PerOla Norrby BigChem Training School 9 May 2019 Chemical reaction selectivity forward prediction Reduction Chemoselectivity Regioselectivity Enantioselectivity Chemical reaction ID: 1042130
Download Presentation The PPT/PDF document "Methods for predicting reaction selectiv..." is the property of its rightful owner. Permission is granted to download and print the materials on this web site for personal, non-commercial use only, and to display it on your personal computer provided you do not modify the materials and that you retain all copyright notices contained in the materials. By downloading content from our website, you accept the terms of this agreement.
1. Methods for predicting reaction selectivityPer-Ola NorrbyBigChem Training School9 May 2019
2. Chemical reaction selectivity, forward predictionReductionChemoselectivityRegioselectivityEnantioselectivity
3. Chemical reaction selectivity, cont.Late Stage FunctionalizationC-H activationRadical brominationPd-catalyzed C-H activationElectrophilic Aromatic bromination
4. 4Calculating reaction selectivityG‡Boltzmann summation1-10 paths1000’s possible!
5. 5Modeling catalyst stereoselectivity (I): QSARQSSR: Quantitative Structure Selectivity RelationshipNorrby, Lipkowits, Kozlowski, Bo, Harvey, Fey, ..., Sigman, Doyle, …
6. 6Modeling catalyst stereoselectivity (II): QM(R)- vs. (S)-energy, DFT-based screeningMaeda, Morokuma, Zimmerman, ..., Wheeler27% ee → 97% eeMUE 4 kJ/mol
7. 7Modeling catalyst stereoselectivity (III): Q2MMMolecular Mechanics @ TS, conformational search
8. 8Molecular Mechanics for Reactivity19501960197019801990Westheimer1946Sprecher1965De Tar, Paterson, …HoukTransition state force fields, TSFFMenger1989”Fudgit”Using 200 parameters to fit20 data points is not a good ideaCan the overfitting problem be solved?
9. 9Parameterization from QM dataQM calculations give abundant dataDetermining force fields from quantum mechanical data:Hagler, Kollman, Goddard, …Bond lengths, angles from optimized structuresCharges from electrostatic potentialsTorsions from calculated rotation profilesForce constants from calculated vibrations (Hessian data)Problem: wrong curvature
10. 10Reaction coordinate modificationDiagonalization: Eigenvalue replacement:Forming the new Hessian:
11. 11Q2MM resultsMUE = 2.8 kJ/mol
12. 12Virtual screening development and workflow
13. 13Virtual screening of catalysts: CatVSfilenamecatalystproR energyproS energydeltaEeeer[kJ/mol][kJ/mol][kJ/mol][%]confsearch00147Ru_SDifluorPHOS_RDAIPEN704.8700.9-3.9-65.20.21confsearch00254Ru_SXylPhanePHOS_RIPBAN838.6816.3-22.3-100.00.00012confsearch00255Ru_bisPPh3_RIPBAN571.0552.5-18.4-99.90.00059confsearch00256Ru_SXylBINAP_RRDACH602.3597.3-5.0-76.70.13confsearch00257Ru_RBINAP_RRDACH621.2589.4-31.8-100.02.6e-06confsearch00399Ru_RXylPPHOS_SIPBAN392.5384.0-8.5-93.80.032confsearch00398Ru_RXylBINAP_SIPBAN640.5626.3-14.2-99.40.0032confsearch00252Ru_STolBINAP_RIPBAN621.2613.0-8.2-92.90.037confsearch00253Ru_SXylPPHOS_RIPBAN379.5360.2-19.3-99.90.00041confsearch00395Ru_RPhanePHOS_SIPBAN850.9839.7-11.2-97.80.011confsearch00394Ru_RDifluorPHOS_SIPBAN625.5616.2-9.3-95.40.023confsearch00397Ru_RTolBINAP_SIPBAN644.5626.6-17.9-99.90.00074confsearch00396Ru_RSegPHOS_SIPBAN644.4641.9-2.4-45.60.37confsearch00258Ru_RDifluorPHOS_RRDACH598.4570.1-28.3-100.01.1e-05confsearch00259Ru_RPhanePHOS_RRDACH798.5796.8-1.7-33.60.5confsearch00393Ru_RBINAP_SIPBAN655.3632.9-22.4-100.00.00012confsearch00392Ru_SXylBINAP_SIPBAN633.8622.5-11.2-97.90.011confsearch00474Ru_SXylPPHOS_TMEN358.8368.49.695.847confsearch00475Ru_SXylPhanePHOS_TMEN838.1829.1-9.0-94.90.026confsearch00476Ru_bisPPh3_TMEN555.2562.67.490.420confsearch00470Ru_SDifluorPHOS_TMEN594.9602.07.189.117confsearch00471Ru_SPhanePHOS_TMEN837.0827.2-9.8-96.20.019confsearch00472Ru_SSegPHOS_TMEN635.2648.913.799.22.5e+02confsearch00473Ru_STolBINAP_TMEN620.5620.2-0.3-6.50.88confsearch00179Ru_RXylPhanePHOS_RDMAPEN906.9896.4-10.5-97.10.014confsearch00071Ru_RDifluorPHOS_EN554.1559.04.976.07.3confsearch00072 Ru_RPhanePHOS_EN766.8788.621.8100.06.6e+03 Build structures Conformational search Calculate e.r.
14. 14Rh-catalyzed enamide hydrogenationCollaboration with Paul Helquist and Olaf Wiest, Notre Dame, INQ2MM force field for TSPaul HelquistOlaf Wiest
15. 15Asymmetric hydrogenation – Test I Developed for:Tested on:Underlying program changed - Energies no longer the sameCurrent version: no model solventEnrichment, 4 of 5!ΔΔG* = 10 kJ/mole.r. = 50 98:2 e.e. = 96%
16. 16Asymmetric hydrogenation – Test 21 good, 1 fair, 1 false negativeAcceptableΔΔG* = 10 kJ/mole.r. = 50 98:2 e.e. = 96%Developed for:Tested on:
17. 17CatVS - SummaryNeeds Q2MM force fields; significant development time Development tools on www.github.com/Q2MM/Significant enrichment Makes experimental testing more efficient Faster routes to chiral compoundsChemical selectivity can now be reliably predicted!Draw substrate&Click ”submit”Combine with ligands libraryDetermine all low energy TSEmail results for all ligands
18. 18OngoingAsymmetric Suzuki-MiyauraChristian SköldMaria VergouLiam ByrnePeter SmithRachel Munday
19. 19Hybrid ML-mechanistic modelMachine learningMechanistic modelHybrid modelElectrophilic aromatic substitution: Tomberg, Johansson, Norrby, J. Org. Chem. 2019, 4695
20. 20AcknowledgmentUppsalaChristian SköldDenmarkTommy LiljeforsTorben RasmussenPeter BrandtPeter FristrupLundUlf RydePatrik RydbergStockholmBjörn Åkermark$$$AstraZeneca, The University of Gothenburg, COST, FP7/SYNFLOW, The Swedish Research Council, C3SE/GothenburgAstraZenecaTobias ReinPer RybergSimone TomasiElaine LiméDavid ButtarRachel MundayMaria VergouPeter SmithLiam ByrneAnna TombergNotre DamePaul HelquistOlaf WiestPatrick DonoghueElsa KiekenVincenzo VerdolinoAaron ForbesEric HansenBrandon TutkowskiAnthony RosalesTaylor QuinnJessica WahlersXin Zhang (Shenzhen)
21. 21
22. 22