasses. Nowadays, large-scale computer simulations have been playing a significant role in understanding glass formation. Although the atomic-scale features can be well captured, the simulations themselves are constrained to a limited timescale. To overcome these issues, we propose to explore the glass-forming ability of the modeled alloys from computer simulations by supervised machine learning. We aim to gain insights into the key features determining Rc and found that the non-linear couplings of the geometrical and energetic factors are of great importance. An optimized machine learning model is then established to predict new glass formers with a timescale beyond the current simulation capability. This study will shed new light on both unveiling the glass formation mechanism and guiding new alloy design in practice.

Ternary Fe–W–B bulk metallic glasses with ultrahigh thermal stabilities

Fe-based bulk metallic glasses (BMGs) with broad application prospects often exhibit good comprehensive properties. However, they are facing the bottleneck problem of structural instability during the long-term service process due to their own metastable state. In this work, a new ternary system of Fe–W–B BMGs with ultrahigh thermal stabilities, where the glass transition temperature of Fe59W23B18 BMG reached to 923 K and the onset crystallization temperature above 970 K, were successfully developed. In addition, the Fe–W–B BMGs exhibit very high micro-hardness of 1280–1380 HV and high density about 9.6 g/cm3. The ultrahigh thermal stability and hardness are closely related to the prominent covalent bonds in the Fe–W–B BMGs. The ternary Fe–W–B BMGs with outstanding thermal properties and simple chemical components are promising materials for use in extreme service environment and model materials to understand the glass formation mechanism of Fe-based BMGs.

Glass formation mechanism in aqueous formic acid solutions

The glass-formation mechanism in the HCOOH–H2O system, which occurred at component molar ratios from 1:2 to 3:2, was studied by calorimetry and quantum chemistry. The results were analyzed and interpreted in light of the laws that guide the formation of stable H-bonded hetero associates in binary systems. The thermo analytical curves for 1:2, 1:1, and 3:2 glasses each showed three thermal events due to glass transition at about –130°С, crystallization, and melting of crystals. The structures and energy parameters were calculated in terms of density functional theory for polymeric chain moieties comprised of one or two types of the HCOOH⋅(H2O)2, (HCOOH)2⋅(H2O)2, and (HCOOH)3⋅(H2O)2 hetero associates. Glass formation was shown to occur precisely when [HCOOH]:[H2O] = 1:2–3:2 due to the ability of those hetero associates to be linked into strongly bonded, unlimitedly long polymeric chains. The energy parameters of 1:2, 1:1, and 3:2 samples derived from the calorimetric experiment were compared to those gained from the calculations of moieties that model polymeric structures in these samples, so all experimentally observed thermal events were interpreted at the molecular level. The glass-formation mechanism in the HCOOH–H2O system where the key role is played by hetero associates and not by individual molecules was elucidated.

A comprehensive study of lead telluride (PbTe)-based amorphous alloys: Glass formation and thermoelectric properties

Vitrification enables to create complex structure with a large degree of disorder, which provides a new solution to improve the performance of lead telluride (PbTe)-based thermoelectric materials by reducing the lattice thermal conductivity. Yet, it is an extreme challenge to vitrify PbTe-based alloys due to the incomplete knowledge of the glass formation mechanism. Here, the glass formation of PbTe-based alloys is studied by jointly concerning thermodynamics and kinetics. The comparison of (PbTe)x(Ga2Te3)100-x and (PbTe)x(In2Te3)100-x alloys showed that low entropy of fusion(∆Sm) and deep eutectic fashion are the two keys in glass forming ability(GFA). Novel (PbTe)x(Ga2Te3)100-x glasses are prepared and the glass-forming region is identified to be the composition range of x = 52 ~ 64. An extremely low thermal conductivity of the best glass component (PbTe)61(Ga2Te3)39 is determined with the value of ~ 0.1 Wm−1K−1 at 300 K.

Impact of Orientational Glass Formation and Local Strain on Photo-Induced Halide Segregation in Hybrid Metal-Halide Perovskites.

Band gap tuning of hybrid metal–halide perovskites by halide substitution holds promise for tailored light absorption in tandem solar cells and emission in light-emitting diodes. However, the impact of halide substitution on the crystal structure and the fundamental mechanism of photo-induced halide segregation remain open questions. Here, using a combination of temperature-dependent X-ray diffraction and calorimetry measurements, we report the emergence of a disorder- and frustration-driven orientational glass for a wide range of compositions in CH3NH3Pb(ClxBr1–x)3. Using temperature-dependent photoluminescence measurements, we find a correlation between halide segregation under illumination and local strains from the orientational glass. We observe no glassy behavior in CsPb(ClxBr1–x)3, highlighting the importance of the A-site cation for the structure and optoelectronic properties. Using first-principles calculations, we identify the local preferential alignment of the organic cations as the glass formation mechanism. Our findings rationalize the superior photostability of mixed-cation metal–halide perovskites and provide guidelines for further stabilization strategies.

Ultrahigh-field 67Zn NMR reveals short-range disorder in zeolitic imidazolate framework glasses.

The structure of melt-quenched zeolitic imidazole framework (ZIF) glasses can provide insights into their glass-formation mechanism. We directly detected short-range disorder in ZIF glasses using ultrahigh-field zinc-67 solid-state nuclear magnetic resonance spectroscopy. Two distinct Zn sites characteristic of the parent crystals transformed upon melting into a single tetrahedral site with a broad distribution of structural parameters. Moreover, the ligand chemistry in ZIFs appeared to have no controlling effect on the short-range disorder, although the former affected their phase-transition behavior. These findings reveal structure-property relations and could help design metal-organic framework glasses.

On the correlation between glass forming ability (GFA) and soft magnetism of Ni-substituted Fe-based metallic glassy alloys

The present work aims to investigate the mechanism of glass formation and correlates it to the soft magnetism of Ni-substituted Fe-B-Nb alloys. The impact of ferrous Ni on the mechanism of glass formation and soft magnetism of Fe-based bulk metallic glasses was critically analyzed. By quantifying glass forming ability and soft magnetic characteristic for varying degrees of substitution, we observe a maximum in glass forming ability together with a minimum coercivity, which we suggest is due to an increased atomic packing density of the glassy phase. Interestingly, a monotonic increase of Curie temperature with increasing substitution of Ni was observed, which could be attributed to a reduction of an antiferromagnetic Fe-Fe interaction in the glassy iron-rich matrix. The overarching goal of this study is to explore the underlying mechanisms of enhanced glass forming ability, improved soft magnetic properties, and increased Tc of Ni-substituted Fe-based glassy alloys.

Impact of hybridization on metallic-glass formation and design

The formation mechanism of glass at the atomic scale has been under debate over centuries. In this work, we demonstrate that hybridization, as manifested by Mott’s pseudogap, has a strong influence on the bond length as well as atomic packing, which can potentially tailor the formation of metallic glasses at microscopic time and length scales. A p–d orbital hybridization between the post-transition metal Al and the transition metal was shown by the 27 Al isotropic shifts and the spin–lattice relaxation time of Zr–Co–Al alloys using nuclear magnetic resonance. These bonds lead to a charge transfer between the specific atomic pairs and the shrinkage of interatomic distances. Such chemical bonding favors the formation of metallic glasses by introducing a string-like structure and further stabilizes metallic glasses via a reduction in the density of states at the Fermi level. Our work has implications for understanding the glass formation mechanism at the electronic level and may open up new possibilities on the design of glass from the perspective of atomic interactions.

Recent Topics on the Structure and Crystallization of Al-based Glassy Alloys

Al-based glassy alloys (metallic glasses) have been of great scientific and technological interest as a high specific strength and high corrosion resistant materials. However, the low glass-forming ability (GFA) is still a choke point that greatly influences the applications. In order to further comprehend the glass formation mechanism and investigate the possibility of enhancing the GFA, it is important to obtain the definite information on the atomic scale glassy structures in correlation with properties and to study the composition dependence of crystallization and properties of Al-based metallic glasses. The purpose of this paper is to summarize the structure characteristics and crystallization behavior in conjunction with alloy component in Al-based metallic glasses and to investigate the possibility of synthesizing an Al-based bulk metallic glass with better engineering performances. This may assist to learn the structural features systematically and understand the development and current status for Al-based metallic glasses.

Glass forming, crystallization, and physical properties of MgO-Al2O3-SiO2-B2O3 glass-ceramics modified by ZnO replacing MgO

This study focused on the glass forming, crystallization, and physical properties of ZnO doped MgO-Al2O3-SiO2-B2O3 glass-ceramics. The results show that the glass forming ability enhances first with ZnO increasing from 0 to 0.5 mol%, and then weakens with further addition of ZnO which acted as network modifier. No nucleating agent was used and the crystallization of studied glasses is controlled by a surface crystallization mechanism. The predominant phase in glass-ceramics changed from α-cordierite to spinel/gahnite as ZnO gradually replaced MgO. The phase type did not change; however, the crystallinity and grain size in glass-ceramics increased when the glasses were treated from 1030 °C to 1100 °C. The introduction of ZnO can improve the thermal, mechanical, and dielectric properties of the glass-ceramics. The results reveal a rational mechanism of glass formation, crystal precipitation, and evolution between structure and performance in the xZnO-(20-x)MgO-20Al2O3-57SiO2-3B2O3 (0 ≤ x ≤ 20 mol%) system.

Different scenarios of dynamic coupling in glassy colloidal mixtures.

Colloidal mixtures represent a versatile model system to study transport in complex environments. They allow for a systematic variation of the control parameters, namely size ratio, total volume fraction and composition. We study the effects of these parameters on the dynamics of dense suspensions using molecular dynamics simulations and differential dynamic microscopy experiments. We investigate the motion of small particles through the matrix of large particles as well as the motion of large particles. A particular focus is on the coupling of the collective dynamics of small and large particles and on the different mechanisms leading to this coupling. For large size ratios, of about 1 : 5, and an increasing fraction of small particles, the dynamics of the two species become increasingly coupled and reflect the structure of the large particles. This is attributed to the dominant effect of the large particles on the motion of the small particles, which is mediated by the increasing crowding of the small particles. Furthermore, for moderate size ratios of about 1 : 3 and sufficiently high fractions of small particles, mixed cages are formed and hence the dynamics are also strongly coupled. Again, the coupling becomes weaker as the fraction of small particles is decreased. In this case, however, the collective intermediate scattering function of the small particles shows a logarithmic decay corresponding to a broad range of relaxation times.

Abnormal correlation between phase transformation and cooling rate for pure metals.

This work aims to achieve deep insight into the phenomenon of phase transformation upon rapid cooling in metal systems and reveal the physical meaning of scatter in the time taken to reach crystallization. The total number of pure metals considered in this work accounts for 14. Taking pure copper as an example, the correlation between phase selection of crystal or glass and cooling rate was investigated using molecular dynamic simulations. The obtained results demonstrate that there exists a cooling rate region of 6.3 × 1011–16.6 × 1011 K/s, in which crystalline fractions largely fluctuate along with cooling rates. Glass transformation in this cooling rate region is determined by atomic structure fluctuation, which is controlled by thermodynamic factors. According to the feature of bond-orientation order at different cooling rates, we propose two mechanisms of glass formation: (i) kinetic retardation of atom rearrangement or structural relaxation at a high cooling rate; and (ii) competition of icosahedral order against crystal order near the critical cooling rate.

Amorphization of Ni61Nb39 Alloy by Laser Surface Treatment

The surface of Ni61 Nb39 crystalline ingot was treated by laser surface melting with different processing parameters. A fully amorphous layer with a thickness of approximately 10 μm could be produced on the top surface under optimal parameters. An amorphous-crystalline composite layer with the depth from 10 to 50 μm, consisting of amorphous matrix and intermetallic phases of Ni3Nb and Ni6Nb7, could be formed. The micro-hardness (about 831 HV) of the treated surface was remarkably improved by nearly 100% compared with the value of the crystalline substrate caused by the formation of the fully amorphous structure. A finite volume simulation was adopted to evaluate the temperature distribution in the laser-affected zone of Ni61 Nb39 alloys and to reveal the mechanism of glass formation in the laser-affected zone.

Identifying icosahedron-like clusters in metallic glasses

Since the discovery of the first metallic glass (MG) with the composition of Au75Si25 in 1960, vast efforts have been devoted to understanding the mechanisms of glass formation in metals, because this class of glassy alloy usually possesses unique properties that may have the potential application as engineering material. As is well known, structure determines properties of material. Therefore, understanding the glass formation of MG from the structural perspective is helpful for guiding researchers in developing more MGs. So far, icosahedral clusters are regarded as the preferred clusters contributing to the formation of amorphous structure due to its five-fold symmetrical feature and high atomic packing. However, it has been found that an ideal icosahedron usually does not have a high concentration in many MG compositions. Thus, we wonder whether icosahedral clusters are popular in microstructures of amorphous alloys. In this work, a feasible scheme for identifying the icosahedron-like clusters in MGs is developed to address this issue. It is found that icosahedron-like clusters are popular structural units in amorphous structure indeed, contributing to the glass formation in alloy. A projection method of reflecting the styles of shell-atom connections in Voronoi-tessellation indexed clusters is developed in detail, so that all clusters can be further geometrically indexed as different projected types of polyhedra. It is revealed that there are three kinds of clusters (0, 2, 8, 1, 0, 2, 8, 2 I-type, and 0, 1, 10, 2) which have the most similar geometrical features to that of the so-called ideal icosahedron, 0, 0, 12, 0. Therefore, besides the ideal icosahedron, these three types of clusters can be regarded as the icosahedron-like clusters. The ideal icoshahedron (0, 0, 12, 0) has a coordination number (i.e., the number of shell atoms) of 12, while these three icosahedron-like clusters have coordination numbers ranging from 11 to 13, so that structural balance between the geometrical atomic stacking and the chemical interactions among various elements in MGs (especially multicomponent MGs) is more easy to achieve. Furthermore, structural models of three selected ZrCu compositions are studied, which are obtained by systematic experimental measurements combined with reverse Monte Carlo simulation. It is found that both the icosahedron-like cluster and the ideal icosahedron have the similar values of some structural parameters, in terms of high atomic packing efficiency, high cluster regularity, fruitful five-fold symmetrical feature, etc. In addition, it is revealed that these ideal icosahedra and icosahedron-like clusters can contain almost all the atoms in these structural models, enhancing the space filling efficiency. In conclusion, these identified icosahedron-like clusters should be the popular building blocks, contributing to the glass formation in alloy. This work provides an insight into the glass formation in alloy from the cluster-level structural angle and will shed light on developing more MGs.

Atomistic Design of Favored Compositions for Synthesizing the Al-Ni-Y Metallic Glasses.

For a ternary alloy system promising for obtaining the so-called bulk metallic glasses (BMGs), the first priority issue is to predict the favored compositions, which could then serve as guidance for the appropriate alloy design. Taking the Al-Ni-Y system as an example, here we show an atomistic approach, which is developed based on a recently constructed and proven realistic interatomic potential of the system. Applying the Al-Ni-Y potential, series simulations not only clarify the glass formation mechanism, but also predict in the composition triangle, a hexagonal region, in which a disordered state, i.e., the glassy phase, is favored energetically. The predicted region is defined as glass formation region (GFR) for the ternary alloy system. Moreover, the approach is able to calculate an amorphization driving force (ADF) for each possible glassy alloy located within the GFR. The calculations predict an optimized sub-region nearby a stoichiometry of Al80Ni5Y15, implying that the Al-Ni-Y metallic glasses designed in the sub-region could be the most stable. Interestingly, the atomistic predictions are supported by experimental results observed in the Al-Ni-Y system. In addition, structural origin underlying the stability of the Al-Ni-Y metallic glasses is also discussed in terms of a hybrid packing mode in the medium-range scale.

Molecular simulation of freestanding amorphous nickel thin films

Size effects on glass formation in freestanding Ni thin films have been studied via molecular dynamics simulation with the n-body Gupta interatomic potential. Atomic mechanism of glass formation in the films is determined via analysis of the spatio-temporal arrangements of solid-like atoms occurred upon cooling from the melt. Solid-like atoms are detected via the Lindemann ratio. We find that solid-like atoms initiate and grow mainly in the interior of the film and grow outward. Their number increases with decreasing temperature and at a glass transition temperature they dominate in the system to form a relatively rigid glassy state of a thin film shape. We find the existence of a mobile surface layer in both liquid and glassy states which can play an important role in various surface properties of amorphous Ni thin films. We find that glass formation is size independent for models containing 4000 to 108,000 atoms. Moreover, structure of amorphous Ni thin films has been studied in details via coordination number, Honeycutt–Andersen analysis, and density profile which reveal that amorphous thin films exhibit two different parts: interior and surface layer. The former exhibits almost the same structure like that found for the bulk while the latter behaves a more porous structure containing a large amount of undercoordinated sites which are the origin of various surface behaviors of the amorphous Ni or Ni-based thin films found in practice.