Figure - available from: Advanced Materials
This content is subject to copyright. Terms and conditions apply.
![a) Demonstration of an optimal catalyst based on Sabatier's postulate of chemical catalysis based on unstable intermediates. b) The temperature at which log of reaction rate = 0.8 versus the formate adsorption energy showing the “volcano” curve.[249]](publication/340073257/figure/fig2/AS:11431281173395075@1688861998249/a-Demonstration-of-an-optimal-catalyst-based-on-Sabatiers-postulate-of-chemical.png)
a) Demonstration of an optimal catalyst based on Sabatier's postulate of chemical catalysis based on unstable intermediates. b) The temperature at which log of reaction rate = 0.8 versus the formate adsorption energy showing the “volcano” curve.[249]
Source publication
The chemical conversion of small molecules such as H2, H2O, O2, N2, CO2, and CH4 to energy and chemicals is critical for a sustainable energy future. However, the high chemical stability of these molecules poses grand challenges to the practical implementation of these processes. In this regard, computational approaches such as density functional t...
Similar publications
![The DIKW structure for the engineering decisions [5]](publication/351578108/figure/fig1/AS:11431281117190164@1675396563232/The-DIKW-structure-for-the-engineering-decisions-5_Q320.jpg)
![Illustration of preposterior decision analysis in decision tree [37]](publication/351578108/figure/fig2/AS:11431281117204873@1675396563301/Illustration-of-preposterior-decision-analysis-in-decision-tree-37_Q320.jpg)
The concept of Value of Information (VoI) has attracted significant attentions within the civil engineering community over especially the last decade. Triggered by the increasing focus on structural health monitoring, availability of data and emerging techniques of Big Data analysis and Artificial Intelligence, important insights on how to take ben...
With the development of computer science, more and more hardware implementations can be reproduced by software programming, bringing compact, cheap, and fast components to imaging instrumentation. In recent years, computational methods have been introduced into spectral detection, and computational spectrum acquisition implementations have emerged....
Cassava production and productivity in Eastern, Central, and Southern Africa are ravaged by cassava brown streak disease (CBSD), causing yield losses of up to 100% when susceptible varieties are grown. Efforts to develop CBSD-resistant clones are underway. However, the methods for screening CBSD resistance currently vary between breeders and pathol...
Citations
... The goal is to identify complex patterns and correlations within the known data, which can then be extrapolated to accurately forecast unexplored data. Once the training set is sufficiently large and representative of the catalytic property of interest, the trained model can be used to predict this quantity also for systems not considered during the training process, 29,[34][35][36][37] without the need of further DFT simulations. This approach has seen applications in various chemical fields, [38][39][40] including catalysis. ...
The electrochemical reduction of CO 2 (CO 2 RR) has the potential to allay the greenhouse gas effect while also addressing global energy challenges by producing value-added fuels and chemicals (mostly C 2 molecules...
... In the field of fuel cells, there have been relevant studies reporting that ML can be employed in optimizing materials and their membrane electrode assemblies, utilizing data related to practical applications [31][32][33][34]. Also, in selecting algorithms, their interpretability is crucial to ensure that the analysis output can be effectively translated into guidance for material preparation and enhance the understanding of reaction mechanisms [35][36][37][38]. It requires the application of black-box explanation methods to directly demonstrate the insights of ML models regarding important parameters and to reveal the influence of each parameter in complex systems [39][40][41]. ...
... The predication of CO 2 reduction properties and catalyst potentials can be accelerated by machine learning (ML) by high throughput DFT. In order to learn from supplied data, recognize patterns, and anticipate performance, well-trained algorithms known as machine learning (ML) have been established for the study of materials [120,121]. A simple instrument to automatically analyze the structure and physical and chemical performance in a highly effective and accurate manner is provided by machine learning [122,123]. ...
Production of ethylene (C 2 H 4) from Carbon dioxide (CO 2) is a promising and effective approach for mitigating the adverse greenhouse effect of anthropogenic CO 2 emission from human activities including chemical industries along with energy deficiency caused by the burning of nonrenewable fossil fuels. The C 2 H 4 , one of the utmost extensively employed chemical compounds, is readily converted into various valuable products that can substitute those derived from fossil fuels. Electrochemical reduction is a process of converting CO 2 into C 2 H 4 in which the electrocatalysts, electrolyzer and electrolyte all play the critical role. Despite considerable improvements in electrochemical CO 2 reduction but its efficacy is still hindered by a number of complications, including the deprived mass transfer efficiency of the elec-trolyzer, the high cost, low activity, poor stability and limited selectivity of electrocatalysts towards C 2 H 4. Accomplishments to date in CO 2 electroreduction to C 2 H 4 have been inspiring, while extensive advances are still required for it to develop a profitable technology able to move society towards renewable energy sources. The current review demonstrates a unified discussion on recent progress in the area of CO 2 reduction to C 2 H 4. The role of a class of promising advanced electrocatalysts, electrolytes and the design of electrolyzer in the electrochemical reduction method is thoroughly reviewed. We also highlight the current developments in photocatalytic CO 2 reduction to C 2 H 4. Lastly, this review is concluded by deliberating the current challenges that are encountered during the production of C 2 H 4 from CO 2 electro-chemical reduction with an objective of revolutionary futuristic developments. This article will provide comprehensive knowledge to the readers on CO 2 electrochemical reduction to C 2 H 4 production.
... The three important intermediates are shown in Figure 1. In theoretical studies, the Gibbs free energy of H* formation, G(H*), has been used to determine the favorable reaction between the hydrogen evolution reaction and the CO2RR [30], while the Gibbs free energies G(C*OOH) and G(O*CHO) can be used to predict the favored products [31][32][33][34][35][36]. Thus, based on theoretical catalytic studies, G(H*), G(C*OOH), and G(O*CHO) can be calculated and compared to determine the product selectivity. ...
The quantitative structure–electrochemistry relationship (QSER) method was applied to a series of transition-metal-coordinated porphyrins to relate their structural properties to their electrochemical CO2 reduction activity. Since the reactions mainly occur within the core of the metalloporphyrin catalysts, the cluster model was used to calculate their structural and electronic properties using density functional theory with the M06L exchange–correlation functional. Three dependent variables were employed in this work: the Gibbs free energies of H*, C*OOH, and O*CHO. QSER, with the genetic algorithm combined with multiple linear regression (GA–MLR), was used to manipulate the mathematical models of all three Gibbs free energies. The obtained statistical values resulted in a good predictive ability (R2 value) greater than 0.945. Based on our QSER models, both the electronic properties (charges of the metal and porphyrin) and the structural properties (bond lengths between the metal center and the nitrogen atoms of the porphyrin) play a significant role in the three Gibbs free energies. This finding was further applied to estimate the CO2 reduction activities of the metal–monoamino–porphyrins, which will prove beneficial in further experimental developments.
... Electrochemical nitrogen reduction reaction (NRR) to produce ammonia in a renewable manner is one of the most promising approaches and has been the focus of research over the last decade. These works cited here clearly show a summary of the progress in the field, new insights, and current challenges (Cui et al., 2018;Wang et al., 2018;Andersen et al., 2019;Chen et al., 2020;Duan et al., 2020;Gu et al., 2020;Liu et al., 2020;Manjunatha et al., 2020;Qing et al., 2020;Chanda et al., 2021;Choi et al., 2021;Du et al., 2021;Li et al., 2021;Rehman et al., 2021;Shen et al., 2021;Yang et al., 2021;Azofra, 2022;Chen et al., 2022;Fuller et al., 2022;Liao et al., 2022;Zhang et al., 2022) , (MacLaughlin, 2019;Singh et al., 2019;Choi et al., 2020;Dražević and Skúlason, 2020;Hanifpour et al., 2020;MacFarlane et al., 2020;Hanifpour et al., 2022a;Li et al., 2022;Wan et al., 2022;Yao et al., 2022). In catalysis, the utilization of nitrogen-rich transition metals is particularly promising where transition metal nitrides (TMNs) were predicted to be good for making ammonia chemically (Michalsky et al., 2015a;Zeinalipour-Yazdi et al., 2015) and electrochemically (Abghoui et al., 2015;Michalsky et al., 2015b;Abghoui et al., 2016;Abghoui and Skúlason, 2017a;Abghoui, 2017;Abghoui and Skúlason, 2017b;Garden et al., 2018;Gudmundsson et al., 2022) through a Mars-van Krevelen mechanism. ...
Following our previous report on N2 reduction reaction (NRR) on the surface of nitrides, we investigated the influence of incorporation of titanium nitride as a stable and inactive-NRR material into the structure of DFT-predicted NRR-active surfaces of chromium, vanadium, niobium, and zirconium nitrides. The outcome of our density functional theory (DFT) based analyses suggests that combination of titanium nitride with vanadium nitride can enhance the potential-determining step of the reaction with up to 20% compared to pure vanadium nitride while maintaining similar number of proton-electron transfer steps for formation of two ammonia molecules. The influence of titanium nitride on chromium nitride is expected to be more pronounced as rate-determining step associated with nitrogen adsorption on the vacancy and regeneration of the catalyst improves by around 90% compared to the pure chromium nitride. This effect on niobium and zirconium nitride is, however, negative as the potential-determining step becomes larger for the case of niobium nitride, and the reaction pathway changes from nitrogen reduction to hydrogen evolution for the case of zirconium nitride. These results not only encourage experimentalists to explore these overlayered structures further in experiments, but it also opens up the avenue for considering the alloys and dopants of these nitrides via both density functional theory modelling and experiments.
... 2-10 Unfortunately, SQM accuracy for barriers is rather limited, 11-13 but because of their high efficiency, they are used in reaction explorations, 7,8,11,14,15 often followed by a subsequent refinement with higher-level QM methods. 16 Machine learning (ML) can be used to attain both high accuracy and low computational cost by replacing or improving QM methods (see some of the plentiful recent reviews and textbooks [17][18][19][20][21][22][23][24][25][26][27][28][29][30], and much effort is put into developing ML models for reactive properties, e.g., barrier heights, [31][32][33][34][35][36][37][38][39][40][41][42][43][44][45][46] rate constants, 44,[47][48][49][50] reactivity, 36,51-53 reaction yields 54 and conditions, 55-60 and catalysts, 38,41,45,[61][62][63][64][65][66][67][68][69] regio- [70][71][72][73][74][75] and enantioselectivity, 76 modeling, discovering and planning reactions and reaction paths, and finding transition-state (TS) geometries [113][114][115][116][117][118][119][120] (see also recent reviews 32,[121][122][123][124][125][126][127][128]. Most of such models are special-purpose, i.e., they were trained only to predict either specific reactive property or were trained for a specific reaction type. ...
Artificial intelligence-enhanced quantum mechanical method 1 (AIQM1) is a general-purpose method that was shown to achieve high accuracy for many applications with a speed close to its baseline semiempirical quantum mechanical (SQM) method ODM2*. Here, we evaluate the, hitherto unknown, performance of out-of-the-box AIQM1 without any refitting for reaction barrier heights on 8 datasets including in total ca. 24 thousand reactions. This evaluation shows that AIQM1's accuracy strongly depends on the type of transition state and ranges from excellent for rotation barriers to poor for, e.g., pericyclic reactions. AIQM1 clearly outperforms its baseline ODM2* method and even more so, a popular universal potential ANI-1ccx. Overall, however, AIQM1 accuracy largely remains at the level of SQM methods suggesting that it is desirable to focus on improving AIQM1 performance for barrier heights in the future. We also show that the built-in uncertainty quantification helps in identifying confident predictions. The accuracy of confident AIQM1 predictions is approaching the level of popular DFT methods for most reaction types. Encouragingly, AIQM1 is rather robust for transition state optimizations even for the type of reactions it struggles with the most. Single-point calculations with high-level methods on AIQM1-optimized geometries can be used to significantly improve barrier heights which cannot be said for its baseline ODM2* method.
... Theoretical studies were added to those experimental results, to better understand the reaction mechanism behind the capture and reduction of CO 2 , as well as proposing a systematic way to improve the existing FLP playing with the intrinsic properties of the atoms involved. [18,22,[37][38][39][40][41][42][43][44][45][46][47][48][49][50][51] Considering phosphorus as a LB, boron as a LA, and that the distance between both atoms, may be imposed by a ring structure and/or large substituents, it is possible to design FLP systems to capture CO 2 . Among potential intramolecular FLPs, the tricyclic derivatives of 9,10-dihydroanthracene (DHA) with different substituents in positions 9 and 10 should be mentioned ( Figure 1). ...
The reactivity of 5,10‐disubstituted dibenzoazaborines and dibenzophosphaborines towards carbon dioxide was studied at the DFT, M06‐2X/def2‐TZVP, computational level. The profile of this reaction comprises of three stationary points: the pre‐reactive complex and adduct minima and the transition state(TS) linking both minima. Initial results show that dibenzoazaborines derivatives are less suitable to form adducts with CO2 than dibenzophosphaborine systems. The influence of the basicity on the P atom and the acidity on the B center of the dibenzophosphaborine in the reaction with CO2 was also explored. Thus, an equation was developed relating the properties (acidity, basicity and boron hybridization) of the isolated dibenzophosphaborine derivatives with the adduct energy. We found that modulation of the boron acidity allows to obtain more stable adducts than the pre‐reactive complexes and isolated monomers.
... [9][10][11] More importantly, apart from Pt electron density, their energy in the partial density of states also demonstrates profound influences on the adsorption and activation of reactants according to the d-band theory as shown in Figure 1a, which is somewhat neglected for this reaction due to the challenge in measuring it directly. [12][13][14][15] Hence, it is highly desirable to decipher the enigmas of electron transfer and/or electron energy change that underlie the Pt electronic effects, thus providing generalizable guidelines of Pt catalysts to inhibit CO poisoning. ...
Carbon monoxide (CO) is notorious for its strong adsorption to poison platinum group metal catalysts in the chemical industry. Here, we conceptually distinguish and quantify the effects of the occupancy and energy of d electrons, emerging as the two vital factors in d‐band theory, for CO poisoning of Pt nanocatalysts. The stepwise defunctionalization of carbon support is adopted to fine‐tune the 5d electronic structure of supported Pt nanoparticles. Excluding other promotional mechanisms, the increase of Pt 5d band energy strengthens the competitive adsorption of hydrogen against CO for the preferential oxidation of CO, affording the scaling relationship between Pt 5d band energy and CO/H2 adsorption energy difference. The decrease of Pt 5d band occupancy lowers CO site coverage to promote its association with oxygen for the total oxidation of CO, giving the scaling relationship between Pt 5d occupancy and activation energy. The above insights outline a molecular‐level understanding of CO poisoning.
... There also exists a fair amount of literature concerning ML predictions of the adsorption energies of small molecules onto catalytic surfaces, [150][151][152][153][154][155][156] but these are beyond the scope of the discussion here and a review on ML for heterogeneous small molecule activation has been published very recently. 157 We also do not include works where ML has been used for predicting activation energies in heterogeneous systems, [158][159][160][161] only homogeneous reactions are covered here. There are also a couple of works exploring ML for activation barriers from molecular dynamics simulations, 162,163 which are also not covered here. ...
Application of machine learning (ML) to the prediction of reaction activation barriers is a new and exciting field for these algorithms. The works covered here are specifically those in which ML is trained to predict the activation energies of homogeneous chemical reactions, where the activation energy is given by the energy difference between the reactants and transition state of a reaction. Particular attention is paid to works that have applied ML to directly predict reaction activation energies, the limitations that may be found in these studies, and where comparisons of different types of chemical features for ML models have been made. Also explored are models that have been able to obtain high predictive accuracies, but with reduced datasets, using the Gaussian process regression ML model. In these studies, the chemical reactions for which activation barriers are modeled include those involving small organic molecules, aromatic rings, and organometallic catalysts. Also provided are brief explanations of some of the most popular types of ML models used in chemistry, as a beginner's guide for those unfamiliar.
This article is categorized under: Structure and Mechanism > Reaction Mechanisms and Catalysis
Computer and Information Science > Visualization
... 1,2 In materials science, discovery of new materials, optimization of processes, and enhancement of performances have been achieved by datadriven methods. [3][4][5][6][7][8][9][10][11][12][13][14][15] In chemistry, new functional molecules and catalysts have been found using ML. Combination of ML and robotic equipment further accelerates discovery. ...
Data-scientific approaches have permeated into chemistry and materials science. In general, these approaches are not easily applied to small data, such as experimental data in laboratories. Our group has focused on sparse modeling (SpM) for small data in materials science and chemistry. The controlled synthesis of 2D materials, involving improvement of the yield and control of the size, was achieved by SpM coupled with our chemical perspectives for small data (SpM-S). In the present work, the conceptual and methodological advantages of SpM-S were studied using real experimental datasets to enable comparison with other machine learning (ML) methods, such as neural networks. The training datasets consisted of ca. 40 explanatory variables (xn) and 50 objective variables (y) regarding the yield, size, and size-distribution of exfoliated nanosheets. SpM-S provided more straightforward, generalizable, and interpretable prediction models and better prediction accuracy for new experiments as an unknown test dataset. The results indicate that machine learning coupled with our experience, intuition, and perspective can be applied to small data in a variety of fields.
