29
© Materials Design, SARL 2017 A Cleaner World with MedeA ® Xavier Rozanska Computational Chemistry for Pollutant Mitigation IFPEN, Rueil-Malmaison, March 13 th , 2017

New Xavier Rozanska Computational Chemistry for Pollutant …projet.ifpen.fr/Projet/upload/docs/application/pdf/2017... · 2017. 3. 31. · 9 High-throughput: Normal Boiling Point

  • Upload
    others

  • View
    0

  • Download
    0

Embed Size (px)

Citation preview

Page 1: New Xavier Rozanska Computational Chemistry for Pollutant …projet.ifpen.fr/Projet/upload/docs/application/pdf/2017... · 2017. 3. 31. · 9 High-throughput: Normal Boiling Point

© Materials Design, SARL 2017

A Cleaner World with MedeA®

Xavier Rozanska

Computational Chemistry for Pollutant MitigationIFPEN, Rueil-Malmaison, March 13th, 2017

Page 2: New Xavier Rozanska Computational Chemistry for Pollutant …projet.ifpen.fr/Projet/upload/docs/application/pdf/2017... · 2017. 3. 31. · 9 High-throughput: Normal Boiling Point

OutlineMaterials Design

High-throughput in MedeA• Molecular and Fluids Properties

Integration in MedeA• Reactions in Fluid

• Metal organic framework

• Gas separation

Summary

2© Materials Design, SARL 2017

Page 3: New Xavier Rozanska Computational Chemistry for Pollutant …projet.ifpen.fr/Projet/upload/docs/application/pdf/2017... · 2017. 3. 31. · 9 High-throughput: Normal Boiling Point

Introducing Materials Design, Inc.

Leader in atomic-scale materials modeling− MedeA® modeling software

− Materials properties service

− Research services

Founded in 1998 by a team of experts Success with 400+ global companies and institutions

Leader in atomic-scale materials modeling− MedeA® modeling software

− Materials properties service

− Research services

Founded in 1998 by a team of experts Success with 400+ global companies and institutions

© Materials Design, SARL 2017

Page 4: New Xavier Rozanska Computational Chemistry for Pollutant …projet.ifpen.fr/Projet/upload/docs/application/pdf/2017... · 2017. 3. 31. · 9 High-throughput: Normal Boiling Point

Integrated Modeling With MedeA®

Designed for productivity in engineering applications

Unique integration of quantum (ab initio) and classical forcefield methods

Foundation of multi-scalemodeling

High throughputscreening

Computationalefficiency (HPC)

Designed for productivity in engineering applications

Unique integration of quantum (ab initio) and classical forcefield methods

Foundation of multi-scalemodeling

High throughputscreening

Computationalefficiency (HPC)

© Materials Design, SARL 2017

Page 5: New Xavier Rozanska Computational Chemistry for Pollutant …projet.ifpen.fr/Projet/upload/docs/application/pdf/2017... · 2017. 3. 31. · 9 High-throughput: Normal Boiling Point

High-throughput:Molecular and Fluids Properties

5

Page 6: New Xavier Rozanska Computational Chemistry for Pollutant …projet.ifpen.fr/Projet/upload/docs/application/pdf/2017... · 2017. 3. 31. · 9 High-throughput: Normal Boiling Point

Research projects: High throughput property calculations for 100s of compounds

• Published results:

Big Data & Atomistic Simulations: Validation

6© Materials Design, SARL 2017

Page 7: New Xavier Rozanska Computational Chemistry for Pollutant …projet.ifpen.fr/Projet/upload/docs/application/pdf/2017... · 2017. 3. 31. · 9 High-throughput: Normal Boiling Point

Simplification and standardization: Use of Flowcharts & Structure lists for Cp,id calculation

7© Materials Design, SARL 2017

Page 8: New Xavier Rozanska Computational Chemistry for Pollutant …projet.ifpen.fr/Projet/upload/docs/application/pdf/2017... · 2017. 3. 31. · 9 High-throughput: Normal Boiling Point

High-throughput: Heat Capacity

8

, , , ,Ideal gas 

heat capacityResidual heat 

capacity

From quantum chemistry (MOPAC) on isolated molecule 

From MD or MC simulation of a condensed phase

Total heat capacity

, ,

© Materials Design, SARL 2017

Page 9: New Xavier Rozanska Computational Chemistry for Pollutant …projet.ifpen.fr/Projet/upload/docs/application/pdf/2017... · 2017. 3. 31. · 9 High-throughput: Normal Boiling Point

9

High-throughput: Normal Boiling Point Temperature (pure compounds)

Samples shown in the plot: ~100

AAE of the Tb calculated by GEMC simulation from the DIPPR* data, is 1.4%.

More than half of the compounds have an Absolute Deviation (AD) of the Tb below 1.0%.

MedeA®-GIBBS simulations

Exp. data: Wilding WV, Rowley RL, Oscarson JL. DIPPR project 801 evaluated process design data. Fluid Phase Equil. 1998;150–151:413–420

© Materials Design, SARL 2017

Page 10: New Xavier Rozanska Computational Chemistry for Pollutant …projet.ifpen.fr/Projet/upload/docs/application/pdf/2017... · 2017. 3. 31. · 9 High-throughput: Normal Boiling Point

Integration:Reactions in Fluid

10

Page 11: New Xavier Rozanska Computational Chemistry for Pollutant …projet.ifpen.fr/Projet/upload/docs/application/pdf/2017... · 2017. 3. 31. · 9 High-throughput: Normal Boiling Point

MedeA Integration: Communication between Tools and Methods

11

VASP FFO LAMMPS

GIBBSElectrostaticpotential of

(porous) solids

ForcesEnergiesStresses

FF parameters

GaussianESP

charges

FF inter-operability

© Materials Design, SARL 2017

Page 12: New Xavier Rozanska Computational Chemistry for Pollutant …projet.ifpen.fr/Projet/upload/docs/application/pdf/2017... · 2017. 3. 31. · 9 High-throughput: Normal Boiling Point

Catalytic Ethylene FormylationReaction

Reaction conditions • CO/H2+C2H4 200-300 bar

• T=383-473 K

Catalyst• Tetracarbonyl cobalt

hydride

− HCo(CO)4

12

O

CO/H2

Catalyst

Reaction mechanisms

© Materials Design, SARL 2017

Page 13: New Xavier Rozanska Computational Chemistry for Pollutant …projet.ifpen.fr/Projet/upload/docs/application/pdf/2017... · 2017. 3. 31. · 9 High-throughput: Normal Boiling Point

Molecular modeling analysis

13

1

2

3

4

CCSD(T)/CBS//B3LYP/aug-cc-pVTZ + anharmonic frequency calculations/B3LYP/aug-cc-pVTZ

© Materials Design, SARL 2017

Page 14: New Xavier Rozanska Computational Chemistry for Pollutant …projet.ifpen.fr/Projet/upload/docs/application/pdf/2017... · 2017. 3. 31. · 9 High-throughput: Normal Boiling Point

Molecular modeling analysis

14

Reaction fluid thermophysical properties with MD and MC at 200 bar Density of supercritical H2/CO/C2H4 Residual Cp of supercritical H2/CO/C2H4

Diffusion of H2/CO/C2H4 at 473 K and 200 barMolc. log10(D)H2 ‐2.13CO ‐2.58C2H4 ‐2.66D is in cm2 s‐1

Diffusion of C2H4

at different T

© Materials Design, SARL 2017

Page 15: New Xavier Rozanska Computational Chemistry for Pollutant …projet.ifpen.fr/Projet/upload/docs/application/pdf/2017... · 2017. 3. 31. · 9 High-throughput: Normal Boiling Point

Integration:Characterization of Solids and their Interaction with Fluids

15

Page 16: New Xavier Rozanska Computational Chemistry for Pollutant …projet.ifpen.fr/Projet/upload/docs/application/pdf/2017... · 2017. 3. 31. · 9 High-throughput: Normal Boiling Point

Integration: Communication betweenTools and Methods

16

VASP FFO LAMMPS

GIBBSElectrostaticpotential of

solids

ForcesEnergiesStresses

FF parameters

GaussianESP

charges

FF inter-operability

© Materials Design, SARL 2017

Page 17: New Xavier Rozanska Computational Chemistry for Pollutant …projet.ifpen.fr/Projet/upload/docs/application/pdf/2017... · 2017. 3. 31. · 9 High-throughput: Normal Boiling Point

MOF functionalization with amino-acids

17

Applications:-enantioselective catalysis-adsorption/diffusion selectivity

Are the amino-acid groups fully free?(hopefully not!)If they are constrained: what is their conformation?

Considered functional groups:-GlycineProline and -Proline

T. Todorova, X. Rozanska, C. Gervais, A. Legrand, L. N. Ho, P. Berruyer, A. Lesage, L. Emsley, D. Farrusseng, J. Canivet, and C. Mellot-Draznieks, Chem. Eur. J. 2016, 22, 16531-16538

Page 18: New Xavier Rozanska Computational Chemistry for Pollutant …projet.ifpen.fr/Projet/upload/docs/application/pdf/2017... · 2017. 3. 31. · 9 High-throughput: Normal Boiling Point

Conformations in peptide-grafted MOF

18

Developping missing FF parameters from VASP to MM

Validating the FF parameters with adsorption isotherm (MC) and density (NPT MD)

T. Todorova, X. Rozanska, C. Gervais, A. Legrand, L. N. Ho, P. Berruyer, A. Lesage, L. Emsley, D. Farrusseng, J. Canivet, and C. Mellot-Draznieks, Chem. Eur. J. 2016, 22, 16531-16538

Page 19: New Xavier Rozanska Computational Chemistry for Pollutant …projet.ifpen.fr/Projet/upload/docs/application/pdf/2017... · 2017. 3. 31. · 9 High-throughput: Normal Boiling Point

MD+ab initio+experimental data

19

Proline GlycineProline

Page 20: New Xavier Rozanska Computational Chemistry for Pollutant …projet.ifpen.fr/Projet/upload/docs/application/pdf/2017... · 2017. 3. 31. · 9 High-throughput: Normal Boiling Point

Integration: Communication betweenTools and Methods

20

VASP FFO LAMMPS

GIBBSElectrostaticpotential of

solids

ForcesEnergiesStresses

FF parameters

GaussianESP

charges

FF inter-operability

© Materials Design, SARL 2017

Page 21: New Xavier Rozanska Computational Chemistry for Pollutant …projet.ifpen.fr/Projet/upload/docs/application/pdf/2017... · 2017. 3. 31. · 9 High-throughput: Normal Boiling Point

21

Application: CO2/N2 separation by ALPO and modified ALPO (defects)

Aluminophosphatematerial: ZON

1: SiO22: AlPO43: Al30Si4P30O1284: NaAl32SiP31O128

The cells are optimized using VASP. The ab initio electrostatic potentials are computed with VASP. They are used in Grand Canonical Monte Carlo simulation (GIBBS).

Source: Rozanska et al. Oil Gas Sci. Tech. 2013, 68, 299-307

Page 22: New Xavier Rozanska Computational Chemistry for Pollutant …projet.ifpen.fr/Projet/upload/docs/application/pdf/2017... · 2017. 3. 31. · 9 High-throughput: Normal Boiling Point

22

Application: CO2/N2 separation by ALPO and modified ALPO

1: SiO22: AlPO43: Al30Si4P30O1284: NaAl32SiP31O128

Molecular CO2/N2 ratio from the Adsorption Isotherms of CO2 and N2 in ZON.

Source: Rozanska et al. Oil Gas Sci. Tech. 2013, 68, 299-307

Page 23: New Xavier Rozanska Computational Chemistry for Pollutant …projet.ifpen.fr/Projet/upload/docs/application/pdf/2017... · 2017. 3. 31. · 9 High-throughput: Normal Boiling Point

SummaryHigh-throughput & Integrated Atomistic Simulations

• Validation

• Simplification

• Standardization

• Integration and information exchange

• Multiscaling

Applications examples

• Properties & Big Data

• Catalysis and Reactions

• Characterization

• Separation

23© Materials Design, SARL 2017

Page 24: New Xavier Rozanska Computational Chemistry for Pollutant …projet.ifpen.fr/Projet/upload/docs/application/pdf/2017... · 2017. 3. 31. · 9 High-throughput: Normal Boiling Point

Acknowledgements

24

Caroline Mellot-Draznieks and Tanya Todorova (et al.), Collège de France –CNRS (MOF)

Materials Design’s colleagues (MedeA-Gibbs of organic fluids: Cp,res and Tb)

Dr Marianna Yiannourakou

Dr Philippe Ungerer

Page 25: New Xavier Rozanska Computational Chemistry for Pollutant …projet.ifpen.fr/Projet/upload/docs/application/pdf/2017... · 2017. 3. 31. · 9 High-throughput: Normal Boiling Point

Credits

© Materials Design, Inc. 2017 25

VASP: The VASP Group, Theoretical Physics Department, University of Vienna • Kresse, G., & Hafner, J. (1993). Ab initio molecular dynamics for liquid metals. Physical Review

B, 47(1), 558.• Kresse, G., & Furthmüller, J. (1996). Efficiency of ab-initio total energy calculations for metals

and semiconductors using a plane-wave basis set. Computational Materials Science, 6(1), 15-50.

LAMMPS: S. Plimpton, Fast Parallel Algorithms for Short-Range Molecular Dynamics, J Comp Phys, 117, 1-19 (1995), www.lammps.sandia.govGIBBS: License IFP-EN – LCP (CNRS – Université Paris Sud)• P Ungerer, C Beauvais, J Delhommelle, A Boutin, B Rousseau, AH Fuchs, The Journal of

Chemical Physics 112 (12), 5499-5510• E. Bourasseau, M. Haboudou, A. Boutin, A.H. Fuchs, P Ungerer, The Journal of chemical

physics 118 (7), 3020-3034• A.D. Mackie, .B Tavitian, A. Boutin, A.H. Fuchs, Molecular Simulation 19 (1), 1-15• M. Lagache, P. Ungerer, A. Boutin, A.H. Fuchs, Physical Chemistry Chemical Physics 3 (19),

4333-4339• E. Bourasseau, P. Ungerer, A. Boutin, A.H. Fuchs, Molecular Simulation 28 (4), 317-336• N. Ferrando, A. Boutin and V. Lachet, Journal of Physical Chemistry 114: 8680-8688, (2010).

MOPAC2012: James J. P. Stewart, Stewart Computational Chemistry, Colorado Springs, CO, USA, HTTP://OpenMOPAC.net (2012) MedeA®: MedeA: Materials Exploration and Design Analysis; Materials Design, Inc. 2017. www.materialsdesign.com

Page 26: New Xavier Rozanska Computational Chemistry for Pollutant …projet.ifpen.fr/Projet/upload/docs/application/pdf/2017... · 2017. 3. 31. · 9 High-throughput: Normal Boiling Point

Backup slides

26

Page 27: New Xavier Rozanska Computational Chemistry for Pollutant …projet.ifpen.fr/Projet/upload/docs/application/pdf/2017... · 2017. 3. 31. · 9 High-throughput: Normal Boiling Point

High-throughput: SimplificationMedeA-LAMMPS Flowchart• 4 modules ~ 4 parameters)

27

LAMMPS input command file• 130 lines

© Materials Design, SARL 2017

Page 28: New Xavier Rozanska Computational Chemistry for Pollutant …projet.ifpen.fr/Projet/upload/docs/application/pdf/2017... · 2017. 3. 31. · 9 High-throughput: Normal Boiling Point

High-throughput: Standardization

28© Materials Design, SARL 2017

Page 29: New Xavier Rozanska Computational Chemistry for Pollutant …projet.ifpen.fr/Projet/upload/docs/application/pdf/2017... · 2017. 3. 31. · 9 High-throughput: Normal Boiling Point

High-throughput: Standardization

29© Materials Design, SARL 2017