universality in heterogeneous catalysis in relation to catalysis on zeotype materials Łukasz Mentel January 12, 2015 VU University Amsterdam 

Paper summary 1 

Ideas J. Nørskov et al., J. Catal. 2002, 209, 275–278 For a certain class of reactions on transition metal surfaces a criterion describing the optimal catalyst can be found. 2 

Ideas J. Nørskov et al., J. Catal. 2002, 209, 275–278 For a certain class of reactions on transition metal surfaces a criterion describing the optimal catalyst can be found. It is based on a universal, reaction independent relationship between activation energies and the stability of important intermediates in the reaction. 2 

The scaling relation The reactions considered are of the form1 A2 + • ⇄ A2• adsorption A2 • + • ⇄ 2A• activation } dissociative chemisorption A• +B ⇄ AB + • product formation and desorption 1• denotes the active site of the catalyst 3 

The scaling relation The reactions considered are of the form1 A2 + • ⇄ A2• adsorption A2 • + • ⇄ 2A• activation } dissociative chemisorption A• +B ⇄ AB + • product formation and desorption Ea = α · ∆E + β where: Ea is the activation barrier for dissociation ∆E is the dissociative chemisorption energy α, β are parameters 1• denotes the active site of the catalyst 3 

BEP relation Brønsted-Evans-Polanyi (BEP) relation2 Ea = α · ∆E + β 2T. Bligaard et al., J. Catal. 2004, 224, 206–217 . 4 

BEP relation Brønsted-Evans-Polanyi (BEP) relation2 Ea = α · ∆E + β case 1 when Ea is high the adsorption/dissociation is the rate limiting step case 2 when ∆E is high the forming of products/desorption will be the limiting step 2T. Bligaard et al., J. Catal. 2004, 224, 206–217 . 4 

BEP relation The BEP relationship leads directly to the so called volcano-curve for the overall reaction rate3 volcano-curve is obtained when the activity of a catalyst for a certain reaction is plotted as a function of a parameter quantifying the adsorbate(intermediates)-catalyst bond strength BEP combined with microkinetic modelling helps to identify rate limiting step and understand the kinetics in terms of reaction energetics 3T. Bligaard et al., J. Catal. 2004, 224, 206–217 . 5 

Rationalization Why does the scaling relation hold? ∙ energetic similarity of intermediate state to product4 ∙ the strength of the adsorbate-surface interaction is largely determined by coupling of the metal d-band with the valence state of the adsorbate ∙ activation barrier and reaction energy must depend in the same way on the external variables e.g. electronic structure parameters 4F. Calle-Vallejo et al., Phys. Rev. Lett. 2012, 108, 116103 . 6 

Conclusions ∙ scaling relations provide a robust descriptor for catalytic activity allowing discrimination of catalysts based on their performance 7 

Conclusions ∙ scaling relations provide a robust descriptor for catalytic activity allowing discrimination of catalysts based on their performance ∙ the universal relation allows optimization of catalysts by searching for the materials with the optimal strength of the surface-adsorbate interaction 7 

Conclusions ∙ scaling relations provide a robust descriptor for catalytic activity allowing discrimination of catalysts based on their performance ∙ the universal relation allows optimization of catalysts by searching for the materials with the optimal strength of the surface-adsorbate interaction ∙ can reduce computational cost since only adsorption energy is required 7 

Conclusions ∙ scaling relations provide a robust descriptor for catalytic activity allowing discrimination of catalysts based on their performance ∙ the universal relation allows optimization of catalysts by searching for the materials with the optimal strength of the surface-adsorbate interaction ∙ can reduce computational cost since only adsorption energy is required ∙ BEP relation combined with microkinetic modelling can provide insight into which step steps control catalytic performance as well as which adsorption energies control these steps 7 

Conclusions ∙ scaling relations provide a robust descriptor for catalytic activity allowing discrimination of catalysts based on their performance ∙ the universal relation allows optimization of catalysts by searching for the materials with the optimal strength of the surface-adsorbate interaction ∙ can reduce computational cost since only adsorption energy is required ∙ BEP relation combined with microkinetic modelling can provide insight into which step steps control catalytic performance as well as which adsorption energies control these steps ∙ shed light on into how molecular structure controls adsorption energy 7 

Conclusions ∙ scaling relations provide a robust descriptor for catalytic activity allowing discrimination of catalysts based on their performance ∙ the universal relation allows optimization of catalysts by searching for the materials with the optimal strength of the surface-adsorbate interaction ∙ can reduce computational cost since only adsorption energy is required ∙ BEP relation combined with microkinetic modelling can provide insight into which step steps control catalytic performance as well as which adsorption energies control these steps ∙ shed light on into how molecular structure controls adsorption energy ∙ with quantitative DFT calculations we are able to reproduce the trends in agreement with empirical observations quite well 7 

limitations 

Limitations Geometric ∙ transition metals and their alloys form highly symmetric lattices, more uniform than other heterogeneous catalyst 9 

Limitations Geometric ∙ transition metals and their alloys form highly symmetric lattices, more uniform than other heterogeneous catalyst ∙ the scaling relations already change for different sites on the same surface (flat surface, steps, edges) which makes practical practical application challenging 9 

Limitations Geometric ∙ transition metals and their alloys form highly symmetric lattices, more uniform than other heterogeneous catalyst ∙ the scaling relations already change for different sites on the same surface (flat surface, steps, edges) which makes practical practical application challenging 9 

Limitations Geometric ∙ transition metals and their alloys form highly symmetric lattices, more uniform than other heterogeneous catalyst ∙ the scaling relations already change for different sites on the same surface (flat surface, steps, edges) which makes practical practical application challenging Energetic ∙ it is an open question if the scaling relation will hold if the nature of the catalyst-adsorbate interaction changes (chemisorption vs. physisorption, additional repulsion, charged surface or intermediates) 9 

Limitations Geometric ∙ transition metals and their alloys form highly symmetric lattices, more uniform than other heterogeneous catalyst ∙ the scaling relations already change for different sites on the same surface (flat surface, steps, edges) which makes practical practical application challenging Energetic ∙ it is an open question if the scaling relation will hold if the nature of the catalyst-adsorbate interaction changes (chemisorption vs. physisorption, additional repulsion, charged surface or intermediates) ∙ due to the d-band coupling with the adsorbate the scaling might be altered on surfaces/catalysts with different electronic structure 9 

Limitations ∙ the universality was only demonstrated for a simple class of reactions, composed of atoms and diatomics, 10 

Limitations ∙ the universality was only demonstrated for a simple class of reactions, composed of atoms and diatomics, ∙ the outlined approach focuses only on catalytic activity but not other important characteristics of optimal catalysts: selectivity, stability, catalyst poisoning, etc. which are important theoretically, practically, and economically 10 

Limitations ∙ the universality was only demonstrated for a simple class of reactions, composed of atoms and diatomics, ∙ the outlined approach focuses only on catalytic activity but not other important characteristics of optimal catalysts: selectivity, stability, catalyst poisoning, etc. which are important theoretically, practically, and economically ∙ the approach relies heavily on existence of a database of calculated reaction energies, chemisorption energies for reactions on surfaces 10 

towards application in zeolite catalysis 

Possible applications to zeolite catalysis ∙ assessment of the scaling relations on zeotype materials should be performed to establish if those relations hold 5C.-M. Wang et al., J. Phys. Chem. Lett. 2014, 5, 1516–1521 , M. M. Montemore, J. W. Medlin, Catal. Sci. Technol. 2014, 4, 3748–3761 . 6R. Shah et al., J. Phys. Chem. 1996, 100, 11688–11697 . 12 

Possible applications to zeolite catalysis ∙ assessment of the scaling relations on zeotype materials should be performed to establish if those relations hold ∙ next to BEP there are other scaling relations that can be used as descriptors of catalytic activity5 5C.-M. Wang et al., J. Phys. Chem. Lett. 2014, 5, 1516–1521 , M. M. Montemore, J. W. Medlin, Catal. Sci. Technol. 2014, 4, 3748–3761 . 6R. Shah et al., J. Phys. Chem. 1996, 100, 11688–11697 . 12 

Possible applications to zeolite catalysis ∙ assessment of the scaling relations on zeotype materials should be performed to establish if those relations hold ∙ next to BEP there are other scaling relations that can be used as descriptors of catalytic activity5 ∙ a simple energetic criterion can be a suitable descriptor for screening databases of available zeolite structures to find the optimal catalyst, 5C.-M. Wang et al., J. Phys. Chem. Lett. 2014, 5, 1516–1521 , M. M. Montemore, J. W. Medlin, Catal. Sci. Technol. 2014, 4, 3748–3761 . 6R. Shah et al., J. Phys. Chem. 1996, 100, 11688–11697 . 12 

Possible applications to zeolite catalysis ∙ assessment of the scaling relations on zeotype materials should be performed to establish if those relations hold ∙ next to BEP there are other scaling relations that can be used as descriptors of catalytic activity5 ∙ a simple energetic criterion can be a suitable descriptor for screening databases of available zeolite structures to find the optimal catalyst, ∙ strong correlation between the neighborhood of the active site in the zeolite and the adsorbed state of the methanol6 suggests that energetic descriptors should be coupled with geometrical descriptors (such as ring size and structure) quantifying the local topology of the active site or zeolite material 5C.-M. Wang et al., J. Phys. Chem. Lett. 2014, 5, 1516–1521 , M. M. Montemore, J. W. Medlin, Catal. Sci. Technol. 2014, 4, 3748–3761 . 6R. Shah et al., J. Phys. Chem. 1996, 100, 11688–11697 . 12 

Challenges ∙ the diverse structure of the zeolites might result in several different active site with different characteristics (e.g. acidity, local geometry) altering the scaling relations7 7M. Boronat et al., J. Phys. Chem. B 2001, 105, 11169–11177 . 13 

Challenges ∙ the diverse structure of the zeolites might result in several different active site with different characteristics (e.g. acidity, local geometry) altering the scaling relations7 ∙ universality was established for surfaces that are topologically similar but for zeotype materials topology can vary significantly 7M. Boronat et al., J. Phys. Chem. B 2001, 105, 11169–11177 . 13 

Challenges ∙ the diverse structure of the zeolites might result in several different active site with different characteristics (e.g. acidity, local geometry) altering the scaling relations7 ∙ universality was established for surfaces that are topologically similar but for zeotype materials topology can vary significantly ∙ electronic structure of zeolites is quite different from bulk transition metals, that might alter the relations 7M. Boronat et al., J. Phys. Chem. B 2001, 105, 11169–11177 . 13 

Challenges ∙ the diverse structure of the zeolites might result in several different active site with different characteristics (e.g. acidity, local geometry) altering the scaling relations7 ∙ universality was established for surfaces that are topologically similar but for zeotype materials topology can vary significantly ∙ electronic structure of zeolites is quite different from bulk transition metals, that might alter the relations ∙ due to the rich structure of zoelites the surface/adsorbate interaction can depend on additional components (vdW, hydrogen bonds, local charges, etc.) 7M. Boronat et al., J. Phys. Chem. B 2001, 105, 11169–11177 . 13 

Challenges ∙ the diverse structure of the zeolites might result in several different active site with different characteristics (e.g. acidity, local geometry) altering the scaling relations7 ∙ universality was established for surfaces that are topologically similar but for zeotype materials topology can vary significantly ∙ electronic structure of zeolites is quite different from bulk transition metals, that might alter the relations ∙ due to the rich structure of zoelites the surface/adsorbate interaction can depend on additional components (vdW, hydrogen bonds, local charges, etc.) ∙ the DFT might have to be reevaluated for accuracy and be augmented with additional corrections (e.g. vdW term, BEEF-vdW) 7M. Boronat et al., J. Phys. Chem. B 2001, 105, 11169–11177 . 13 

conclusions 

Conclusions ∙ a procedure for discriminating different zeotype materials based on energetic and geometric descriptors would be an invaluable tool for catalysts design and optimization ∙ it would allow searching databases of zeolite structures for an optimal catalyst or suggest what kind of reaction would be optimally catalysed by a given structure ∙ with a compiled library of zeolites (and their properties/attributes) it is possible to find features that will allow discrimination between different catalytic performance ∙ encouraging results suggest that such descriptors can be established for zeotype materials 15 

Questions? 16 

References I [1] J. Nørskov, T. Bligaard, A. Logadottir, S. Bahn, L. B. Hansen, M. Bollinger, H. Benegaard, B. Hammer, Z. Slijvancanin, M. Mavrikakis, Y. Xu, S. Dahl, C. J. H. Jacobsen, J. Catal. 2002, 209, 275–278. [2] T. Bligaard, J. Nørskov, S. Dahl, J. Matthiesen, C. Christensen, J. Sehested, J. Catal. 2004, 224, 206–217. [3] F. Calle-Vallejo, J. I. Martínez, J. M. García-Lastra, J. Rossmeisl, M. T. M. Koper, Phys. Rev. Lett. 2012, 108, 116103. [4] C.-M. Wang, R. Y. Brogaard, B. M. Weckhuysen, J. K. Nørskov, F. Studt, J. Phys. Chem. Lett. 2014, 5, 1516–1521. 17 

References II [5] M. M. Montemore, J. W. Medlin, Catal. Sci. Technol. 2014, 4, 3748–3761. [6] R. Shah, J. D. Gale, M. C. Payne, J. Phys. Chem. 1996, 100, 11688–11697. [7] M. Boronat, C. M. Zicovich-Wilson, P. Viruela, A. Corma, J. Phys. Chem. B 2001, 105, 11169–11177. 18