Ternary Melts Research Articles

Viscosity and melting temperature prediction of mold fluxes based on explainable machine learning and SHapley additive exPlanations

Viscosity and melting temperature are indispensable properties for mold flux design and evaluation, significant investment is required for measurement. In this paper, viscosity and melting temperature predictive model for mold fluxes were established and trained based on 3300 groups of data and four representative machine learning algorithms. The gradient boosting regression tree (GBRT) algorithm-based model performed best prediction with the determination coefficient R2 of 0.969 and 0.900 for viscosity and melting temperature. It also far outperforms the widely used viscosity and melting temperature prediction models, with a least mean deviation of 7.06% and 0.8%. SHapley Additive exPlanations (SHAP) analysis revealed that Al2O3, followed by SiO2 showed the strongest positive correlation with viscosity, Na2O has the greatest negative contribution to melting temperature. Ternary viscosity and melting temperature distribution diagrams were constructed to visualize the prediction results. Insights from this study will highly benefit computer-aided design of mold fluxes with desired properties.

Ab initio molecular dynamics study on the local structures and solid/liquid interface in liquid Al-Ti and Al-B-Ti alloys

High-purity aluminum, owing to its excellent physical and chemical properties, finds extensive applications across various fields. A comprehensive understanding of the local structure of liquid Al-based alloys and the characteristics of solute atoms at the forefront of the solid-liquid interface is highly advantageous for the purification of aluminum. Ab initio molecular dynamics (AIMD) simulations were employed to study the local structures, element interactions and electronic structures in Al-Ti and Al-B-Ti melts, together with the behaviors of Ti atom and B-Ti atom pair at the solid/liquid interface forefront of Al. In the Al-B-Ti ternary melt, there are stronger interactions between B and Ti atoms than those between Al and B atoms as well as between Al and Ti atoms. The introduction of B atoms leads to the disruption of Al-Ti bonds and the formation of B-Ti bonds, thereby altering the state of Ti atoms in the melt and then influencing their diffusion properties. The dynamic process of B-Ti bond formation in the melt has been elucidated. The bond angle distribution function shows that the geometric configuration around Ti atoms in the Al melt undergoes significant changes with the addition of B atoms. A strong hybridization occurs between the 3d orbital of Ti atom and the 2p orbital of B atom. Moreover, a notable charge accumulation between B and Ti atoms, indicates the strong interaction between them. The addition of Ti atom can promote the migration of the Al solid/liquid interface. The B-Ti bond is very stable at the forefront of the Al solid/liquid interface, and the formation of B-Ti bond significantly drag the migration of the Al solid/liquid interface.

The tin content of lead inclusions in ancient tin-bronze artifacts: a time-dependent process?

In antiquity, Pb was a common element added in the production of large bronze artifacts, especially large statues, to impart fluidity to the casting process. As Pb does not form a solid solution with pure Cu or with the Sn–Cu alloy phases, it is normally observed in the metal matrix as globular droplets embedded within or in interstitial positions among the crystals of Sn-bronze (normally the α phase) as the last crystallizing phase during the cooling process of the Cu–Sn–Pb ternary melt. The disequilibrium Sn content of the Pb droplets has recently been suggested as a viable parameter to detect modern materials [Shilstein, Berner, Feldman, Shalev & Rosenberg (2019). STAR Sci. Tech. Archaeol. Res. 5, 29–35]. The application assumes a time-dependent process, with a timescale of hundreds of years, estimated on the basis of the diffusion coefficient of Sn in Pb over a length of a few micrometres [Oberschmidt, Kim & Gupta (1982). J. Appl. Phys. 53, 5672–5677]. Therefore, Pb inclusions in recent Sn-bronze artifacts are actually a metastable solid solution of Pb–Sn containing ∼3% atomic Sn. In contrast, in ancient artifacts, unmixing processes and diffusion of Sn from the micro- and nano-inclusions of Pb to the matrix occur, resulting in the Pb inclusions containing a substantially lower or negligible amount of Sn. The Sn content in the Pb inclusions relies on accurate measurement of the lattice parameter of the phase in the Pb–Sn solid solution, since for low Sn values it closely follows Vegard's law. Here, several new measurements on modern and ancient samples are presented and discussed in order to verify the applicability of the method to the detection of modern artwork pretending to be ancient.

Machine learning model to efficiently predict the structure and properties of MgCl2–NaCl–KCl melts

MgCl2–NaCl–KCl ternary melts hold great promise as thermal energy storage and heat transfer fluid for the next generation of concentrating solar power (CSP), offering major advantages such as high-temperature storage capacity, adjustable temperature ranges, and renewable usability. However, a critical challenge is the lack of an accurate model to analyze the structural and property aspects of this complex melt. In response, a machine learning-based model with high accuracy and applicability has been developed for this purpose. Based on DPMD techniques, the trained model is used to investigate the structure information, transport and thermal properties of the melt. Temperature and composition effects are introduced. Radial distribution functions, coordination numbers, and angular distribution functions are applied to investigate the microscopic ionic environment within the melt. Transport properties studied include density, self-diffusion coefficient, ionic conductivity, and viscosity, while thermal properties include heat capacity at constant pressure and volume and thermal conductivity. Our model accurately reproduces the results obtained from FPMD simulations and accurately predicts the macroscopic properties of the melt through validation. The high stability of the Mg2+-Cl- interactions is introduced, which remains relatively unaffected by temperature and concentration changes. The presence of MgCl53− species in the melt was observed. We are devoted to the development of AI for the CSP science field. Due to the application of machine learning strategies, our models have proven to have exceptional accuracy and ease of study for MgCl2–NaCl–KCl ternary melts, and they can provide valuable insights into their properties and ionic behaviors.

Low-Temperature Molten Salt Electrochemical CO2 Upcycling for Advanced Energy Materials.

One strategy for addressing the climate crisis caused by CO2 emissions is to efficiently convert CO2 to advanced materials suited for green and clean energy technology applications. Porous carbon is widely used as an advanced energy storage material because of its enhanced energy storage capabilities as an anode. Herein, we report electrochemical CO2 upcycling to solid carbon with a controlled microstructure and porosity in a ternary molten carbonate melt at 450 °C. Controlling the electrochemical parameters (voltage, temperature, cathode material) enabled the conversion of CO2 to porous carbon with a tunable morphology and porosity for the first time at such a low temperature. Additionally, a well-controlled morphology and porosity are beneficial for reversible energy storage. In fact, these carbon materials delivered high specific capacity, stable cycling performances, and exceptional rate capability even under extremely fast charging conditions when integrated as an anode in lithium-ion batteries (LIBs). The present approach not only demonstrated efficient upcycling of CO2 into porous carbon suitable for enhanced energy storage but can also contribute to a clean and green energy technology that can reduce carbon emissions to achieve sustainable energy goals.

Coordination Structure of Si(II) Ions in Binary Melts (LiF-KF, LiF-NaF, and NaF-KF) and Ternary Melt (LiF-NaF-KF).

The formation of Si(II) by the comproportionation reaction between electrodeposited silicon and Si(IV) ions [Si(IV) + Si0 ⇄ 2Si(II)] in fluoride melts has been discussed in many previous reports on the electrochemical processing of silicon. In this study, the coordination structure of Si(II) ions in LiF-KF, LiF-NaF, NaF-KF, and LiF-NaF-KF melts was investigated by high-temperature Raman spectroscopy and quantum chemical calculation based on density functional theory. The formation of Si(II) ions was induced by mixing metallic silicon powder as the source of Si0, with the melts containing hexafluoro silicate or silicon dioxide as the source of Si(IV) ions. The simulated vibrational modes of the Si(II) species and the experimental Raman bands showed good agreement. It was found that both Si(II) and Si(IV) species were detected in either fluoride melt, demonstrating that the proposed comproportionation reaction between Si0 and Si(IV) occurred. In addition, the interaction between Si(II) and F-, O2-, Li+, Na+, and K+ ions significantly changed depending on the composition of fluoride melts. The coordination structure of the Si(II) ions reported in this study will provide clues for controlling the electrodeposition process of silicon in fluoride melts.

Laser induced reduction of iron ore by silicon

Iron ore powder accompanied by Si-powder as a reducing agent, was melted using a high-power laser beam. During laser melting of the two different powder beds placed next to each other, silicon merged and diffused into the iron ore, forming a ternary melt phase Fe-O-Si of around 30–60–10 at%. High speed imaging of the laser melting process as well as subsequent SEM-analysis of the generated nuggets showed the formation of droplets that merge with the surrounding Si- or ore-powder and form distinct domains. Under certain circumstances, the solidifying nuggets, of the order of 0.5–5 mm in size, generated numerous small domains, up to 25 µm, of high purity iron, 90 + at%, surrounded by a matrix of the above mentioned slag. Many of these Fe-domains were created in the vicinity of regions of high Si-content.

Glutaric acid purification by coupled solution crystallization and melt crystallization

The mixed dibasic acids (DBA) is a by-product during adipic acid (ADA) production by oxidation of cyclohexanol or cyclohexanone, which contains large amount of glutaric acid (GA) with ADA and succinic acid (SA). In this work, we have explored the feasibility of the coupled solution crystallization and melt crystallization for the purification of GA. To obtain fundamental thermodynamic properties for the process design, the gravimetric method was used to measure the solubility of the three dibasic acids in water from 293.15 K to 345.15 K at atmospheric pressure. The results showed that GA was the most soluble component while all solubility increased with increasing temperature. Therefore, it is reasonable to concentrate GA in the liquid phase by cooling crystallization. Similarly, the ternary solid-liquid melt equilibrium phase diagram was constructed based on the experimental data of the three binary subsystems (GA-ADA, GA-SA, ADA-SA). It was found that the concentrated GA after solvent evaporation could be purified by melt crystallization efficiently. Thus a coupled solution crystallization and melt crystallization process was designed. With the operated operating parameters, the purity of GA could reach 99.06% by secondary melt crystallization. Overall, the coupled solution crystallization and melt crystallization method to separate DBA has proven to be effective.

Viscosity database for ternary Cu–Cr–X (X=Ni, Si, Zr) alloys based on CALPHAD-type modeling

A viscosity model with composition and temperature dependence for the liquid alloys of the Cu–Cr–X (X = Ni, Si, Zr) systems was developed using the CALPHAD (CALculation of PHAse Diagrams) approach. Based on the critical review of the available experimental data for pure metal viscosity, the viscosities of pure liquid Cu, Cr, Zr, Ni and Si were simulated with the Arrhenius formula. The viscosities of the binary liquid alloys of the Cu–Ni, Cu–Si, Cu–Zr, Cr–Ni, Cr–Si and Ni–Zr systems at different temperatures were evaluated using the CALPHAD-type formalism. The non-linear viscosity behavior resulting from chemical short-range order was described using temperature-dependent binary interaction parameters. The viscosities of the Cu–Cr and Cr–Zr liquid alloys were predicted employing the Hirai equation. The viscosities of the Cu–Cr–Zr, Cu–Cr–Ni and Cu–Cr–Si ternary liquids were directly extrapolated using the Redlish-Kister-Muggianu equation from the parameters of the binary systems. A comprehensive understanding of the temperature-composition-solidification phase-viscosity relationship was established for ternary melts. The satisfactory agreement observed between calculations and experimental data validates the use of CALPHAD-type viscosity modeling, supporting the suitability of the constructed viscosity database for predicting the viscosity of high-performance Cu-based alloys.

Electrochemical extraction of uranium on the gallium and cadmium reactive electrodes in molten salt

The electrochemical extraction of uranium in ternary low melting LiCl–KCl–CsCl eutectic on inert and reactive electrodes via different electrochemical techniques was investigated. It was established that the electrochemical reduction process of U(III) ions on the inert W electrode was irreversible and proceeded in one stage. On reactive liquid Ga and liquid Cd electrodes the reduction of uranium ions took place with the considerable depolarization with the formation of UGa2, UGa3 and UCd11 intermetallic compounds. Thermodynamic characteristics of uranium compounds and alloys were calculated. The conditions for the extraction of uranium from the electrolyte in the form of alloys on both liquid reactive electrodes via potentiostatic electrolysis were found.

Carbon Storage in Earth's Deep Interior Implied by Carbonate‐Silicate‐Iron Melt Miscibility

AbstractCarbonate melts have been proposed to exist in the lower mantle, but their interaction with other lower mantle melt compositions is poorly understood. To understand miscibility in the carbonate‐silicate‐metal melt system, we simulate endmember, binary, and ternary melt mixtures and study how their Gibbs free energies of mixing evolve with pressure. We find that carbonate‐metal and carbonate‐silicate melts have miscibility gaps that close with increasing pressure, while silicate‐metal melts are immiscible at all lower‐mantle pressures. Extending this analysis to the core‐mantle boundary, we suggest three miscible melt fields near the endmember carbonate, silicate, and iron melt compositions. Analysis of the densities of these miscible melt compositions indicates that some carbonate‐rich and some silicate‐rich melt compositions are gravitationally stable at the core‐mantle boundary and could be candidate compositions to explain ultra‐low velocity zones. Additionally, we evaluate the speciation of an example immiscible melt composition at various pressures throughout the mantle and identify reduced carbon species that would be expected to form in the melt. Our analysis reveals that a majority of Earth's carbon could have been transported to the core during core‐mantle differentiation and that much of Earth's carbon may be stored in the deep interior today.

Formation of the Short‐Range Order in Liquid Al−Ni−Sn Alloys

AbstractShort‐range order (SRO) in the liquid Al−Ni−Sn alloys has been investigated using the X‐ray diffraction method and Reverse Monte Carlo simulations. Obtained results point out a micro‐inhomogeneous structure of Al−Sn−Ni melts with Sn content between 10 and 65 at.% due to the presence of atomic clusters with the structure of liquid Sn and atomic clusters with Ni−Sn coordination. The influence of the atomic size ratio (size factor) and pair atomic interactions (energy factor) on the short‐range order of the investigated ternary melts are discussed. There is a competition between Al and Sn atoms in the formation of the SRO around Ni atoms due to the difference in bond energy in Ni−Al and Ni−Sn pairs. This competition, as well as a bigger radius of Sn atoms, results in the formation of atomic clusters with liquid Sn‐like local atomic structure.

Термодинамічні властивості розплавів системи Cu—In—La

Partial for the components and integral enthalpies of mixing of the ternary melts of the Cu—In—La system were determined for the first time by the method of calorimetry on five radial sections with a constant ratio of two components:, and (up to xIn = 0,02, 0,14, and 0,42, respectively) and and (up to xCu = 0,15 and 0,2, respectively) at 1220—1450 K. It is shown that when indium 1 is added to the CuxLa1-x melt, the thermal effect of its dissolution increases, which is caused by the formation of strong bonds between In and La. In the other two sections (and), the enthalpies of mixing of ternary melts decrease during dissolution. Using the reliable mixing enthalpies of the melts of the dual systems Cu—In(La) and In—La, similar parameters for liquid alloys of the Cu—In—La system were calculated according to various “geometric” and “analytical” models. It was found that the values calculated by the Redlich—Kister—Mujianu model with the triple contribution –200 agree with the determined enthalpies of mixing of melts of the Cu—In—La system within the limits of experimental errors. It was established that the minimum enthalpy of mixing of melts of the Cu—In—La system falls on the alloy of the In—La subsystem (−43,4 ± 2,1) at xLa = 0,4 at T = 1450 K, i. e., the largest contribution to the enthalpy of mixing of melts of the Cu—In—La system is made by boundary subsystem In—La. The activity of the components in the melts of this system was calculated according to the same model. It is shown that they exhibit small negative deviations from ideal solutions at 1450 K. From these data, G, S of melts of the Cu—In—La system were estimated. It was established that Gmin = –26 kJ/mol, Smin = –12 J/mol∙K, which are attributed to the In0.6La0.4 alloy. Keywords: calorimetry, copper, lanthanum, indium, thermodynamic properties, modeling, mixing enthalpies, activity of components.

C–O–H fluid-melt-rock interaction in graphitic granulites and problems of quantifying carbon budget in the lower continental crust

Estimates on the geological carbon cycle are subject to large uncertainties that can be reduced by thorough observation of rocks. In this contribution, we focus specifically on C–O–H fluid-melt-rock interactions in graphitic metapelitic granulites and on their bearing to the carbon budget of granulitic roots of continents. We provide robust microstructural and thermometric constraints on the coexistence of anatectic silicate melts and C–O–H fluids up to near ultrahigh temperature conditions in the archetypal crustal section of Ivrea-Verbano Zone (IVZ, Italian Alps). Fluid inclusions in garnet are investigated before and after high-temperature experiments, and contain considerable proportions of CO2, CH4, N2, but lower H2O than predicted for graphitic systems at granulite facies. When comparing and contrasting the melt compositions obtained by Perple_X and rhyolite-MELTS with natural melts from IVZ, a much better match is obtained by the former, questioning the choice of rhyolite-MELTS for modelling melting equilibria of metasedimentary rocks and for quantifying carbon budget of the lower crust. Overall, data show that assuming only a limited extent of fluid-melt immiscibility in the deep crust contradicts the evidence from natural rocks and prompts to an incomplete view of actual carbon behavior and carbonic fluids. The available experimental dataset on CO2 solubility in felsic melts cannot be used to interpret the volatile budget of melt inclusions in graphitic migmatites and granulites, as most solubility experiments were conducted under carbonate-saturated (i.e. highly oxidizing) conditions which maximize CO2 content of melt, compared to graphitic (i.e. more reducing) protoliths. As a consequence, thermodynamic models still cannot account for all the complexities related with interactions among H2O–CO2–CH4 ternary fluids, H2O- and CO2-bearing anatectic melts and graphite-bearing residues in graphitic metapelites. Targeted experimental studies are therefore crucial to boost substantial computational efforts, before any precise estimates on carbon budget and fluxes in the lower anatectic crust can be made.

Mechanism of Metallic Uranium and Bimetallic U-Ga, U-Cd Alloys Electrodeposition in Molten LiCl–KCl–CsCl Eutectic

Speciation and behavior of uranium (III) chloride in the ternary low melting LiCl–KCl–CsCl eutectic was studied by electrochemistry and spectroscopy techniques. Cathodic reduction of U(III) ions on inert (tungsten) and reactive (gallium, cadmium) electrodes was investigated at 623–923 K using cyclic and square wave voltammetry. The potential scan rate was changed from 0.075 to 0.5 V s−1 in all experiments. It was established that the electrochemical reduction process on the inert electrode was irreversible, proceeded in one stage, and was controlled by the charge transfer. Formation of uranium alloys with gallium and cadmium was studied using active liquid Ga and Cd electrodes. Reduction of uranium ions of the reactive electrodes proceeded with considerable depolarization. The effect of current density on the composition of the cathodic product was considered. Conditions for the electrochemical production of alloys of a given composition were determined.

Antiferromagnetism of CeCd0.67As2 existing deep inside the narrow gap semiconducting state

Single crystals of $R$Cd$_{0.67}$As$_2$ ($R$ = La and Ce) have been synthesized by high temperature ternary melt and their physical properties have been explored by means of magnetization, specific heat, electrical resistivity, Hall coefficient, and thermoelectric power measurements. $R$Cd$_{0.67}$As$_2$ compounds indicate a (structural) phase transition at high temperatures, accompanied by a remarkable increase of the electrical resistivity with an extremely low carrier concentration. CeCd$_{0.67}$As$_2$ exhibits a large magnetic anisotropy and an antiferromagnetic (AFM) order below $T_{N} = 4$~K. Magnetic susceptibility curves, together with magnetization isotherms and specific heat, are analyzed by the point charge model of crystalline electric field (CEF). In the paramagnetic state, the observed magnetic properties can be well explained by the CEF effects, implying that the 4$f$ moments remain localized. Electrical resistivity measurements, together with Hall resistivity and thermoelectric power, also suggest highly localized 4$f$ electrons, where Kondo contributions are negligible. The low temperature physical properties manifest strong magnetic field dependencies. For $H \perp c$, $T_{N}$ shifts to lower temperature as magnetic field increases, and eventually disappears at $H_{c} \sim 60$~ kOe. Inside the AFM state, three metamagnetic transitions are clearly evidenced from the magnetization isotherms. The RKKY interaction may be responsible for the AFM ordering in CeCd$_{0.67}$As$_2$, however it would have to be mediated by extremely low charge carriers. Although the AFM ordering temperature in CeCd$_{0.67}$As$_2$ can be continuously suppressed to zero, no AFM quantum phase transition is expected due to the lack of conduction electron clouds to screen the 4$f$ moments.

Selection Criterion of Stable Dendritic Growth for a Ternary (Multicomponent) Melt with a Forced Convective Flow

A stable growth mode of a single dendritic crystal solidifying in an undercooled ternary (multicomponent) melt is studied with allowance for a forced convective flow. The steady-state temperature, solute concentrations and fluid velocity components are found for two- and three-dimensional problems. The stability criterion and the total undercooling balance are derived accounting for surface tension anisotropy at the solid-melt interface. The theory under consideration is compared with experimental data and phase-field modeling for Ni98Zr1Al1 alloy.

Quantum oscillations and physical properties of LaCu2Si2

High quality single crystals of LaCu2Si2 are successfully grown from a high temperature ternary melt. The physical properties are investigated by measurements of specific heat, electrical resistivity, thermoelectric power, and magnetization. The electrical resistivity follows an ordinary metallic behavior with a residual resistivity ρ0 = 0.87 μΩ cm and residual resistivity ratio (RRR) ρ(300 K)/ρ0 ∼ 13. Quantum oscillations for H || ab are clearly observed from electrical resistivity, magnetization, and thermoelectric power measurements in magnetic fields as low as 30 kOe and at temperatures as high as 20 K. In the earlier studies no quantum oscillations have been observed from both single crystal and polycrystalline samples. The small value of ρ0 and quantum oscillations indicate a high quality of LaCu2Si2 samples used in the present study. The cyclotron effective mass obtained from the analysis of the de Haas – van Alphen (dHvA) oscillations is small m0 ∼ 0.14me, which is consistent with the small electronic specific heat coefficient γ ∼ 3.4 mJ/mol K2. The detected dHvA frequencies of LaCu2Si2 are compared with those of the RCu2Si2 (R = Y, Th, and Ce) family.

Anisotropic magnetism and antiferromagnetic ordering in the Kondo-lattice compound YbCuAs2

Single crystals of $R\mathrm{Cu}{\mathrm{As}}_{2}$ ($R=\mathrm{Yb}$ and Y) have been synthesized from high temperature ternary melts and their physical properties investigated by magnetization, electrical resistivity, specific heat, and $\ensuremath{\mu}\mathrm{SR}$ measurements. Magnetization measurements of $\mathrm{Yb}\mathrm{Cu}{\mathrm{As}}_{2}$ suggest that $4f$ electrons are almost localized and the observed magnetic properties are influenced by the Kondo and Ruderman-Kittel-Kasuya-Yosida interactions as well as crystalline electric field (CEF) effect. Electrical resistivity measurements show typical Kondo-lattice behavior with a broad local maximum around $\ensuremath{\sim}125\phantom{\rule{0.16em}{0ex}}\mathrm{K}$ and a secondary hump around $\ensuremath{\sim}20\phantom{\rule{0.16em}{0ex}}\mathrm{K}$, suggesting a strong hybridization between $4f$ and conduction electrons. A small residual resistivity $(\ensuremath{\sim}0.68\phantom{\rule{4pt}{0ex}}\ensuremath{\mu}\mathrm{\ensuremath{\Omega}}\phantom{\rule{0.16em}{0ex}}\mathrm{cm})$ and the large residual resistivity ratio ($\mathrm{RRR}\ensuremath{\sim}150$) indicate a high quality single crystal. An analysis of the specific heat data indicates that the ground state doublet is well separated from excited CEF levels. Although an earlier neutron scattering experiment on powdered $\mathrm{Yb}\mathrm{Cu}{\mathrm{As}}_{2}$ samples observed no magnetic Bragg peaks associated with magnetic ordering, thermodynamic and transport property measurements in this study clearly indicate antiferromagnetic ordering below ${T}_{N}=3.7\phantom{\rule{0.16em}{0ex}}\mathrm{K}$ which is also supported by a muon spin relaxation investigation of $\mathrm{Yb}\mathrm{Cu}{\mathrm{As}}_{2}$ single crystals. The Kondo temperature (${T}_{K}$) estimated from the magnetic entropy and magnetic susceptibility is in the range $5--7.5\phantom{\rule{0.16em}{0ex}}\mathrm{K}$, which is of similar magnitude as ${T}_{N}$. Interestingly, magnetization, specific heat, and electrical resistivity measurements as a function of magnetic field show a hysteresis loop both below and above ${T}_{N}$.

Amphoteric behavior of component and microstructure feature on CaO-Al2O3-TiO2 ternary melt by molecular dynamics simulation

Amphoteric oxides, Al2O3 and TiO2, are essential compositions in metallurgical melts and have an important role in regulating slag performance. In this study, the structural characteristics of the CaO-Al2O3-TiO2 ternary melt, such as local structural order and particle coordination changes, are investigated by molecular dynamics simulation. The effects of Al2O3 and TiO2 on the microstructure of various melts and their role differences are analyzed. The results show that the amphoteric behavior of TiO2 dominate with low titanium concentration ranging from 3 mol% to 18 mol%. It promotes the change of Al2O3 from acidic to alkaline. Meanwhile, the value of (NBO/T) increases and the melt’s viscosity decreases. In addition, the amphoteric behavior of TiO2 dominate in the high titanium system, ranging from 52 mol% to 66 mol% of TiO2 concentration. It promotes the conversion of TiO2 to alkaline. Meanwhile, the melt polymerization and viscosity are enhanced. As for the ‘aluminum avoidance principle’, it only applies to low-titanium system and invalid in high-titanium system.