1 / 61

Chemistry with Computers

Chemistry with Computers. Yingbin Ge Iowa State University. Central Washington University October 13, 2007. coupled-cluster CCSD(T). Perturbation theory MP2. density functional theory (DFT). Accuracy. Hartree Fock (HF). molecular mechanics. Computer time. What has been done?.

marv
Télécharger la présentation

Chemistry with Computers

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript


  1. Chemistry with Computers Yingbin Ge Iowa State University Central Washington University October 13, 2007

  2. coupled-cluster CCSD(T) Perturbation theory MP2 density functional theory (DFT) Accuracy Hartree Fock (HF) molecular mechanics Computer time

  3. What has been done? • Global optimization of silicon nanoclusters. • Chemical vapor deposition of silicon carbide. Si14H20

  4. Global optimization of silicon nanoclusters • Why Si nanoclusters? • Si nanoclusters exhibit bright room-temperature photoluminescence which could be used in light-emitting devices. A. Meldrum group, Adv. Mater. 17, 845 (2005) • To model the excitation and emission of the Si nanoclusters, we need to know their thermodynamically stable structures.

  5. Global vs. local optimization local optimization energy  Energy local minimum local minimum global minimum conformations

  6. Why is global optimization difficult? Tsai and Jordan, JPC 97, 11227 (1993)

  7. Global optimization strategies • Exhaustive search: too many minima to sample. • Random sampling:”But there’s one I always miss.” • Genetic algorithm is based on “the fittest survive” principle. It has been proven efficient for the global optimization of clusters and molecules.* *Applications of evolutionary computation in chemistry, Structure and Bonding, Vol. 110 (2004)

  8. Genetic algorithm based global optimization Produce random structures as initial population. Evaluate energy (fitness) for each individual. Repeat following steps until convergence: Perform competitive selection. Apply genetic operators* to produce new clusters. Lower energy clusters replace higher-energy ones. *Genetic operators: crossover and mutation.

  9. Biological crossover and mutation Crossover of 2 DNA strings after crossover normal Mutation: 1 missing nucleotide normal missing nucleotide after mutation

  10. Crossover: silicon hydrides local opt. crossover

  11. Mutation methods Hydrogen shift Partial rotation

  12. Mutation methods SiH2 SiH3 a. initial geometry b. after mutation c. final structure SiH2 SiH3

  13. Diamond-lattice SixHy global minima Si10H16 Si14H20 Si18H24 MP2 & DFT SixHy-2 global minima Si10H14 Si18H22 Si14H18

  14. MP2 & DFT SixHy global minima Si7H14 Si8H14 Si10H16 Si10H14 SixFy global minima DFT Si7F14 Si8F14 Si10F16 Si10F14

  15. Ligand effect L= H CH3 OH F L2Si=SiL2 L3Si-SiL MP2 global minimum

  16. Ligand effect • Si10(CH3)16 and Si10H16 adopt the same diamond-lattice Si core. • Si10(OH)16 and Si10F16 adopt same Si core with a 4-membered Si ring. • Ligand electronegativity affects the Si core structures. • -SiF3 and -Si(OH)3 are preferred at expenses of forming small 4-membered Si rings.

  17. What did we learn? • GA is efficient, scaling O(N4-5). • Well H-passivated Si clusters adopt diamond-lattice Si cores. • Si core can be tuned with # ligands. • Si core can be tuned with ligand electronegativity. SixCly and SixBry? • Further study the excitation and photon-emitting mechanism of Si nanoclusters. • Questions and comments?

  18. Questions?

  19. May 18, 2007 HomeStead Road, Sunnyvale, CA http://www.opentravelinfo.com/north-america/gas-price-hike

  20. Nuclear Energy • Additional energy source: less fight on oil. • No SO2 - less acid rains. • No CO2 - less global warming. Let’s try to keep New York & Shanghai above sea. http://globalwarming--awareness2007.com/globalwarming-awareness2007/

  21. What about the safety? Layer 1.Porous carbon to accommodate fission products and kernel swelling. Layer 2. Pyrolytic carbon to trap fission products. UO2 kernel Layer 3. Silicon carbide is impervious to fission products and serves as a pressure vessel. Layer 4. Pyrolytic carbon to protect SiC. http://www.iaea.org/inis/aws/htgr/fulltext/xa54410.08.pdf

  22. Chemical vapor deposition inlets • CVD: gas phase molecules break down at high T; fragments deposit on a substrate to account for the solid growth. diamond growth outlet substrate CH4 C H2 http://www.ieee-virtual-museum.org/collection/tech.php?taid=&id=2345958&lid=1

  23. Silicon carbide (SiC) coating process Coater Wall Uranium Particles Annealing Zone Deposition Zone precursors

  24. Why silicon carbide? • High melting point: 2700 C. • Mohs’ hardness: 9.3/10. • Imperviousness to fission products. • Lower reactivity at high temperature. • Low cost. • SiC made by chemical vapor deposition is ideal material for the protective layer of nuclear energy pellets.

  25. P: Defects in the SiC layer cause cracks on the surfaces of nuclear energy pellets. Q: How to reduce defects in SiC? A: Understand the mechanism of the SiC chemical vapor deposition. Propose ideal production condition.

  26. Detailed Reaction Kinetics for Modeling of Nuclear Fuel Pellet Coating for High Temperature Reactors. • Drs. Gordon and Ge from the chemistry department. • Drs. Fox and Gao from the chemical engineering department. • Drs. Battaglia and Vedula from the mechanical engineering department.

  27. Chemical vapor deposition of SiC Precursors: CH3SiCl3 (methyltrichlorosilane) Temperature: 1000-2000 K Pressure: ~1 atm Complex gas-phase and surface chemistry CH3SiCl3  SiC (solid) + 3HCl

  28. CH3SiCl3 decomposition pathways G = H - TS in kcal/mol at 0 K (left) and 1400 K (right)

  29. 50 gas phase species Cl, Cl2, H, H2, HCl, C2H, C2H2, C2H3, C2H3Cl, C2H4, C2H5, C2H5Cl, C2H6(e), C2H6, 1CH2, 3CH2, CH2C, CH2Cl, CH2Cl2, CH3, CH3CH(s), CH3Cl, CH4, HCHC, Si2Cl4, Si2Cl5, Si2Cl6, SiCl2, SiCl3, SiCl4, SiH2Cl, SiH2Cl2, SiH3Cl, SiHCl, SiHCl2, SiHCl3, CH2SiCl2, CH2SiCl3, CH2SiHCl, CH2SiHCl2, CH3SiCl, CH3SiCl2, CH3SiCl2Cl, CH3SiCl3, CH3SiH2Cl, CH3SiHCl, CH3SiHCl2, HCSiCl, 1CHSiCl3, 3CHSiCl3

  30. 41 reactions without a transition state To be continued …

  31. 73 reactions with a transition state

  32. Reduced mechanism • Our collaborators, including the chemical engineers and mechanical engineers, also complained about the long lists. • How to reduce it? • Remove the species whose concentration is very low at high temperatures. • Keep important species such as 3CH2, CH3, SiCl2, and SiCl3 as target molecules. • Remove 1 species at a time and compare the reduced and full mechanisms. • Reduced to 28 species and 29 reactions.

  33. [C2H3] Time (s)

  34. [SiHCl] Time (s)

  35. Surface reactions: deposition • Surface reactions involve thousands of atoms. • Hybrid quantum mechanics/molecular mechanics (QM/MM) method.

  36. (bulk)-C3SiCl QM region C H Si QM + MM regions Cl

  37. 1). Production of Si*. 2). Si-C growth. H attacks Cl HCl leaving H3C attacks Si* Forming H3C-Si bond MM region MM region MM region MM region

  38. What did we learn? • A gas phase mechanism was proposed in the silicon carbide chemical vapor deposition. • The gas phase mechanism was reduced to 28 species and 29 reactions. • How temperature and precursor concentration affect gas phase chemistry. • Surface chemistry under investigation. • Questions and comments?

  39. Research plan • Atomic layer deposition of Al2O3, TiO2, and SiO2. • Global optimization of protein structures. • Astrochemistry in ice. • Chemical vapor deposition of diamond C, pyrolytic C, and bulk Si. • Fast global optimization of large silicon clusters.

  40. Atomic layer deposition • ALD is based on sequential, self-limiting surface chemical reactions. • Precise atomic layer control: no defects! A repeat B http://www.colorado.edu/chemistry/GeorgeResearchGroup/intro/aldcartoon.GIF

  41. Vanadium oxide (VxOy) catalyzed oxidative dehydrogenation • Experimental energy barrier: 20-30 kcal/mol. • Theoretical energy barrier: 45-80 kcal/mol. • What’s wrong? Vanadium oxide is supported by the ALD produced Al2O3, SiO2, or TiO2 surfaces. • How to model an ALD surface? • How does the ALD surface help lower the energy barrier of C3H8 + 1/2O2 C3H6 + H2O?

  42. Global optimization of protein structures: important for drug design primary structure secondary structure tertiary structure quaternary structure

  43. Global optimization methods • Random sampling: 30 dihedral angles each with 5 possible values. 530 (~1 billion trillion) conformations. • Molecular dynamics: some proteins fold in minutes; energy and force need to be evaluated 1018 times (t=10-15s). • Genetic algorithm + Tabu + In situ adaptive tabulation.

  44. dihedral angles crossover mutation • Genetic algorithm. • Tabu (taboo): to penalize the moves to previously visited conformations. • In situ adaptive tabulation. {1… N} -> E a). Enew Eold b). EnewwiEiold c). compute Enew 2 1

  45. Astrochemistry in ice ? Callisto Europa Ganymede

  46. Jupiter’s Magnetic Field

  47. Potential energy surface of 1H2O2 CCSD(T) (kcal/mol)

  48. Probable Reaction Paths to HOOH • 1O + H2O 1H2O-O  HOOH • 1O2 + H21H2O-O  HOOH • 1O (3O) + H2O 2OH + 2OH  HOOH

  49. Future work • Study the reaction paths at higher level of theories. • Study the potential energy surfaces that involves cations such as 2O+. • Reaction rate constant calculations. • Molecular dynamics calculations. • Elucidation of H2O2 formation mechanism. • Study of H2O2 reaction paths in a biological environment.

  50. Acknowledgements Prof. John D. Head at University of Hawaii Prof. Mark S. Gordon at Iowa State University Department of Energy Grant# DE-FC07-05ID14661

More Related