First-principles Research Articles

Integration of Neural Networks and First-Principles Model for Optimizing l-Lactide Branched Polymerization.

Addressing the growing demand for sustainable materials, this research paves the way for the efficient consumption and sustainable production of branched polylactide (PLA). A novel hybrid modeling approach combines first-principles (FP) model with artificial neural network (ANN) for ring-opening polymerization (ROP). The hybrid ANN, trained with FP model data, demonstrated optimal performance with a hidden layer of 20 neurons, achieving a root mean square error (RMSE) of 0.004 and a regression coefficient (R2) of 0.99. The hybrid model accurately predicted key polymer properties, including average molecular weights (Mn and Mw), polydispersity index (PDI), degree of branching (DB), monomer conversion, and polymerization time. Validation was performed on various branched PLA compositions (PLLH80, PLLH94, and PLLH97). Multiobjective optimization (MOO) using NSGA-II showed strong agreement between FP model and hybrid ANN across six case studies, highlighting their effectiveness in predicting polymerization outcomes.

Development of a Direct Electron Transfer-Type Enzymatic Sensor for the Spermine in Saliva, a Marker of Pancreatic Cancer

Introduction Pancreatic cancer (PC) has one of the worst prognoses among cancers because of the scarcity of early symptoms and the delay of early diagnosis. Current most available and advanced diagnosis methods of PC are based on imaging technologies such as computed tomography, positron emission tomography, magnetic resonance imaging, and endoscopic ultrasonography. However, such imaging methods cannot diagnose PC at an early stage. Patients diagnosed with PC often have metastasized, and only less than 30% of PC patients can undergo surgical resection. Therefore, there is an urgent need for more accurate, simple, and innovative methods for PC early diagnosis. Recently, salivary spermine has attracted attention as a new biomarker for PC <Asai et.al., Cancers, 2018>, and studies aimed at detecting spermine are attempted. Thus, we focused on the electrochemical measurements of spermine using enzyme.Electrochemical enzymatic biosensors are categorized into three generation principles based on the electron acceptors used for the oxidative half reaction. The 1st generation principle uses oxygen as the electron acceptor. However, the signal is affected by the oxygen concentration and a high applied potential is required to detect hydrogen peroxide. The 2nd generation principle requires the use of synthetic electron acceptors or mediators. The 3rd generation principle utilizes enzymes which are capable of direct electron transfer (DET) to the electrode, does not require any additional electron acceptors. Notably, the 3rd generation principle-based sensor can monitor target by applying relatively lower potentials, therefore will not be influenced by redox-active interferents. Considering the salivary sample contains variety of ingredient potentially interferants for enzymatic spermine monitoring, the DET-type enzymatic sensors are ideal. However, those have never been reported. In this work, we present the development of the 3rd generation principle based enzymatic sensor for spermine by employing a unique Methods SpdH was recombinantly produced using E. coli and purified by Ni2+ affinity chromatography. An electrochemical enzymatic sensor was constructed by immobilizing SpdH on gold electrodes. The DET properties of the enzyme-immobilized electrode were evaluated by cyclic voltammetry with an Ag/AgCl reference electrode and a Pt wire counter electrode, respectively. Chronoamperometry measurement was performed with SpdH immobilized gold electrodes using artificial saliva or phosphate buffer, where spermine was spiked with different concentrations. Results and Discussions For spermine measurement, we selected SpdH which was expected to be capable of DET with electrode, which was predicted from its unique protein structure. The crystal structure of SpdH shows the presence of heme b and FAD as the co-factors of the enzyme with a distinctive character. The heme b locates on the surface of the enzyme and is expected to accept electron from FAD and can transfer electrons to the electrode. The cyclic voltammetry observation showed the spermine-concentration dependent peak current increase in the absence of any additional electron acceptor, revealing that the enzyme harbors DET ability with electrode. The spermine concentration dependent currents increase was also observed by chronoamperometry measurement. The limit of detection of the enzymatic electrode was sub-micromolar level, which covers the physiologically relevant ranges of spermine. Furthermore, the electrode did not respond to electrochemically active interferants due to its characteristic of low redox potentials of heme being measurable at low potentials.This is the first report to construct a spermine sensor with a DET-type enzyme. DET-type enzymes allow simple electrode configurations operating at lower potentials. These features are less susceptible to redox-active interferents. The current achievements promise the future realization of a simple diagnosis method for PC using a non-invasive specimen, saliva. Figure 1

Grey‐box modelling for estimation of optimum cut point temperature of crude distillation column

AbstractA grey‐box modelling framework was developed for the estimation of cut point temperature of a crude distillation unit (CDU) under uncertainty in crude composition and process conditions. First principle (FP) model of CDU was developed for Pakistani crudes from Zamzama and Kunnar fields. A hybrid methodology based on the integration of Taguchi method and genetic algorithm (GA) was employed to estimate the optimal cut point temperature for various sets of process variables. Optimised datasets were utilised to develop an artificial neural networks (ANN) model for the prediction of optimum values of cut points. The ANN model was then used to replace the hybrid framework of the Taguchi method and the GA. The integration of the ANN and FP model makes it a grey‐box (GB) model. For the case of Zamama crude, the GB model helped in the decrease of up to 38.93% in energy required per kilo barrel of diesel and an 8.2% increase in diesel production compared to the stand‐alone FP model under uncertainty. Similarly, for Kunnar crude, up to 18.87% decrease in energy required per kilo barrel of diesel and a 33.96% increase in diesel production was observed in comparison to the stand‐alone FP model.

Modelling a chemical plant using grey‐box models employing the support vector regression and artificial neural network

AbstractIn this work, the performances of a nonlinear dynamic industrial process are examined using grey‐box (GB) models. To understand the dynamics of the system, the transient state is targeted. A white‐box (WB) model holds the prevailing knowledge using some assumptions. The performance of this model is limited. Artificial neural network (ANN) and support vector regression (SVR), which are techniques employed in numerous chemical engineering applications, are employed to construct the associated black‐box (BB) models. GA is used to optimize the SVR parameters. Dimensional and range extrapolations of different manipulated inputs, feed concentrations, feed temperatures, and cooling temperatures of the GB model and BB model are discussed. The different inputs extrapolation has different results because each input's effectiveness in the system is different. The results are compared, and the best model is suggested among the models, ANN, SVR, first principle (FP)‐ANN serial structure, FP‐ANN parallel structure, FP‐SVR serial structure, and FP‐SVR parallel structure.

Performance Comparisons of NequIP and DPMD Machine Learning Interatomic Potentials for Tobermorites

Machine Learning Interatomic Potentials (MLIPs) present a significant advancement in fitting molecular potential energy surfaces and predicting molecular crystal structures and mechanical properties by closely approximating first-principles (FP) calculations results while substantially reducing computational time. In this work, utilizing datasets derived from FP calculations, neural equivariant interatomic potentials (NequIP) and Deep Potential Molecular Dynamics (DPMD) was implemented to develop MLIPs for tobermorite 9, 11, 14 Å. The accuracy and efficiency of NequIP and DPMD were assessed by predicting energies and forces as well as performing molecular dynamics (MD) simulations. The findings indicate that NequIP closely fits the results of FP calculations, with both accuracy and efficiency improved by 1∼2 orders of magnitude compared to the DPMD. This demonstrates that MLIPs developed through NequIP for tobermorites have the capacity to enhance the capability of large-scale molecular dynamics simulations.

Investigation for the hydrogen storage properties of XCrH3 (X=Na, K) and KYH3(Y[dbnd]Mo, W) perovskite type hydrides based on first-principles

Perovskite type hydrides have emerged as a focal point in hydrogen storage applications as potential candidates in recent years. However, the currently synthesized perovskite hydrogen storage materials suffer from high hydrogen desorption temperatures and unstable reactions. To identify promising perovskite hydrogen storage materials, this study explores the hydrogen storage properties of XCrH3 (X = Na, K) and KYH3 (YMo, W) based on first-principles (FPs) calculations. The results indicate that NaCrH3 and KCrH3 exhibit half-metallic properties, while KMoH3 and KWH3 are metallic materials with high hydrogen storage safety. Mechanical property analysis reveals that NaCrH3 exhibits the highest hardness, indicating better stability for hydrogen storage. Thermodynamic analysis shows that KCrH3 has a lower hydrogen desorption temperature than the other three compounds, with NaCrH3 having the second-lowest. The gravimetric hydrogen storage capacities of NaCrH3, KCrH3, KMoH3, and KWH3 are 3.85 wt%, 3.21 wt%, 2.19 wt%, and 1.34 wt%, respectively. Overall, NaCrH3 demonstrates superior gravimetric hydrogen storage capacity, stability, and safety, making it a promising candidate for new perovskite type hydrogen storage materials.

EVALUATION OF ENERGY BAND STRUCTURE OF HALF-HEUSLER ALLOY LiZnX (X = As, P, and Sb) USING FIRST PRINCIPLE CALCULATION

The evaluation of energy band structure plays a vital role in understanding the electronic properties of materials. This research, we investigate the energy band structure of Half-Heusler alloys LiZn(X = As, P, and Sb) using a first principle approach based on Density Functional Theory (DFT). These alloys are of particular interest due to their potential applications in thermoelectric and spintronics devices. The corresponding Density of States (DOS) for the tripartite compounds LiZnX (X=As, P, and Sb) have been calculated and the contributions of the Li, Zn, As, P and Sb orbital to the Density of States at ambient pressure. This also confirmed that LiZnX (X=As, P, and Sb) is a semi-conductor with a narrow band-gap between the occupied and unoccupied regions around the Fermi level. The orbitals Li -1s, As-4p, As- 4s, Zn-3d has the highest contributions. The dominant of the orbitals P-1s and P-2p before the Fermi- level and Zn-2p after the Fermi-level are observed. We observed the dominant of the orbitals Sb-1s, Sb-3d, Li-1s, Li-2s, Zn-3d shows weak hybridization and low contribution. This features indicates that the covalent bond between these two atoms is weak, and could be responsible for the mechanical instability observed in the calculation. Meanwhile the band structure calculated and presented has narrow band-gab of 0.625. 0.937 and 0.313 respectively for the tripartite compound LiZnX(X=As, P, and Sb) and its a direct band-gap semiconductor. The obtained energy band structures provide valuable information about the electronic properties of LiZn (X = As, P, and Sb) alloys. The presence of band gaps is crucial for thermoelectric applications, as it indicates the presence of regions where electrons and holes are confined, enabling efficient charge transport.

Detailed and reduced chemical kinetic model of CH4/air mixtures combustion based on high-precision first-principles molecular dynamics simulation

CH4/air combustion reaction kinetics has attracted many research interests because of its wide application in engineering. In this work, the CH4 oxidation in O2 at high temperatures is simulated based on the first-principles molecular dynamics method first time. Two new intermediates, HCOOH and O3, and their effect on the CH4 oxidation are revealed. Specifically, adding HCOOH-related reactions into current model GRI 3.0, NUIG 1.1 and USC II may significantly improve their predictions on the premixed laminar flame speed of CH4/air mixtures. Elementary reaction analysis also discovered that radicals, such as OH, HO2, and others play key roles in CH4 oxidation. These radicals act as highly active oxidants and promote final product formation. Combining elementary reaction with simplifying technique, a novel First-Principle (FP) and 30-step Reduced-FP (R-FP) chemical kinetic model are constructed. By using our two models, the ignition delay time, species concentration in flow reactor and premixed laminar flame speed of premixed reactive mixture (CH4/O2/Ar, CH4/O2/H2O/N2 and CH4/air) combustion are successfully predicted and verified by comparing with experimental results. In general, the FP and R-FP models will be advantageous for application of engineering in the future.

The Structural, stability, electronic and optoelectronics properties of Y2AgBiX6 (Y = K, Na, li; X = I, Br, Cl) halide double Perovskites: A First-Principles study

The crystal structure, electronic and optoelectronic properties of halide double perovskites Y2AgBiX6 (Y = K, Na, Li; X = I, Br, Cl) have been studied by first principles (FPs) in this paper. According to the elastic modulus and formation energy of crystal Li2AgBiX6, its structural stability is poor. In addition, the dependence of their electronic and mechanical properties, the effective mass of carriers and light absorption on elemental composition has been analyzed, and it is inferred that these materials have good electron transport properties. The photovoltaic characteristics of halide perovskite materials (PMs) based on potassium and sodium are studied. The photovoltaic properties of perovskite films with thickness of 1 μm, 5 μm and 10 μm were further analyzed, and it was found that the absorption was the highest when the film thickness was 10 μm. The current–voltage curve of the solar cell with above perovskite as absorbing material is calculated and analyzed by first principles. The result shows that Na2AgBiBr6, Na2AgBiCl6, K2AgBiI6 and K2AgBiCl6 are predicted to have excellent photovoltaic properties. Compared to other commonly used PMs, perovskite Y2AgBiX6 (Y = K, Na, Li; X = I, Br, Cl) has a smaller effective mass, which may contribute to good carrier transport and high performance of its optoelectronic devices. The absorption curves of K2AgBiBr6, K2AgBiCl6, Na2AgBiBr6, Li2AgBiBr6, and Li2AgBiCl6 matched well with the spectral irradiance of AM 1.5. Taken together, K2AgBiBr6 and Na2AgBiBr6 are the most suitable candidates for solar cell absorption layers because of their appropriate band gap and good stability.

Correlation between Local Structure and Reaction Mechanism in Disordered Rock-Salt-Type Lithium–Manganese Oxides

First-principles (FP) calculations of disordered rock-salt-type lithium–manganese oxides were performed by assuming a quasi-random structure model with the composition of Li16Mn16O32. The effects of randomness in the crystal structure and the changes in local atomic configurations during the Li extraction process on electronic structures, phase stability, and electrochemical properties were investigated by the FP calculations. The calculations indicate that the randomness of local atomic configurations, which originates from the disordered structure, makes it possible for Mn and O sites to take on a variety of electronic states. Analyses of densities of states, magnetic moments of Mn atoms, and Bader charges indicate the possibility that not only Mn but also oxygen atoms contribute to the redox reaction. The migration of Li and Mn is found to appear from octahedral to tetrahedral sites at a high state of charge. The FP calculations suggest that the change in local atomic configurations and electronic structures causes the transition in voltage profiles from sloping to 3 V plateau shape as the cycle progresses, which is characteristic of spinel-related positive electrode materials.

First principles molecular dynamics study of proton disorder in [formula omitted] phase of H2 hydrate

The proton-disordered and -ordered crystal structures proposed based on X-ray diffraction experiments for the C1′ phase of H2 hydrate were investigated using first-principles (FP) path-integral ring-polymer molecular dynamics (RPMD) simulations. Our simulated Raman spectra obtained from the FP-RPMD trajectories provide unambiguous evidence for static orientational disorder of H2O molecules forming stacked H-bonded rings. This feature is essential to reproduce the characteristics of the observed Raman spectra. Furthermore, the present study shows that van der Waals interactions need to be treated carefully to achieve satisfactory computational results. The findings of this study can help for better understanding of icy materials.

Fast atomic structure optimization with on-the-fly sparse Gaussian process potentials *

We apply on-the-fly machine learning potentials (MLPs) using the sparse Gaussian process regression (SGPR) algorithm for fast optimization of atomic structures. Great acceleration is achieved even in the context of a single local optimization. Although for finding the exact local minimum, due to limited accuracy of MLPs, switching to another algorithm may be needed. For random gold clusters, the forces are reduced to ∼0.1 eV Å−1 within less than ten first-principles (FP) calculations. Because of highly transferable MLPs, this algorithm is specially suitable for global optimization methods such as random or evolutionary structure searching or basin hopping. This is demonstrated by sequential optimization of random gold clusters for which, after only a few optimizations, FP calculations were rarely needed.

Band structure of molybdenum disulfide: from first principle to analytical band model

A simple band model such as the effective mass approximation (EMA) can be used to quickly obtain the lower-energy region for the band structure of monolayer molybdenum disulfide. But the EMA band model cannot give the correct description for the band structure in the higher-energy region. To address this major issue, we propose an analytical band calculation (ABC) model to study monolayer molybdenum disulfide. Important parameters of the ABC model are obtained by fitting the three-direction band structure of monolayer molybdenum disulfide obtained from the first-principles (FP) method. The proposed ABC model fits well with the FP band structure calculation result for monolayer molybdenum disulfide. We also use the ABC model to calculate physical quantities used in carrier transport such as density of states and group velocity. Our ABC model can be extended and further utilized for calculating the key physical quantities of ballistic transport of 2D semiconductor materials.

ORIGEN: ON FIRST PRINCIPLES: A READER’S EDITION. Translated by John Behr. Oxford: Oxford University Press, 2019. Pp. lxxxviii + 357. Paper, $35.00.

Religious Studies ReviewVolume 48, Issue 1 p. 106-106 Short Reviews of Recent Publications: Theology ORIGEN: ON FIRST PRINCIPLES: A READER’S EDITION. Translated by John Behr. Oxford: Oxford University Press, 2019. Pp. lxxxviii + 357. Paper, $35.00. First published: 28 April 2022 https://doi.org/10.1111/rsr.15642Read the full textAboutPDF ToolsRequest permissionExport citationAdd to favoritesTrack citation ShareShare Give accessShare full text accessShare full-text accessPlease review our Terms and Conditions of Use and check box below to share full-text version of article.I have read and accept the Wiley Online Library Terms and Conditions of UseShareable LinkUse the link below to share a full-text version of this article with your friends and colleagues. Learn more.Copy URL Share a linkShare onFacebookTwitterLinked InRedditWechat No abstract is available for this article. Volume48, Issue1Georges Bataille and the SacredMarch 2022Pages 106-106 RelatedInformation

Combined first-principles and machine learning study of the initial growth of carbon nanomaterials on metal surfaces

To understand the effects of H of hydrocarbon precursors and alloying elements of catalysts on the initial growth of carbon nanomaterials, this work combined first-principles (FP) calculation and machine learning (ML) methods to study the adsorption of complete hydrocarbon molecules on various alloy surfaces. Specifically, we calculated the energetics and geometries of the adsorption of acetylene (C2H2) molecules on the pristine Ni, Co, and Fe metal and a benzene molecule on Ni-X [X = thirty 3d, 4d, and 5d transition metal (TM) elements] alloy surfaces. The results suggest that benzene has larger adsorption energies on Ni-X (X = Y, Cd, Sc, Lu, Ag, Ti, Hg, Pd, Zn, V, Cu, and Au) than that on the pristine Ni surface. Furthermore, the ML models were constructed using the surface “Center-Environment” (SCE) features to predict the FP results. The ML models suggest that the features critical to adsorption energies include the size, electronegativity, and valence of the doping element. We proposed an “add and dehydrogenate” mechanism underlying the initial growth of nanocarbon with hydrocarbon precursors. These results provide a theoretical basis in the design of alloy catalysts including in-plane single atom catalysts.

Simulating Raman spectra of hydrogen hydrates using first-principles path-integral ring-polymer molecular dynamics

A variant of first-principles (FP) path-integral ring-polymer molecular dynamics (RPMD) simulations was used to investigate the C0,C1, and C2 H2 hydrates at 280 K aiming at reproducing their experimental Raman spectra. We find that our implementation based on white-noise Langevin thermostat applied to the H2O sub-system only in the formulation of FP-RPMD is able to reproduce with high accuracy the characteristic features of the measured Raman spectra of H2 hydrates in all C0,C1, and C2 phases, demonstrating potential of our computational approach for investigating H-containing materials under pressure.

ORIGEN: ON FIRST PRINCIPLES, 2 Vols. By Origen. Translated by John Behr. Oxford: Oxford University Press, 2017. Pp. cv + 664. Hardcover, $210.00.

Religious Studies ReviewVolume 47, Issue 4 p. 516-517 Short Reviews of Recent Publications: Theology ORIGEN: ON FIRST PRINCIPLES, 2 Vols. By Origen. Translated by John Behr. Oxford: Oxford University Press, 2017. Pp. cv + 664. Hardcover, $210.00. First published: 19 January 2022 https://doi.org/10.1111/rsr.15510Read the full textAboutPDF ToolsRequest permissionExport citationAdd to favoritesTrack citation ShareShare Give accessShare full text accessShare full-text accessPlease review our Terms and Conditions of Use and check box below to share full-text version of article.I have read and accept the Wiley Online Library Terms and Conditions of UseShareable LinkUse the link below to share a full-text version of this article with your friends and colleagues. Learn more.Copy URL Share a linkShare onFacebookTwitterLinked InRedditWechat No abstract is available for this article. Volume47, Issue4New Religious Movements and Documentary FilmsDecember 2021Pages 516-517 RelatedInformation

Energy and exergy analysis of an absorption system with working pairs LiBr-H2O and Carrol-H2O at applications of cooling and heating

In this work, an air-cooled, single effect solar-driven absorption system is being evaluated from the point of view of 1st and 2nd thermodynamic principles for two different applications: absorption chiller and heat pump. One of the most widely used working pairs, LiBr-H2O, is applied in this study due to its high performance in the absorption cycle. Their performance is compared with another working pair Carrol-H2O (Carrol contains LiBr and EG -Ethylene glycol- with a mass ratio of 4.5:1). The Carrol solution has the advantage of reducing the crystallization risk at the high concentration solution that enters the absorber. The numerical modelling was implemented on a modular object-oriented simulation platform (NEST platform tool), which allows linking different components, considered objects. In the simulations performed, the heat source temperature in the system is in the range of 70–90 ∘C, and the inlet temperature at evaporator secondary circuit at chiller application is fixed in two values, 9∘C and 14∘C, and for heat pump application in 0∘C and -5∘C. Moreover, EG is added to the evaporator at heat pump application to prevent the refrigerant water from freezing below zero. The studied mass concentration range of EG of 10–40%. The result shows the COP of an absorption chiller and heat pump are around 0.7 and 1.6, respectively, and the COPEX values are 0.2-0.6 at chiller application and 0.5-1.5 at heat pump application. When compared with LiBr system, Carrol system has about 6.4% higher COP, about 6.3% higher COPEX, and a decrease of about 19% of cooling capacity. In the heat pump application, the heat source temperature should be lower than 90∘C, and EG concentration at evaporator has been chosen as 30% as an optimal value. According to the operation condition, this EG concentration has been determined to avoid freezing in the evaporator in the studied working range. However, too much EG significantly decreases the pressure in the evaporator and increases the viscosity, hence will increase the maintenance of equipment as more vacuum tightness is required.

A comparison of two methodologies for radiotherapy treatment plan optimization and QA for clinical trials.

Background and purposeThe efficacy of clinical trials and the outcome of patient treatment are dependent on the quality assurance (QA) of radiation therapy (RT) plans. There are two widely utilized approaches that include plan optimization guidance created based on patient‐specific anatomy. This study examined these two techniques for dose‐volume histogram predictions, RT plan optimizations, and prospective QA processes, namely the knowledge‐based planning (KBP) technique and another first principle (FP) technique.MethodsThis analysis included 60, 44, and 10 RT plans from three Radiation Therapy Oncology Group (RTOG) multi‐institutional trials: RTOG 0631 (Spine SRS), RTOG 1308 (NSCLC), and RTOG 0522 (H&N), respectively. Both approaches were compared in terms of dose prediction and plan optimization. The dose predictions were also compared to the original plan submitted to the trials for the QA procedure.ResultsFor the RTOG 0631 (Spine SRS) and RTOG 0522 (H&N) plans, the dose predictions from both techniques have correlation coefficients of >0.9. The RT plans that were re‐optimized based on the predictions from both techniques showed similar quality, with no statistically significant differences in target coverage or organ‐at‐risk sparing. The predictions of mean lung and heart doses from both methods for RTOG1308 patients, on the other hand, have a discrepancy of up to 14 Gy.ConclusionsBoth methods are valuable tools for optimization guidance of RT plans for Spine SRS and Head and Neck cases, as well as for QA purposes. On the other hand, the findings suggest that KBP may be more feasible in the case of inoperable lung cancer patients who are treated with IMRT plans that have spatially unevenly distributed beam angles.

Machine learning aided first-principles studies of structure stability of Co3(Al, X) doped with transition metal elements

To understand the alloying effects on the stability of Co3Al precipitate phase in Co-based superalloy, the energetic stability and structure of ternary alloy Co3(Al, X) doped with the thirty 3d, 4d, and 5d transition metal (TM) elements were studied in this work using first-principles (FP) computation and machine learning (ML) methods. Our FP computation indicated that Hf, Ta, and Ti doping were thermodynamically most stable. Based on the FP computation data, the ML models with three types of chemical composition (CC) and Center-Environment (CE) features, were developed to predict the formation energies and lattice constants of Co3(Al, X) (X = 3d, 4d, and 5d TM elements). The results show that these ML models all had good prediction accuracy with averaged mean absolute errors (<MAE > ) ~ 0.02 eV/atom for formation energies and ~ 0.01 Å for lattice constants. Then, the effects of important features were discussed on the energy and geometry properties. To study the alloying effects on the stability of Co3(Al, W) precipitate phase, the ML models of Co3(Al, X) were found to be equally accurate to predict the untrained structures of Co3(Al, WX3) where X = 3d, 4d, and 5d TM elements excluding Co and W. We found that the Co3(Al, WX3) structures became most stable when X are IVB and VB group TM elements. This work show that machine learning methods can efficiently extend the capabilities of first-principles predictions on the structure stabilities in multi-component alloy design.