The application of first-principles based multiscale modelling to microkinetic analysis of heterogeneous catalytic processes is hampered by the inherent complication and complexity of catalytic systems, especially at the microscopic scale. In this framework, complication is brought in by the necessity of achieving an accurate description of the reacting events involved in catalytic mechanisms at different reaction conditions. While DFT- based methods are the most suitable for the description of potential energy surfaces, real processes require huge computational costs for carrying on microkinetic modelling. Therefore, less computationally demanding semi-empirical methodologies such as the UBI-QEP method emerge as more appealing for a preliminary screening of complex reaction networks in catalytic systems but its accuracy remains in general uncertain. This work provides a systematic benchmark of UBI-QEP semi-empirical method against first-principles calculations, aiming at validating the reliability of such approach as an effective way to escape the trap of complication. The CO2 activation via r-WGS reaction (CO2 + H2 → CO + H2O) on a Rh(211) surface at different levels of carbon monoxide coverage is chosen as case study since the effect of weakly adsorbed molecules such as CO2 on model performances has not been assessed at the level of full reaction mechanisms. In particular, three mechanisms are studied: the direct dissociation and two hydrogen mediated paths involving respectively trans-COOH and HCOO as intermediates. The analysis of single elementary steps shows potentials and limitations of the UBI-QEP method which proves capable to predict coverage-dependent activation energies based on coverage dependent binding energies when both reactants and products are adsorbed on the surface while fails when either a reactant or a product is weakly interacting with the surface. The analysis of the full activation paths as predicted DFT points out that the hydrogenation to trans-COOH mechanisms dominates r-WGS reaction at 1/6 and 2/6 CO ML, while at zero coverage it is comparable to the direct dissociation mechanism. Differently, the UBI-QEP semi-empirical method conceives the direct dissociation as energetic prevalent path for the CO2 activation at all coverages. This discrepancy is due to the intrinsic limitations of the UBI-QEP semi-empirical method, specifically to (1) the unreliability of the method when dealing with physisorbed molecules such as CO2, and (2) the blindness of the method to the nature of the TS and therefore to the physics of the reaction. Finally, a critical analysis of the limitations of the UBI-QEP semi-empirical method and how to overcome them is provided as future developments for this work
L’applicazione di modelli multi-scala a base first-principle all’analisi di modelli microcinetici di processi di catalisi eterogenea è limitata dall’intrinseca complicazione e complessità dei sistemi catalitici, specialmente su scala microscopica. In questo contesto, la complicazione viene introdotta dalla necessità di fornire una descrizione accurata degli eventi reattivi coinvolti nei meccanismi catalitici in differenti condizioni dell’ambiente reattivo. Sebbene metodi DFT siano i più adatti alla descrizione di superfici di energia potenziale, processi reali richiedono ingenti costi computazionali per condurre la loro modellazione microcinetica. Pertanto, metodi semi-empirici a minor costo computazionale quali lo UBI-QEP diventano attraenti per una selezione preliminare di complessi schemi di reazione. In questo lavoro, viene condotta un’analisi sistematica del metodo semi-empirico UBI-QEP attraverso calcoli first-principles con l’obiettivo di verificare la validità e l’attendibilità di questo approccio come effettiva via d’uscita dalla trappola della complicazione nella modellazione microcinetica di sistemi catalitici. Come caso studio è stata scelta l’attivazione della CO2 via r-WGS su una superficie Rh(211) per la validazione del modello semi empirico poiché l’effetto di specie debolmente assorbite come la CO2 sulle capacità predittive del modello al livello di meccanismi di reazione completi non è stato precedentemente e investigato. Più precisamente, vengono considerati tre differenti meccanismi partendo da CO2 e H2 in fase gas, ciascuno a diverse condizioni di ricoprimento superificiale di monossido di carbonio. La dissocaizione diretta, dando CO e O, e due percorsi di dissociazione mediati dall’idrogeno coivolngenti due diversi intermedi, rispettivamente t-COOH e HCOO. L’analisi delle singole reazioni elementari rivela potenzialità e limiti del metodo UBI-QEP che si osserva essere in grado di prevedere energie di attivazione dipendenti dal ricoprimento superficiale sulla base di energie di adsorbimento dipendente dal ricoprimento superficiale quando sia i reagenti che i prodotti sono chemisorbiti sulla superficie, mentre fallisce quando uno dei reagenti o prodotti interagisce debolmente con la superficie catalitica. L’analisi dei cammini di attivazione predetti da DFT e UBI-QEP con divere parametrizzazioni permette di mostrare nuove scoperte sui percorsi dominanti e stati cineticamente determinanti della r-WGS su catalizzatori a base di Rh ed anche un ulteriore verifica del metodo semi-empirico. Inoltre, è stato analizzato in dettaglio l’effetto delle interazioni adsorbato-adsorbato sull’energetica reattiva e di adsorbimento. Infine, viene condotta un’analisi critica delle limitazioni del metodo UBI-QEP e di come superarle come prospettive future di questo lavoro.
First-principles assessment of UBI-QEP semi-empirical method for predicting catalytic mechanisms : an application to CO2 activation via r-WGS reaction
CANNIZZARO, FRANCESCO
2018/2019
Abstract
The application of first-principles based multiscale modelling to microkinetic analysis of heterogeneous catalytic processes is hampered by the inherent complication and complexity of catalytic systems, especially at the microscopic scale. In this framework, complication is brought in by the necessity of achieving an accurate description of the reacting events involved in catalytic mechanisms at different reaction conditions. While DFT- based methods are the most suitable for the description of potential energy surfaces, real processes require huge computational costs for carrying on microkinetic modelling. Therefore, less computationally demanding semi-empirical methodologies such as the UBI-QEP method emerge as more appealing for a preliminary screening of complex reaction networks in catalytic systems but its accuracy remains in general uncertain. This work provides a systematic benchmark of UBI-QEP semi-empirical method against first-principles calculations, aiming at validating the reliability of such approach as an effective way to escape the trap of complication. The CO2 activation via r-WGS reaction (CO2 + H2 → CO + H2O) on a Rh(211) surface at different levels of carbon monoxide coverage is chosen as case study since the effect of weakly adsorbed molecules such as CO2 on model performances has not been assessed at the level of full reaction mechanisms. In particular, three mechanisms are studied: the direct dissociation and two hydrogen mediated paths involving respectively trans-COOH and HCOO as intermediates. The analysis of single elementary steps shows potentials and limitations of the UBI-QEP method which proves capable to predict coverage-dependent activation energies based on coverage dependent binding energies when both reactants and products are adsorbed on the surface while fails when either a reactant or a product is weakly interacting with the surface. The analysis of the full activation paths as predicted DFT points out that the hydrogenation to trans-COOH mechanisms dominates r-WGS reaction at 1/6 and 2/6 CO ML, while at zero coverage it is comparable to the direct dissociation mechanism. Differently, the UBI-QEP semi-empirical method conceives the direct dissociation as energetic prevalent path for the CO2 activation at all coverages. This discrepancy is due to the intrinsic limitations of the UBI-QEP semi-empirical method, specifically to (1) the unreliability of the method when dealing with physisorbed molecules such as CO2, and (2) the blindness of the method to the nature of the TS and therefore to the physics of the reaction. Finally, a critical analysis of the limitations of the UBI-QEP semi-empirical method and how to overcome them is provided as future developments for this workFile | Dimensione | Formato | |
---|---|---|---|
04_12_2018_Cannizzaro.pdf
non accessibile
Descrizione: Tesi del 04/12/218
Dimensione
13.71 MB
Formato
Adobe PDF
|
13.71 MB | Adobe PDF | Visualizza/Apri |
I documenti in POLITesi sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
https://hdl.handle.net/10589/145337