Contents for Abstracts Conference on Current Trends in Computational Chemistry 2003 11
P1 William H. Adams Intermolecular Perturbation Theory: Radii of Convergence at 19 Infinite Separation P1 Lovell Agwaramgbo, Marcus Harris, and Kimberly Bernard Experimental Analysis 20 of the Reactions of Silyl and Non-Silyl Epoxides with Basic Nucleophiles like LiAlH4 P1 R. N. Allen, Pawel Kedzierski, Andrzej Sokalski, and Jerzy Leszczynski Interactions 21 within the Active Site of Urate Oxidase P1 Valentine P. Ananikov Theoretical Study of Bicyclic Aromatic System Formation 22 from Unsaturated Linear Precursors P1 T. Aversa, C. L. Coblitz, K. A. Flanagan, M. D. Kasner, K. VanArsdale, and M. L. 23 Kasner Contributions of the Endo- and Exo- Anomeric Effect to the Conformational Energies of Substituted Cyclohexanes, 2-Oxanes, and 2-Thianes P1 Jon Baker, László Füsti-Molnar and Peter Pulay The Fourier-Transform Coulomb 26 Method P1 Anu Bamgbelu, Suely Black, and Jaroslaw J. Symczak Theoretical Study of Band 27 Gaps of Conjugated Polymer Materials with Donor-Acceptor Architectures P1 Tunna Baruah, Mark R. Pederson, Rajendra R. Zope, and Steven L. Richardson 28 Electronic Structure, Stability and Bonding of As@Ni12@As20 Cage P1 Judge Brown, Diwakar M. Pawar, and Eric A. Noe Conformational Study of 29 Cyclotridecane by Dynamic NMR Spectroscopy and Computational Methods P1 Jaroslav V. Burda, Jiří Šponer, Jana Hrabáková, Michal Zeizinger, and Jerzy 30 Leszczynski The Influence of N7 Guanine Modifications on the Strength of Watson- Crick Base Pairing and Guanine N1 Acidity: Comparison of Gas Phase and Condensed Phase Trends P1 Jaroslav Burda, Michal Zeizinger, and Jerzy Leszczynski Activation Barriers and 31 Rate Constants for Hydration of Platinum and Palladium Square Complexes: An ab Initio Study P1 Michael Cato, Jesse Edwards, Henry Joung Lee, Zhengqing You Computational 32 Studies on Novel AZT-Derivatives P1 Anthony Chuma, Dong Hee Kim, and Peter Pulay Minimizing Cost in Calculations 33 for Conformational Preferences of Small Polypeptides: Number of Residues to Consider in a Protein Model, Attenuated Basis Sets, Solvent Effects … P1 David M. Close Calculation of Hyperfine Couplings from Non-Optimized Structures 34 P1 Crystal B. Coghlan, Shelley S. Huskey, and David H. Magers Conventional Strain 37 Energy in Small Heterocycles of Carbon and Silicon P1 Lonnie D. Crosby and Henry A. Kurtz Study of Amorphous Silicon Dioxide Clusters 38 for Use in an Additive Polarizability Model P1 Jennifer L. Curry and Joseph A. Bentley The Accurate Calculation of Ro- 41 Vibrational Eigenenergies of HOD P1 Y. Daoudi, P.J. Bonifassi Correlation of Dynamic Calculations on Chiral Nonlinear 42 Optical Molecules with Measurements by Hyper-Rayleigh Scattering(HRS) and Conformational Effects 12 Conference on Current Trends in Computational Chemistry 2003 Contents for Abstracts
P1 Dalephine Davis, Diwakar Pawar, and Eric A. Noe Conformational Studies of 44 Triphenylmethyl Formate and N-Triphenylmethyl Formamide P1 Yuanjian Deng and Ming-Ju Huang Capillary Electrophoretic Separation and 45 Theoretical Study of Inclusion Complexes of Sulfobutyl Ether β-Cyclodextrin with Estrogens P1 Claudia Eybl, Brian Johnson, Jesse Edwards The Correlation between Impact 47 Sensitivity and Heat of Detonation in a Unique Class of Energetic Materials P1 Peng-Dong Fan, Karol Kowalski, Maricris Lodriguito, and Piotr Piecuch New 48 Alternatives for Accurate ab Initio Calculations P1 Antonio M. Ferreira, Bob M. Moore, II, and Henry A. Kurtz Exploring the 49 Electronic Structure of Novel Cannabinoid Derivatives: New Approaches to Rational Drug Design P1 Eric W. Fisher The Distribution of Water Geometries about Polar Surface Residues 51 in Proteins, as Studied by Molecular Dynamics P1 Aviane Flood, Jacquelin McCuller, Gareth Forde, Glake Hill, and Jerzy 52 Leszczynski The Effects of Metalation on Methylated Adenine Thymine Watson- Crick Base Pair P1 Alan Ford and Peter Pulay MP2 Computational Study of van der Waals Interactions 53 Between Graphene Sheets P1 Gareth Forde, Aviane Flood, Angela Fortner, Adrian Ford, Alejandro Nazario, 54 Curinetha Hubbard, Glake Hill, Leonid Gorb, and Jerzy Leszczynski Comprehensive Study of the Effects of Methylation on Tautomeric Equilibria of Nucleic Acid Bases P1 Jason Ford-Green, Deborah Bryan, Jesse Edwards, John West, Ben M. Dunn 55 Active Site-Inhibitor Modeling Using a Customized HIV-Protease Polypeptide P1 A. D. Fortner, A. Michalkova, J. Leszczynski Theoretical Study of Adsorption of 56 Methyl-Cytosine on Dickite P1 Fillmore Freeman, Christine Fang, David Hoang, Angela C. Huang, Katie Le, 57 Thuy D. Mai, and Khue Trinhk Density Functional Theory (DFT) Study of Stannacyclohexanes and Distannacyclohexanes: Conformational Interconversions, Relative Energies, Stereoelectronic Effects, and Structures P1 Al’ona Furmanchuk, Olexandr Isayev, Leonid Gorb and Jerzy Leszczynski 60 Theoretical Investigation of 3-Methyl-Cytosine Hydration P1 Kurt R. Glaesemann and Laurence E. Fried An Improved Thermodynamic Energy 61 Estimator for Path Integral Simulations P1 Sharye Glenn, Brian Johnson, Jesse Edwards Bond Dissociation Processes in 62 Various Energetic Materials P1 Sławomir J. Grabowski, W. Andrzej Sokalski, and Jerzy Leszczynski π...H+...π 63 Hydrogen Bonds P1 Jiande Gu, Jing Wang, and Jerzy Leszczynski Hydrogen Bonding in 5-Bromouracil- 64 Adenine-5-Bromouracil-Adenine (T+AT+A) Tetrads
Contents for Abstracts Conference on Current Trends in Computational Chemistry 2003 13
P1 Nasser L. Hadipour, Nasser Zamand Ab Initio Calculation of 14N NQR Parameters 65 13 1 and C, H NMR of C18H12N6 P1 Frank Hagelberg, Chuanyun Xiao, Ahmed M. El-Nahas Structures and 67 Dissociation Channels of Metal Dications Solvated by Acetonitrile Ligands P1 John A.W. Harklessa and Karl K. Irikura Multi-determinant Trial Functions in the 68 Determination of the Dissociation Energy of the Beryllium Dimer: A Quantum Monte Carlo Study P1 Robert H. Higgins Alkylation of Cytosine by cis-1-Methyl-3-hydroxyazetidinium 69 Ions: Transition States, Tautomerism, and Hydrogen Bonding P1 Glake Hill, Alex Kollias, Tomekia Simeon, Gareth Forde, William Lester, and Jerzy 70 Leszczynski Using Variational Monte Carlo for Excited States Calculations in Biological and Material Calculations P2 Patricia L. Honea, Ashley L. Ringer, and David H. Magers Conventional Strain 71 Energy in the Diazetidines and the Diphosphetanes P2 Ming-Ju Huang Theoretical Study of the Stereoisomers of Salsolinol 72 P2 Ming-Ju Huang and Manyin Yi Theoretical AM1 Studies of Inclusion Complexes of 73 Heptakis(2-o-hydroxypropyl)-β-Cyclodextrin with Alkylated Phenol P2 Danielle L. Hudson, Jeffrey A. Hinkley, Thomas C. Clancy, Melissa S. Reeves 74 Molecular Modeling of Helium Diffusion in Isomeric Polyimides P2 Shelley S. Huskey and David H. Magers Conventional Ring Strain in Unsaturated 76 Four-Membered Rings P2 Olexandr K. Isayev, Leonid Gorb, Igor Zilberberg and Jerzy Leszczynski 77 Mechanism of Nitrobenzene Reduction by Iron (II) Compounds: Density Functional Theory Study P2 Olexandr K. Isayev, Leonid Gorb, Bakhtiyor Rasulev and Jerzy Leszczynski 78 Theoretical Investigations and Structure-Toxicity Relationships of Nitroaromatic Compounds P2 Cynthia Jeffries, Glake Hill, Jerzy Leszczynski Insight into the Dispersion Energies 79 of Hydrogen and Carbon Dimer Interactions S6 Bogumil Jeziorski, Alston J. Misquitta, and Krzysztof Szalewicz Density-Functional 80 Theory Approach to van der Waals Interactions via Symmetry-Adapted Perturbation Expansion P2 Adria Johnson, Noel Matthews, and David H. Magers Computation of Conventional 81 Strain Energy in the Thiazetidines P2 Candace L. Jones, Morgan S. Ponder and Tracy P. Hamilton Computational 82 Analysis of Endo vs. Exo Retinoid Compounds P2 Dwayne C. Joseph and Bidhan C. Saha Low-Energy Single-Electron Capture in 84 5+ B and H2 Collisions P2 Isabella Karle Women in Science and Engineering. The Untapped Resource in Many 86 Countries
14 Conference on Current Trends in Computational Chemistry 2003 Contents for Abstracts
P2 I.V. Kochikov, G.M. Kuramshina, D.A. Sharapov, S.A. Yagola Data Base of 87 Quantum Mechanical and Regularized Force Constants in Redundant Internal Coordinates P2 I.V. Kochikov, G.M. Kuramshina, D.A. Sharapov, S.A. Yagola Self-Consistent 91 Model for the Joint Treatment of Spectroscopic and Electron Diffraction Data S6 Walter Kohn Van der Waals Energies and Time-Dependent Density Functional 95 Theory P2 V.V. Kukueva Quantum-Chemical Research of Chemical of Elementary Radical 96 Reactions
P2 H. A. Kurtz and N. P. Labello Modeling ZrxSi1-xO2: The Effects of Unique Bonding 98 Arrangements P2 Charles H. Langley and Eric A. Noe Ab Initio Studies of Performic Acid, Peracetic 101 Acid and Methyl Performate P2 Chittima Laohpongspaisan, Atchara Wijitkosoom, Surapong Pinitglang, Vudhichai 102 Parasuk, Supot Hannongbua Quantum Chemical Calculations on the Structure and Binding of Water Molecules in the HIV-1 Protease (PR) Enzyme P2 M.G. Levkovich, N.J. Abdullaev, D.N. Dalimov Modelling of Intra- and 103 Intermolecular Interactions of Glycyrrhizinic Acid P2 L. Jami Lewis and David H. Magers Binding Energies of Monovalent and Divalent 104 Cations with TNT P2 Tia Lewis, Jesse Edwards, Desiree Paramore, Henry Joung Lee, Zhengqing You 106 Solvation Studies on Novel Steroid–Nucleoside Conjugates: Alkylated Derivatives S4 Jan Linderberg The Reaction Simplex a Computational and Conceptual Tool 107 P2 M. Jeanann Lovell, G. Reid Bishop, and David H. Magers Conformational 110 Energetics of Naphthylquinolines P2 Yuguang Ma and Peter Politzer Calculation of Energies of Noncovalent Interactions 111 P2 Arnaldo E. Marrero, Dalephine Davis, Judge Brown, Diwakar M. Pawar, Eric A. 112 Noe Conformational Study of Cyclopentadecane P2 Massimo Malagoli and Jon Baker The Effect of Grid Quality and Weight 113 Derivatives in Density Functional Calculations of Harmonic Vibrational Frequencies P2 Artem Masunov and Sergei Tretiak Prediction of Two Photon Absorption Properties 114 for Large Organic Molecules Using Time-Dependent Density Functional Theory P2 Kareem Mckinney, Henry J. Lee and John S. Cooperwood Homology Studies of the 118 Pacific Electric Ray and Mouse Acetylcholinesterase as a Means of Bioterrorism Defense P2 James L. Meeks Ab Initio Studies of Boron Polyhedron Molecules 119 P2 A. Michalkova, M. Ilchenko, L. Gorb, J. Leszczynski Theoretical Study of 120 Adsorption and Decomposition of Nerve Agents on Metal Oxides P2 A. Michalkova and J. Leszczynski Theoretical Study of Adsorption and 121 Decomposition of Methyl tert-buthyl Ether on Dickite Contents for Abstracts Conference on Current Trends in Computational Chemistry 2003 15
P2 Ashley Moorer, Jesse Edwards, Desiree Paramore, Henry Joung Lee, Zhengqing 122 You Solvation Studies of Anti-HIV Prodrugs P2 Jane S. Murray, Pat Lane, Monica C. Concha and Peter Politzer The Effect of the 123 Cap Closure Method Upon the Surface Electrostatic Potentials of Some (6,0) Carbon, Boron/Nitrogen and Carbon/Boron/Nitrogen Nanotube Models P2 Edmund Moses N. Ndip, Shukla Manoj and Jerzy Leszczynski A Theoretical Study 124 of the Ground State Unimolecular Decomposition Channels of Propynoic Acid P2 S.I. Okovytyy, T.V.Petrova, L.I. Kasyan Quantum-Chemical Study of the Mechanism 126 of Norbornane Row Epoxyamidoacids Intramolecular Cyclization P2 S.I. Okovyty, L.K. Umrikhina, J. Leszczynski Modeling Through-Space Proton 128 Magnetic Shielding Over Oxirane Ring P2 R.B. Pandey and Marek Urban Film Growth in Reactive Aqueous Solution of 129 Hydrophobic and Polar Components on Adsorbing Substrate: A Computer Simulation Study P2 Reena R. Patel, Rajendran Mohanraj, and Charles U. Pittman, Jr Molecular 130 Dynamics Simulations of Polymer Nanocomposites Containing POSS S3 Mark R. Pederson Simulation of Molecular Magnets within Density Functional 134 Theory P2 Zenaida Peralta-Inga, Ping Jin, Jane S. Murray and Peter Politzer Prediction of 135 Free Energies of Aqueous Solvation P2 Yevgeniy Podolyan, Leonid Gorb, Jerzy Leszczynski Comparison ab Initio Study of 136 the Double-Proton Transfer in Methylated and Non-Methylated DNA Base Pairs P2 Jillian Pope, Jesse Edwards, Henry Joung Lee, Zhengqing You Conformational 137 Properties of Anti-HIV Prodrgus: AZT Derivatived P2 Arnaldo O. M. Porrata, Dalephine Davis, Judge Brown, Diwakar M. Pawar, Eric 138 A. Noe Conformational Study of Cyclopentadecane P2 Harry L. Price Stereoelectronic Interactions and Chemical Reactivity of a Masked 139 Electrophile P3 Mohammad (Mo) Qasim, Herbert L. Fredrickson, John Furey, Chris McGrath 142 Theoretical Predictions as to Initial/Intermediate Steps in Chemical Transformation of Cyclic and Cage Cyclic Nitramines Are Supported by UV/VIS/FTIR Spectrophotometry, Using CL-20 as the Model P3 B.F. Rasulev, N.D. Abdullaev, V.N. Sirov, J. Leszczynski Quantitative Structure- 144 Activity Relationship (QSAR) Study by GA-MLR Analysis of the Antioxidant Activity of Flavonoids P3 Pornpun Rattananakin, Yevgeniy Podolyan, Svein Saebø, and Charles U. Pittman, 145 Jr. Ab Initio Study of the Rotation around the C=C Bond in Push-Pull Systems P3 Paresh Chandra Ray Molecular Design for Ionic Nonlinear Optical Systems 147 P3 Kamila Reblova, Nada Spackova, Richard Stefl, Kristina Csaszar, Jaroslav Koca, 148 Neocles B. Leontis and Jiri Sponer Molecular Dynamics of 5S rRNA Loop E P3 Demarcio Reed, Yinghong Sheng, and Jerzy Leszczynski Theoretical Study on 149 Thermal Conversion of Benzopyran: Substituent and Solvent Effects 16 Conference on Current Trends in Computational Chemistry 2003 Contents for Abstracts
P3 Ashley L. Ringer and David H. Magers gem-Dimethyl Substituents of Cyclopropane 150 and Cyclobutane P3 Burke Ritchie and Charles A. Weatherford Time-Dependent Non-Wavepacket 152 Theory of Electron Scattering P3 Jamar Robinson, Jesse Edwards, Henry Joung Lee, Zhengqing You Bond 153 Dissociation of Novel AZT Derivative Prodrugs P3 Amy Rowe An ab Initio Study of C-H···O Hydrogen Bonding in Peptide Models 154 P3 Amar Saal, Yahia Moussaoui, and Ourida Ouamerali Hydrogen Bond Effect on 155 Electronic and Vibrational (Hyper)Polarizability of The Dimmers (HF)2, (H2O)2, and (NH3)2: Theoretical Study P3 Svein Saebo, Krzysztof Wolinski, Peter Pulay, and Jon Baker Efficient AO- 156 Formulation of MP2-Gradients P3 Julia Saloni, Szczepan Roszak, and Jerzy Leszczynski Theoretical Studies of Neutral 157 and Ionized Lanthanide Halides P3 L. M. Salter, G. M. Chaban, and J. Leszczynski Theoretical Study of Gas Phase Tau- 158 tomerization Reactions for the Ground and First Excited Electronic States of Adenine S5 Joachim Sauer Calculations on Transition Metal Oxides – From Gas Phase Clusters 159 to Solid Catalysts P3 Yinghong Sheng, Ray Butcher, Bakhtiyor Rasulev, Jerome Karle and Jerzy 160 Leszczynski “Aromatic” 4n π-system Diazapentalenes P3 M.K. Shukla and Jerzy Leszczynski MCSCF Study on Excited State Properties of 162 Thiouracil and Its Comparison with Uracil P3 M.K. Shukla and Jerzy Leszczynski MCSCF Investigation of Singlet Excited State 164 Geometries of Guanine Tautomers P3 S. A. de Silva, M. L. Kasner, M. A. Whitener and S. L. Pathirana Computational 166 Study of a Fluorescent Photoinduced Electron Transfer (PET) Sensor for Cations P3 Tomekia M. Simeon, Chi-Cobi Speaks, Glake Hill, and Jerzy Leszczynski 168 Theoretical Study of Oxygen Absorption on Single-Walled Carbon Nanotubes P3 Vitaly Solkan and Jerzy Leszczynski The Density Functional Theory Study of the 169 Superacid Acidity Scale P3 Vitaly Solkan and Jerzy Leszczynski The Nature of Superacid Electrophilic Species 170 in HF/SbF5 and FSO3H/SbF5: The Ab Initio and Density Functional Theory Studies Taking Into Account the Solvation Effects P1 Chi-Cobi Speaks, Tomekia Simeon, Glake Hill, Jerzy Leszczynski A Theoretical 171 Study of Carbon Nanotube Interactions P3 J. Sponer, J. E. Sponer, J.V. Burda, B. Lippert, and J. Leszczynski The Effects of 172 Metal Cations on Nucleic Acid Bases P3 E. Štefáneková, J. Urban, P. Mach and T. Hianik Interaction of Model α-Helical 173 Peptides with Lipid Bilayers P3 A. Sykes and S. M. Black Effects of the Binding of Molybdenum on the Electrostatic 174 Charge, Solvation and Solute energy of 4-hydroxyl-2, 6-dicarboxylate-pyridine Anion Contents for Abstracts Conference on Current Trends in Computational Chemistry 2003 17
P3 Jaroslaw J. Szymczak, Szczepan Roszak, and Jerzy Leszczynski Molecular 175 + Structures and Nature of Interactions in OCH (Ne)n (n=1-17) Complexes P3 Jaroslaw J. Szymczak, Szczepan Roszak, and Jerzy Leszczynski The Computational 176 Studies of the Intermo-lecular Interactions in Donor-Acceptor XL3NH3 (X=B, Al, Ga; L=H, Cl) Complexes P3 G. J. Tawa, I. A. Topol, S. K. Burt, R. A. Caldwell, A. Rashin The Structure and 177 Thermodynamics of Solvated Monatomic Ions Using a Hybrid Model of Solvation P3 Kanchana S. Thanthiriwatte and K.M. Nalin de Silva Non-Linear Optical Properties 178 of Novel Fluorenyl Derivatives: Ab Initio Quantum Chemical Calculations P3 S. B. Tripathi Debye Temperature of Ionic Solids Revisited 179 P3 M. Tulyasheva, B. F. Rasulev, N.D. Abdullaev, K.K. Turgunov, Kh. M. 181 Shakhidoyatov and J. Leszczynski Quantum-Chemical and Experimental Study of Cyano-(2-3H-quinazolin-4-ylidene)-acetic Acid Ethyl Ester P3 R. K. Vadapalli, J. W. Mintmire and B. I. Dunlap Object Oriented Approaches for 182 Parallelizing Electronic and Structure Calculations of Nano & Biomaterials P3 V.F. Vargalyuk and V.A. Seredyuk Modeling of Electroreduction Reaction of Some 185 Zn2+ and Mn2+ Heterocomplexes P3 Ramaiyer Venkatraman and Paresh Chandra Ray Self-Assembly Based on the 187 Levulinic Acid-Melamine Lattice S2 Chen Wang and Chunli Bai Weak Forces in Building the Molecular Nanostructures 188 P3 Jing Wang, Jiande Gu, Jerzy Leszczynski Modeling Interactions of Fasciculin 2 with 189 AChE: A DFT-based Molecular Dynamics and Quantum Chemical Study P3 Andrzej Wierzbicki, Pranav Dalal, Jeffry D. Madura, and Herman S. Cheung 192 Molecular Dynamics Simulations of Crystal-Induced Membranolysis P3 Chuanyun Xiao and Frank Hagelberg Substitutional and Endohedral Structures 193 Based on Polyhedral Oligomeric Silsesquioxane (POSS) Molecules P3 Ilya Yanov and Jerzy Leszczynski Quantum Transport in Biosystems 194 P3 N.U. Zhanpeisov, J. Leszczynski, M. Anpo Ab Initio and DFT Study of Intrinsic 195 Band Gap in Semiconductor Oxides and Ti-silicalite P3 N.U. Zhanpeisov A Density Functional Theory Study of the Oxidation of Methanol 196 to Formaldehyde over Vanadia Supported on Titania P3 N. U. Zhanpeisov, H. Ohmukai, M. Anpo IR Bands Originated from the Adsorption 197 of CO on Platinum Supported Zeolites: A Density Functional Theory Study S5 G. M. Zhidomirov Structure, Chemisorption and Catalytic Activity of Multivalent 198 Metal Oxide Species in High Silica Zeolites P3 Igor Zilberberg, Sergey F. Ruzankin Analysis of Spin-Unrestricted Wave Function 199 by Means of the Expansion of β Orbitals in the Basis of α Orbitals P3 M. A. Zottola and J. Karle Activation and Binding of Qinghaosu and Its Derivatives 200 P3 A.Yu. Rubina, Yu. V. Rubin, M. K. Shukla, J. Leszczynski Investigation of Adenine 201 Tautomer Complexes with Metal Ions: Stability and Molecular Structure
Conference on Current Trends in Computational Chemistry 2003 19
Intermolecular Perturbation Theory: Radii of Convergence at Infinite Separation
William H. Adams
Rutgers University, 1208 Ursulines Ave. New Orleans, LA 70116
We explain how the radius of convergence ρ of the Rayleigh-Schrödinger perturbation theory applied to the interaction between two atoms may be calculated at infinite separation by calculations on the individual atoms. This is simpler and more straightforward than deriving ρ from perturbation calculations carried to very high order or from searches for singularities of complex energy surfaces. A ρ greater than one is required for intermolecular perturbation theories to converge to physically meaningful results. We find for the interaction between a ground state H atom and a ground state Li atom ρ ≤ 0.702. This is smaller than the best published value: 0.725 at 20.0 bohr. We have also calculated ρ for the interaction of Li in its first excited state with ground state H and for the interaction between two ground state Li atoms. For the latter we find ρ ≤ 0.529. There are no published results for either the excited states of LiH or the ground state of Li2. For Be2 we find a still smaller radius, 0.476. Results for several other systems will also be presented. It is possible to reach some general qualitative conclusions by our method. For example, the radius of convergence for any hydride A-H must be larger than that for A-A. The same can be said for the ρ of any A-He system. Another conclusion is that the divergence of the perturbation expansion is not caused by the Coulomb singularity. The method is applicable to the interaction between molecules as well as between atoms. We also consider how the method can be extended to symmetry adapted perturbation theories.
20 Conference on Current Trends in Computational Chemistry 2003
Experimental Analysis of the Reactions of Silyl and Non-Silyl Epoxides with Basic Nucleophiles like LiAlH4
Lovell Agwaramgbo*, Marcus Harris, and Kimberly Bernard
Dillard University, New Orleans, Louisiana
It has been documented that silicon stabilizes α-anions and β-cations and that many α,ß- epoxysilanes undergo ring-opening α to silicon. However, it has not been very clear why α,ß- epoxysilanes open α to silicon even under acidic conditions that facilitate SN1 - like process which should lead to ß-opening since cations ß to silicon are stabilized. Unsubstituted styrene oxide underwent ring-opening to give mixtures of α and ß opening products. Trimethylsilystyrene oxide underwent regiospecific ring-opening α to silicon under experimental conditions. This was attributed to the inductive effect of silicon. To compare silicon and carbon, in the ring-opening reactions, this proposal investigated the reactions of TrimethylsilyEthylene oxide (1) and tertiarybutyl Ethylene Oxide (2) with Nucleophilic reagents like lithium aluminum hydride under experimental and theoretical conditions.
O R1 A HO R1 R B OH NU NU R R OMe MeO R1
If opening occurs at carbon A above Alpha opening If opening occurs at carbon B above Beta opening
1 1, R = Me3Si, R = H 1 2, R = Me3C, R = H
Nu = H
Conference on Current Trends in Computational Chemistry 2003 21
Interactions within the Active Site of Urate Oxidase
R. N. Allen1, Pawel Kedzierski2, Andrzej Sokalski2, and Jerzy Leszczynski1
1Computational Center for Molecular Structure and Interactions, Jackson State University, Jackson, MS 39217 2Institute of Physical and Theoretical Chemistry, Wroclaw University of Technology, Wyb. Wyspianskiego 27, 50-370 Wroclaw Poland
The physical nature of interactions within the active site of urate oxidase from Aspergillus flavus was studied using a variation-perturbation energy decomposition scheme defining sequence of approximate intermolecular interaction energy models. These models have been used to analyze the role of residues constituting urate oxidase active site in binding of substrate, intermediate and product. Some literature reports assign essential role of Phe159, Arg176, Gln228, Thr57, Asn254, and HOH2 residues while recent study indicates that two other residues, Lys10 and Asp58 might play a role in the degradation of urate. In this study we examine the magnitude of substrate, intermediate, product, and inhibitor stabilization with all active site residues. 22 Conference on Current Trends in Computational Chemistry 2003
Theoretical Study of Bicyclic Aromatic System Formation from Unsaturated Linear Precursors
Valentine P. Ananikov
ND Zelinsky Institute of Organic Chemistry, Russian Academy of Sciences, Leninsky Prospect 47, Moscow 119991, Russia. http://nmr.ioc.ac.ru/Staff/AnanikovVP/
Cycloaddition reactions play an important role in modern organic chemistry, especially for constructing aromatic and heteroaromatic systems. Recently, a tunable practical approach has been developed to prepare both five-membered furan derivatives (4) and six-membered benzene derivatives (5) based on the intramolecular cycloaddition reactions of the same substrate (1, X=O, CH2) [1]. The theoretical study clarified several interesting features of the reaction mechanism and located the appropriate transition states [2].
X=O X=CH2 ∆E≠/∆E ∆E≠/∆E [3+2] 1,2-H shift cycloaddition XX v i 29.4/-10.7 51.7/-5.9 i 24X=O ii 32.6/-22.3 29.7/-21.5 X iii 50.9/-11.6 26.0/-15.6 iii iv iv 62.5/11.6 41.6/15.6 1 ii v 12.6/-62.3 9.9/-67.6 vi - 6.5/-91.0 X=O,CH2 [4+2] 1,2-H shift cycloaddition X vi X 3 X=CH 5
Activation (∆E≠) and reaction (∆E) energies at B3LYP/6-31G(d) level (in kcal/mol).
Particularly, it was shown that for oxygen-containing substrates (1, X=O) [3+2] cycloaddition reaction followed by 1,2-H shift should be preferably observed (4). In contrast, hydrocarbons undergo [4+2] cycloaddition with the final benzene ring formation (5). In principle, 1,2-Carbon may take place in the system linking both pathways. The further study of the system has revealed several new properties of the system, which will be presented and discussed.
References:
1. (a) R.L.Danheiser, A.E.Gould, R.F. de la Pradilla, A.L.Helgason, J. Org. Chem., 1994, 59, 5514; (b) M.S.B.Wills, R.L.Danheiser, J. Am. Chem. Soc., 1998, 120, 9378. 2. (a) V.P.Ananikov, J. Phys. Org. Chem., 2001, 14, 109; (b) V.P.Ananikov, J. Phys. Org. Chem., 2003, 16, 253. Conference on Current Trends in Computational Chemistry 2003 23
Contributions of the Endo- and Exo- Anomeric Effect to the Conformational Energies of Substituted Cyclohexanes, 2-Oxanes, and 2-Thianes
T. Aversa, C. L. Coblitz, K. A. Flanagan, M. D. Kasner, K. VanArsdale, and M. L. Kasner
Department of Chemistry and Biochemistry Montclair State University, Upper Montclair, NJ 07043
The Anomeric Effect In 1955, it was proposed that the axial orientation for alkoxy groups adjacent to the oxygen in the pyranose ring is generally more preferred than the equatorial orientation (1):
H
H
O H H H H H H H H (1) H O H H O H O H H H H H H H H H The interpretation was that the axial preference was related to the orientation of the lone pairs on the ring oxygen. Today, the importance of stereoelectronic interactions in determining molecular structures is recognized and computational studies are a relatively recent mechanism to study these interactions in detail.
Exo- versus Endo-Anomeric Effect Although the concept of an exo- and endo-anomeric effect appears to have been generally accepted, there is considerable doubt about the magnitude of the effects and their respective contributions to the conformational energy. The most common explanation for the endo- and exo-anomeric interactions is in terms of the stabilizing effects of charge transfer processes between a donor nonbonded pair of electrons and an adjacent antiperiplanar σ* orbital. Compounds such as 2-methoxytetrahydro-2H-pyran (1) have two centers of nonbonded electrons: the endocyclic oxygen atom at position 1 and the exocyclic oxygen atom at position 7. In Scheme 1, the exocyclic oxygen atom [O(7)] in axial 2-methoxytetrahydro-2H-pyran (1a, 1b) is the electron donor to the C(2)-O(1) σ* orbital. Consequently, the oxygen atom at position 7 develops a partial positive charge, the oxygen atom at position 1 develops a partial negative charge and the O(1)-C(2) bond is lengthened relative to O(1)-C(6). This is an example of the exo-anomeric effect.
24 Conference on Current Trends in Computational Chemistry 2003
δ+
δ−
1a 1b Scheme 1
It is recognized that an endo-anomeric effect also occurs in the axial conformer of 2- methoxytetrahydro-2H-pyran (1). The endocyclic oxygen atom donates electron density to the axially oriented σ*C-O molecular orbital (1c, 1d, Scheme 2).
δ−
δ+
1c 1d Scheme 2
Electron density donated in this manner will cause the oxygen atom at position 1 to have a greater positive charge and the oxygen atom at position 7 will have greater negative charge (1d). In addition, the O(1)-C(2) bond will shorten relative to the O(1)-C(6) bond. Although steric effects are considered to be responsible for the equatorial preferences in alkylcyclohexanes, natural bond orbital (NBO) analysis, suggests the bond-antibond interactions of the exocyclic C(1)-C(7) bond with the σ*C-C and σ*C-H bonds are mainly responsible for the equatorial preference in methylcyclohexane while steric factors are responsible for the equatorial preference in 2-methyltetrahydro-2H-pyran. It appears that these same hyperconjugative orbital interaction can occur with any axially oriented σ* orbital and recent computational studies have shown that the axial C-H bond adjacent to the oxygen is 7-12% longer than the equatorial C-H bond in tetrahydro-2H-pyran. In addition, it has been shown that in 2-alkyltetrahydro-2H- pyrans, the C(2)-Hax bond is longer than the C(4)-Hax when the alkyl substituent is equatorial. Conference on Current Trends in Computational Chemistry 2003 25
Computational Methods Ab initio Hartree-Fock and Density Functional Theory calculations were used to obtain optimum geometries and relative energies of the various axial and equatorial rotamers of a series of substituted cyclohexanes, 2-oxanes, (tetrahydro-2H-pyrans) and 2-thianes (tetrahydro-2H- thiopyrans):
Y
X X Y
X = CH2, O, S Y = CH3, CH2CH3, OH, OCH3, SH, SCH3
Geometry optimizations on all structures were done at the HF/6-31G(d,p) and B3LYP/6- 31G(d,p) levels. Additionally, second-order Møller-Plesset (MP2) single point energies were determined on the optimized geometries. In principle, DFT methods take into account the full correlation energy: the mixing of the ground-state (HF) wavefunction with “excited state” wave functions. As a check, MP2 energies were also determined for the B3LYP/6-31G(d,p) optimized geometries. In order to verify that the structures are in fact minima and not transition states, vibrational frequencies were determined on these optimized geometries. The thermodynamic parameters were used to convert the relative energy differences to conformational free energies. A study of the relative energies allowed an evaluation of the relative contributions of the exo- and endo- anomeric effects to the conformational energy. A study of the changes in the bond angles of the heteroatoms and the various C-X and C-Y bond lengths provided information about the degree of the steroelectronic interactions.
Acknowledgements Financial support was provided by Montclair State University and the Petroleum Research Fund of the American Chemical Society. 26 Conference on Current Trends in Computational Chemistry 2003
The Fourier-Transform Coulomb Method
Jon Baker1,2, László Füsti-Molnar2 and Peter Pulay1,2
1Parallel Quantum Solutions, 2013 Green Acres Road, Suite A Fayetteville, Arkansas 72703
2Department of Chemistry, University of Arkansas Fayetteville, Arkansas 72701
The recently described Fourier Transform Coulomb (FTC) algorithm for fast and accurate Density Functional Theory (DFT) calculations (L. Füsti-Molnar and P. Pulay, J. Chem. Phys. 117 (2002) 7827) is briefly discussed. We present results showing the speed and accuracy of DFT energy computations, including parallel job runs, comparing the performance of the FTC method with our standard all-integral DFT code. The current parallel FTC algorithm is up to six times faster in total than our classical all-integral algorithm, and well over an order of magnitude faster for computation of the Coulomb terms, with essentially no loss in accuracy. Proposed improvements should more than double these factors. The Coulomb energy in non-hybrid DFT energy calculations can now be computed accurately for large molecules and/or basis sets faster than the exchange-correlation energy. Extension to gradients is straight-forward and should result in even greater savings. Scaling with increasing system size at constant basis set quality is near-linear, and - unlike almost all other fast-DFT methods - scaling with increasing basis set size at constant system size is also genuinely near-linear, and allows for significant savings over conventional algorithms even for relatively small systems (e.g., aspirin). Conference on Current Trends in Computational Chemistry 2003 27
Theoretical Study of Band Gaps of Conjugated Polymer Materials with Donor-Acceptor Architectures
Anu Bamgbelu1, Suely Black1, and Jaroslaw J. Symczak2
1Center for Materials Research, Norfolk State University 2Computational Center for Molecular Structures and Interactions, Jackson State University
Photovoltaic devices provide a promising option to fossil fuels as an alternative for generation of electrical power. However, a novel concept to be applied requires a suitable material. One of the possible approaches is the rational design of the required materials at the molecular level. In the present communication, we report on the ab initio calculations of the molecule suitable as an electron donor for polymeric photovoltaic devices: benzene, 1,1 – [2- butene-1,4-diyl bis [oxy (1-methylene – 2,1 – ethane diyl. All computations were performed based on the density functional theory with B3LYP hybrid functional and the AM1 geometry optimizing theoretical method. For the calculations we employed the 3-21G and 6-31G Gaussian basis sets. The computations provide information on the energy difference between the highest occupied molecular orbital [HOMO] and the lowest unoccupied molecular orbital [LUMO] for the donor molecules of a possible photovoltaic polymer. In our simple approach, the oligomer band gap is equal to the HOMO-LUMO energy difference. Extrapolation of the DFT HOMO- LUMO energy differences for the monomer, dimer and trimer give reasonable band gap predictions. 28 Conference on Current Trends in Computational Chemistry 2003
Electronic Structure, Stability and Bonding of As@Ni12@As20 Cage
Tunna Baruah1,2, Mark R. Pederson2, Rajendra R. Zope3, and Steven L. Richardson2,4
1Georgetown University, Washington DC, 20057 2Naval Research Laboratory, Washington DC, 20375 3George Mason Unversity, Fairfax, Virginia 22030 4Howard University, School of Engineering, Washington DC 20059
We present our density-functional based study of the electronic structure of the recently synthesized As@Ni12@As20 cluster. This cluster consists of an inner icosahedral As@Ni12 core which in turn is encapsulated within a dodecahedral As20 cluster. The vibrational stability of the cluster is confirmed and the infrared and Raman spectra of the molecule are calculated. The bonding of the cluster is examined from its vibrational modes vis-à-vis those of its subunits. The bonding pattern changes as we go from the isolated As20 to the encapsulated As@Ni12@As20. 3 The As atoms are sp hybridized and in the isolated As20 cage they form a network of σ-bonds. In the encapsulated form, strong As-Ni bonds form at the expense of the As-As bonds. The existence of the As lone pairs can be clearly seen from the picture of the electron localization function. Localized Wannier functions are calculated from the atomic orbitals. The Wannier functions bring out the bonding of the system more clearly and these pictures are compared with that of the electron localization function. Conference on Current Trends in Computational Chemistry 2003 29
Conformational Study of Cyclotridecane by Dynamic NMR Spectroscopy and Computational Methods
Judge Brown, Diwakar M. Pawar, and Eric A. Noe
Department of Chemistry, Jackson State University 1400 J. R. Lynch Street, Jackson, MS 39217
Low - temperature 13C NMR spectra of a 2% solution of cyclotridecane (1) in propane showed the presence of eight overlapping peaks. The carbon NMR spectrum at -174.5 ºC was analyzed in terms of the conformations predicted by Allinger's molecular mechanics program and suggested that conformations of C1 and C2 symmetry point groups could be populated. The MM3 calculations for this hydrocarbon predicted that the C1 conformation is more stable than the C2 conformation by 1.772 kcal/mol. Free energies were calculated at the HF/ 6-31G* level for fifteen conformations and compared with our MM3 and other published results. Free energies and chemical shifts are also planned to be obtained for the five lowest-energy conformations of 1 at the MP2/6-31G* level by the GIAO method. This work was supported by NSF - CREST Grant No. HRD - 9805465. 30 Conference on Current Trends in Computational Chemistry 2003
The Influence of N7 Guanine Modifications on the Strength of Watson-Crick Base Pairing and Guanine N1 Acidity: Comparison of Gas Phase and Condensed Phase Trends
Jaroslav V. Burda,a* Jiří Šponer,b,c* Jana Hrabáková,a Michal Zeizinger,a and Jerzy Leszczynskid
aDepartment of Chemical Physics and Optics, Faculty of Mathematics and Physics, Charles University, Ke Karlovu 3, 121 16 Prague 2, Czech Republic bInstitute of Biophysics, Academy of Sciences of the Czech Republic and National Center for Biomolecular Research, Kralovopolska 135, 612 65 Brno, Czech Republic. cJ. Heyrovský Institute of Physical Chemistry, Academy of Sciences of the Czech Republic, Dolejškova 3, 182 23 Prague 8, Czech Republic dDepartment of Chemistry, Jackson State University, 1325 J. R. Lynch Street, Jackson, Mississippi 39217-0510, USA
Ab initio quantum-chemical calculations have been carried out to investigate the correlation between N1-H1 deprotonation energy and base pairing of modified guanines (neutral 8-oxoguanine (8OG) and monocationic 7,9-dimethylguanineH+ (DMG)) and dicationic platinated guanines. The calculated intrinsic gas phase trends are compared with available solution data. In the gas phase, the stability of the base pair increases with acidification of the N1(H1) position, however, the relation is not linear. The guanine N1 gas phase deprotonation energies are primarily determined by the total charge of the system, however, there is also a non- negligible contribution caused by polarization effects, which is especially significant for the DMG. The order of gas phase deprotonation energies differs from solution pK values. aHO, 2 However, when assuming that the polar environment annihilates the ionic-electrostatic contribution, qualitative agreement is seen between the gas phase and condensed phase data. The pK is primarily determined by polarization effects. In contrast, very poor correlation has aHO, 2 been found between the intrinsic Guanine – Cytosine Watson-Crick (GC WC) base pairing energies and the KGC,DMSO condensed phase data, even when separately weighting the ionic- electrostatic and polarization contributions. The bell-shaped correlation between the solution N1H acidity of the guanine derivative and the association constant K does not reflect the intrinsic gas phase trends. The lack of correlation between gas phase and solution data may be for example due to some specific interference with the GC WC base pairing caused by counter- anions or presence of structures competing with the desired base pairing. Conference on Current Trends in Computational Chemistry 2003 31
Activation Barriers and Rate Constants for Hydration of Platinum and Palladium Square Complexes: An ab Initio Study
Jaroslav Burdaa, Michal Zeizingera, and Jerzy Leszczynskib
aDepartment of Chemical Physics and Optics, Faculty of Mathematics and Physics,Charles University, Ke Karlovu 3, 121 16 Prague 2, Czech Republic bDepartment of Chemistry, Jackson State University, 1325 J. R. Lynch Street, Jackson, Mississippi 39217-0510, USA
In present work, ab initio study on hydration of cis- and transplatin and their palladium analogues was performed within a neutral pseudomolecule approach (e.g., metal-complex + water as reactant complex). Subsequent replacement of two ligands was considered. Optimizations were performed at MP2/6-31+G(d) level with single-point energy evaluation using the CCSD(T)/6-31++G(d,p) approach. For the obtained structures of reactants, transition states (TS), and products, both thermodynamic (reaction energies and Gibbs energies) and kinetic (rate constants) characteristics were estimated. It was found that all the hydration processes are mildly endothermic reactions — in the first step they require 8.7 and 10.2 kcal/mol for ammonium and chloride replacement in cisplatin and 13.8 and 17.8 kcal/mol in the transplatin case, respectively. Corresponding energies for cispalladium amount to 5.2 and 9.8 kcal/mol, and 11.0 and 17.7 kcal/mol for transpalladium. Based on vibrational analyses at MP2/6-31+G(d) level, TST rate constants were computed for all the hydration reactions. A qualitative agreement between the predicted and known experimental data was achieved. It was also found that the close similarities in reaction thermodynamics of both Pd(II) and Pt(II) complexes (average difference for all the hydration reactions ca 1.8 kcal/mol) do not correspond to the TS characteristics. The TS energies for examined Pd(II) complexes are about 9.7 kcal/mol 6 lower in comparison with the Pt-analogues. This leads to 10 times faster reaction course in the Pd cases. This is by 1 or 2 orders of magnitude more than the results based on experimental measurements.
32 Conference on Current Trends in Computational Chemistry 2003
Computational Studies on Novel AZT-Derivatives
1Michael Cato*, 1Jesse Edwards, 2Henry Joung Lee, 2Zhengqing You
1Department of Chemistry/AHPCRC Florida A&M Tallahassee, Florida,USA 32307 2College of Pharmacy and Pharmaceutical Sciences, Florida A&M Tallahassee, Florida,USA 32307
H.J. Lee et al. have synthesized several potential drug candidates along the lines of an AZT derivative. The potential uses each of these drugs is numerous and varied. However, due to the novelty of each compound there have been no or very little high level computational studies performed on these compounds. In order to verify the structure the accuracy of the lower level molecular mechanics calculations, higher level quantum mechanics calculations need to be performed. In this study semi-empirical, PM3 and AM1, and DFT calculations will be performed in order to compare the theories. These results will also be compared to Molecular Mechanics results using the Tripos Forcefield. Conference on Current Trends in Computational Chemistry 2003 33
Minimizing Cost in Calculations for Conformational Preferences of Small Polypeptides: Number of Residues to Consider in a Protein Model, Attenuated Basis Sets, Solvent Effects …
Anthony Chuma, Dong Hee Kim, and Peter Pulay
Department of Chemistry and Biochemistry, University of Arkansas, Fayetteville, Arkansas 72701
First we were interested in establishing a lower computational level to explain the effects of the neighboring residues. It has been demonstrated that shielding patterns are fairly insensitive to geometry optimization in low-energy regions;[iii] we therefore used standard geometries from HyperChem molecular modeling program[iv] to draw the polypeptides (β-sheet); GGGXGGG (X=Gly, Ala, Phe). N-terminal is protected by a formyl group and C-terminal is protected by an amino group. All of φ and ψ angles were fixed at 180°to exclude backbone angle effects. We calculated 15N, 1HN, 13 Cα, 1Hα, and 13C' chemical shifts within the residue of interest (X). Also we calculated 15N, 1HN, and 13C' chemical shifts for AAAAAAA peptide series to investigate a local sequence dependence of 15N, 1H, and 13C'. To further reduce computational cost we investigated the effect of Chesnut’s attenuated basis set approach on the chemical shifts of some residues of interest. Investigations on more efficient ways to include solvent effects are under way.
34 Conference on Current Trends in Computational Chemistry 2003
Calculation of Hyperfine Couplings from Non-Optimized Structures
David M. Close
Department of Physics, East Tennessee State University, Johnson City, TN, 37614
Geimer, Beckert, Naumov, and Lh, in Leipzig, have a series of articles reporting in situ photolytic experiments which produced reduction and oxidation products in pyrimidine bases [1- 3]. From the highly resolved FT EPR experiments the authors have obtained accurate isotropic hyperfine coupling constants of even the small proton couplings that are usually hidden in the linewidth of solid-state EPR spectra. In these articles there are discussions of several non-planar radicals. The authors calculated spin densities and isotropic hyperfine couplings of the pyrimidine base reduction and oxidation products by Density Functional Theory to support these non-planar radical assignments [4-6]. The theoretical calculations discussed here include estimates of spin densities and isotropic and anisotropic hyperfine couplings which can be compared with the experimental results. In many cases theoretical calculations can assist in making free radical assignments. The theoretical and experimental results often agree rather well. However, in other cases there are discrepancies between the theoretical and experimental results. The successes and failures of DFT to calculate spin densities and hyperfine couplings of more than twenty primary radiation induced radicals observed in the nucleobases have recently been reviewed [7]. The work of Geimer et al. involves transient free radicals in aqueous solutions. The theoretical calculations were performed on the free radicals in the gas phase. This presents one obvious problem. One knows that calculations performed to optimize the structures of gas phase anions tend to produce rather non-planar structures. One cannot however know the geometry of a radical anion in aqueous solution if one has only knowledge of the isotropic hyperfine couplings. Consider for example the thymine radical anion observed by Geimer et al. The major site of unpaired spin density is at C6 which gives rise to an isotropic hyperfine coupling of -33.0 MHz. The same radical observed in a single crystal of thymidine also has an isotropic hyperfine coupling –33.1 MHz (see Table I). From analysis of the single crystal data it is also possible to determine the C6-H" aniosotropic hyperfine coupling, from which the orientation of the C6-H" o can be determined. In Table I, one sees that the direction associated with Amid deviates only 1.1 from the computed perpendicular to the ring plane (the direction of the π-orbital), while the o direction of Amin deviates only 2.5 from the computed direction of the C6-H bond. This is very clear evidence that the radical shown here is planar in the solid state. Now then, does the remarkable similarity of the C6-H isotropic hyperfine couplings for the thymine anion in aqueous solution as in the solid state tell us anything? Since the radical in the solid state is planar, does it follow that the radical in solution is planar? Of course not. One can site important examples, such as biphenyl, where the molecular structures differ in solution and in the solid state. On the other hand, one cannot simply say that since the DFT gas phase calculations predict a non-planar radical geometry, that the thymine anion observed in aqueous solution is non-planar.
Conference on Current Trends in Computational Chemistry 2003 35
Table I: Hyperfine Coupling Parameters for Pyrimidine Anions Principal Isotropic Dipolar Direction Cosines Coupling Values Value Values
There was a time when Geimer et al. agreed that this was so. Not long ago they said “the gas phase optimizations leads to a non-planar radical structure with N1-C2-N3-C4 dihedral angle of around 40o, whereas in the dielectric surroundings a nearly planar radical structure is obtained” [4]. When it was pointed out that with these large distortions from planarity lead to C6-H isotropic hyperfine coupling that are far from the observed values, it should have given some credence to the claim that the thymine anion radicals observed in solution are essentially planar [8]. This is not the case however [9]. A new paper by Naumov et al. has returned to the discussion of non-planar radical anions [6]. Again the basic argument is that geometry optimization calculations produce non-planar molecules. However they have not been able to reproduce the C6-H isotropic hyperfine coupling for the thymine anion radical. They claim that the radical anions show pyrimidality at the radical center C6 connected with a deviation of the C6-H atom from the molecular plane up to around 12o. To have some numbers to refer to, one needs to see the authors Fig. 5 on the thymine anions, where one sees degree of non-planarity in vacuum and in water with different basis sets. The variation is from 30o to 25o in vacuum and 23o to 15o degrees in water. The C6-H isotropic hyperfine coupling varies from +5 to –2 MHz in vacuum and from –12 to –23 MHz in water. The authors optimized the structure at B3LYP/6-311+G(d,p)/CPCM, followed by a single point CPCM calculation at 6-311+G(d,p) gives a C6-H coupling of –23 MHz which is far from the experimental value of –33 MHz. For the present discussion, it is important to look at variations in the hyperfine couplings for a variety of thymine anions. Bernhard and Patrzalek showed a spread of C6-H isotropic hyperfine couplings for various thymine anions in various LiCl glasses, from 34.3 to 38.5 MHz [10]. In the solid state the variation is from –33.1 MHz in thymidine [11] to –39.8 MHz for anhydrous thymine single crystals [12]. It is significant to note that the higher values here are easily reproduced by DFT calculations. For example, the C6-H hyperfine coupling using B3LYP/6-311+G(2df,p) is computed to be –40.3 MHz, for a planar geometry. Thus, one must take exception to the remark by Naumov et al. “vacuum structures, even optimized with the large 6-311+G(2df,p) basis set, are not suitable to calculate reliable A(H6) coupling constants” [6]. One has to also consider that the geometries of radicals in a single crystal or in solution are not minima on the gas phase potential energy surface. One difference may be that the geometry optimized structures of cytosine and guanine the –NH2 groups are rather non-planar. Also there may be subtle differences in the ring structures between geometry optimized structures, and the structures found in the hydrogen bonded network of a single crystal. To study the effects of using non-optimized structures, calculations have been performed on several pyrimidine reduction products observed in the single crystal systems mentioned above. The geometries were taken from the single crystal x-ray diffraction study on the parent compound. Best results were obtained by using B3LYP/6-31+G(d). The calculated results shown in Table II are for essentially planar geometries, and correctly account for the observed C6-H hyperfine coupling. 36 Conference on Current Trends in Computational Chemistry 2003
Table II. Calculated Results Using Structures from Single Crystal X-Ray Analysis Crystal (C6-H" Experimental Theoretical (in MHz) Refs. Coupling) Thymidine -33.1 -33.2 [11,13] Thymine (Anhydrous) -39.8 -38.4 [12,14] 5’-UMP -36.1 -35.7 [15,16] Co-crystal MU:EA 39.1 -38.8 [17,18]
Acknowledgements: This work is supported by PHS Grant CA36810-16 awarded by the National Cancer Institute, DHHS. Thanks to Leonid Gorb (JSU) for helpful discussions.
References: 1. J. Geimer, O. Brede, and D. Beckert, Chem. Phys. Letts., 1997, 276, 411. 2. J. Geimer and D. Beckert, J. Phys. Chem., 1999, 103, 3991. 3. J. Geimer, K. Hildenbrand, S. Naumov, and D. Beckert, Phys. Chem. Chem. Phys., 2000, 2, 4199. 4. S. Naumov, A. Barthel, J. Reinhold, F. Dietz, J. Geimer, and D. Beckert, Phys. Chem. Chem. Phys., 2000, 2, 4207. 5. J. M. Lh, J. Geimer, S. Naumov, and D. Beckert, Phys. Chem. Chem. Phys., 2001, 3, 952. 6. S. Naumov, J. Reinhold, and D. Beckert, Phys. Chem. Chem. Phys., 2003, 5, 64. 7. D. M. Close, ‘Model Calculations of Radiation Induced Damage in DNA Constituents Using Density Functional Theory’, Computational Chemistry, Vol.8. ed. J. Leszczynski, Singapore, 2003. 8. D. M. Close, Phys. Chem. Chem. Phys., 2001, 3, 43. 9. S. Naumov and D. Beckert, Phys. Chem. Chem. Phys., 2002, 4, 43. 10. W. A. Bernhard and A. Z. Patrzalek, Radiat. Res., 1989, 117, 379. 11. E. O. Hole, E. Sagstuen, W. H. Nelson, and D. M. Close, J. Phys. Chem., 1991, 95, 1494. 12. E. Sagstuen, E. O. Hole, W. H. Nelson, and D. M. Close, J. Phys. Chem., 1992, 96, 1121. 13. D. W. Young, P. Tollin, and H. R. Wilson, Acta Cryst., 1969, B25, 1423. 14. K. Ozeki, N. Sakabe, and J. Tanaka, Acta Cryst., 1969, B25, 1038. 15. H. C. Box, W. R. Potter, and E. E. Budzinski, J. Chem. Phys., 1975, 62, 3476. 16. E. Shefter and K. N. Trueblood, Acta Cryst., 1965, 18, 1067. 17. E. Sagstuen, E. O. Hole, W. H. Nelson, and D. M. Close, Radiat. Res., 1998, 149, 102. 18. F. S. Mathews and A. Rich, J. Mol. Biol., 1964, 8, 89. Conference on Current Trends in Computational Chemistry 2003 37
Conventional Strain Energy in Small Heterocycles of Carbon and Silicon
Crystal B. Coghlan, Shelley S. Huskey, and David H. Magers
Department of Chemistry and Biochemistry Mississippi College, Clinton, Mississippi
The conventional strain energies for three- and four-membered heterocycles of carbon and silicon are determined within the isodesmic, homodesmotic, and hyperhomodesmotic models. Optimum equilibrium geometries, harmonic vibrational frequencies, and corresponding electronic energies are computed for all pertinent molecular systems using SCF theory, second- order perturbation theory (MP2), and density functional theory. The DFT functional employed is Becke’s three-parameter hybrid functional using the LYP correlation functional. Two basis sets, both of triple zeta quality on valence electrons, are employed: 6-311G (d,p) and 6- 311+G(2df,2pd). Finally, the calculated strain energies are compared to those of cyclopropane, cyclobutane, oxaziridine, azetidine, phosphetane, diazetidine, and diphosphetane.
We gratefully acknowledge support from NSF EPSCoR (EPS-0132618). 38 Conference on Current Trends in Computational Chemistry 2003
Study of Amorphous Silicon Dioxide Clusters for Use in an Additive Polarizability Model
Lonnie D. Crosby and Henry A. Kurtz
Department of Chemistry, University of Memphis, Memphis, TN 38152-3550
Computational Study of amorphous silicon dioxide is of great benefit in research of MOS systems and development of new high-k dielectrics. In the development of high-k dielectrics, the property of interest is the dielectric constant. The dielectric constant is a bulk property that is a function of two molecular properties. These properties are the molar alpha or linear polarizability and the molar volume of the material. The Clausius-Mossotti equation gives the fundamental basis for this relationship. Using this approach, it should be possible to predict bulk dielectric constants via calculations that can quantify the molar polarizability and molar volume of materials of interest. The materials of initial interest are mixed-oxides of silicon. This is done as a starting point, since silicon dioxide is already employed as a gate dielectric. Also, such materials would be highly compatible with current technology. Initially and throughout this study attention is given to zirconium-silicon mixed oxides. Approaches to quantify the molar polarizability of zirconium-silicon mixed oxides will be explored. This is done via ab initio quantum mechanical calculations using terminated cluster models. The main drawback to using cluster models is their obvious non-representative nature. It cannot be expected that a small cluster model of these systems would yield accurate properties. Especially, since these systems are amorphous in nature and form a continuous random network. To account for this deficiency in the models chosen, an additive polarizability model is developed and used to eliminate non-representative contributions to the total polarizability of these models. In the development of an additive polarizability model, the major assumption that has to be made is that the total polarizability is an intensive quantity which can be partitioned into regions of contribution. Equation 1 shows the mathematical formulation of this assumption. N αtotal = ∑ Piαi Eqn. 1 i Where alpha denotes the polarizability, P is the integer which defines the number of occurrences of a particular parameter, the index i denotes the contribution from the ith parameter in the model, and N denotes the total number of parameters incorporated into the model. It is apparent from this equation that the polarizability contributions are highly dependent on the parameterization scheme used to obtain them. Also evident, is the fact that this equation has N unknowns and therefore requires at least N linearly independent equations to be solved. This last observation can transform this equation into a matrix equation, as shown in equation 2. C ∑ Pijα j = αi Eqn. 2 j Where the indices i and j denote the ith cluster and parameter respectively. The symbols P and alpha represent similar quantities as in equation 1. Also, the symbol C is the number of clusters used in the fit. To solve this matrix equation, the most obvious route is to multiply both sides by matrix P, as shown in equation 3. N −1 α j = ∑()P jiαi Eqn. 3 i Conference on Current Trends in Computational Chemistry 2003 39
Therefore, the ability to solve this matrix is dependent on the ability to take the inverse of the P matrix. This is often more difficult than it appears to be. The standard inverse only exists if P is a square matrix, C = N. Furthermore the matrix must be non-singular or equivocally be linearly independent. In the case of these systems, P most definitely is to some extent singular. This requires another method in order to take the inverse. The method employed takes the inverse of this matrix by first decomposing it into a product of two unitary matrices and one diagonal matrix. In the diagonal matrix, singularities show up as zero’s or near zeros in any one of the elements. These singularities are removed by setting the inverse of such elements to zero. The decomposition and inverse are shown in equation 4. Q T Pij = ∑(U )ir DrrU rj r Q −1 T −1 (P ) ji = ∑(U ) jr Drr U ri Eqn. 4 r This method turns out to be equivalent to a least squares fit if C > N. Therefore, a solution can be found and a root mean square error of that solution can be determined. This realization can give a quantitative measure of the appropriateness of any such parameterization as long as the number of clusters available is greater than the number of parameters in the model. To utilize this method, several hydrogen terminated clusters of silicon dioxide, and zirconium doped silicon dioxide are prepared. The polarizabilities of these clusters are then calculated using ab initio methods. Once the polarizabilities of the clusters are determined, a parameterization scheme is devised and the above method is employed to calculate contributions from those parameters. Several such parameterization schemes were used, such as atom centered and bond centered schemes. Due to the properties of the cluster models, it is necessary to include nearest neighbor interactions in the parameterization scheme. This ensures that each parameter only corresponds to a single contribution as opposed to an average over many potentially different contributions. A side effect of this nearest neighbor scheme is that it can divide the parameters into those used for bulk predictions, and those that are dependent on the choice of terminator. If you assume that those contributions from the terminators of each cluster are non-representative of the bulk system then this is very beneficial. In order to affectively use an additive polarizability model, however, it is necessary to determine what is or is not representative in these terminated clusters. The non-representative portions may not be confined to the single atom terminators. The appropriate parameterization scheme in such a polarizability model is mostly determined by these representative and non- representative portions of the cluster models. Since the terminated clusters used can be thought of as zirconium doped silicon dioxide, this portion of the study makes use of silicon dioxide clusters exclusively. A series of terminated silicon dioxide clusters of differing size, structure, and termination are employed for this purpose. Ab initio techniques are employed on these clusters to gain insight into their electronic structure. By comparing the changes in electronic structure as the size and terminators are changed, it is possible to determine the extent to which the cluster is a suitable model for the bulk system. In the determination of the properties of these terminated clusters, it is the goal of this investigation to suggest a parameterization scheme for an additive polarizability approach. In order to achieve this goal it is necessary to answer some fundamental questions. These include: • What are the necessary properties for terminators? • What are the effects of termination on the cluster? • How much of a given cluster is representative of the bulk? • Is it possible to maximize this representative portion by a different choice of terminator? 40 Conference on Current Trends in Computational Chemistry 2003
The first two questions are answered via some suggestions taken from a hydrogen terminator scheme and expanded upon taking into account known properties of the bulk system. The arguments used follow some fairly elementary considerations, such as steric, electronic, and inductive interactions between the clusters and their terminators. The remaining questions are explored by performing ab initio quantum mechanical calculations on a set of clusters composed from differing termination schemes as suggested by taking into account the mentioned interactions with the cluster. It is the results of these last two questions which will form the basis of the suggested parameterization scheme. At this point, some assumptions are made to simplify the amorphous nature of these terminated clusters and interpretation of the results possible. This is achieved by using only symmetrically built clusters in the calculations and also by reducing the amorphous properties of these clusters into average properties with associated error limits. It is within these limits of error that the amorphous dependency is collapsed. The property of interest in this study is the electronic structure. Lowdin population analysis has been implemented to give partial charge distributions of these clusters and forms a basis to quantify the average electronic structure. This basis is viewed with the assumption that the terminators of a cluster are the sole contributing factor in the formation of non-representative partitions. The mechanism that is responsible for this partitioning is the inductive effects on the cluster. This assertion is made on the premise that all other interactions have been taken into account and minimized in the choice of termination scheme. The inductive interactions are measured indirectly in a comparative technique in which similar clusters with differing termination are examined. The change in the portions of the electronic structure relative to the terminator change is compared to changes in the cluster portions. This gives a shielding constant that describes the dependence of the cluster’s electronic structure on termination scheme. This is used to define a limit in which terminator dependence approximately vanishes and the cluster can be considered representative. When this method is applied to the sample of terminated clusters involved in the study, the answers to the remaining questions can be answered. The results of this study will show that a hydrogen type termination scheme yields clusters with larger representative portions than do more electronegative halide type schemes. It also suggests that only the first and second nearest bonded neighbors to this terminator is affected by inductive interactions. Furthermore, these results are very consistent across the sample of hydrogen terminated clusters used in the study, as well as, a cluster not present in the study. Using these results, a parameterization scheme is suggested that takes into account the non-representative nature of the first and second nearest neighbors to the hydrogen terminator. The results from the additive polarizability model strongly suggest that the non- representative portions are not confined to the hydrogen terminators. In fact, the fits are improved if the suggested parameterization scheme is used. Also, in general, it will be shown that an additive polarizability model is sufficient in quantizing bulk susceptibilities. Conference on Current Trends in Computational Chemistry 2003 41
The Accurate Calculation of Ro-Vibrational Eigenenergies of HOD
Jennifer L. Curry and Joseph A. Bentley
Department of Physical Sciences, Delta State University, Cleveland, MS, 38733
We report the accurate calculation of ro-vibrational eigenenergies of the ground electronic state of the HOD molecule. The Radau coordinate system (R1, R2, θ) is employed. The discrete variable representation (DVR) [Z. Bacic and J. C. Light, Annu. Rev. Phys. Chem. 40, 469 (1989)] is used as a basis for both radial coordinates – this leads to a sparse Hamiltonian matrix. A primitive angular basis set is constructed which diagonalizes the rotational (J > 0) kinetic energy operator; subsequently, this basis is contracted through a series of diagonalizations of smaller Hamiltonian matrices. The final Hamiltonian matrix is then constructed out of this contracted angular basis set. The eigenvalues of this matrix are then obtained by using the Implicitly Restarted Arnoldi Method (IRAM) which is part of a recently developed numerical package (ARPACK) designed to solve large scale eigenvalue problems.
42 Conference on Current Trends in Computational Chemistry 2003
Correlation of Dynamic Calculations on Chiral Nonlinear Optical Molecules with Measurements by Hyper-Rayleigh Scattering(HRS) and Conformational Effects
Y. Daoudia, P.J. Bonifassib
a)Institut de chimie industrielle, Université des Sciences et de la Technologie Houari Boumedienne, Bab Ezzouar, Alger, Algérie. b)Laboratoire de synthèse organique, Faculté des sciences, Université du Maine, Avenue Olivier Messiaen, Le Mans, 72017, France.
The interpretation of the noncoherent HRS experiment has the ability to connect microscopic molecular properties such as the β tensor to their macroscopic manifestation (1). Thus, the measurements of the components of the hyperpolarizability tensor using Hyper- Rayleigh scattering are compared with the Time Dependent Hartree Fock (TDHF) (2,3) dynamic hyperpolarizabilities computations. Indeed, a pseudo tensor contribution to the β tensor in chiral dyes can be expected to lead to macroscopic second harmonic generation in poled polymer materials axialled aligned by Corona effect (1). According to group theory, this tensor can be described in part by rotationally invariant scalar figures of merit and this hyperpolarizability tensor β can be decomposed into four rotationally irreducibles tensor parts : β1ss, β1mm, β2mm and β3ss. We have in cartesian representation of the rank 3, the irreducible form of β: β =β(3s)+β(2m)+β(1s)+β(1m) ijk ijk ijk ijk ijk where the superscript indicates the tensor rank and the permutation symmetry(s=symmetric and m=mixed symmetry).Each of these components can be obtained by HRS experimental technique with an elliptically polarized laser beam to the 1340 nm wave length. The studied molecules (4,5,6) are showed below (Fig 1). The full optimized geometries were obtained by DFT-BP86-6-31G** techniques. In this work, we have studied the correlation between the sum of these 4 components and the theoretical β values obtained by TDHF computations with DFT-BP86-6-31G** geometry to 1340 nm(0.918eV) included in the table below. We obtain an acceptable correlation and we remark a conformational effect between the conformations A and B in all the cases.
Table 1 Measurements of the components Method for TDHF calculation of the Hyperpolarizability tensor full optimized SHG β i-30 esu to 1340 nm geometry β in 10-30 esu β 1ss β 1mm β 2mm β 3ss β sum to 1340 nm AQNH2 DFT/BP86 25,7 12 0 3 0 15 BQNH2 DFT/BP86 28,6 MolA DFT/BP86 171 78 26 0 49 153 MolB DFT/BP86 71 45 10 15 31 101 AQBNH DFT/BP86 28,9 16 0 13 0 29 BQBNH DFT/BP86 36,9 AQKCH DFT/BP86 46,5 35 0 10 0 45 BQKCH DFT/BP86 28,2 Conference on Current Trends in Computational Chemistry 2003 43
Acknowledgements: We acknowledge P.Delage and S.Bourdais for full access to computers.
References 1. S.Brasselet and J.Zyss, J.Opt.Soc.Am.B, Vol 15, 1, (1998) 257-288. 2. H.Sekino and R.Bartlett, J.Chem.Phys, 98, 4, (1993) 3022-3037. 3. S.P.Karna and M.Dupuis, Journal of Computational Chemistry, Vol 12, 4, (1991) 487-504. 4. S.F.Hubbard, R.G.Petschek and K.D.Singer, J.Opt.Soc.Am.B, Vol 15, 1, (1998) 289-301. 5. V.Ostroverkhov,O.Ostroverkhova, R.G.Petschek and K.D.Singer, Chemical Physics, 257, (2000) 263-274. 6. S.F.Hubbard, R.G.Petschek and K.D.Singer, J.Opt.Soc.Am.B, Vol 17, 9, (2000) 1531-1542.
CONFORMATION A
R NH2 H C 3 H3C R = C(CN)2 1 Mol A C X = CH 6 X H3C 2 R = O AQNH2 3 5 X = N 4 CONFORMATION B
R NH2 H C 3 R = C(CN)2 Mol B 6 1 C X = CH X H3C 2 O 3 R = BQNH2 5 X = N 4 CH3
H O N H3C H C 3 1 C 6 N N H3C 2 3 5 4 O AQBNH : conformation A : CH3 group on atom 6 H3C CH3 BQBNH : conformation B : CH3 group on atom 3 H3C
C O C C C C C C C C
O N H3C H C 3 1 C C C C C C C 6 O C C X C H3C 2 3 5 AQKCH : conformation A : CH3 group on atom 6
4 BQKCH : conformation B : CH3 group on atom 3
Figure 1 44 Conference on Current Trends in Computational Chemistry 2003
Conformational Studies of Triphenylmethyl Formate and N- Triphenylmethyl Formamide
Dalephine Davis, Diwakar Pawar, and Eric A. Noe
Department of Chemistry, Jackson State University, Jackson, MS 39217
Triphenylmethyl formate was synthesized from dry potassium formate and triphenylmethyl bromide in the presence of toluene. The 1H NMR spectrum of a 5% sample of triphenylmethyl formate in acetone-d6 showed formyl protons at 8.63 ppm for the E and 8.43 ‡ ‡ ppm for the Z at -105.3°C. ∆G Z→E and ∆G E→Z were determined to be 8.92 and 8.69 kcal/mol, respectively, at -93.4°C by lineshape matching. Ab initio calculations were carried out for triphenylmethyl formate. At the HF/6-31G* level, the E conformation was calculated to be 0.38 kcal/mole higher in free energy than the Z, and a Z-to-E free-energy barrier of 6.30 kcal/mole was calculated. Free-energy barriers to rotation about the bond between oxygen and the triphenylmethyl group were also calculated. The 13C NMR spectrum of a 1% sample of N-triphenylmethyl formamide in 50:50 1 CH2Cl2/CD2Cl2 showed two carbonyl peaks at 166.1 ppm (E) and 160.6 ppm (Z). The H spectrum of N-triphenylmethyl formamide showed two doublets for the formyl protons. The doublet from the formyl proton cis to the NH proton (Z conformation) had a smaller coupling constant and absorbed at lower field strength than the trans formyl proton of the E conformation. A population of 37.4% was found for the Z conformation for a 1% solution of N-triphenylmethyl formamide in CH2Cl2/CD2Cl2. This work was supported by NSF-CREST Grant No. HRD – 980 5465. Conference on Current Trends in Computational Chemistry 2003 45
Capillary Electrophoretic Separation and Theoretical Study of Inclusion Complexes of Sulfobutyl Ether β-Cyclodextrin with Estrogens
1 2 Yuanjian Deng and Ming-Ju Huang
1 Department of Chemistry, Texas Southern University, 3100 Cleburne Ave., Houston, TX 77004 2 Computational Center for Molecular Structure and Interactions, Department of Chemistry, Jackson State University, 1325 J.R. Lynch Street, Jackson, MS 39217
Estrogens are sex hormones produced in the female body and have been used as therapeutic drugs to treat hormonal deficiency diseases. The analysis of estrogens is important for the diagnosis of reproductive diseases and for monitoring the normal pregnancy process. The complete separation of five estrogens (2-hydroxyestrone, 2-methoxyestradiol, estrone, estriol, and 17α -estradiol) and two stereoisomers (17α- and 17β-estradiol) has been achieved using capillary electrophoresis with anionic sulfobutyl ether β-cyclodextrin (SBE-β-CD) [1]. The SBE-β-CD used in the separation was a mixture of positional and regional isomers containing from one to ten sulfobutyl ether groups with an average degree of substitution of four per cyclodextrin (CD); thus, a tetra-SBE substituted CD, SBE[4]-β-CD, has been chosen as the model host molecule in the present study. With SBE[4]-β-CD as the buffer additive, the migration times of each estrogen in capillary electrophoresis reflect the stability of their SBE-β- CD inclusion complexes. In the present study the four SBE groups are assumed to be on the 6 position of the glucose moieties of β-CD. Initially three arrangements which are based on the relative orientation between the sulfobutyl chain and the axis of the cavity of β-CD are considered: four SBE parallel, four SBE perpendicular, and two SBE parallel and two SBE perpendicular to the axis of the cavity. In addition, due to the unique structure of SBE[4]-β-CD, two types of inclusion complexes, cis- and trans-orientations, are possibly formed with each estrogen. In the cis- structure the SBE group of the cyclodextrin and the phenol group of the estrogen are on the same side, while in the trans- structure the two groups are on opposite sides. Unlike other coordination compounds, no covalent bonds exist between the CD and its guest molecules; complexation is a dynamic process. It is thus reasonable to assume that each estrogen-CD complex consists of a mixture of cis- and trans-orientations. The objective of the present study is to perform PM3 semi-empirical molecular orbital calculations on the six inclusion complexes (12 different orientations) for each SBE[4]-β-CD to correlate the migration order with their relative stability [2]. Theoretical calculations have shown that only when all four SBE chains are perpendicular to the axis of CD, the two openings of the cavity are completely open for incoming guest molecules. A correlation has been obtained between the stability of the estrogen-CD inclusion complexes and the electrophoretic migration time, although the relationship is not linear. The PM3 optimized geometries for the other two arrangements of SBE on the CD indicate that the smaller opening of the cavity is partially capped because of the formation of intra hydrogen bonds among three of the four SBE groups. This capped SBE[4]-β-CD host may prevent a guest molecule from entering through the capped opening; thus a guest molecule can enter the cavity from the larger opening of the cavity only. As a result the estrogen molecules protrude from the cavity. Thus the heats of formation for PM3-optimized geometries of the inclusion complexes have no correlation at all with the migration time. 46 Conference on Current Trends in Computational Chemistry 2003
O OH O CH3 CH3 CH3
HO CH3O
HO 2-Hydroxyestrone HO 2-methoxyestradiol HO Estrone
OH CH OH CH OH CH3 3 3
OH
HO Estriol HO 17α-Estradiol HO 17β-Estradiol
References 1. Deng, Y. J., Zhou, J. X., Perkins, M. D., and Lunte, S. M., Anal. Commun., 1997, 34, 129- 131. 2. Huang, M.-J., Watts, J. D., and Bodor, N., Int. J. Quant. Chem., 1997, 65, 1135-1152.
Conference on Current Trends in Computational Chemistry 2003 47
The Correlation between Impact Sensitivity and Heat of Detonation in a Unique Class of Energetic Materials
Claudia Eybl1*, Brian Johnson2, Jesse Edwards3
1College of Pharmacy and Pharmaceutical Sciences/AHPCRC 2Computer Information Systems/AHPCRC 3Department of Chemistry/AHPCRC Florida A&M Tallahassee, Florida, USA 32307
Current computational capabilities and advances in quantum theory are recognized as cost-effective and time-saving. Theoretical screening allows for the identification of promising candidates for further study and the elimination of poor candidates. Quantum mechanical calculations of individual energetic molecules result in the prediction of structures, stabilities, vibrational spectra and other properties. In this study, we describe the use of the Density Functional Theory predictions of heats of formation of explosives in calculating their heats of detonation, which is used to assess a candidate’s detonation performance. We will present correlations between the measured impact sensitivity and the approximated heats of detonation.
48 Conference on Current Trends in Computational Chemistry 2003
New Alternatives for Accurate ab Initio Calculations
Peng-Dong Fan, Karol Kowalski, Maricris Lodriguito, and Piotr Piecuch
Department of Chemistry, Michigan State University, East Lansing, Michigan, 48824
The applicability of exponential cluster expansions involving one- and two-electron operators in highly accurate ab initio calculations is discussed. First, an evidence is presented that one may be able to represent the exact or virtually exact ground-state wave functions of many-electron systems by exponential cluster expansions employing general two-body operators [1,2]. Calculations for a few many-electron systems indicate the existence of finite two-body parameters that produce numerically exact wave functions. This result may have a significant impact on future quantum calculations for many-electron systems, since normally one needs triply excited, quadruply excited, and other highly excited Slater determinants in addition to all singly and doubly excited determinants to obtain the exact or virtually exact wave functions. Another two-body theory, the extended coupled-cluster method with singles and doubles (ECCSD), is tested in the most demanding studies of the potential energy surfaces involving multiple bond breaking. The numerical results for the triple bond breaking in N2 show that the single-reference ECCSD method is capable of providing a qualitatively correct description of the entire potential energy curve of N2, eliminating, in particular, the failures and unphysical behavior of all standard coupled-cluster methods in similar cases [2,3]. It is also demonstrated [4] that one can obtain the entire potential energy curve of N2 with a millihartree accuracy by combining the ECCSD theory with the noniterative a posteriori corrections obtained using the method of moments of coupled-cluster equations (MMCC) [4,5]. This is the first time when the simple single-reference formalism employing one- and two-body clusters only provides a highly accurate description of the dynamic and significant nondynamic correlation effects.
1. P. Piecuch, K. Kowalski, P.-D. Fan, and K. Jedziniak, Phys. Rev. Lett. 90, 113001 (2003). 2. P. Piecuch, I.S.O. Pimienta, P.-D. Fan, and K. Kowalski, in: Progress in Theoretical Chemistry and Physics, Vol. 12, Topics in Theoretical Chemical Physics, edited by Jean Maruani, Roland Lefebvre, and Erkki Brändas (Kluwer, Dordrecht, 2004), p. 119-206. 3. P. Piecuch, I.S.O. Pimienta, P.-D. Fan, and K. Kowalski, in: Recent Progress in Electron Correlation Methodology, ACS Symposium Series, Vol. XXX, edited by A.K. Wilson (American Chemican Society, Washington, D.C., 2003), p. XX-XXX. 4. P. Piecuch, K. Kowalski, I.S.O. Pimienta, P.-D. Fan, M. Lodriguito, M.J. McGuire, S.A. Kucharski, T. Kuś, and M. Musiał, Theor. Chem. Acc., submitted. 5. K. Kowalski and P. Piecuch, J. Chem. Phys. 113, 18 (2000); P. Piecuch and K. Kowalski, in: Computational Chemistry: Reviews of Current Trends, edited by J. Leszczynski (World Scientific, Singapore, 2000), Vol. 5, pp. 1-104; P. Piecuch, K. Kowalski, I.S.O. Pimienta, and M.J. McGuire, Int. Rev. Phys. Chem. 21, 527-655 (2002). Conference on Current Trends in Computational Chemistry 2003 49
Exploring the Electronic Structure of Novel Cannabinoid Derivatives: New Approaches to Rational Drug Design
Antonio M. Ferreira a, Bob M. Moore, II b, and Henry A. Kurtz a
a Computational Research on Materials Institute, Department of Chemistry, The University of Memphis, Memphis, Tennessee 38152, USA. b Department of Pharmaceutical Sciences, College of Pharmacy, University of Tennessee Health Science Center, Memphis, Tennessee 38103, USA.
The cannabinoids are a diverse class of compounds that have gained considerable attention due to their potential indication in the treatment of obesity, cancer, glaucoma and several other pathological conditions. These compounds exert their biological effects via interaction with the cannabinoid receptor 1 (CB-1) and cannabinoid receptor 2 (CB-2). The differential expression of the receptor subtypes in cells and tissues has provided a set of unique targets for the treatment of a variety of diseases. To exploit the CB-1 and CB-2 receptor biochemical pathways, a diverse set of chemical entities have been synthesized in efforts to develop receptor subtype selective agents. These compounds can be broadly grouped into four distinct chemical classes, the classical cannabinoids, non-classical cannabinoids, alkyl amino indoles, and pyrazoles. The classical cannabinoids are, by far, the most extensively studied group in terms of structure activity relationships and pharmacology. Considerable efforts have been made to elucidate the SAR of these compounds with respect to modifications of the B and C rings as well as the side chain substituents. To assess the latter, researchers have synthesized a variety of analogs that have incorporated substituents such as branched chain alkyls, unsaturated alkyls and alkyls containing 1', 1' cyclic functionality. To date no X-ray or NMR structure of either the CB-1 or CB-2 receptors has been reported. However, this information is essential to better understand the structural requirements of the ligand-binding pocket and allow for the continued development of the classical cannabinoids. In the absence protein structural data researchers must utilize site directed mutagenesis, molecular modeling studies and traditional medicinal chemistry to provide insights into the geometric and chemical requirements of the ligand-binding pocket (LBP). As part of our efforts to characterize the structural requirements of the CB-1 and the CB-2 LBP, we had previously synthesized a series of cycloalkyl side chain analogues of ∆8–THC. The series consisted of cyclopentyl, hexyl and heptyl side chains with 1',1'–dimethyl or dithiolane functionalities. The 1',1'–dimethyl series exhibited, on an average, a 100-fold increase in binding affinity relative to ∆8-THC for both receptor subtypes. In an extension of this series of analogs, we later synthesized as set of C-1' substituted phenyl derivatives that contained either a dimethyl (KM-223), dithiolane (KM-222), methylene (KM-224) and ketone (KM-233) at the C-1' position.
50 Conference on Current Trends in Computational Chemistry 2003
OH OH
O O SS
KM-222 KM-223
OH OH
O O O
KM-224 KM-233
We hypothesized that the introduction of a phenyl side chain could significantly alter the electronic properties of both the side chain and A ring while maintaining steric bulk and conformational restraint. Furthermore, this substitution was also designed to study potential π-π interactions of the ligands with aromatic amino acids in the LBP. Hypothesizing that there was an interaction between the π-electron density associated with Ring A and some residue in the binding pocket, the goal was then to develop a new class of cannabinoid derivatives that would provide an electronic mechanism to stabilize this π-electron density. Although there is no classical π-overlap between Ring A and the front-side phenyl group, the hope was that interactions similar to those seen with cyclohexyl derivatives would lead to interaction between the rings. In an attempt to understand the relationship between the electronic structure of this important class of compounds and their interactions with the CB-1 and CB-2 receptors, we have carried out a series of semi-empirical and density functional theory (DFT) calculations. Potential energy surfaces for each of these novel compounds were calculated using the PM3 semi- empirical Hamiltonian and show a wide range of conformational freedom for all the compounds in the study. The electronic structure for each of these derivatives has been studied using B3LYP/6-31G(p,d) calculations. Effects of the electronic structure on NMR experiments have been analyzed and clearly demonstrate that an understanding of these molecules requires a detailed description of their electronic structure. Additionally, we show that there is significant potential for these compounds to exhibit fluorescence, making them possible candidates for probing receptor sites in vivo. Conference on Current Trends in Computational Chemistry 2003 51
The Distribution of Water Geometries about Polar Surface Residues in Proteins, as Studied by Molecular Dynamics
Eric W. Fisher
Department of Chemistry, University of Illinois at Springfield Springfield, Illinois 62703-5404
Despite the importance of water in mediating protein dynamics and protein-ligand recognition, and the large numbers of water molecules in protein crystals and solvated protein complexes, many individual water molecules potentially playing important chemical and physical roles in protein function remain uncharacterized. In crystallographic models developed by refinement of x-ray data, several water molecules are sufficiently tightly bound that their locations can be identified, but the locations of more distant waters are lost. In solution NMR, all but a few water molecules exchange rapidly enough with bulk solvent that the locations of waters near the surface of the protein are identified even more rarely. Currently, the best ideas of the locations and orientations of water molecules about proteins are derived from spectroscopic data indicating their geometric distributions, and from molecular dynamics simulations. Unfortunately, despite the intricate molecular detail provided by molecular dynamics, no well characterized code of water geometry about protein surfaces has been developed to date. The present investigation examines the geometry of interaction between water molecules and polar groups on model protein surfaces, and between water molecules interacting with one another near these surfaces. Polar groups, such as the amide tails of the sidechains in the amino acid residues asparagine (Asn) and glutamine (Gln), interact directly and specifically with individual water molecules in sufficiently restrained geometries that these interactions may be possible to categorize in quantitative detail. The survey of distributions of water geometries about Asn and Gln sidechains was conducted by first constructing a model peptide system, in which the B- and C-helices of the four-helix bundle in the granulocyte colony-stimulating factor (GCSF), derived from the NMR structure 1GNC (Protein Data Bank), were employed as an α- helical hairpin which would maintain a stable conformation in molecular dynamics simulations. The sidechains of all amino acid residues in the hairpin were modified to strengthen the interactions between the helices and place evenly spaced, non-interacting Asn and Gln residues along the helices and pointing into the bulk solution. The interactions between helices were optimized by placing hydrophobic sidechains between the helices, and the termini of the helices were joined by a disulfide bond between two cysteine residues. One nanosecond of molecular dynamics simulation was carried out in a box of water molecules, using the molecular dynamics software package NAMD2 to ensure that the helical hairpin structure was stable; then, geometric configurations of water molecules in spheres of hydration about each Asn and Gln amide tail were calculated, and distributions of hydrogen-bond lengths and angles were constructed at regular points along the dynamics trajectories. These distributions were compared with the known crystalline structures of ice, and regions of geometry corresponding most closely with a particular isoform of ice were identified. The thorough description of dynamic water geometry about protein surfaces will assist in developing more robust structural models, in which water molecules can be reliably included in higher-level quantum calculations where hydration plays key structural or functional roles. 52 Conference on Current Trends in Computational Chemistry 2003
The Effects of Metalation on Methylated Adenine Thymine Watson- Crick Base Pair
Aviane Flood, Jacquelin McCuller, Gareth Forde, Glake Hill, and Jerzy Leszczynski
Computational Center for Molecular Structure and Interactions Jackson State University, Jackson, MS 39217
Mg2+ is a common ion in biological systems. One of the major roles of magnesium ions in nature is stabilizing DNA in the active site of enzymes. In addition, divalent magnesium ions have also been found to bind to DNA through the formation of coordinate bonds. In this investigation, we have performed ab initio calculations to explore the binding of Mg2+ ions to Watson-Crick A:T base pairs. Additionally, since DNA is ubiquitously being methylated, both endogenously and exogenously, the synergistic effects of these processes have been assessed. The geometries of the local minima were optimized without symmetry restrictions by the gradient procedure at Hartree-Fock (HF) levels of theory and were verified by energy second derivative calculations. The standard 6-31G* basis set was used. The single-point calculations have been performed at the DFT/6-31G(d,p) and MP2/6-31G(d,p) levels of theory. The geometrical parameters and counterpoise corrected interaction energies are reported.
Conference on Current Trends in Computational Chemistry 2003 53
MP2 Computational Study of van der Waals Interactions Between Graphene Sheets
Alan Ford and Peter Pulay
Department of Chemistry and Biochemistry, University of Arkansas, Fayetteville, AR 72701
Systems of parallel graphene sheets can exhibit a considerable amount of dispersion energy, which can be modeled computationally with some difficulty. Large graphene sheets should be used, but the relatively inexpensive density functional theory (DFT) is a poor model for van der Waals forces. Instead, more expensive methods such as MP2 that model van der Waals forces better must be used at the cost of the model size. As a result, coronene dimer and supercoronene dimer are used in the present study with MP2 theory and medium basis sets. Some MP2 potential results of coronene dimer are shown below. For the calculations, the planar coronene molecules are vertically separated by a fixed distance of 3.3504 Å. The lowest potential occurs at a horizontal displacement of 1.4 Å, which compares well with experimental measurements on graphite. Calculations have also been performed on supercoronene dimmer, which gives similar results. Furthermore, scaled MP2 energies [1] have been acquired for coronene dimer.
Interaction Energy vs. Horizontal Displacement for Coronene Dimer
-0.025
-0.027
-0.029
-0.031 ) h -0.033
-0.035
-0.037 Interaction Energy (E
-0.039
-0.041
-0.043
-0.045 00.511.522.533.5 Displacement (Å)
1. Grimme, S. J. Chem. Phys. 118, 2003, 9095.
54 Conference on Current Trends in Computational Chemistry 2003
Comprehensive Study of the Effects of Methylation on Tautomeric Equilibria of Nucleic Acid Bases
Gareth Forde1, Aviane Flood1, Angela Fortner1, Adrian Ford1, Alejandro Nazario2, Curinetha Hubbard1, Glake Hill1, Leonid Gorb1, and Jerzy Leszczynski1*
1Computational Center for Molecular Structure and Interactions Jackson State University, Jackson, MS 39217 2Department of Biology, Universidad Metropolitana, Rio Piedras, Puerto Rico
Minor tautomers of nucleic acid bases can result from the inter-molecular displacement of a proton. We believe that these rare tautomers could be stabilized through the covalent addition of methyl groups to DNA bases. Methylation of DNA occurs most readily at N(3), N(7), and O(6) of purine bases and N(3), O(4) and O(2) of pyrimidines. The results of a theoretical ab initio study on the effects of methylation on the stability of rare tautomers of the pyrimidines, cytosine, uracil, and thymine, and the purines, guanine and adenine are presented. The geometries of the local minima were optimized without symmetry restrictions by the gradient procedure at DFT and MP2 levels of theory and were verified by energy second derivative calculations. The standard 6-31G(d,p) basis set was used. The single point calculations were carried out at CCSD(T) level of theory with the 6-311G(d,p) basis set. For the MP2 and CCSD( T) predictions, complete basis set estimates were obtained extrapolation techniques. The relative stability and relevant geometrical parameters are reported.
*Corresponding Author Conference on Current Trends in Computational Chemistry 2003 55
Active Site-Inhibitor Modeling Using a Customized HIV-Protease Polypeptide
1Jason Ford-Green, 2Deborah Bryan, 2Jesse Edwards, 2John West, 3Ben M. Dunn
1Department of Biology/AHPCRC 2Department of Chemistry/AHPCRC Florida A&M University Tallahassee, Florida, USA 3Department of Molecular Biology and Biochemistry, University of Florida, Gainesville, Florida, USA 32608
In an effort to develop unique HIV protease inhibitors Dunn et al. have synthesized customized polypeptides. A series of these polypeptide inhibitors is capable of adapting to mutations in the HIV protease. Using molecular modeling we will explore whether these inhibitors induce a stable conformation in the HIV protease. In particular, quantum mechanics, and molecular mechanics methods will be used to accomplish this task. This work will discuss those results.
56 Conference on Current Trends in Computational Chemistry 2003
Theoretical Study of Adsorption of Methyl-Cytosine on Dickite
A. D. Fortnera, A. Michalkovaa,b, J. Leszczynskia
aComputational Center of Molecular Structure and Interactions, Department of Chemistry, Jackson State University, 1400 J. R. Lynch Street, P. O. Box 17910, Jackson, MS 39217, USA bInstitute of Inorganic Chemistry, Slovak Academy of Sciences, Dubravska cesta 9, 842 36 Bratislava, Slovak Republic
The adsorption of methyl-cytosine (MC) on the hydrated (two water molecules with the Na+ cation) and non-hydrated surface of dickite have been studied. Methyl-cytosine is one of the four major basic components of DNA and RNA. It is a pyrimidine derivative consisting of a single six member ring, containing both nitrogen and carbon atoms, and an amino group. Cytosine is intimately involved in the preservation and transfer of genetic information. Dickite, a chosen model of clay minerals, has the same formula as kaolinite, nacrite and halloysite Al2Si2O5(OH)4 but a different crystal structure. It is a kaolinite polymorph and, as such, is a 1:1 aluminous dioctahedral phyllosilicate (clay) mineral. The B3LYP level of theory with 6-31G(d) basis set have been used to calculate adsorption systems. Representative cluster models have been used to simulate the surface of the mineral. The MC molecule in several initial positions on the surface was fully optimized while geometry of the mineral part was kept frozen. We have calculated the interaction energy corrected by the basis set superposition error. We have studied the interactions between MC and the mineral, changes in the geometrical parameters and electron density during adsorption. Conference on Current Trends in Computational Chemistry 2003 57
Density Functional Theory (DFT) Study of Stannacyclohexanes and Distannacyclohexanes: Conformational Interconversions, Relative Energies, Stereoelectronic Effects, and Structures
Fillmore Freeman, Christine Fang, David Hoang, Angela C. Huang, Katie Le, Thuy D. Mai, and Khue Trinhk
Department of Chemistry, University of California, Irvine, Irvine, CA 92697-2025
Stereoelectronic hyperconjugative interactions, relative energies, and structures of conformers, conformations, and transition states of stannacyclohexane (stanne), 1-methyl- stannacyclohexane (1-methylstanne), 1,1-dimethylstannacyclohexane (1,1-dimethylstanne), 1,2- distannacyclohexane (1,2-distanne), 1,3-distannacyclohexane (1,3-distanne), and 1,4-distanna- cyclohexane (1,4-distanne) have been calculated using density functional theory (BLYP, B3LYP, B3P86, B3PW91, BHandHLYP) with the SDD basis set. In general, the five density functional models gave similar energy differences (∆E) with B3LYP giving slightly larger ∆E values. Similar equilibrium geometries were calculated by the five models with BLYP generally giving longer bond lengths and BHandHLYP) generally giving the shorter bond lengths. Intrinsic reaction coordinate [IRC, minimum-energy path (MEP)] calculations have been used to connect the transition states between respective chair and twist conformers and between respective twist enantiomers and to construct conformational interconversion diagrams. The chair conformer (1a) of stannacyclohexane is 2.91, 3.34, and 4.23 kcal/mol, respectively, more stable than the 1,4-twist conformer (1b), 2,5-twist conformer (1c), and 2,5- boat transition state [1e]≠. The 1,4-boat transition state [1d]≠ was not located.
H H H H H H H H H H H H H H H H H H H Sn H H H H H Sn H Sn H H H H H H H H H
H H
1a 1b 1c The axial chair conformer (2a) of 1-methylstannacyclohexane is 0.17 kcal/mol higher in energy than the equatorial chair conformer (3a). The conformational free energy (∆G°) of 1- methylstannacyclohexane at 298.15 K is smaller than that of methylcyclohexane (toluene) owing in part to the long Sn—C bond lengths. The axial chair conformer (2a) of 1-methylstanna- cyclohexane is 2.70, and 3.26, respectively, more stable than the 1,4-twist (2b) and 2,5-twist (2c) conformers.
58 Conference on Current Trends in Computational Chemistry 2003
H H H H H H H H H H H H H H H H H H H H Sn H H H H Sn Sn H H H H H H H H H
H H H H H H H H H
2a 2b 2c
The chair conformer (3a) of equatorial 1-methylstannacyclohexane is 2.86 and 3.37 kcal/mol, respectively, more stable than the 1,4-twist (3b) and 2,5-twist (3c) conformers.
H H H H H H H H H H H H H H H H H H H H H H H H H H H Sn H H H Sn Sn H H H H H H H H H H
H H
3a 3b 3c
The chair conformer (4a) of 1,1-methylstannacyclohexane is 2.75, 3.35, and 4.15 kcal/mol, respectively, more stable than the 1,4-twist conformer (4b), 2,5-twist conformer (4c), and 2,5-boat transition state [4e]≠. The 1,4-boat transition state [4d]≠was not located.
H
H H H H H H H H H H H H H H H H H H H H H H H H H Sn H H H H Sn Sn H H H H H H H H H H H H H H H H H H
4a 4b 4c
The 3,6-half-chair conformer (5g, no imaginary frequencies) of 1,2-distannacyclohexane is more stable than the chair conformer (5a) and 2.39 kcal/mol more stable than the 1,4-twist conformer (5b).
Conference on Current Trends in Computational Chemistry 2003 59
H H H H H H H H H H H H H H H H H H H H H Sn Sn Sn Sn Sn Sn H H H H H H H H H H H H H H H
5a 5b 5g
The chair (6a) and 2,5-boat conformer (6e, no imaginary frequencies) of 1,3-distanna- cyclohexane are of comparable energy and are 2.32 kcal/mol more stable than the 2,5-twist conformer (6c).
H H H H H H H H H H H H Sn H H H H H H H Sn H Sn H H H Sn Sn H H H Sn H H H H H H H
H H H
6a 6c 6e
The chair (7a) and 1,4-twist (7b) conformers of 1,4-distannacyclohexane are of comparable energy and are more stable than the 2,5-twist conformer (7c). The 1,4-boat transition state [7d]≠was not located.
H
H H H H H H H H H H H Sn H H H H H Sn Sn H H H H H Sn H Sn H H H H H Sn H H H H H H H
H
7a 7b 7c
Intrinsic reaction coordinate [IRC, minimum energy path (MEP)] calculations were used to identify the transition states connecting the chair (7a) and 1,4-twist (7b) conformers of 1,4- distannacyclohexane. Two paths were located for the chair-chair interconversion of 1,4- distannacyclohexane. Neither path involves a boat transition state, but each proceeds via the 1,4- twist intermediate (conformer, (7b). The lower energy path proceeds via the 1,4-half-chair transition state [7f]≠ which is 4.27 kcal/mol higher in energy than the chair conformer (7a). The higher energy path for the chair-chair interconversion involves a 2,5-sofa transition state [7i]≠ which is 4.47 kcal/mol higher in energy than the chair conformer (7a).
60 Conference on Current Trends in Computational Chemistry 2003
Theoretical Investigation of 3-Methyl-Cytosine Hydration
Al’ona Furmanchuk, Olexandr Isayev, Leonid Gorb and Jerzy Leszczynski
Computational Center for Molecular Structure and Interactions, Department of Chemistry, Jackson State University, Jackson, MS 39217, USA
DNA, like most organic compounds, is rather unstable, and spontaneously breaks down. There are several factors which influence the stability of DNA structure. Among them is the influence of water molecules, which is very specific. The presence of water molecules preserves the 3-dimensional structure of A-, B-, Z-, etc. DNA molecules. However, water molecules could also initiate spontaneous hydrolysis of C-O, C-N, O-P bonds. Therefore, the knowledge of the details of water – DNA interaction is very important. On another hand the analysis of literary sources suggests that available data on this subject are rather scare. For example, we did not find any solid data on the amount of the water molecules placed in the first solvation shell of the DNA bases. To shed a light on Methylcytosine with the three water this problem we have performed theoretical molecules in the first hydration shell investigation based on the analysis of thermodynamics of stepwise addition of water molecules to 3-methyl-cytosine. All calculations have been performed at the DFT (B3LYP functional) level of theory with standard 6-31G(d,p) basis set. The nature of minima and transition states were proved by the calculation of vibrational frequencies. We have found that 3-methilcytosine, like cytosine itself, exists in gas phase as the mixture of tautomeres which are canonic and its imino-form. Therefore, we have calculated the thermodynamis values of stepwise hydration for both these species. Basing on the analysis of Gibbs free energies we obtain two the most important conclusions. 1. The interaction with water molecules destabilizes the structure of 3-methil-cytosine inimo- form over canonic structure. It results in the change of equilibrium constants for the equilibrium between canonic and imino forms of 3-methyl-cytosine from 2.9*10-4 (in the case of nonhydrated species) to 1.2*10-6 (in the case of three hydrated species). The addition of first water molecule has the most important influence for this transformation. 2. The structure of the first solvation shell of 3-methyl-cytosine depends on the temperatute. At room temperature it consists of 3 water molecules. All of them are located in the places which are later substituted by guanine during the formation of Watson–Crick base pair. Therefore, we can predict that, being immersed in water surrounding at room temperature, 3-methyl-cytosine will be hydrated specifically by three water molecules. The remaining part of 3-methyl-cytosine will be hydrated unspecifically
Conference on Current Trends in Computational Chemistry 2003 61
An Improved Thermodynamic Energy Estimator for Path Integral Simulations
Kurt R. Glaesemann and Laurence E. Fried
University of California, Lawrence Livermore National Laboratory Chemistry and Material Science, Livermore, CA 94550 USA
A new path integral energy estimator is presented that improves upon the thermodynamic energy estimator via a free particle projection.[1] The new centroid thermodynamic estimator has accuracy close to the centroid virial estimator. New path integral Monte Carlo constant volume specific heat (CV) estimators are also presented that improve upon the thermodynamic, virial, and centroid virial CV estimators via a free particle projection. A double virial estimator is derived for real space path integrals and comparisons are made to it.[2] Projected estimators significantly reduce the numerical noise of the traditional estimators. The new projected thermodynamic estimators have particular advantages when derivatives of the potential are expensive to evaluate. The centroid virial estimators are found to be significantly better than the noncentroid virial estimators.
[1] An improved thermodynamic energy estimator for path integral simulations, K. R. Glaesemann & L. E. Fried, Journal of Chemical Physics 116 (14): 5951-5955 APR 8 2002
[2] Improved heat capacity estimator for path integral simulations, K. R. Glaesemann & L. E. Fried, Journal of Chemical Physics 117 (7): 3020-3026 AUG 15 2002
This work was performed under the auspices of the U.S. Department of Energy by the University of California Lawrence Livermore National Laboratory under contract No. W-7405-Eng-48.
62 Conference on Current Trends in Computational Chemistry 2003
Bond Dissociation Processes in Various Energetic Materials 1Sharye Glenn*, 2Brian Johnson, 1Jesse Edwards
1Department of Chemistry/AHPCRC, 2Computer Information Systems/AHPCRC Florida A&M Tallahassee, Florida,USA 32307,
An objective of this work is to establish relationships between impact sensitivity for energetic materials and molecular and thermodynamic properties. We chose as our materials the nitramine group, which have been known to be the most sensitive of the energetic materials. A quantum mechanical method is applied to calculating their bond dissociation energies of the nitro groups of these compounds. The nitro groups have been speculated to be the point of initiation of the detonation process. Relationships between the bond dissociations and impact sensitivity of this class of compound are examined. Conference on Current Trends in Computational Chemistry 2003 63
π...H+...π Hydrogen Bonds
Sławomir J. Grabowski a,c, W. Andrzej Sokalski b, and Jerzy Leszczynski c
a Department of Crystallography and Crystal Chemistry, University of Łódź, 90-236 Łódź, ul.Pomorska 149/153, Poland b Institute of Physical and Theoretical Chemistry, Wroclaw University of Technology, Wyb. Wyspiańskiego 27, 50-370 Wroclaw, Poland c Computational Center for Molecular Structure and Interactions, Department of Chemistry, Jackson State University, Jackson, MS 39217, USA
+ + C2H2...H ...C2H2 and C2H4...H ...C2H2 complexes are studied theoretically. The full geometry optimizations were performed up to MP2/aug-cc-pVDZ level of theory. The geometrical and energetic characteristics of π...H+...π systems are compared with characteristics of the other conventional and unconventional H-bonds. This comparison shows that π...H+...π + complexes may be classified as hydrogen bonded. The H-bond energies for C2H2...H ...C2H2 and + C2H4...H ...C2H2 calculated at MP2/aug-cc-pVDZ level are equal to 15.0 and 10.4 kcal/mol respectively (BSSE included). The electron densities at H+...π bond critical points and their Laplacians also indicate that these systems possess hydrogen bonds of the medium strength. The variation-perturbation approach was also applied to evaluate components of the + + interaction energy of C2H2...H ...C2H2 and C2H4...H ...C2H2 complexes and the results show the unique character of these interactions.
+ The molecular graph of the C2H2...H ...C2H2 complex, attractors, bond critical point and bond paths are shown 64 Conference on Current Trends in Computational Chemistry 2003
Hydrogen Bonding in 5-Bromouracil-Adenine-5-Bromouracil- Adenine (T+AT+A) Tetrads
Jiande Gu,a Jing Wang,b and Jerzy Leszczynskib
aDrug Design & Discovery Center, State Key Laboratory of Drug Research, Shanghai Institute of Materia Medica, Shanghai Institutes for Biological Sciences, Chinese Academy of Sciences, Shanghai 200031 P. R. China bComputational Center for Molecular Structure and Interactions, Department of Chemistry, Jackson State University, Jackson, MS 39217 U. S. A.
In order to understand the role of 5-bromouracil-adenine-5-bromouracil-adenine (T+AT+A) tetrads in the formation of the tetraplexes, the relative stability of different conformers of the tetrad and their bonding characteristics have been studied by the quantum chemistry methods. The influence of bromine in the formation of the T+AT+A tetrads has been revealed by the comparative studies of thymine-adenine-thymine-adenine (TATA) and uracil-adenine-uracil- adenine (UAUA) tetrads. The stabilization energy of T+AT+A has been evaluated to be around 40 kcal/mol, compatible to those of TATA and UAUA. The role of Br playing in the stabilization of the tetrads is two folded: by improving the ability of proton-donating on its N3 position it reinforces the H-bonding between A and T+, while through electrostatic repulsion with N7 or N1 of A, it destabilizes the binding between the AT+ pairs. The increase of the intra-base-pair binding energy compensates the decrease of the inter-base-pair interaction. This bifurcated H- bond consisting of Br(T+), O4(T+), and H6’(A) that binds two AT+ pairs to form the stable T+AT+A tetrads has been revealed through the atoms-in-molecules (AIM) theory, the complementary method of electron-localization function (ELF), and the electron density difference analysis. The results of this study suggest that TATA should exist in the dimeric intermolecular tetraplexes formed from 12-nucleotide repeat sequences from human telomeres. Conference on Current Trends in Computational Chemistry 2003 65
Ab Initio Calculation of 14N NQR Parameters and 13C, 1H NMR of 1 C18H12N6
Nasser L. Hadipour, Nasser Zamand
Department of Chemistry, Tarbiat Modaress University, P. O. Box 14115-175 Tehran, Iran
The electrostatic interaction of a nuclear electric quadrupole moment and the electron charge cloud surrounding the nucleus can give rise to the observation of pure Nuclear Quadrupole Resonance (NQR) [1]. Hamiltonian of this interaction is given:
) 2 ) 2 ) 2 ) 2 ) 2 H e Qq / 4I 2I 1 3I I / 2 I I Q = zz ( − )[ z − −η ( + + − )]
2 The term e Qqzz / h is known as nuclear quadruple coupling constant (χ). In this work we have computed χ via calculating qzz around nitrogen atoms in C18H12N6. For this aim ab initio HF/6- 31+G*, B3LYP/6-31+G* and MP2/6-31+G* have been employed. XRD crystallography data were our guideline in order to find out we had performed a proper molecular optimization. NQR parameters found by this work are summarized in table 1. The accurate detection of frequencies is made possible by cross-polarization transfer[*]. On the bases of optimized molecular structure the chemical shifts of 13C and 1H are calculated (table 2). The experimental chemical shifts of 13C and 1H of this molecule are available [2]. Comparison of calculated chemical shifts with the experimental values shows no significant structural change in solution, in fact this result implies that the calculated chemical shifts satisfy the experimental values.
26
25 3 28 2 6
24 35 N 1 5 27 4 N 36 21 N 23 78 N 20 15 9 29 N 12 19 10 17 N 13 34 18 16 11 30 31 14 33 32
1 Cyclotrisazobenzene [2].
66 Conference on Current Trends in Computational Chemistry 2003
Table 1. NQR parameters for C18H12N6 Center χ(MHz) η% υ0 (kHz) υ+ (kHz) υ- (kHz) N7 5.121 56.4 1445.90 4562.81 3118.69 N8 4.305 40.4 869.61 3663.56 2793.95 N15 5.047 54.8 1382.66 4476.69 3093.81 N16 4.459 48.4 1222.57 3883.79 2804.71 N23 4.846 47.2 1392.81 4206.33 3062.68 N24 4.379 41.6 1012.87 3739.67 2828.83
13 1 Table 2. Predicted and experimental C, H chemical shifts of C18H12N6 Atoms C(I) C(II) C(III) H(I) H(II) Methods 2 Calcs. 144.92 121.49 127.31 8.16 7.91 Exp. 146.5 122.0 130.1 7.71 7.59
References:
1. Molecular Spectroscopy by Jack D. Graybeal, (1988), MacGraw-Hill.
2. Ander S. Dreiding et al. 38.Reduction von 1,2-Bis[(z)-(2-…, Helv. Chem. Acta, 68, 325 (1985).
2 The methods and basis sets are HF/6-31G* Conference on Current Trends in Computational Chemistry 2003 67
Structures and Dissociation Channels of Metal Dications Solvated by Acetonitrile Ligands
Frank Hagelberg1, Chuanyun Xiao1, Ahmed M. El-Nahas2
1Computational Center for Molecular Structure and Interactions Department of Physics, Atmospheric Sciences and General Science Jackson State University, Jackson, MS 39217, USA 2Chemistry Department, Faculty of Science El-Menoufia University, Shebin El-Kom, Egypt
The structures and dissociation pathways of metal dications solvated by one or two 2+ acetonitrile ligands, M (CH3CN)n (n=1, 2 for M=Be, and n=1 for M=Mg, Ca, Fe, Cu and Zn), were studied by density functional theory at the B3LYP/6-311+G(d, p) level. The analyzed dissociation processes include the loss of a neutral ligand, the dissociative electron transfer, and + the cleavage of neutral and charged methyl (CH3 and CH3 ). For the diligated Be complex, the dissociative proton transfer is considered in addition to the processes indicated. The equilibrium structures of these complexes were found to be linear with the metal atom attached to the N-end of the CH3CN ligands for all metals except Cu, for which the structure is slightly bent. The calculated dissociation energies indicate that the complexes are thermodynamically stable with respect to all considered processes for M=Be, Mg, Ca, and Fe, but are thermodynamically unstable with respect to the dissociative electron transfer process for M=Cu and Zn and to the + cleavage of CH3 for M=Cu. The energy barriers for the processes of dissociative electron + transfer and the cleavage of CH3 are determined for all units, which suggests that the 2+ 2+ Zn CH3CN and Cu CH3CN species are kinetically metastable with long lifetimes. The loss of neutral CH3 is energetically unfavorable for all species, while the loss of a neutral acetonitrile ligand is energetically unfavorable for all metals except Ca, where the loss of a neutral ligand competes with the dissociative electron transfer. The theoretical results agree well with available experimental observations [1].
[1] A.A. Shvartsburg, K. W. M. Siu, J. Am. Chem. Soc. 123 (2001) 10071. 68 Conference on Current Trends in Computational Chemistry 2003
Multi-Determinant Trial Functions in the Determination of the Dissociation Energy of the Beryllium Dimer: A Quantum Monte Carlo Study
John A.W. Harklessa and Karl K. Irikurab
aDepartment of Chemistry, Howard University, 525 College St. NW, Washington, DC 20059
bPhysical and Chemical Properties Division, National Institute of Standards and Technology, 100 Bureau Drive, Stop 8380, Gaithersburg, MD 20899-8380
Both variational Monte Carlo (VMC) and fixed-node diffusion Monte Carlo (DMC) are used to estimate the dissociation energy of Be2. The effect of using single and multireference trial functions on the quality of the Monte Carlo estimates is investigated, with independent- particle wavefunctions ranging from a restricted Hartree-Fock (RHF) calculation up to a complete active space self-consistent field (CASSCF) with four valence electrons in twelve active orbitals. It was determined that the best trial function for DMC had a high cutoff for inclusion and included all 2s and 2p orbitals in a CASSCF(4,8) calculation. The best DMC estimate, 829(64)cm-1, is found to compare well to computed values of similar quality as well as an experimental value of 839(10)cm-1.
Conference on Current Trends in Computational Chemistry 2003 69
Alkylation of Cytosine by cis-1-Methyl-3-hydroxyazetidinium Ions: Transition States, Tautomerism, and Hydrogen Bonding
Robert H. Higgins
Department of Natural Sciences, Fayetteville State University, Fayetteville, NC 28301
Alkylation of nucleic acids is currently an area of intense interest. While strained heterocycles are not novel alkylating agents, their use is often overlooked. Autopsies of victims of mustard gas, bis-(2-chloroethyl)sulfide (BCES), poisonings during World War I indicated that BCES alkylated nucleic acids (presumably through alkyl episulfonium ions). Similarly alkylations of nucleic acids by epoxides are known and are presumed to be responsible for much, if not all, of their carcinogenic nature. In addition, the patent literature from the 1970’s contains numerous examples of alkylations of phenols by 1-alkyl-3-azetidinols for the preparation aryloxypropanolamines structurally related to propranolol (Inderol). Having completed a laboratory and theoretical investigation of the latter reaction, we felt it would be a logical extension of our research with strained 4-membered heterocyclic alkylating agents to investigate the feasibility of employing these to affect alkylations of nucleic acids.
(CH3)2CH Cl Cl N O S S H HO CH2CH2Cl bis-(2-Chloroethyl)sulfide (BCES) 1-(2-Chloroethyl) episulfonium cation
Propranolol Nucleic acids exist as equilibrium mixtures of numerous rapidly equilibrating tautomers, although a single tautomer of each nucleic acid is predominant. While the rate equations for alkylations will depend upon the concentrations of each tautomer, these rates will also depend upon the energies of the transition states. Consequently, our current efforts have been directed towards MP2/6-31G** ab initio calculations of some of the transition states considered likely for alkylation of the N-1, O-2, N-3, and 4-NH2 substituents of the various tautomers of cytosine by cis-1-methyl-3-hydroxyazetidinium ion. While our study is incomplete, sufficient data has been obtained on the electronic energies of many of these transition states to assess their nature and the importance of hydrogen bonding in these.
Acknowlegements The author gratefully acknowledges the support provided by the Computational Center for Molecular Structure and Interactions at Jackson State University and to Dr. Jerzy Leszczynski, its Director. In addition thanks are due to Drs. Leonid Gorb, Manoj Shukla, and Pawel Kedzierski for their suggestions and patient guidance. 70 Conference on Current Trends in Computational Chemistry 2003
Using Variational Monte Carlo for Excited States Calculations in Biological and Material Calculations
Glake Hill 1,2, Alex Kollias 2, Tomekia Simeon 1, Gareth Forde 1, William Lester 2 , and Jerzy Leszczynski 1
1 Computational Center for Molecular Structure and Interactions Jackson State University, Jackson, MS 39217 2 Kenneth S. Pitzer Center for Theoretical Chemistry, Department of Chemistry, University of California at Berkeley, Berkeley, California, 94720-1460
The need to calculate accurate excited states is prevalent in a number of areas. In biological systems, excitations of nucleic acids and other biomolecules lead to immense changes and many possible diseases, notable cancers. In material science, the photo-excitation of materials is important to the elucidation of nanotubes for many important purposes. However, many theories have not been able to accurately describe these excitations. Using Variational Monte Carlo (VMC), these systems are being addressed. In this presentation, these results will be discussed.
Conference on Current Trends in Computational Chemistry 2003 71
Conventional Strain Energy in the Diazetidines and the Diphosphetanes
Patricia L. Honea, Ashley L. Ringer, and David H. Magers
Department of Chemistry and Biochemistry Mississippi College, Clinton, MS 39058
The conventional strain energies for the cis and trans conformations of 1,2-diazetidine, 1,3- diazetidine, 1,2-diphosphetane, and 1,3-diphosphetane together with the 1,2- and the 1,3- isomers of thiaphosphetane are determined within the isodesmic, homodesmotic, and hyperhomodesmotic models. Optimum equilibrium geometries, harmonic vibrational frequencies, and corresponding electronic energies are computed for all pertinent molecular systems using SCF theory, second-order perturbation theory, and density functional theory (DFT). The DFT functional employed is Becke’s three parameter hybrid functional using the LYP correlation functional. Two basis sets, both of triple-zeta quality on valence electrons, are employed: 6-311G(d,p) and 6-311+G(2df,2pd). Additionally, single point coupled-clustered calculations using the optimized MP2 geometries and the larger of the two basis sets, are used to investigate the effects of higher-order electron correlation. Finally, the calculated strain energies are compared to those of cyclopropane, cyclobutane, azetidine, phosphetane, 1,2-oxazetidine, and 1,3-oxazetidine. We gratefully acknowledge support from NSF EPSCoR (EPS-0132618).
trans-1,2-diazetidine trans-1,3-diazetidine
cis-1,2-diazetidine cis-1,3-diazetidine 72 Conference on Current Trends in Computational Chemistry 2003
Theoretical Study of the Stereoisomers of Salsolinol
Ming-Ju Huang
The Computational Center for Molecular Structure and Interactions, Department of Chemistry, P. O. Box 17910, Jackson State University, 1400 J. R. Lynch Street, Jackson, MS 39217
Salsolinol (6,7-dihydroxy-1-methyl-1, 2, 3, 4-tetrahydroisoquinoline; Sal) is the dopamine (DA)-derived tetrahydroisoquinoline (TIQ), supposed to be a potent dopaminergic neurotoxin, similar to 1-methyl-4-phenyl-1, 2, 3, 6-tetrahydropyridine (MPTP), an exogenous neurotoxin causing Parkinsonism in humans, monkeys, and various animals. Different neurotoxicological properties were observed for (R)- and (S)-enantiomers of Sal. The structure of Sal has one chiral center, a pyramidal nitrogen next to this chiral center, and four possible orientations from the two hydroxyl groups so there are eight conformers for each of the stereoisomers. An extensive theoretical study of the stereoisomers of Sal has been performed at the semi-empirical AM1, ab initio HF/6-31G**, HF/6-311G**, MP2/6-31G**, MP4SDQ//6- 31G**, and B3LYP/6-31G** levels. The most stable conformer structure for the (S)-enantiomer of Sal has lower energy than the most stable conformer structure of the (R)-enantiomer for all the methods. Conference on Current Trends in Computational Chemistry 2003 73
Theoretical AM1 Studies of Inclusion Complexes of Heptakis(2-o- hydroxypropyl)-β-Cyclodextrin with Alkylated Phenol
Ming-Ju Huang and Manyin Yi
CCMSI, Department of Chemistry Jackson State University, Jackson, MS 39217
Semi-empirical AM1 calculations have been performed on a family of inclusion complexes of Heptakis(2- or 6-o-hydroxypropyl)-β-Cyclodextrin (HPβCD) with alkylated phenol derivatives. The study shows that the van der Waals attraction is the dominating force that dictates the structural details, under the steric or size-fitting restriction. The para-substituted phenol body is too large to penetrate into the CD’s parent cavity, but could fit well within the region surrounded by the seven HP threads. An ortho-substituent on the phenol can be inserted into the CD’s parent ring, which surely enhances the van der Waals contacts by a considerable amount. The Inter- molecular hydrogen bonding and point charge interactions are distinguishable, but there are too few of them, and thus are not dependable for the purpose of stabilizing the complex. Dipole- dipole interaction does not have appearent influence. 74 Conference on Current Trends in Computational Chemistry 2003
Molecular Modeling of Helium Diffusion in Isomeric Polyimides
Danielle L. Hudson1, Jeffrey A. Hinkley2, Thomas C. Clancy3, Melissa S. Reeves1
1Tuskegee University Department of Chemistry Tuskegee, Alabama 2Advance Materials and Processing Branch NASA Langley Research Center 3National Research Council
Polyimides (PIs) are being used in many technological applications such as semiconductor devices, high temperature adhesives, and high performance composite materials [1]. These composites could replace metals in aircraft and spacecraft structure. PIs are thermally resistance and possess excellent mechanical properties. They also maintain excellent properties at cryogenic temperatures [2]. In recent studies PI membranes have shown high gas
selectivities for PH2/PN2, PO2/PN2, and PCO2/PCH4 [1]. The purpose of this study was to determine the effect of positional isomerism on diffusion, structural and mechanical properties. The three related PIs studied differed in the connections of the oxygen ether groups on the aromatic rings. (See Figure 1.)
O O O
N (A) (C) O (B) O N
n O Figure 1. Repeat unit for polyimides in this study where positional differences occur at rings A, B, and C. The three polyimides are PI-1 (meta-meta-meta), PI-2 (para-para-para) and PI-3 (para- meta-para).
The three PIs were constructed in periodic amorphous cells (ACs) using a coarse-grain reverse mapping technique to produce well-equilibrated amorphous structures while avoiding ring catenation and spearing [3]. Materials Studio and InsightII were used to perform molecular dynamics (MD) on ACs containing penetrants and the three isomeric BPDA (biphenyl dianhydride) PIs. First the ACs were energy minimized using the Discover_3 modules until the -1 -1 maximum derivative was 0.001 kcal mol Å , then equilibrated at 500K for 300ps without any penetrants. The final 100ps of the trajectory was examined for the conformation with the lowest energy. This conformation was minimized, then used to calculate mechanical properties, including Young’s, Bulk, and Shear moduli and Poisson’s ratio. PI-3 was the stiffest polymer according to the Young’ s modulus (see Table 1), which is the resistance of a polymer to uniaxial tension or stretching [4]. Young’s modulus normally ranges from 0.01 to 10 GPa for polymers.
Conference on Current Trends in Computational Chemistry 2003 75
Table 1. The summary of the mechanical properties for polyimides calculated at 0 K. Polymer Young’s Modulus Bulk Modulus Shear Modulus Poisson’s Ratio (GPa) (GPa) (GPa) PI-1 3.4 6.1 1.2 0.41 PI-3 5.5 8.7 1.9 0.39
The range of calculated values for these PIs was 3-6, which indicates that these are moderately stiff polymers. Bulk Modulus is the inverse of Compressibility. The values in Table 1 indicate that the material that is softest in tension, PI-1, is also the most compressible. The Shear modulus follows the same trends as the Young’s and Bulk moduli where the softest PI is the easiest to shear as well. The penetrants He and H2 were randomly inserted into the equilibrated AC followed by energy minization. MD were run on each minimized AC with penetrants for 300ps at 500K with a time step of 1fs. The final 100ps were used to determine the diffusion coefficients of the penetrants. The positions of the gas molecules as a function of time were obtained from the trajectory and used to determine the diffusion coefficients in each PI. The diffusion coefficient for He in each PI is listed in Table 2.
Table 2 Helium diffusion coefficients for each PI. 2 -1 Polymer DHe (cm s ) PI-1 6.2 x 10-6 PI-2 3.8 x 10-6 PI-3 2.3 x 10-6
The diffusion coefficient of He is larger in PI-1, which is the least stiff. During diffusion, thermal fluctuations allow the formation of voids and channels for the penetrants to travel through [5]. If PI-1 has more voids and channels being formed then the atoms have more room to travel. However, if there are no voids and channels being formed then the opposite would be true. The He atoms diffused at a slower pace throughout the PI-2 and PI-3, which had a para- para-para and para-meta-para connection, respectively, between the nitrogen and oxygen ether groups on the aromatic rings. This could be an indication that these PIs have fewer or smaller voids and channels, which the atoms cannot travel though. In conclusion, the positions of connections of the aromatic rings have an effect on structural and mechanical properties and on diffusion. Further studies will be done with Ar to see if these trends in diffusion continue with larger penetrants.
1. H. Kawakami, M. M. Mikawa, S. Nagaoka., Formation of surface skin layer of asymmetric polyimide membranes and their gas transport properties, J. Membranes Sci, 137 (1997) 241- 250. 2. Parker, E. R.; Materials Data Book For Engineers and Scientists McGraw-Hill Book Company, Inc., New York, 1967, Plastics, pg. 314. 3. Clancy, T.C.; Jang, J.H.; Dhinojwala, A.; Mattice, W. L. J. Phys. Chem. B. 2001 105 11493. 4. Bicerano, J., 1996. Prediction of Polymer Properties, 2nd Edition, New York: Marcel Dekker, Inc. 5. Mattice, W. L.; Zhang, R. J. Memb. Sci. 1995 108 15. 76 Conference on Current Trends in Computational Chemistry 2003
Conventional Ring Strain in Unsaturated Four-Membered Rings
Shelley S. Huskey and David H. Magers
Department of Chemistry and Biochemistry Mississippi College, Clinton, MS 39058
In order to study the effect of unsaturation on the ring strain in small cyclic molecules, the conventional strain energies for cyclobutene, azetidine-1-ene (Figure 1), phosphetane-1-ene (Figure 2), azetidine-2-ene (Figure 3), and phosphetane-2-ene (Figure 4) are determined within the isodesmic, homodesmotic, and hyperhomodesmotic models. Optimum equilibrium geometries, harmonic vibrational frequencies, and corresponding electronic energies are computed for all pertinent molecular system using SCF theory, second-order perturbation theory, and density functional theory (DFT). The DFT functional employed is Becke’s three parameter hybrid functional using the LYP correlation functional. Two basis sets, both of triple-zeta quality on valence electrons, are employed: 6-311G(d,p) and 6-311+G(2df,2pd). Finally, the calculated strain energies are compared to those of cyclopropane, cyclobutane, azetidine, and phosphetane. We gratefully acknowledge support from NSF EPSCoR (EPS-0132618).
Figure 1 Figure 2
Figure 3 Figure 4 Conference on Current Trends in Computational Chemistry 2003 77
Mechanism of Nitrobenzene Reduction by Iron (II) Compounds: Density Functional Theory Study
Olexandr K. Isayev1, Leonid Gorb1, Igor Zilberberg2 and Jerzy Leszczynski1
1Computational Center for Molecular Structure and Interactions Department of Chemistry, Jackson State University, 1400 J.R. Lynch Street, Jackson, MS 39217 2Boreskov Institute of Catalysis, Novosibirsk, 63090, Russia
The development of cleanup technologies for a disposal of nitroaromatics (NACs) is a challenge for environmental science. Such development involves the coordination of experimental and theoretical investigations to integrate both technological and fundamental aspects of key process. The Fe(II) hydroxide is one of the most attractive NACs-reducing agent being extremely cheap and environmentally clean. Although the major processes affecting the iron treatment of NACs have been investigated qualitatively, many issues regarding a reaction mechanism remain unsolved. In the present work an activity of the ferrous iron hydroxide in the remediation of nitroaromatics was studied by means of density functional theory. The Fe(OH)2 molecule has been chosen as a simplest Fe(II)-hydroxide model while NACs were modeled by nitrobenzene (NB) molecule. The pure BLYP and the hybrid B3LYP functionals have been applied in this work. The standard 6-311++G(d) basis set has been used. All geometry minima have been verified by the absence of imaginary frequencies. The analysis on the internal instability of DFT solution has been also performed. Reaction of nitrobenzene with iron (II) hydroxide is formally a six-electron reduction of nitrobenzene to aniline. During the first step of the reaction a nitrobenzene-ferrous-iron complex is formed. Our DFT study has revealed that the electronic structure of the NB-Fe(OH)2 complex could be described as the result of intramolecular electron transfer. This finding predicts in fact a non-barrier process for the transfer of the first electron from Fe(OH)2 to nitrobenzene molecule. At the next stage, proton is shifted from a water molecule to form an intermediate containing the -NO(OH) group. This scenario seems to be the major motif of the whole process. Unlike reduction with FeO, for the system nitrobenzene-iron (II) hydroxide is more feasible to transfer two hydrogen atoms and then eliminate water molecule, than remove oxygen directly. Hydration is found to play an important role in the reaction. The Gibbs free energy is dramatically decreased with an increase in the number of water molecules participating in the reaction. 78 Conference on Current Trends in Computational Chemistry 2003
Theoretical Investigations and Structure-Toxicity Relationships of Nitroaromatic Compounds
Olexandr K. Isayev1, Leonid Gorb1, Bakhtiyor Rasulev2 and Jerzy Leszczynski1
1Computational Center for Molecular Structure and Interactions Department of Chemistry, Jackson State University, 1400 J.R. Lynch Street, Jackson, MS 39217 2Institute of Chemistry of Plant Substances AS RUz, Kh. Abdullaev Str., 77, Tashkent, 700170, Uzbekistan
Nitroaromatic compounds are released into the biosphere almost exclusively from anthropogenic sources. Some compounds are produced by incomplete combustion of fossil fuels; others are used as synthetic intermediates, dyes, pesticides, and explosives. Many nitroaromatics have also been known to be toxic or mutagenic. In addition the electron withdrawing nitro groups provide the compounds resistance to typical advanced oxidation destruction techniques as well as generally protecting the compound from chemical and biological attack. Therefore, the understanding of the mechanisms of nitrocompounds chemical transformation is very important from both fundamental and practical points of view. In order to distinguish some tendencies in reactivity of nitrocompounds we decided to perform extensive investigation of them by the methods of computational chemistry. The molecular geometries of all compounds have been optimized at B3LYP / 6-311+G(d,p) and MP2 / AUG-cc-pVDZ levels. The following characteristics have been predicted. 1. Gas phase molecular geometries 2. Electronic distribution and electrostatic potential 3. Adiabatic ionization potentials. 4. Dipole moments. 5. Shape and population of molecular orbitals. We have also performed the topological analysis of electronic density (AIM). In the framework of this theoretical treatment we have shown the presence of intramolecular interaction of nitro group oxygen atom with the hydrogen atom of substituent in some of the molecules. This kind of interactions is believed to be an important at the first step of explosive detonation. To determine the physicochemical parameters that influence on toxicity properties of nitrobenzenes we have applied Quantitative Structure-Activity Relationship (QSAR) method. For these purposes the Genetic Algorithm (GA) and Multiple Linear Regression Analysis (MLRA) were used to select the descriptors and to generate the equations that relate the structural features to the toxicity values. As result the models with satisfactory predictive power of toxicity have been obtained. We predict the toxicity properties also for other aromatic nitrocompounds which are not included in current theoretical consideration.
NO2
n = 0, 1, 2, 3
R = -CH3, -NH2, -F, -COOH
NO2
(NO2)n R
Conference on Current Trends in Computational Chemistry 2003 79
Insight into the Dispersion Energies of Hydrogen and Carbon Dimer Interactions
Cynthia Jeffries1, Glake Hill2, Jerzy Leszczynski1
1Computational Center for Molecular Structure and Interactions, Jackson State University, Jackson, MS 39217 2Pitzer Center for Theoretical Chemistry, Department of Chemistry, University of California, Berkeley, CA 95401
Dispersion has been of great interest to those that are studying weak interactions in a number of systems. Scientists of nanostructures, computational biology, and fuel cell research are often limited because of an inadequate description of this term. In the present research, the interaction energies of hydrogen and carbon dimers at different angles and distances are elucidated and studied to provide an empirical formula to better predict its magnitude. The idea is to see it at multiple angles and distances and compare the energies of each one to see how much change of dispersion energy has occur between each molecule. These calculations were run on Gaussian 98 using HF and CCSDT levels of theory using aug-cc-pVDZ, aug-cc-pVTZ and aug-cc-pVQZ. Results will be discussed.
80 Conference on Current Trends in Computational Chemistry 2003
Density-Functional Theory Approach to van der Waals Interactions via Symmetry-Adapted Perturbation Expansion
Bogumil Jeziorskia, Alston J. Misquittab, and Krzysztof Szalewiczb
aDepartment of Chemistry, University of Warsaw, Pasteura 1, 02-093 Warsaw, Poland bDepartment of Physics and Astronomy, University of Delaware, Newark, DE 19716, USA
Symmetry-adapted perturbation theory (SAPT) has been shown to provide accurate intermolecular potentials which performed very well in numerous applications to spectroscopy and to simulation of bulk properties. The high accuracy of the SAPT approach requires, however, that the effect of the intra-monomer electron correlation on the major components of the intermolecular potential (electrostatic, exchange, induction, and dispersion) is properly included. In present SAPT codes these effects are accounted for by computationally costly expansion in powers of the monomer fluctuation potentials or by, even more costly, coupled cluster iterative techniques. We show that by adopting the density-functional theory (DFT) description of monomers, it is possible to include the intra-monomer correlation effects with great accuracy and at a very modest computational cost. Specifically, by evaluating the simplest zeroth-order SAPT expressions with the Kohn-Sham orbitals and orbital energies, we effectively perform an infinite-order summation of SAPT expansion in powers of monomers' fluctuation potentials and obtain very accurate values of the electrostatic, exchange, and induction energies. This accuracy is possible, however, only if the exchange correlation potential Vxc(r) is asymptotically corrected at large r to compensate for the lack of the derivative discontinuity in the bulk region. The dispersion part of the interaction energy is obtained from a generalized Casimir- Polder formula evaluated with monomers' dynamic density susceptibilities provided by the time- dependent DFT approach (also employing asymptotically corrected exchange correlation potentials). The method recovers the dispersion energies of He, Ne or H2O dimers to within 3% or better and leads to very accurate interaction potentials in the whole van der Waals region. It appears that except for the helium dimer, where an explicitly correlated approach is possible, the obtained accuracy is better than that achievable using currently most advanced wave-function techniques. The method has been implemented using the density-fitting procedure, which leads to cubic dependence of the computational effort on the employed basis set size. Due to this favorable scaling, the resulting computational procedure, referred to as SAPT(DFT), is applicable to study weak interactions of large polyatomic molecules.
Conference on Current Trends in Computational Chemistry 2003 81
Computation of Conventional Strain Energy in the Thiazetidines
Adria Johnson1, Noel Matthews2, and David H. Magers3
1Department of Chemistry, Tougaloo College 2Department of Chemistry, Fort Valley State University 3Department of Chemistry and Biochemistry, Mississippi College
The conventional strain energies 1,2-thiazetidine (Figure 1) and 1,3-thiazetidine (Figure 2) are determined within the isodesmic, homodesmotic, and hyperhomodesmotic models. Optimum equilibrium geometries, harmonic vibrational frequencies, and corresponding electronic energies are computed for all pertinent molecular systems using SCF theory, second- order perturbation theory, and density functional theory (DFT). The DFT functional employed is Becke’s three parameter hybrid functional using the LYP correlation functional. Two basis sets, both of triple-zeta quality on valence electrons, are employed: 6-311G(d,p) and 6- 311+G(2df,2pd). Finally, the calculated strain energies are compared to those of cyclopropane, cyclobutane, 1,2-oxazetidine, and 1,3-oxazetidine. We gratefully acknowledge support from NSF EPSCoR (EPS-0132618) and the 2003 CCMSI Summer Institute.
Figure 1. Figure 2. 1,2-thiazetidine 1,3-thiazetidine
82 Conference on Current Trends in Computational Chemistry 2003
Computational Analysis of Endo vs. Exo Retinoid Compounds
Candace L. Jones,1,2 Morgan S. Ponder2 and Tracy P. Hamilton1
1Department of Chemistry, University of Alabama at Birmingham, Birmingham, AL 35294 2Department of Chemistry, Samford University, Birmingham, AL 35229-2236
(9Z)-Retinoic acid (1) is the natural ligand for one class of nuclear retinoid receptors, the RXRs.
7 9 5 11
13 1
15 O
Structural analogs of retinoic acid have received considerable recent attention because of their potential use as chemotherapeutic and chemopreventive drugs for cancer. However, their further use is hindered by undesirable side effects. One possible cause for these side effects is the ability of these molecules to bind at multiple sites on the receptors. This ability to bind at multiple sites may stem from the torsional flexibility of the molecules. To reduce this flexibility, rotationally hindered retinoids have been synthesized. Compounds represented by structure 2 are intermediates in the synthesis of one series of these rotationally hindered retinoids. Possible side products in the formation of these intermediates are compounds represented by structure 3.
( )n ()n
7 910 7 9 10 6 6 8 8
O O 2 3
The goal of the work presented here is to calculate the relative energies of compounds represented by 2 (exo compounds) and of compounds represented by 3 (endo compounds). Compounds containing five-, six-, and seven-member aliphatic rings (n = 0, 1, and 2 respectively) were studied. Spartan was used to construct the molecules and perform conformational analysis with the Merck Force Field (MMFF). One of the dihedral angles in the side chain (7-8-9-10 in the exo compounds and 6-7-8-9 in the endo compounds) was rotated in increments of 10ْ and minimized at each step to ensure that no low-energy conformers were overlooked. Each conformer identified in this manner was then optimized at the B3LYP/6-31G* level of theory using Gaussian 98 or PQS version 2.5. The lowest-energy conformers that we Conference on Current Trends in Computational Chemistry 2003 83
found for each intermediate were taken as a base structure, and the side chain was extended to form the full retinoic acid analog. These full retinoic acids were then re-optimized at the B3LYP/6-31G* level of theory. Results are shown below.
Molecule Conformer Energy (au) Relative E % Dihedral (kcal/mol) Distributiona Angle 5 Exo Conf # 1 -617.117481 0.0 70.3% -51.2 Conf # 2 -617.116647 0.52 29.2% 150.2 5 Endo Conf # 1 -617.112219 3.30 0.3% -34.1 Conf # 2 -617.111958 3.47 0.2% 17.1 6 Exo Conf # 1 -656.432208 0.0 89.0% 52.9 Conf # 2 -656.430227 1.25 10.8% -64.0 6 Endo Conf # 1 -656.426160 3.80 0.1% -48.8 Conf # 2 -656.424691 4.72 <0.1% 54.9 7 Exo Conf # 1 -695.735399 0.0 51.6% 61.9 Conf # 2 -695.734200 0.12 42.1% -77.3 7 Endo Conf # 1 -695.732802 1.63 3.3% 62.9 Conf # 2 -695.732704 1.69 3.0% -136.9 Conf # 3 -695.724186 7.04 <0.1% -71.6 aCalculated from Boltzmann distribution.
Molecule Conformer Energy (au) Relative E Dihedral Angle (kcal/mol) 5 Exo Conf # 1 -886.492125 1.09 -52.5 Conf # 2 -886.493874 0.0 180.9 6 Exo Conf # 1 -925.806896 0.0 53.6 Conf # 2 -925.804564 1.46 -66.7 7 Exo Conf # 1 -965.110144 0.0 59.7 Conf # 2 -965.108379 1.11 -91.6
84 Conference on Current Trends in Computational Chemistry 2003
5+ Low-Energy Single-Electron Capture in B and H2 Collisions
Dwayne C. Joseph and Bidhan C. Saha
Department of Physics, Florida A. & M. University Tallahassee, FL 32307
There are not many theoretical investigations for charge transfer involving ions and H2. The presence of additional degree of freedom for molecular targets poses many practical difficulties than their atomic counterpart [1]. Collision of multiply charged ions with molecules is common in gastro-physical plasmas. The charge-exchange plays an important role in ionizing balance, in the production of radiative energy-loss and cooling from impurity ions. H2 is the simplest molecules. The charge-exchange with H2 provides a re-combination mechanism for multiply charged ions in X-ray ionized astronomical environments [2]. The recent X-ray emission data from comets has drawn considerable attention on the role of charge exchange. It is now believed that the charge exchange between solar wind and cometary’s gases is the most likely mechanism for X-ray emission. To understand this one must have the detailed information on the collision cross sections. In collisions involving neutral targets and multiply charged ions the process of single electron capture dominates over the other competing channels [3]. In addition to the total electron capture cross sections, their l-distribution demands a very through investigation since they can provide valuable information about the structure of the projectile core. We report calculations of state-selective electron capture by B5+ from molecular hydrogen at low energies. We use, in the impact parameter formalism, the semiclassical close-coupling method with molecular-state expansion embodying the electronic translation factors. The electronic wave functions and the electronic energies ε(R) are obtained by the LCAO (Linear Combination of Atomic Orbitals) method using the Slater-type-orbitals (STO). During the collision, a transient quasi-molecule comprising the entrance and various exit channels of the charge-transfer reaction is considered. Freezing the molecular details of the target, the effective binding of the active electron inside the transient molecule is simulated by pseudo-potential. In order to solve the time-dependent Schrödinger equation for the system, we expand the total wave function in terms of the product of electronic wave functions and ETF (Electron translation factors) [4]. The ETF is essential not only to satisfy the Galilean invariance but also to account for the physical transfer of the electron during the course of the collision from one core to the other. To obtain the coupled equations for the amplitudes an (t) we substitute the wave function into the time dependent Schrödinger equation; multiplying by its complex conjugate from the left we integrate the resulting equation with respect to electronic coordinates. Expanding the ETF in powers of V and retaining the first order terms we obtain the following close-coupled equations:
r r r iaÝ n = ∑ V .(P + A ) kn ak +ε n an (2) k≠n Here P and A represent the nonadiabatic coupling matrix elements and the ETF corrections, respectively. P and A are the sum of radial and angular couplings. For each contributing impact parameters (b) these equations are solved numerically to obtain the transition probability from the initial state to a particular (k th) final state: 2 Pk (E, b) = ak (+∞,b) (3) The cross section s are evaluated using the following relation: Conference on Current Trends in Computational Chemistry 2003 85
Qk (E) = 2π∫ db.b.Pk (E,b) (4)
Details results will be presented at the conference.
This work was supported, in part, by the Research Corporation, the NASA, and the Army High Performance Computing Research Center under the auspices of the Department of the Army, Army Research Laboratory, the content of which does not necessarily reflect the position or the policy of the government, and no official endorsement should be inferred.
References 1. B. H. Bransden and M. R. C. McDowell, “Charge-exchange and the Theory of Ion-Atom Collisions” (Oxford Science, NY) 1992. 2. B. C. Saha and A. Kumar, J. Mol. Struct. Theochem 487, 11 (1999). 3. A. Kumar and B. C. Saha, Phys. Rev. A 59, 1237 (1999). 4. D. R. Bates and R. McCarroll, R. Proc. Roy. Soc. (London) A 247, 158 (1958). 86 Conference on Current Trends in Computational Chemistry 2003
Women in Science and Engineering. The Untapped Resource in Many Countries
Isabella Karle
Naval Research Laboratory Washington, D.C.
The participation of women in the technological aspects of our world has had a long history, beginning in ancient times, albeit a sparse and frustrating history. I will share with you a number of vignettes to illustrate that through the ages women had succeeded, despite unfavorable or unusual circumstances, to make significant contributions to theoretical and practical aspects of science and technology. Conference on Current Trends in Computational Chemistry 2003 87
Data Base of Quantum Mechanical and Regularized Force Constants in Redundant Internal Coordinates
I.V. Kochikova, G.M. Kuramshinab , D.A. Sharapovc, S.A. Yagolac
aScientific Research Computer Centre, Moscow State University bDepartment of Physical Chemistry, Chemical Faculty, Moscow State University, cDepartment of Mathematics, Physical Faculty, Moscow State University Moscow 119992, Russia
The main problems in organizing a data base of molecular force constants are connected with the properties of the inverse vibrational problem. It is a problem of determining parameters of the molecular force filed (force constants which constitute a positive definite matrix F) from given experimental data (vibrational frequencies for isotopomers, Coriolis constants, centrifugal distortion constants, etc.). Data on molecular force fields are used for different theoretical aims, e.g. to check the validity of various model assumptions commonly used by spectroscopists for approximating the potential function and helping to understand the origin and nature of interatomic forces. They are also useful for practical purposes of predicting vibrational frequencies of other molecules including those not yet observed. It is well known that the inverse vibrational problem has non-pleasant mathematical properties related to its ill-posedness [1], i.e. it may have non-unique solution or no solutions at all, and solution is unstable with respect to the errors in experimental data. In a general case for the N-atomic molecule, it is necessary to determine n(n+1)/2 elements (n = 3N - 6) of the force constant matrix from n vibrational frequencies. The nonuniqueness and instability of the solution often lead to significant differences in the force field parameters obtained in different investigations, as well as to difficulties in comparing and transferring force constants. In the modern approaches for the calculation of empirical molecular force fields based on Tikhonov regularization theory [1] an inverse vibrational problem is formulated as a problem of finding the so-called normal solution (or normal pseudo(quasi)solution in the case of incompatibility of input data) of a nonlinear operator equation. The desired solution is a matrix F that reproduces experimental data within given error level and is the nearest in the Euclidean metrics to some given matrix F0 ∈ Z (Z is a set of possible solutions). All explicit and implicit model considerations concerning the form of force field may be taken into account by the choice of certain matrix F0 (or stabilizer of Tikhonov's functional) and a preassigned set D of a priori constraints on the values of the force constants. This set defines a form of matrix F in the framework of the desired force field model (i.e., with specified zero elements, equality of some force constants, etc.). If no a priori data constrains the form of solution, then D coincides with the set Z. In the Tikhonov regularizing procedure, one could increase the stability and accuracy of the calculated solution Fα by using: a) an extended set of experimental data (including, e.g., Coriolis constants, mean square amplitudes, etc.); b) an improved choice of the stabilizer matrix F0; c) an improved choice of the constraints set D. As a particularly effective choice of stabilizer, we have proposed [7] to use an ab initio quantum mechanical F0 matrix in the regularizing procedure. This leads to the concept of regularized quantum mechanical force field (RQM FF), defined as the force constant matrix that 88 Conference on Current Trends in Computational Chemistry 2003
is nearest to a corresponding quantum mechanical matrix F0 and reproduces experimental frequencies within given error level. The rapid development of quantum mechanical methods in the past decades (higher levels of ab initio approach and very successful DFT methods) greatly accelerates applications to larger molecules and opens a possibility to organize a data base of theoretical and regularized force constants for many classes of compounds. For this purpose the choice of a system of generalized coordinates is very important. Regularizing algorithms permit use of any system of generalized coordinates – Cartesian, internal or symmetry coordinates, including redundant systems of internal coordinates. The latter choice considerably facilitates the transfer and comparison of force constants between related molecules. The theoretical basis and practical aspects of using redundant coordinates were previously discussed and the examples of effective application of these algorithms for the calculation of RQMFF for different types of molecules were presented elsewhere [2, 3 and references there]. The effective regularizing algorithm for finding a scaled force constant matrix in redundant system of internal coordinates was proposed in Ref. [4]. One of the most important features of the organized database is the uniform principle of choosing the internal coordinates and corresponding force constants. The redundant system includes all internal coordinates for bonds stretching, valence angles and other vibrational degrees of freedom defined in accordance with Ref. [5]. The well known numerical problems arise when redundant coordinates are used: while in the independent system of generalized coordinates, the matrices T (of kinetic energy) and F are determined in a unique way, in the redundant system of coordinates there exist classes of these matrices which may be determined as
S *FS = F , S *TS = T , (1)
where S is n × n matrix relating the independent and redundant coordinates (q1,..., qn;n = 3N − 6) and redundant (q1,..., qn;n > 3N − 6 ) coordinates:
q = Sq . (2)
Linear dependencies between coordinates following from Eq.(2) may be expressed as
Wq = 0
where W is the m × n matrix, with m = n − (3N − 6) . The class of matrices satisfying conditions (1) is well known [6] and may be expressed in the form
F = F0 + B *W +W * B , (3)
where F0 is any of the matrices F satisfying Eq. (1), and B is an arbitrary matrix having the same dimensions as W. A similar formula may be written for kinetic energy. It has been shown earlier [4] that for preserving the standard form of vibrational calculations in the case of redundant coordinates, it is sufficient to satisfy the condition WGF = 0 (4) where G is any pseudo-inverse matrix to T. Among all possible sets of matrices G and F in redundant coordinates there are two special so-called canonical matrices, which have rank 3N - 6. In the case of calculations in internal coordinates, the canonical (m-fold degenerate) matrix G Conference on Current Trends in Computational Chemistry 2003 89
is always used, and thus no limitations on the F matrix are imposed. We may define a canonical matrix F in internal coordinates by the equation WF = 0 . (5)
This condition along with Eq. (1) or (3) determines the matrix F in a unique way. This may not always be a preferable choice of the unique matrix because matrix F determined in this way is a normal one on the set of all possible matrices F – that is, having a minimal norm 1/ 2 n F = f 2 . ∑ ij i, j=1 Certain force field models are aimed at collecting the maximum possible part of the interaction energy in a limited number of force constants fij , while a large fraction of force constants corresponding to remote interactions in a molecule is taken to be zero (MVFF). Such F matrices are far from being normal, and it is necessary to transform the canonical matrix to the one corresponding to the chosen model of force field. In our software package SPECTRUM [2], we have implemented the following algorithm of transforming the quantum mechanical force constant matrix in Cartesian coordinates to a canonical F matrix in redundant system of internal coordinates. In independent coordinates, such transformation is possible using matrix A such that n r ∆ri = ∑ Ak qk (i = 1, ..., N ) (6) k =1 or β n β ∆r = A q (i = 1, ..., N ) , i ∑ ik k k =1 where index β = 1, 2, 3 corresponds to x, y, z. The force constant matrix in internal coordinates is expressed as N 3 N 3 β γ F = A A F , (7) kl ∑∑∑∑ ik jl iβ , jγ i====1βγ11j 1 where ∂2F Fiβ , jγ = β γ eq (β ,γ = 1,2,3) . ∂∆ri ∂∆rj Matrix A may be obtained from a relation r 1 n r Aik = ∑ Bi,lTlk (8) mi l=1 These Cartesian displacements satisfy the Eckart’s conditions. If we use the redundant system, then formulas (6)-(8) are still true if in Eq. (8) the matrix T is understood as the normal pseudoinverse to matrix G (which has a rank 3N-6). In this case one achieves only one of many possible matrices F satisfying conditions 1-3), namely, the canonical F0, its rank being equal to 3N-6. Apart from obtaining canonical matrix F, in the SPECTRUM we have implemented two more possibilities for transforming the matrix of ab initio force constants in Cartesian coordinates to a matrix in a redundant system of internal coordinates. Both of them are based on Eq. (3) which allows to reproduce exactly the ab initio frequencies, while satisfying one of the two possible model assumptions. 90 Conference on Current Trends in Computational Chemistry 2003
1. A matrix in internal coordinates is a normal matrix with respect to a given F (0) . To obtain this matrix it is necessary to solve the following system of equations for coefficients Bki (k = 1,...,m; i = 1,...,n) : n m ∑∑Qlj,kiBki = Dlj (l = 1,..., m; j = 1,..., n) , (9) i=1k =1 where n Qlj,ki = WliWkj + δij ∑WlβWkβ , β =1 n (0) Dlj = − ∑ Wli ( fij − fij ) . i=1
(0) (0) Here fij are elements of matrix F ; fij are elements of F0, and δij is a Kronecker symbol. In a case of searching for the canonical matrix we choose matrix F (0) as equal to zero.
2. Another choice is to search for a matrix with the minimal off-diagonal norm. This is an attempt to collect the largest possible part of the interaction energy in diagonal elements of the matrix F. In this case we solve the same system (9) with elements
Qlj,ki = WliWkj (i ≠ j) ,
Qlj,ki = ∑WiβWkβ (i = j), Dlj = − ∑Wlβ fβj . β ≠ j β ≠ j
Obviously, the difference between these two sorts of force constant matrix is concentrated in the elements related to the redundant internal coordinates for which the matrix with the minimal off- diagonal norm is close to the quasi-diagonal. This form of matrix F might be preferable in a case of constructing force fields for the large molecular systems.
Acknowledgement. This work is partly supported by the RFBR grants 03-07-96842 (region 2003 Yugra) and 01-03-32412.
References 1. Tikhonov A.N., Leonov A.S., Yagola A.G.// Nonlinear Ill-posed Problems. London: Chapman and Hall, 1998, 387 p. 2. Yagola A.G., Kochikov I.V., Kuramshina G.M., Pentin Yu.A. //Inverse Problems of Vibrational Spectroscopy. Zeist: VSP, 1999, 297 p. 3. Kuramshina G.M., Weinhold F., Kochikov I.V., Pentin Yu.A., Yagola A.G. //J. Chem. Phys. 1994, 100, 1414. 4. Stepanova A.V., Kochikov I.V., Kuramshina G.M., Yagola A.G. //Regularizing scale factor method for molecular force field calculations. Computer Assistance for Chemical Research. International Symposium CACR-96. 1996. Moscow, p.52. 5. Pulay P., Fogarasi G., Pang F., Boggs J. //J. Am. Chem. Soc. 1979, 101, 2550. 6. Groner P., Guntard M. S.H. // J. Mol. Spectrosc., 1976, 61, 151. 7. Kuramshina G.M., Weinhold F. // J. Molec. Struct. 1997, 410-411, 457. Conference on Current Trends in Computational Chemistry 2003 91
Self-Consistent Model for the Joint Treatment of Spectroscopic and Electron Diffraction Data
I.V. Kochikova, G.M. Kuramshinab , D.A. Sharapovc, S.A. Yagolac
aScientific Research Computer Centre, Moscow State University bDepartment of Physical Chemistry, Chemical Faculty, Moscow State University, cDepartment of Mathematics, Physical Faculty, Moscow State University Moscow 119992, Russia
We consider the general problem of extracting internal molecular parameters (such as geometrical structure and properties of the intramolecular force field) from the available experimental data on infrared spectroscopy, electron diffraction analysis, microwave spectroscopy, etc. and guided by the ab initio calculations. An implementation of such generalized approach is hindered by many well-known problems:
1. Analysis of different experimental data is often carried out using different molecular models. For example, ab initio calculations and most spectroscopic studies use equilibrium geometry data, while ED studies directly provide only thermally averaged values. Hence the results obtained for the same molecule using different experimental techniques may prove incompatible. 2. Spectroscopic data is often insufficient to restore the complete force field, thus making it necessary to introduce model assumptions restricting the molecular force fields. 3. Electron diffraction data is often insufficient to determine all structural parameters, especially when a molecule possesses a set of similar interatomic distances; this also implies necessity of introducing external constraints on molecular geometry. 4. Quantum mechanical calculations often lack accuracy to match the experimentally measured values. For example, an approach of scaling an ab initio force field is widely used to achieve a reasonable agreement between calculated and measured frequencies. 5. Data on vibrations anharmonicity cannot be readily obtained from experimental data unless for very small and simple molecules. This leads to different model evaluations (or to the usage of ab initio values) that may lack necessary precision. 6. A general problem of involving quantum mechanical data in analysis gives rise to methodological problems: we have to rely on them when experimental evidence is insufficient and, on the other hand, it is obvious that the results of these calculations should not be given the same priority as the experimental data.
We suggest the scheme for the analysis of the molecules with relatively small atomic excursions that takes into account anharmonicity and successfully deals with curvilinear motions. It is referred to as a “small amplitude” approximation, to distinguish between this scheme and a more elaborate one that is capable of dealing with bending motions and internal rotation with fairly large amplitudes. The essence of this scheme is as follows:
1. A common molecular model is created that connects molecular parameters and experimentally measured values. Within this model, the molecular parameters to be defined are equilibrium geometry and force field parameters. All experimentally measured values are calculated using the same set of parameters, and these parameters are adjusted so as to fit the experimental evidence. It is important that all experimentally measured values are determined from the same 92 Conference on Current Trends in Computational Chemistry 2003
set of parameters. For example, the ED analysis will therefore obtain amplitudes compatible with spectroscopic evidence, and spectroscopy analysis will obtain frequencies compatible with the equilibrium geometry provided by the ED and microwave data.
2. Every time when experimental data is insufficient for the unique determination of some or all molecular parameters, we should employ some kind of external knowledge or experience. In accordance with the basics of regularization theory, we suggest to choose the solution that is in a certain sense nearest to some a priori chosen parameter set. This set may not necessarily conform to the experiment, but should be based on data complementary to the experiment. The external evidence may be derived from some general ideas (for example, molecular force field models, or data on similar molecular structures), or, preferably, be based on ab initio data. Within this approach, the results will tend to be as close to quantum-mechanical data as the experiment allows. From mathematical point of view, the algorithm should provide approximations to the solution that tend to the exact solution when experimental data becomes more extensive and accurate.
These “soft” constraints may be combined with a more rigid set of constraints imposed on the solution to obtain the unique solution. For example, when there are close interatomic distances in a molecule, it is a common practice to determine only one of them from an ED experiment, fixing all differences between the distances at ab initio values. This may be called a “rigid” approach. Within a “soft” approach, it is possible to find a solution that will have required properties unless it does not contradict to experiment (and if it does, we shall find a solution which is the nearest to the one with that properties).
Let us for simplicity limit the set of available experimental data by the vibrational frequencies (ω) and normalized electron scattering intensity (M). In a general case there may be more experimental data (rotational constants, etc.). All experimentally measured values depend on both equilibrium geometry (R) and force field (F) parameters. Harmonic vibrational frequencies can be obtained from the force constant matrix (the matrix of the second derivatives of the molecular potential at the point of equilibrium). To calculate anharmonic corrections to frequencies, we need cubic and quartic terms in the potential expansion. In many cases, however, anharmonic effects are relatively small and may be neglected during a normal coordinate analysis. The ED intensity shows only a moderate dependence on force field, while geometry changes are of much greater importance. However, the ED experiment is directly related to the thermally averaged interatomic distances rather than to their equilibrium values, and difference between these two kinds of distances depends on the anharmonic terms of the potential. Hence, calculation of ED intensities requires knowledge of the harmonic and anharmonic terms of the molecular potential (at least cubic terms).
Before the formulating a problem of determining the molecular parameters, it makes sense to analyze what amount of data we are capable to find from a given set of measured data. Indeed, if we introduce a very limited set of parameters (and thus create a very “rigid” model), we are likely to fail achieving a good fit between experimental and calculated data. On the other hand, if a model is very flexible (that is, contains too many adjustable parameters), we are likely to find a wide variety of solutions that all satisfy the experiment. Even if we employ the concept of a regularized solution, there must exist some kind of optimal parameter set that would correspond to the available experimental data. As for the force field determination, it is a common knowledge that (except for a limited set of small or very symmetrical molecules) we never have enough data to restore a complete force field. The ED data usually provide only a small additional data on force field, so as a rule we are in the situation when there exist a wide range of Conference on Current Trends in Computational Chemistry 2003 93
force fields compatible with spectroscopic experiment. Among the ways to reduce the ambiguity of the force fields, we could mention the following: 1. Introducing model assumptions based on general ideas of molecular structure (e.g. valence force field, etc.): these will result in neglecting some force constants, fixing the others, and/or introducing model potentials that would allow to generate force matrix depending on a small number of parameters. 2. Transferring some force field parameters from similar fragments in related molecules and assuming they are not likely to be significantly changed in a different environment. 3. Applying scale factors technique when all allowed force matrices are obtained from ab initio values by the certain scaling procedure. These factors may be treated as force field parameters to be determined.
All of these ways may be formulated as the “rigid” restrictions. Instead of introducing them, we may generate a force field F0 that possesses the above properties and attempt to find solutions nearest to F0. For the approach based on equilibrium configurations, the force field anharmonicity is of great importance because it defines difference between the equilibrium distances and thermally averaged used in electron diffraction analysis. Taking the mentioned limitations into account, we come to the following conclusions. 1. Force field parameters – at most – should include a matrix of the quadratic terms in potential expansion. This set should be even more constrained by using ab initio data (as a stabilizer matrix F0 or with the use of scaling scheme). 2. Cubic (and quartic) terms cannot be determined from the experimental data under discussion and should be somehow evaluated. It is possible to introduce some anharmonicity models or – preferably – use ab initio data. To maintain consistence with quadratic potential, these terms may require adjustment when quadratic terms are changed during the fitting procedure.
Similar problems exist for the interatomic distances determined from the ED data. Though many successful results have been provided by this technique, it’s evident that the accuracy of the obtained data may be insufficient in cases when a molecule possesses a number of bond lengths that are different in chemical nature but close to each other in their values. Here, again, we need to introduce model assumptions that, at best, are based on the ab initio calculations. Under certain unfavorable conditions, the ED data may be insufficient even to define symmetry of the equilibrium configuration, which in this case should be obtained from alternative sources.
Taking all this into account, we come to the following formulation of the inverse problem. Let Λ be a set of available experimental data, and A(R, F) be a procedure allowing to calculate this data from the set of molecular parameters R and F. We may suppose that Λ is a finite-dimensional vector from the normalized space Rm; parameters (R, F) may also be chosen so as to constitute a vector from Rn, then A is an operator acting from Rn to Rm. Let (R0,F0) be an a priori given set of parameters (e.g. obtained from ab initio calculations), and we know that accuracy of experimental data Λ is such that it deviates from the “ideal” data not more than by a given constant δ. Let us also introduce a set of constraints D in Rn to which our solution should belong. We need to find an approximation (R, F)δ to the exact parameters (R, F) such that: 1) the solution is compatible with the experimental data within the accuracy range
(R, F)δ ∈ Zδ where Zδ = {}(R, F) ∈ D : A(R, F) − Λ ≤ δ ; 2) among all possible solutions, we choose the one most close to (R0,F0) 94 Conference on Current Trends in Computational Chemistry 2003
0 0 (R, F)δ = arg min (R, F) − (R , F ) (R,F )∈Zδ 3) when accuracy increases, we get more accurate approximations to an exact solution
(R, F)δ → (R, F) when δ → 0 . One of the possible implementation of the procedure is to obtain such approximations based on Tikhonov functional technique when we minimize 2 M α (R, F) = A(R, F) − Λ 2 +α (R, F) − (R 0 , F 0 ) (1) on the set D, and regularization parameter α is chosen as a solution of the equation
A(R, F)α − Λ = δ α where (R,F)α delivers a minimum to M (R,F).
Under certain conditions on the operator A(R, F) the existence and uniqueness of the solution can be guaranteed. Eq. (1) shows that within this approach the deviations of solution from an a priori given set may be treated as a penalty. Obviously, when (R0,F0) itself is compatible with experimental data, no further adjustment is necessary. The formulation given above is very general; in practical implementation we may, for example, assume that R is a set of independent equilibrium geometry parameters and F is a set of harmonic force constants (or the Pulay scale factors).
Summary conclusions: 1. The presented approach is aimed at simultaneous determination of the geometry and force field parameters of a molecule. It combines techniques previously used in IR spectroscopy and ED data analysis. In particular, it allows using more flexible force field models when fitting ED data, far beyond the usually employed scaling of the ab initio force field. 2. Ab initio data (or any other external data) is automatically “weighed” so as to serve an additional source of information when data supplied by the experiment proves insufficient. There is no need to supply ab initio data with some kind of assumed errors, etc. 3. Molecular geometry is defined in terms of equilibrium distances thus allowing compatibility with spectroscopic models and ab initio calculations. Besides, the self- consistency of geometrical configuration is automatically maintained at all stages of the analysis.
Acknowledgement. This work is partly supported by the RFBR grants 01-03-32412 and 03-07- 96842 (region 2003 Yugra). Conference on Current Trends in Computational Chemistry 2003 95
Van der Waals Energies and Time-Dependent Density Functional Theory
Walter Kohn
Department of Physics, University of California, Santa Barbara Broida Hall, Building 572, Santa Barbara, CA 93106 USA
DFT, in principle, includes VdW energies, but approximations rooted in the LDA, such as GGA’s and WDA’s do not. This talk will describe recent and ongoing work to use TDDFT to calculate VdW attractions between two systems of interacting atoms (molecules, clusters, solids, etc., of arbitrary sizes and shapes. 96 Conference on Current Trends in Computational Chemistry 2003
Quantum-Chemical Research of Chemical of Elementary Radical Reactions
V.V. Kukueva
Fire Safety Institute, Onoprienko str. 8, Cherkassy, 18034, Ukraine
Improvements in techniques to control fires may aid in ameliorating fire problems. Water, the extinguishing of antiquity, remains the major material applied. However, water additives, to produce foams, inert gases such as CO2 and N2, chemical suppressants containing halogens and dry powders of various types have countered increasing utility. Improved knowledge of mechanisms by which these agents operate may help in advancing suppression strategies. The scope of the theoretical research is limited to chemical suppressants acting mainly on the hydrogen flames. It is well known that the flame propagation mechanism on hydrogen-air flames at the pressure above one atmosphere the clue role plays hyper equilibrium concentration atoms and radicals connecting with chain brunching reaction. Addition a few inhibitor amount, which able to catch (scavenge) the radicals can significant to reduce combustion speed. It is known that addition of halogens to flames promotes extinction at least partially through modification of the chemical kinetics. There are experimental and theoretical evidences [1, 2], that reaction products of CF3Br must be directly responsible for this chemical inhibition. Since similar effects are produced by other bromine-containing compounds, it is reasonable to assume that product molecules containing Br interact with the chain mechanism of combustion to slow the overall rate. It would be useful to have overall kinetic parameters that describe the inhibition effect of Br. In this paper C2F4ClF and C2F4ClCl (242) and also well known flame inhibitor C2F4Br2(2402) have been investigated. Choice of objects have been caused such fact that the similarity of the compounds isn’t provided analogical inhibition effect. The quantum-chemical calculations by the SCF MO LCAO in the frame approximation MNDO [3] by program [4] the destruction ways a number of known and studying inhibitors have been carried to study effect of the substance composition and structure on inhibition activity in hydrogen flame and also for investigation inhibition mechanism. At first the destruction ways of all investigated fire suppressants have been calculated for proof the radical mechanism of probable interactions. The analysis has been carried out by the comparison of enthalpies of particles. For the next stage the collision complexes between destruction products and active centers of flame (Н., ОН., О.) have been calculated to research scavenging effectiveness of these substances. The interaction probability has been estimated by the depth of minimum on the potential curves (dependence potential energy from distance between particles). The relative stability of the short living complexes have been determined by the comparative analysis of bind lengths and energies. Conference on Current Trends in Computational Chemistry 2003 97
Destruction products’ enthalpies (∆Hf кJ/mol) of investigated fire suppressants Fire suppressants Destruction products ∆Hf кJ/mol + - I. CF3-CF3 CF3 + CF3 221,68 . . CF3 + CF3 22,16 · · F + C2F5 86,7
· · II. СF2Cl-CF2Cl С2F4Cl + Cl 44,69 · · СF2Cl + CF2Cl 19,29
· · III СF2Br-CF2Br СF2Br + CF2Br 54,63 . . С2F4Br + Br 66,37
· · IV СF3-CF2Br С2F5Br + Br 30,01 . . С2F4Br + F 79,21 · · СF2Br + CF3 16,21
· · V. СF3-CF2Cl C2F5 + Cl 44,84 · · СF3 + CF2Cl 18,39
As we can see from table the homolitical decay much more preferable for all investigated fire suppressants therefore we can to suggest the possibility of participation of these particles in branch radical flame spreading reactions. Among the all substances, C2F4Br2 gives more active inhibiting particles which have smaller values of enthalpies. Apart from that the two bromine atoms can to catch atomic hydrogen by reaction: H. + Br. → HBr. By these facts we can explain the strong inhibiting effect of this suppressant. But contrary for traditional point of view our · research also shows that the most important particle for all investigated compounds is СF3 . The enthalpy has the least values for all cases of destruction. And besides this particle well interacts with all active centers of flame. The binding energy CF3-ОН, for example, is 6, 67 eV. By the substitution only one atom on Cl and Br, lead to inert behavior of CF2Cl and CF2Br to active radicals, which promote to spreading of flame. We can do some conclusions about stabilizing acting of substituents by analyzing of interatomic distances C-halogen. Thus, distance С-Cl in molecule C2F5Cl is R=1,227 Å, аnd С-
F is R=1,346 Å. As we can see, RC−Cl at 0,119 Å less, than С-F, in spite of known fact that radius of chlorine atom some more. Thus, we can conclude that chlorine atom stronger binds in molecule. Fluorine as more mobile substituent can light abstract and scavege atomic hydrogene to stable molecule HF. All this results is in agreement with experimental date.
References: 1. Williams F.A., Fire Safety Journal, 3 (1981) 163-175 2. Noto T, Babushok V., Hamins A., and Tsang W, Inhibition Effectiveness of Halogenated Compounds, Comb. and flame 112: 147-160 (1998) 3. Dewar M.J.S. and Thiel W. Ground states of molecules. 38. The MNDO method. Approximations and parameters // J.Amer. Chem. Soc. – 1977. – V.99, N15. – P.4899-4907. 4. Pilipenko А.Т., Zaetz V.А. Journal of structure chemistry (Russian). – 1987. – Т. 28, № 5 – p. 155-156 98 Conference on Current Trends in Computational Chemistry 2003
Modeling ZrxSi1-xO2: The Effects of Unique Bonding Arrangements
H. A. Kurtz and N. P. Labello
The University of Memphis, Department of Chemistry, Memphis, TN
By increasing the number of circuits on a chip, while improving efficiency and decreasing size, the industry has managed to advance microelectronic technologies at an extreme rate. The improvement trend follows Moore’s Law, and has held for nearly four decades. For most of this time, silicon dioxide has fulfilled the crucial role of gate oxide on the transistors that make up most such devices. The gate oxide serves as an insulator, preventing leaking and tunneling electrons from ruining the transistor. The SiO2 layer is the smallest component of the MOSFET chip. Positioned between the metal layer and the semiconductor, the SiO2 layer is thinned with each scaling of the chip. The layer has shrunk to an atomic scale on recent chips, as demonstrated in Figure 1. Further shrinking, to a layer less than a few atoms across, will lead to the breakdown and loss of electrical insulation, a fundamental physical barrier inherent in pure silicon dioxide technology.
Figure 1. Scaling the SiO2 gate oxide layer any further will soon be impossible, as the atomic structure does not maintain the insulating properties of the bulk material.
One solution to this problem is to replace silicon dioxide with a new material. This would, however, require redesigning the transistor to cope with a new insulator interface. As the industry has decades of experience with silicon dioxide, a more favorable solution would be to extend the ability of SiO2 to insulate without significantly altering its other chemical properties. This can be accomplished by raising the dielectric constant of the material, allowing a thicker layer to be used. The dielectric constant, κ, is a function of molar polarizability and molar volume.
κ = f (αm /Vm ) (1)
Simply stated, the dielectric constant of a material rises if the polarizability increases or the molar volume decreases. Adding certain defects in very low concentration, such as replacing a silicon atom with a zirconium atom, increases the overall polarizability, and therefore, the dielectric. However, Figure 2 shows that in the case of zirconium and some other metals, a considerably higher rate of increase in the dielectric with respect to the concentration of the defect is present.
Conference on Current Trends in Computational Chemistry 2003 99
Figure 2. The lower dashed line is a straight line extrapolation drawn from the dielectric constant of SiO2 to ZrO2. The dotted line accounts for the dielectric enhancement and is verified by the marked experimental points.
G. Lucovsky and G. B. Rayner, Jr, Appl. Phys. Lett 77, 2912 (2000)
One explanation for this large increase in dielectric is due to a possible drop in the molar volume in systems where the defect is present. To examine this possibility, first principle quantum mechanical calculations are being applied to ZrxSi1-xO2 systems, where x is typically less than 0.1. A variety of structures have been discovered that suggest the zirconium defect does not simply replace four coordinate, tetrahedral silicon in the amorphous solid. Lower energy, more stable bonding schemes have been calculated for five, six, and seven coordinate zirconium. One scenario for the extra bonds is explained by VSEPR theory. A five coordinate zirconium takes the trigonal bipyramidal form. A six coordinate structure makes an octahedron. The additional bonds lower the overall energy of the system, while increasing the molecular density. Both of the structures have proven stable enough to survive rigorous QM/MM calculations. The zirconium, its first nearest neighbor oxygens, and second nearest neighbor silicon atoms are modeled with density functional theory. This small group of core atoms is inserted into a cluster of silicon dioxide containing hundreds of additional atoms. The surrounding atoms are modeled with molecular mechanics, and simulate the bending strains and torsions found in the actual amorphous ring systems that make up silicon dioxide. The successful QM/MM calculations prove that the extra bonds of zirconium and 3-coordinate oxygen are indeed stable and capable of holding the system together in a smaller, denser form than silicon alone. Figure 3 demonstrates one such cluster.
Figure 3. A stable 6-coordinate octahedron zirconium structure. The zirconium and immediately surrounding silicon and oxygen are modeled in QM. This core is embedded in amorphous silicon dioxide, modeled with molecular mechanics.
Further study of the small ZrxSi1-xO2 clusters have recently yielded new, very interesting, alternate five and six coordinate zirconium structures. A zirconium atom and a silicon atom can share two oxygens to form a four atom ring as in Figure 4.1. Beginning with a four coordinate zirconium that participates in at least one of the described four atom rings, a fifth and sixth oxygen can be coordinated to the zirconium. In this form, the extra bonds are shorter and stronger than in the octahedron geometry. This structure is also stabilized by very unique pentavalent silicon atoms. A silicon atom attached directly to a three coordinate oxygen can 100 Conference on Current Trends in Computational Chemistry 2003
provide further stabilization by coordinating to and forming a bond with one of the strained oxygen of the 4 atom ring. The additional bonds from the two pentavalent silicon atoms and the hexavalent zirconium in Figure 4.2 below form a small, stable, and very dense planar ring system.
Figure 4.1 Tetrahedral zirconium and silicon Figure 4.2 The zirconium from figure 4.1 share two oxygen instead of the usual one coordinates to six oxygen, creating a planar oxygen to form a 4 atom ring. ring system. On the corners of the system are unique pentavalent silicon atoms.
Both the octahedral six coordinate zirconium and the new structure with the planar ring system have a density that is about 1.5 times that of a 4 coordinate zirconium inside a 5 angstrom sphere. Inside a 3.5 angstrom sphere however, the new six coordinate structure has a density that is 1.5 times higher than that of even the 6-coordinated octahedral form. The increase in density is indicative of a drop in the molar volume. Again, in reference to equation 1, the decrease in Vm will lead to the enhanced dielectric properties found in ZrxSi1-xO2 systems. Conference on Current Trends in Computational Chemistry 2003 101
Ab Initio Studies of Performic Acid, Peracetic Acid and Methyl Performate
Charles H. Langley and Eric A. Noe
Department of Chemistry, Jackson State University Jackson. MS 39217-1510
Ab initio calculations were carried out on the Z,Z, Z,E, E,Z and E,E conformations of performic acid (H(C=O)OOH), peracetic acid (CH3(C=O)OOH) and methyl performate (H(C=O)OOCH3). The levels of calculations ranged from HF/6-31G(d,p) to MP2/6- 311++G(df,pd) and MP2/6-311++G(2d,2p). Dihedral drives for the Z,Z-to-Z,E and E,E-to-E,Z potential barriers of these compounds were carried out at the MP2/6-311++G(df,pd) and MP2/6- 311++G(2d,2p). The lowest-energy conformer of performic acid is Z,Z, with planar (Cs) geometry. The lowest-energy conformer of peracetic acid is also Z,Z, but it is slightly nonplanar with a C-O-O-H dihedral angle of about 10 degrees. The second-lowest energy conformer of both compounds is E,E, with C1 symmetry and C-O-O-H dihedral angles, calculated at the MP2/6-311++G(df,pd) and MP2/6-311++G(2d,2p) levels, of 130.2 and 123.1 degrees for performic acid and 140.2 and 135.6 degrees for peracetic acid, respectively. At these levels, the free energies (∆G) of the E,E conformers, relative to the lowest-energy conformers, are 2.36 and 2.59 kcal/mol for performic acid and 6.23 and 5.94 kcal/mol for peracetic acid. In contrast, the lowest-energy conformer of methyl performate calculated at the MP2/6-311++G(df,pd) level has optimized O=C-O-O and C-O-O-CH3 dihedral angles of –5.0 and 95.2 degrees, respectively. The second-lowest energy conformer calculated at the same level is E,E, with a C-O-O-CH3 dihedral angle of 130.4 degrees and a relative free energy of 0.28 kcal/mol. At the MP2/6-311++G(2d,2p) level, the lowest-energy conformer is E,E, with a C-O-O-CH3 dihedral angle of 120.4 degrees; the second-lowest energy conformer has optimized O=C-O-O and C-O-O-CH3 dihedral angles of –4.7 and 92.5 degrees, and a relative free energy of 0.13 kcal/mol.
102 Conference on Current Trends in Computational Chemistry 2003
Quantum Chemical Calculations on the Structure and Binding of Water Molecules in the HIV-1 Protease (PR) Enzyme
Chittima Laohpongspaisana, Atchara Wijitkosooma, Surapong Pinitglangb, Vudhichai Parasuka, Supot Hannongbuaa,*
aDepartment of Chemistry, Faculty of Science, Chulalongkorn University, Bangkok 10330, Thailand bDepartment of Food Science, Faculty of Science, The University of Thai Chamber of Commerce, Bangkok 10400, Thailand
The role of water molecules in HIV-1 protease (PR) has been investigated, based on quantum mechanical calculations. The data obtained in terms of binding energies and water binding sites were taken into consideration in the modeling and the design of the potent inhibitors in the HIV-1 protease pocket. The interaction energies between the amino acid cluster and water molecules, at locations taken from the X-ray structure of HIV-1 PR from Protein Data Bank (PDB) were evaluated and results from various methods were compared. For each water molecule, the coordinate of the oxygen atom is taken from the PDB while its X-ray geometry with the O-H bond length of 0.975 Å and H-O-H bond angle of 104.5° is applied and kept constant throughout. Here, the position of the oxygen atom of water is fixed but the molecular orientation is optimized. The size of the amino acid cluster was increased as a function of radius centered at the oxygen atom of each water molecule until consistency of the interaction energy was reached. The results show that strategy to increase the size of the model is very useful in the study of water in proteins. The six water molecules in the crystallographic structure of HIV-1 PR, namely W301, W406, W426, W566, W607 and W608 have been taken into consideration. The binding energies obtained for the six water molecules are –11.49, -2.88, -8.11, -14.35, -8.92 and –10.05 kcal/mol, respectively. The results indicate that catalytic water, W607, plays a role not only in the hydrolysis mechanism but also in stabilizing the enzyme structure. The other water molecules were not found to have a substantial role in maintaining the specific conformation of the enzyme structure.
Conference on Current Trends in Computational Chemistry 2003 103
Modelling of Intra- and Intermolecular Interactions of Glycyrrhizinic Acid
M.G. Levkovich, N.J. Abdullaev, D.N. Dalimov*
Institute of Chemistry of Plant Substanses AS RUz, Acad. Kh.Abdullaev str., 77, Tashkent, 700170, Uzbekistan *Institute of Bioorganic Chemistry AS RUz, Tashkent.
The glycyrrhizinic acid (GA) has well-known solubility property. At small concentrations GA in water may promote to solve many medical drugs: hydrocortisone, prednisolone, uracil, nystatin etc. This effect already has sufficiently wide practical application and perspectives of further applications for the complex medical forms of various drugs, which are not solving in water. However mechanism of a solubilization GA remains unknown. In present work by methods of molecular modeling the analysis of the possibilities of various polymolecular systems formation (GA:GA, GA:substance, GA:GA:substance and so on) and influence of water on this systems as dissolvent are study. The estimations were done on energy of hydrogen bondings and enthalpy of formation of these complexes. It was shown that in practical is not necessary to expect the formation of clathrate complexes GA:substance = 1:1. The monomolecular cycle of GA appears sufficiently intense, and the cavity for acceptance of “quest-drug” is too small: ≈1.5 x 3.0 A. So, solubilization of GA is defined not by individual molecule of GA, and the permolecular complexes of several GA with the entrapped drugs. For similar micelleformation of GA with inclusion of “quest-drugs” there are some possibilities. At molecular modelling analysis of lipid like monomolecular layer the scoring in energy of formation per molecule has constituted about 8.0 kcal/mol. At room temperatures it most likely unstable structures. There are probable also formation of reticular, gel-like frames. The availability in one molecule of the GA of three acid groups and one separate carbonyl substituent ensures a possibility of formation of three-dimensional net grids due to hydrogen bondings. The propensity of GA to formation of gels reconfirms this mechanism of intermolecular interactions. Microgelic structures have the much more possibilities for display of solubilization effects. The similar structures with much more diversity may provide both hydrophobic and hydrophilic microvolumes. In this work the energetic characteristics of various variants of intermolecular bindings and possibility of capturing of “guest-drug” in intermolecular cavities of microgelic formations of GA are estimate.
104 Conference on Current Trends in Computational Chemistry 2003
Binding Energies of Monovalent and Divalent Cations with TNT
L. Jami Lewis and David H. Magers
Department of Chemistry and Biochemistry Mississippi College, Clinton, Mississippi
Trinitrotoluene is considered a teratogen and mutagen and therefore a major environmental hazard when it seeps into ground water. And yet, TNT is prevalent at artillery ranges, bomb sights, and anywhere explosives are used for any purpose, military or civil. The only current EPA-approved method for remiaditing TNT from soil is incineration, which is quite expensive. Other treatments are currently being investigated involving base hydrolysis of TNT. However, little is know about the mechanism of this reaction. Base hydrolysis always occurs in the presence of high concentrations of monovalent and divalent cations. Studies have shown that the intermediates of the alkaline hydrolysis of TNT can occur through radicals that have been observed in tight associated with monovalent cations. In the present study, we began our investigation of this process by calculating the binding energy of TNT to such cations using SCF and density functional methods (DFT). The ground-state geometry and the corresponding electronic energy of trinitrotoluene, the energies of the Li, Na, K, Mg, and Ca cations, and the optimum geometries and energies of a TNT dimer with each cation were calculated. In these initial computations, each cation and dimer combination yielded a different optimized structure. Each of these is currently being used as a starting point for further geometry optimizations of the other four cations with TNT dimer, thus yielding up to five possible structures for each system. In each case, the most stable will be used to determine the binding energy. Every computation is being performed at both the SCF and DFT levels of theory with two basis sets: 3-21G(d) and 6- 31G(d,p).
We gratefully acknowledge the support of NSF EPSCoR (EPS-0132618).
Lithium cation with TNT dimer
Conference on Current Trends in Computational Chemistry 2003 105
Sodium cation with TNT dimer
Potassium cation with TNT dimer
Magnesium cation with TNT dimer 106 Conference on Current Trends in Computational Chemistry 2003
Solvation Studies on Novel Steroid–Nucleoside Conjugates: Alkylated Derivatives
1Tia Lewis*, 1Jesse Edwards, 1Desiree Paramore, 2Henry Joung Lee, 2Zhengqing You
1Department of Chemistry/AHPCRC Florida A&M Tallahassee, Florida,USA 32307 2College of Pharmacy and Pharmaceutical Sciences, Florida A&M Tallahassee, Florida,USA 32307
In an attempt to develop anti-HIV agents devoid of serious toxic effects, H.J. Lee et. al. synthesized these three novel compounds along the anti-drug scheme. AZT conjugated to Cholenic Acid (Conjugate1), P-16 acid (Conjugate 2), and P-21-oic acid (Conjugate 3), where P is an abbreviation for Prednisolone. After initial studies on these compounds further modifications were made in order to examine the effect of subtle changes to the structure of these compounds on the activity. Three derivatives of the initial compounds were created synthetically by adding an additional alkyl group between the ester bridge connecting AZT to the steroid or acid. This provided the compound with increased flexibility. This work will examine solvation effects on these compounds using molecular mechanics and conformational searches. In previous computational studies on Conjugates 1 through 3 a dual binding site was hypothesized. A comparison between this work and previous work will be constructed using energy contour diagrams.
Conference on Current Trends in Computational Chemistry 2003 107
The Reaction Simplex a Computational and Conceptual Tool
Jan Linderberg
Department of Chemistry Aarhus University, DK-8000 Aarhus C, Denmark
Reactant, product and intermediate conformations are used to define a minimal space for the quantum mechanical discussion of intramolecular rearrangements. Constraints are imposed on the nuclear and electronic degrees of freedom and are used to “purify” the concepts of bond breaking and bond formation during an isomerization process. Formal reduction of the many dimensional quantum mechanical problem is accomplished through the Rice-Teller [1] measures of conformational distances and the London [2] four orbital, four electron valence bond structure notions. The resulting model offers a modest effort for the consideration of effects beyond the adiabatic description and a starting point for the generation of simple potential energy surfaces. This note concerns processes which can be idealized so that two bonds in the reactant formation are broken and two new bonds are formed to give the product. London saw the generalization of his three atom reaction scheme [3] to four atoms in terms of a redefinition of the appropriate integrals. An explicit derivation is offered here in an overlapping atomic orbital basis and a particular complex representation of the state vectors. Some formal advantages are gained in this way. London considered only the singlet states that may be constructed from a basis of four atomic orbitals, {}a()r ,b()r ,c()r ,d(r ) , with the use of the products
Φ ()r ,r ,r ,r = a()r b()r c()r d()r and proper symmetry adaptation so as to satisfy the Pauli abcd 1 2 3 4 1 2 3 4 exclusion principle. This was accomplished by means of the irreducible representations of the permutation group of four elements, S4, in this case the two-dimensional one. The particular form to be used here is given in the table, 1 i : ω = − 2 + 2 3 Elements of S4 Matrix Elements of S4 Matrix (1)(2)(3)(4) (12)(3)(4) (12)(34) 10 (1)(2)(34) 0 −1 (13)(24) 01 (1324) − 10 (14)(23) (1423) (1)(234) (13)(2)(3) ∗ (143)(2) ω 0 (1)(24)(3) 0 −ω ∗ (124)(3) 0 ω (1234) − ω 0 (132)(4) (1432) (1)(243) (14)(2)(3) ∗ (134)(2) ω 0 (1)(23)(4) 0 −ω ∗ (142)(3) 0 ω (1243) − ω 0 (123)(4) (1342)
Real orbitals ensure that the representation of the hamiltonian and the metric matrix in the two dimensional model space give the secular problem in the form [4] E − QES∗ − J∗ = 0 ES − JE− Q 108 Conference on Current Trends in Computational Chemistry 2003
Normalization of the four electron basis is introduced. The “exchange integral” J and the overlap S are complex. Their direction in an Argand diagram indicate the preferred bonding, e.g. large numerical values along the direction 1 i suggests a pairing of a and c, and of b 2 + 2 3 and d. It is useful to express exchange and overlap in terms of auxiliary quantities: S = σ 1+σ ∗σ and J = Qσ +β so that the roots of the secular problem are 1+σ ∗σ 2 E = Q − 1 σ ∗β+β∗σ ± 1 σ ∗β+β∗σ +β∗β . The parameters can be found from the integrals, e.g. 2 ()4() σ = S . It is to be anticipated that ℜ(σ ∗β)< 0 as in other parametrized models. 1−S∗S
The choice of an electronic model implies that the study concerns a rearrangement from a reactant conformation through an intermediate or transition state to a product form. Born- Oppenheimer calculations provide three sets of nuclear coordinates: r {()raj ,m j j = 1, 2,…n;a = R,T,P}. A suitable, geometrical frame work should describe the relative
positions of the three conformations in a way that is invariant to translations and rotations. Thus we consider the distance measure n 2 n d 2 = min r ′ − r ′ m m ; ab r r ∑ j=1 aj bj j ∑ j=1 j Ωa g aΩb g b r r r r r r raj ′ =Ωa ()raj − g a ;rbj ′ =Ωb ()rbj − gb ; r r r r r r ()e x′ e y′ e z′ =Ω()e x e y e z
2 2 2 2 α 0 +αx −αy −αz 2αxαy −2α0αz 2αxαz +2α0αy r r r 2 2 2 2 = ()e x e y e z 2αxαy +2α0αz α0 −αx +αy −αz 2αyαz −2α0αx 2 2 2 2 2αxαz −2α0αy 2αyαz +2α0αx α0 −αx −αy +αz
Proper rotations are represented in terms of the Euler-Rodrigues parameters [5] and it is only the relative rotation Ω−1Ω to be determined. The optimum vectors gr and gr are the centers a b a b of mass and give the translations to a common reference frame. Parameters for the relative rotation are obtained from a four-by-four, real symmetric eigenvalue problem [6]. Three distances dRT ,dTP , and dPR can be interpreted as the edges of a triangle or a simplex. The requirement of rotational invariance creates a curved space, the quotient [7] between the 3N-3 Euclidean space and the three dimensional parameter space of the proper rotation group, SO(3). Measures for the Rice-Teller principle of least motion are offered by the distances above. It will then be reasonable to consider the dynamics of the reaction as represented by a quantum mechanical problem in the space suggested by the reaction simplex. The three corners determine the essential points of the reaction process and the significant dynamics might be limited to the two dimensional space of the plane of the simplex. Quantum theory in two spatial dimensions of the type designed above is most readily formulated in terms of the baricentric variables of finite element usage. The simplex plane constrains the individual particles to have coordinates of the form rj = rRjρ + rTj τ + rPjϑ;∀j;1= ρ + τ + ϑ.
The wave function in the reduced dimensional domain is conveniently expressed in these variables when proper care is exerted with regard to their linear relation. Schrödinger’s variational functional takes the form Conference on Current Trends in Computational Chemistry 2003 109
N 2 2 2 h m j J Ψ = dυ E −U Ψ − r −r ∂Ψ + r −r ∂Ψ + r −r ∂Ψ ; () ∫ [] ∑ 2M 2 A 2 ()Tj Pj ∂ρ ()Pj Rj ∂τ ()Rj Tj ∂ϑ j=1 U = U()ρτϑ ; Ψ=Ψ()ρτϑ ; 2 A = 1 ()d +d +d ()−d +d +d ()d −d +d ()d +d −d ; 4 RT TP PR RT TP PR RT TP PR RT TP PR A slight generalization admits a matrix form for the potential U with Ψ as a vector. The kinetic energy form derives from definitions of the baricentric coordinates after some algebra. The potential may appear something like the picture here and the quantum mechanical problem is to determine the stationary states,
scattering states, or wave packet propagation using the kinetic form and 20 the potential expression. The development points towards the use of the
finite element method with a suitable tesselation of the domain. 0 Some examples in higher dimensions than two will also be
0.1 0 discussed. -0.1 0 0.1 References 02
1. Rice F. O. and Teller E., J. Chem. Phys. 6(1938)489; Larsson P. E. and Linderberg J. Theor. Chim. Acta 93(1996)79. 2. London F. Zs. F. Elektrochemie 35(1929)552. 3. London F. in Probleme der Modernen Physik (Verlag S. Hirzel, Leipzig 1928, Ed. P. Debye) p. 104. 4. Cf. Kotani M., Amemiya A., Ishiguro E., and Kimura T. Table of Molecular Integrals (Maruzen Co., Ltd, Tokyo 1963) 2nd Ed. pp.21. 5. Biedenharn L. C., Louck J. D. Angular Momentum in Quantum Physics (Addison-Wesley Publ. Co. Reading, Mass 1981) p. 19. 6. Linderberg J. J. Chem. Soc., Faraday Trans., 93 (1997) 893. 7. Kendall D. G. Bull. London Math. Soc. 16 (1984) 81. 110 Conference on Current Trends in Computational Chemistry 2003
Conformational Energetics of Naphthylquinolines
M. Jeanann Lovell, G. Reid Bishop, and David H. Magers
Mississippi College Department of Chemistry and Biochemistry Clinton, Mississippi
Naphthylquinoline derivatives satisfying hypothesized structural criteria for triplex DNA selectivity have been designed and synthesized. Proposed structural characteristic criteria promoting intercalation between bases of triplex DNA include: (i) a large aromatic surface area, (ii) an unfused flexible ring system, (iii) cationic, and (iv) crescent shape. Previous high- throughput competition dialysis experiments provided additional insights by demonstrating that the replacement of the secondary amine function found in the lead compound termed LS8 (Figure 1) with an ether oxygen linkage greatly increased selectivity towards triplex DNA. Those experiments have been extended to include two additional compounds containing either a sulfur containing thiol or amide linkage. Here we present results from computational studies designed to examine the dynamic flexibility of the naphthylquinoline side-chain for the four compounds containing amine, ether, thiol, or amide linkages. Calculations are performed to determine the energy of each compound with varying dihedral angles between the side chain and the naphthylquinoline. Beginning from optimized geometries, the specific dihedral angle is frozen at 10-degree increments for values between 0 and 360 degrees and the rest of the structure is reoptimized to yield the energy barrier of the side-chain rotation and the approximate dihedral angle at which the top of the barrier lies. Calculations are performed using semiempirical theory, SCF theory, and density functional theory.
We gratefully acknowledge the support of NSF EPSCoR (EPS-0132618).
Figure 1: LS8 – amine linkage Figure 2: G106 – amide linkage
Figure 3 : MHQ12 – ether linkage Figure 4 : OZ121 – thiol linkage Conference on Current Trends in Computational Chemistry 2003 111
Calculation of Energies of Noncovalent Interactions
Yuguang Ma and Peter Politzer
Department of Chemistry, University of New Orleans New Orleans, LA 70148 USA
Noncovalent intermolecular interactions, widely found in molecular clusters and bio- molecules, play a key role in many important processes, such as phase changes, folding of proteins and molecular recognition. However, accurate calculation of interaction energies is not an easy task because the interactions are normally very weak. Rigorous expressions for the electrostatic and polarization interaction energies between two molecules A and B, in term of the electronic densities, have been programmed:
1 Z Z Z ρ 0 (r ) Z ρ 0 (r ) ρ 0 (r )ρ 0 (r ) E = ( A B − A B B dr − B A A dr + A A B B dr dr ) (1) es ∑∑ ∑∫ B ∑∫ A ∫∫ A B 4πε 0 AB RAB A RA − rB B RB − rA rA − rB 1 Z ∆ρ ind (r ) Z ∆ρ ind (r ) E = (− A B B dr − B A A dr + pol ∑∫ B ∑∫ A 4πε 0 A RA − rB B RB − rA (2) ρ 0 (r )∆ρ ind (r ) + ρ 0 (r )∆ρ ind (r ) + ∆ρ ind (r )∆ρ ind (r ) A A B B B B A A A A B B dr dr ) ∫∫ A B rA − rB Z is atomic charge, ρ0 is the electron density of the isolated molecule and ∆ρind is the electron density change of the molecule caused by polarization. ε0 is the dielectric constant of vacuum. With some approximations, procedures for electrostatic and polarization energy calculations were developed that involve numerical integration, using a modified version of a method introduced by Gavezzotti. Electrostatic and polarization energies for several bimolecular systems, some of which are hydrogen bonded, were calculated and the results were compared to other theoretical and experimental data. A second method for the computing of intermolecular interaction energies, based on the electrostatic Hellmann-Feynman theorem, has also been developed. It involves a “supermolecule” calculation for the entire system, followed by a partitioning of the overall electric density into the two interacting components and then application of eq. (1) to find the interaction energy. In this approach, all contributions are treated in a unified manner. The advantages of this method are that it avoids treating the supersystem and subsystems separately and no BSSE correction is needed. Interaction energies for several hydrogen- bonded systems are calculated by this method. Compared with the result from experiment and high level ab initio calculation, the results are quite reliable. 112 Conference on Current Trends in Computational Chemistry 2003
Conformational Study of Cyclopentadecane
Arnaldo E. Marrero, Dalephine Davis, Judge Brown, Diwakar M. Pawar, Eric A. Noe
Jackson State University, Department of Chemistry, 1400 J.R. Lynch Street, Jackson, MS 39217-0510. Ph: (601) 979-2922
The 13C spectrum of 1% solution of cyclopentadecane (1) in propane showed the presence of 8 overlapping peaks at –170 °C (dial temperature), although the exact number is uncertain because of a low signal-to-noise ratio. The eight peaks could be consistent with a conformation of C2 symmetry, but the possibility of additional conformations cannot be excluded. The carbon NMR spectra were analyzed in terms of the conformation predicted by Allinger’s molecular mechanics program, which indicated that a conformation of C2 symmetry is of lowest strain energy and is more stable than next (C1) conformation by 0.32 Kcal/mol. The free energies for the first ten conformations were obtained at the HF/6-311*G level and compared with the MM3 results. Carbon chemical shift were obtained for these conformations of 1 by the Gauge-Including-Atomic-Orbital (GIAO) method and compared with experimental values. This work was supported by NSF -CREST Grant No. HRD-980 5465. Conference on Current Trends in Computational Chemistry 2003 113
The Effect of Grid Quality and Weight Derivatives in Density Functional Calculations of Harmonic Vibrational Frequencies
Massimo Malagoli and Jon Baker
Parallel Quantum Solutions, 2013 Green Acres Road Suite A, Fayetteville, Arkansas 72703
In Density Functional Theory (DFT) calculations, the exchange-correlation energy is customarily obtained by a numerical quadrature performed over atom-centered grids. In this context, the calculation of analytical gradients and hessians of the energy with respect to nuclear motion involves additional contributions arising from the derivatives of the quadrature weights. In the case of the energy gradient, it has previously been shown that the weight derivatives contributions can in practice be neglected with minimal effects on the accuracy, provided the integration grid is of sufficient quality [1]. In this paper, we present a study of the effects of including quadrature weight derivatives on the accuracy of computed harmonic vibrational frequencies. Results are given for the diatomic halogen series F2, Cl2, Br2, and I 2, and for CHClBr, Fe(CO)5, and lactic acid, using a variety of different density functionals and basis sets. We show that, unlike the situation with analytical DFT gradients, second derivatives are much more sensitives to the inclusion of weight derivatives, and omitting them can produce nonsensical results unless the numerical grid is of very high quality. Errors introduced by neglecting weight derivatives increase with increasing atomic number and increasing basis set size. The origin of the error is the difficulty of accurately integrating high-order derivatives of basis functions with large exponents around their own atomic center, and it is not the weight derivatives themselves that eliminate the error, but the fact that proper allowance of atomic centered grids that “move” with the atom means that basis functions no longer directly contribute to derivative quantities evaluated on their own grid.
References
[1] J. Baker, J. Andzelm, A. Scheiner and B. Delley, J. Chem. Phys. 101 (1994) 8894.
114 Conference on Current Trends in Computational Chemistry 2003
Prediction of Two Photon Absorption Properties for Large Organic Molecules Using Time-Dependent Density Functional Theory
Artem Masunov and Sergei Tretiak
Theoretical Division, Los Alamos National Lab, MS B268, Los Alamos, NM 87545
Two-photon absorption (TPA) involves electronic excitation of an organic molecule induced by a pair of photons of the same or different energy. Unlike linear absorption (LA), TPA is quadratically proportional to the intensity of the incident light and hence by focusing the beam one can precisely localize TPA in the small area up to one wavelength in size deep inside the bulk of the material. This property of TPA holds a great promise for many useful applications from manufacturing of optoelectronic logical circuits and three-dimensional optical data storage to in-vivo imaging of biological tissues and photodynamic therapy [1]. However, to make these applications practical one has to utilize lasers of lower intensity. Existing materials do not absorb enough at low intensities, which makes the design of new materials with large TPA cross- sections an important task [1]. The accurate theoretical prediction of TPA cross-sections is necessary for rational design of TPA chromophores. While post-Hartree-Fock ab initio methods, such as MR-CI and CASPT2 are capable of accurately describing electronic transitions in principle, they are prohibitively expensive when applied to the molecules of the practical interest. Semiempirical methods do not achieve sufficient precision when applied to two-photon transitions where higher energy levels and double excitations are important. Density response theory, also known as Time Dependent (TD) or Random Phase Approximation (RPA) theory combined with the Kohn-Sham (KS) Hamiltonian of Density Functional Theory (DFT) was shown to be an affordable way to treat the excited states of large molecules [2]. Even with approximate functionals, such as B3LYP, the TD-DFT approach was shown [3] to reproduce the energies of the valence excited states better than CIS and TD-HF. Higher-lying Rydberg states are also well reproduced by DT-DFT if asymptotic corrections to the standard GGA functional are made and large basis sets are used [4]. Here we propose to use TD-DFT for TPA spectra calculation. The necessary RPA formulas [5] are readily applicable to transition densities obtained within the DFT method. We selected the recently published extensive experimental and theoretical studies of bis(styryl)benzene derivatives [1,6] as a benchmark. We shall consider the effects of molecular geometry, and compare predicted values to the experimental ones. Theoretical expressions. The TPA cross-section is proportional to the imaginary part of the frequency-dependent third order polarizability: δ = (4 π2ħω2 / n2 c2 ) L4 Im <γ> where c is a speed of light, n – the refractive index, L – the local field factor (L=1 in vacuum), and <γ> is the orientationally-averaged third polarizability hypertensor: <γ> =1/15 ( Σi γiiii+ Σi≠j (γiijj + γijji ) ), i=x,y,z The third order polarizability is usually obtained using the Sum-Over-States (SOS) formalism. However, transition densities between excited states entering the SOS formula are not defined in the RPA formalism. Instead, we used eq. F55 obtained in ref. 5 as the density response to the third order in the external field. It has a rather bulky expression including ground-to-excited state transition density matrices, their combinations, and two-electron parts of the Fock operators including Coulomb and exchange-correlation interaction acting on those matrices (Coulomb operators). This expression was coded [8] in the CEO suite of programs for use with TD-HF at the semiempirical level. We use this expression, as well as two approximations to calculate the Conference on Current Trends in Computational Chemistry 2003 115
TPA cross-sections. The molecules under investigation are linear and the transition dipoles are nearly parallel to the molecular axis z. As a result, γzzzz is a major component of <γ>, which makes orientation averaging unnecessary. In the first approximation (dubbed the dipole approximation) we neglected only Coulomb operators, leaving all transition dipole moments in place. In the second approximation we retained only one leading term, including the squared product of the largest transition dipole moments from the ground to excited µge and between excited states µee’: 2 4 2 2 2 2 2 2 δ = (16 π L ħ / 5 n c ) (1/2Ege’) µ ee’ µ ge /(Ege - 1/2Ege’) Γge This formula is equivalent to eq. 2 from ref. 1a, obtained using the SOS approach within the 3- level model. We refer to this simplification as the 3-level approximation. The empirical linewidth parameter Γ throughout this study is set to 0.1 eV, like it was done in refs. 1,6. Quantum chemical calculations were performed using Gaussian 98 [7], The molecules considered are shown on Figure 1. Their structures differ from those studied experimentally [1,6] as follows. Alkyl groups (butyl and dodecyl) were replaced with methyl groups, terminal diphenylamino groups were replaced by dimethylamino groups. In this paper we use the letter a (2a,3a, etc.) to mark the molecules with diphenylamino groups. All molecular geometries were optimized starting from conformations analogous to the ones found in the crystal structures of 6a and 11a, which means C2 symmetry for molecules 1, 2, and 2a, and Ci symmetry for the others. Molecules without diphenylamino groups were also constrained to be planar (group C2h). A 6- 31G basis set was used for all the calculations. Solvation e ects were neglected. The geometrical optimization was carried out using Gaussian 98 at the HF and B3LYP levels. Molecules 1, 6 and 11 were also optimized at the AM1 and MP2 levels. The optimized geometry was used to carry out TD-B3LYP calculations for the first 6 singlet excited states. Gaussian 98 was modified to output dipole matrices, transition density matrices between excited states, and Coulomb operators. The CEO program [8] was modified to read these matrices and calculate the third order polarizability. Geometric parameters. The results of TPA calculations inevitably depend on the molecular geometry chosen. Even an excellent method will fail if the geometry used in the calculations is not accurate enough. That is why special care must be taken when selecting a theory level for geometry optimization to make sure it gives a reasonable agreement with experimental geometric parameters. One can select two main parameters affecting the conjugation of the double bonds in the molecules depicted in Figure 1: Bond Length Alternation (BLA=r(C−C)−r(C=C)) and non-planarity (deviation of the torsion angle about single bond from 180o). Both parameters may depend on the chemical substituents, the solvent and other environment factors. We optimized the geometry for the molecules 1, 6, 6a, 11 and 11a at the AM1, HF, B3LYP and MP2 levels. Among these methods, HF gives the best agreement with the experiments6 on BLA values, probably due to cancellation of errors from incomplete basis set and neglect of electron correlation. Regardless of the optimum value of the torsion angle, the torsion potential in stilbene and its analogs is shallow, that allows the molecules to adopt a wide range of torsion angles in solution at room temperature. For this reason we carried the calculations at both planar (HF and DFT) and non-planar (HF optimal) geometries, and found the planar HF geometry to give a better agreement with experiment. TPA cross-sections. Transition densities and Coulomb operator matrices for up to 6 singlet excited states were calculated at the TD-B3LYP/6-31G level for each of the molecules studied. The TPA spectra were obtained from these data using the modified CEO program. Most of them have one maximum (note the exceptions below). Increasing the number of calculated excited states to 30 for molecules 1 and 2 does not affect the results. The cross-section value at the absolute maximum is plotted vs. experimental values on Figure 2. Results obtained with the dipole and 3-level approximations are also presented. 116 Conference on Current Trends in Computational Chemistry 2003
R R N R N R
R N R 2: R=n-Bu R 1 2a:R=Ph N 3: R=n-Bu 3a:R=Ph R R R N N O R O R
R O R O N O N 5 R CN R R 4: R=n-Bu OC12H25 N CN 4a:R=Ph CN R O O CN OC H CN 12 25 R CN O 7 N R N R 6: R=n-Bu Br R 6a:R=Ph O O
OC12H25 N O O N S R Br N S N O O R 10: R=n-Bu N OC12H25 10a:R=Ph R N O O 8 CN R NC O CN
OC12H25 O S R CN O N O O O R S 11: R=n-Bu O 11a:R=Ph OC12H25 n-Bu NC O 9 CN N O n-Bu R N OC12H25 R O n O R N OC12H25 R 12: n=2 n-Bu O 13: n=3 N 16 14: n=4 n-Bu 15: n=5
Figure 1. Molecules investigated in this work and experimentally studied in refs. 1,6.
In a most cases the major LA state is the 5000 first singlet excited state S1, the two-photon TD-DFT calculated vs. experiment state is S2. For the molecules 1 and 14 the TPA state is S4, for some molecules there are 4000
two (S2, S6 for 7, S2, S4 in 10) or three (S2, S4, S6 for 4, S2, S3, S6 for 9), while the molecule 6 has two LA states (S1 and S3), and 3000 two TPA states (S2 and S4). Molecules 4, 7, 10, and 10a have a second, satellite peak in 2000 the calculated spectrum, and molecule 6 has 9 calculated, GM calculated, 16 three maxima of comparable height. The 8 7 4 15 second maximum in the TPA spectrum of 7 1000 5 1 14 3 was indeed observed in the experiment. 1 13 11 6 12 Experimentally unresolved multiple maxima 2 may be responsible for the largest deviations 0 0 1000 2000 3000 4000 5000 from the experimental cross-sections (3- and experimental, GM 2-fold for the molecules 6 and 9 respectively). Figure 2. Calculated TPA cross-sections vs. Ambiguity in the choice of damping factor Γ experimental ones (GM). can be another reason for deviations from experiment. Quantitative comparison with experiment will remain questionable for absolute magnitudes of TPA cross-sections until the vibronic structures and interactions with the solvent are explicitly accounted for. Conference on Current Trends in Computational Chemistry 2003 117
As one can see from the Figure 2, the absolute errors in TPA cross-sections are comparable with the best literature values. 4.5 3.0
4.0 2.5 V)
e 3.5 ( 3.0 2.0 ons
ti a l 2.5 cu
l 1.5
a 2.0 C 1.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 1.0 1.5 2.0 2.5 3.
Figure 3. Calculated TPA (left) and LA (right) transition energies vs. experiment (eV).
Transition energies. The results of our TD-DFT calculations are presented on Figure 3. While the mean errors for semiempirical methods are 13-20%, TD-DFT performs significantly better. For the two-photon transition energies the mean error is 3.8%, 4.1%, and 4.6% at flat HF, optimal HF and flat DFT geometry respectively. For the linear absorption the mean error in the transition energy is 3.1%, 3.9%, and 6.9% at the same geometries. DT-DFT reproduces LA and TPA excitation energies equally well. This is not the case for semiempirical methods, which are usually parameterized for the ground and (at most) one excited state to reproduce either LA or TPA. For centrosymmetric molecules, where the LA and TPA states have different symmetry and TPA state requires double excitations for its correct description. Including double excitations, however, overcorrelates the ground state, distorting the LA transition energies. Conclusions. For the first time, two-photon absorption spectra are calculated at the TD- B3LYP/6-31G//HF/6-31G level for the series of bis(styryl)benzene derivatives. Instead of the habitual SOS approach, the actual third-order response formula for RPA is used. The predicted TPA corss-sections are among the best found in the literature. Uncertainty in the linewidth parameter makes the quantitative comparison of the cross-sections with experiment impossible. Both TPA and LA transition energies are predicted with better than 4% accuracy, which presents a considerable improvement over semiempirical calculations found in the literature. Thus DT- DFT can be used for the rational design of two-photon absorbing materials. Acknowledgements. The authors of this paper would like to thank Richard Martin and Mike Frisch for fruitful discussions and their help with the Gaussian 98 code.
References 1. (a) Albota, M.; Beljonne, D.; Bredas, J.L.; Ehrlich, J.E.; Fu, J.Y.; Heikal, A.A.; Hess, S.E.; Kogej, T.; Levin, M.D.; Marder, S.R.; McCord-Maughon, D.; Perry, J.W.; Rockel, H.; Rumi, M.; Subramaniam, C.; Webb, W.W.; Wu, X.; Xu, C. Science 1998, 281, 1653; (b) Rumi, M.; Ehrlich, J.E.; Heikal, A.A.; Perry, J.W.; Barlow, S; Hu, Z.Y.; McCord-Maughon, D.; Parker, T.C.; Rockel, H.; Thayumanavan, S.; Marder, S.R.; Beljonne, D.; Bredas, J.L. J.Am.Chem.Soc. 2000, 122, 9500; (c) Das, G.P.; Yeates, A.T.; Dudis, D.S. Chem.Phys.Let. 2002, 361, 71; (d) Zhou, X.; Ren, A.M.; Feng, J.K.; Liu, X.J. J.Phys.Chem. 2003, 107, 1850. 2. Casida, M.E. in Recent Advances in Density Functional Methods, Part I, edited by D.P. Chong (Singapore, World Scientific, 1995), p. 155. 3. Andreu, R.; Garin, J.; Orduna, J TETRAHEDRON; 2001; 57, 7883. 4. Casida, M.E.; Salahub, D. J. Chem. Phys. 2000, 113, 8918. 5. Tretiak, S.; Mukamel, S. Cem.Rev. 2002, 102, 3171. 6. Pond, S.J.K.; Rumi, M.; Levin, M.D.; Parker, T.C.; Beljonne, D.; Day, M.W.; Bredas, J.L.; Marder, S.R.; Perry, J.W. J.Phys.Chem. A 2002, 106, 11470; 7. Frisch, M.J., et al. Gaussian 98.A.11, Gaussian, Inc., Pittsburgh PA, 2001. 8. Tretiak, S.; Chernyak, V.; Mukamel, S. J.Am.Chem.Soc. 1997, 119, 11408. 118 Conference on Current Trends in Computational Chemistry 2003
Homology Studies of the Pacific Electric Ray and Mouse Acetylcholinesterase as a Means of Bioterrorism Defense
Kareem Mckinney, Henry J. Lee and John S. Cooperwood
Florida A&M University College of Pharmacy Division of Basic Sciences
In view of the fact that there is no organophosphate – mouse acetylcholinesterase co-crystal structure sited in the literature, we carried out homology studies with the mouse acetylcholinesterase and various organophosphate - pacific electric ray acetylcholinesterase co- crystals to predict conformational changes involved during binding of organophosphate to its mouse counterpart. A comparison of amino acid sequences and morphologies of mouse and electric eel acetylcholinesterase will be carried out. Alignment of the electric eel acetylcholinesterase active to that of the mouse acetylcholinesterase was performed in the Biopolymer module of Sybyl 6.9. PURPOSE: These homology studies will allow us to elucidated active site of mouse acetylcholinesterase and better predict the process of aging. METHODS: Mapping the process by which the aging occurs using FlexiDock module of Sybyl 6.9. RESULTS: The two crystal structures had very similar three-dimensional structures and active site. CONCLUSION: Through homology studies, molecular modeling is a good predictor of the orientation of organophosphates within the active site.
Supported by Army High Performance Computing Research Center
Conference on Current Trends in Computational Chemistry 2003 119
Ab Initio Studies of Boron Polyhedron Molecules
James L. Meeks
Department of Physics, P.O. Box 7380, West Kentucky Community and Technical College, Paducah, KY 42002-7380
The optimized molecular energies and geometries of the novel boronated polyhedron molecules, C15H20, C10H15B5, and C15H15B5, molecules were computed. These computations used ab initio (HF/6-31G**) and DFT (with the B3LYP hybrid functional) of Gaussian 98.
The C10H15B5 molecule is a polyhedron system using the trivalent bonding of Boron and the polyhedron molecule, C15H15B, exhibits a pentavalent Boron structure. The changes of the molecular energies, bond distances and angles of the polyhedron boron molecules are compared. Analyses, by ab initio calculations, of similar polyhedron systems are reported. Geometry optimizations and minimum energies for the different polyhedron boron molecular systems using the different basis sets of Gaussian 98 will be discussed. The optimized (B3LYP/6-31G, 6-31G **, and 6-311++G**) geometrical structures of the polyhedron boron molecules are presented.
120 Conference on Current Trends in Computational Chemistry 2003
Theoretical Study of Adsorption and Decomposition of Nerve Agents on Metal Oxides
A. Michalkovaa,b, M. Ilchenkoa, L. Gorba, J. Leszczynskia
a Computational Center of Molecular Structure and Interactions, Department of Chemistry, Jackson State University, 1400 J. R. Lynch Street, P. O. Box 17910, Jackson, MS 39217, USA b Institute of Inorganic Chemistry, Slovak Academy of Sciences, Dubravska cesta 9, 842 36 Bratislava, Slovak Republic
The adsorption and decomposition of Sarin (GB) and Soman (GD) on the surfaces of metal oxides was investigated at B3LYP/6-31G(d) and MP2/6-31G(d) levels. Sarin (GB) – isopropyl methylphosphonofluoridate (C4H10FO2P) and Soman (GD) – 3,3-dimethyl-2-butyl methylphosphonoflouridate (C7H16FO2P) are very toxic organophosphorus compounds. These substances became known as the nerve agents that are used as lethal warfare chemical weapon. Metal oxides, well known for their industrial application as adsorbents, catalysts, and catalyst supports possess many potential decontamination applications such as protective filtration systems for vehicles, aircraft, and buildings, and the demilitarization of nerve agent munitions and stockpiles. The physisorption of nerve agents on the surface of oxide occurs due to the formation of hydrogen bonds, and ion-dipole and dipole-dipole interactions between adsorbed nerve agent and the surface. The chemisorption occurs due to the formation of covalent bonds between the adsorbed molecule and the surface. Despite of steric reasons, GB and GD are adsorbed in the same way on the CaO surface. The chemisorption of Sarin on specific sites of the MgO surface results in decomposition of Sarin so that the fluorine atom is transferred from Sarin into the binding distance with the Mg atom. Sarin can be also decomposed on MgO in a presence of Mg2+ cation so that this cation forms the isolated MgF+ fragment with the removed fluorine ion from Sarin.
Figure 1. Optimized structure of decomposed Sarin on the surface of MgO.
Conference on Current Trends in Computational Chemistry 2003 121
Theoretical Study of Adsorption and Decomposition of Methyl tert- buthyl Ether on Dickite
A. Michalkovaa,b, J. Leszczynskib
a Institute of Inorganic Chemistry, Slovak Academy of Sciences, Dubravska cesta 9, 842 36 Bratislava, Slovak Republic b Computational Center of Molecular Structure and Interactions, Department of Chemistry, Jackson State University, 1400 J. R. Lynch Street, P. O. Box 17910, Jackson, MS 39217, USA
The adsorption and decomposition of methyl tert-buthyl ether (MTBE) on the substituted surface of dickite (1:1 dioctahedral clay mineral of the kaolinite group) have been studied using the ONIOM(B3LYP/6-31G(d,p):PM3) method. MTBE is an oxygenate additive in reformulated gasoline which is commonly detected as a prominent component of contaminated groundwater and soils. The negative charge of the layer from the substitution of two Si atoms by Al atoms in the tetrahedral side was compensated by the exchangeable Mg2+ cation which was hydrated by two water molecules. Three mutual positions of substituting Al atoms in the tetrahedral ring are possible: (1) adjacent Al atoms (1,2D-MTBE model); (2) one Si atom between two Al atoms (1,3D-MTBE model); (3) two Si atoms between two Al atoms (1,4D-MTBE model). The location and orientation of MTBE on the surface of dickite were found. Different type of substituted surface results in different optimized structures of adsorbed MTBE. The molecule is able to extend above a whole ditrigonal cavity of the 1,2subsituted surface but in the case of the 1,3-substituted surface is on the border of the tetrahedral ring. The orientation of MTBE on the surface corresponds to the formation of attractive interactions with the surface. MTBE is mainly adsorbed due to the formation of multiple weak C-H…O hydrogen bonds between the C-H groups of MTBE (proton-donors) and basal oxygen atoms of the tetrahedral side (proton-acceptors). These attractive interactions are responsible for the stabilization of MTBE on the surface of the mineral. The adsorption on the 1,4-substituted surface leads to the decomposition of MTBE so that one CH3 group of the buthyl part is substituted by the hydrogen atom of one water molecule. Figure 1 displays the decomposed structure of MTBE on the 1,4- substituted surface of dickite with the Mg2+ cation and two water molecules. The adsorption results in changes of conformation of MTBE and in the polarization and the electron density redistribution which are the most significant in the 1,4D-MTBE model as a consequence of the MTBE decomposition.
Figure 1. The optimized structure of decomposed MTBE on the1,4-substituted surface of dickite. 122 Conference on Current Trends in Computational Chemistry 2003
Solvation Studies of Anti-HIV Prodrugs
1Ashley Moorer*, 1Jesse Edwards, 1Desiree Paramore, 2Henry Joung Lee, 2Zhengqing You
1Department of Chemistry/AHPCRC Florida A&M Tallahassee, Florida,USA 32307, 2College of Pharmacy and Pharmaceutical Sciences, Florida A&M Tallahassee, Florida,USA 32307
Newly synthesized prodrugs aimed at delivering the well know nucleoside AZT to the active site of the HIV virus has been developed by Dr. Henry J. Lee et al. Following the prodrug scheme the intact drug will deliver the lethal AZT portion after undergoing a metabolic reaction. In order to proceed with this mechanism the drug must first be made available to the active site by passing through the membrane of the host cell. Therefore, structure of the intact drug becomes critical. This work will examine the lowest energy conformations at various dielectric constants. The change (from 1 to 10) in dielectric constant will mimic the solvation process.
Conference on Current Trends in Computational Chemistry 2003 123
The Effect of the Cap Closure Method Upon the Surface Electrostatic Potentials of Some (6,0) Carbon, Boron/Nitrogen and Carbon/Boron/Nitrogen Nanotube Models
Jane S. Murray, Pat Lane, Monica C. Concha and Peter Politzer
Department of Chemistry, University of New Orleans New Orleans, LA 70148
In an earlier study of (5,5) nanotube models, we closed the tubes at the ends with fullerene-like caps, having alternating hexagons and pentagons. We have more recently been looking at nanotube models with other indices, and have investigated two methods for closing (n,0) type tubes: the first is with a hexagon at the tip, fused with six five-membered rings, while the second has three five-membered rings at the end, which are fused to three more five- membered rings and hexagons. We present, in this poster, the surface electrostatic potentials of (6,0) carbon, boron/nitrogen and carbon/boron/nitrogen nanotube models with caps formed by these two methods. We discuss their similarities and differences and the effects that these may have upon intermolecular interactions.
124 Conference on Current Trends in Computational Chemistry 2003
A Theoretical Study of the Ground State Unimolecular Decomposition Channels of Propynoic Acid
Edmund Moses N. Ndip1, Shukla Manoj2 and Jerzy Leszczynski2
1Department of Chemistry, School of Science, Hampton University, Hampton, VA 23668, USA 2Computational Center for Molecular Structure and Interactions (CCMSI) Department of Chemistry, Jackson State University, Jackson, MS 39217, USA
Acetylene (H-C≡C-H), an abundant interstellar species, and the recently detected species vinyl alcohol (H2C=CH-OH) and hydroxyacetylene (HC≡C-OH) [1] are important species in the discussion of atmospheric processes. While chemical pathways leading to vinyl alcohol are not completely clear yet, the present study focuses on two possible facile chemical transformations of propynoic acid that lead to acetylene and hydroxyacetylene. In the present study, the Gaussian 98 suite of programs [2] was used. All computations of equilibrium and transition- state structures relevant to the two competing unimolecular decomposition channels (decarboxylation and decarbonylation) were performed with ab initio molecular orbital (SCF – MP2) and density functional theory (DFT – B3LYP) [3, 4] calculations using the 6-311++G (d,p) basis set [5-7]. The stationary point geometries were fully optimized and characterized as minima (0 imaginary frequencies) or first-order saddle points (1 imaginary frequency) by full vibration analysis at either MP2/6-311++G** or the B3LYP/6-311++G** level. The decarboxylation channel through a four-center transition state leads to acetylene and carbon dioxide, with a calculated average barrier of 62.1 - 65.1 kcal/mol. Decarbonylation, leading to either hydroxyacetylene (and subsequently the ketene) or carbon monoxide occurs easily either via a direct (one-step) three-center transition state or through a two-step mechanism involving a four-center transition state. The direct (one-step) elimination of carbon monoxide from propynoic acid has a calculated barrier of 79.1 kcal/mol.
References [1] V.A. Basuik and Kobayashi, Lunar and Planetary Science XXXIV (2003), 1085.pdf; Turner B.E. and Apponi A.J. (2001) Astrophys. J., 561, L207. [2] Frisch M. J. et al. (1998) Gaussian 98, Gaussian Inc., Pittsburgh, PA. [3] Becke A. D. (1993) J. Phys. Chem., 98, 5648. [4] Lee C. et al. (1988) Phys. Rev. B, 37, 785. [5] Hariharan P. C. and Pople J. A. (1972) Chem. Phys. Lett., 66, 217. [6] Binkley J. S. and Pople J. A. (1975) Int. J. Quantum Chem., 9, 229. [7] Krishnan R. et al. (1980) J. Chem. Phys., 72, 4244.
Conference on Current Trends in Computational Chemistry 2003 125
Decarboxylation Channel
Imaginary vibrational mode showing formation of acetylene and carbon dioxide
4-Center
s-trans-PA
s-cis-PA
Imaginary vibrational mode showing hydroxyacetylene and carbon monoxide formation
Decarbonylation Channel 3-Center Decarbonylation TS
Scheme. Unimolecular Decomposition Channels for Propynoic Acid 126 Conference on Current Trends in Computational Chemistry 2003
Quantum-Chemical Study of the Mechanism of Norbornane Row Epoxyamidoacids Intramolecular Cyclization
S.I. Okovytyy, T.V. Petrova, L.I. Kasyan
Dnepropetrovsk National University, Nauchny St. 13. 49625, Dnepropetrovsk, Ukraine
One of the most important problem of theoretical organic chemistry is investigation of regio- and chemoselectivity of processes. Interaction of epoxyendic anhydride (I) with amines results in formation of heterocycles (III) with cyclization by carboxylic but carboxyamid group.
O O 1 HO RNH 2 6 O O 2 O C 4 C C6H6 C 3 5 NH NHR C O C 2 O C O O OH O I IIa-c IIIa-c R=H(IIa, IIIa), Ph(IIb, IIIb), CH Ph(IIc, IIIc) 3 The purpose of investigation is theoretical study of potential energy surface (PES) of intramolecular cyclization reaction of epoxyamidoacids (II a-c). Calculations have been performed by quantum-chemical methods PM3 influence of solvent (benzene) has been taken into account using macroscopic approach (method COSMO). According to calculations epoxyamidoacids could exist in several conformations which differ in orientation of carboxylic and carboxyamidic groups relatively to epoxynorbornane frame. In this study we have paid a special attention to conformers (A-D) which able to intramolecular cyclization. Among them the most stable is conformer (A) which includes hydrogen bound between proton of carboxylic and oxygen atom of amid group. For conformations (A-D) we have calculated values of activation barriers for reaction of formation oxa- and azabrendans. It could be seen from calculated values of activation barriers (Table 1) that the most preferred process is cyclization by oxygen atom of carboxyamide grope, which is in line with larger mesomeric effect of amino group but does not agrees with the experimentally obtained product of reaction.
O O O O NHR NHR OH OH C C O O C O O C C C C C O H O O O H O H O N N H R R A B C D Detailed analysis of PES of amidoacids (II a-c) transformation has been shown the alternative mechanisms, which include proton transfer from carboxylic group to oxygen or nitrogen atom of amide group and further intramolecular cyclization of deprotonated carboxylic group (Scheme 1). According to calculations (see Table 2) both of stepwise mechanisms characterize by smaller values activation energy if compare to one-stage cyclization of epoxyamidoacids (II a-c). In summary, on the basis of theoretical study we have can conclude, that chemo-selective by carboxylic grope cyclization of epoxyamidoacids (II a-c) proceeds through stepwise mechanism with preliminary proton transfer from carboxylic group to oxygen or nitrogen atom of amide group. Conference on Current Trends in Computational Chemistry 2003 127
Table 1. Calculated values of enthalpies of formation of the most stable conformers of ≠ amidoacids (ІIа-c) (∆Нf), activation barriers of their intramolecular cyclization reactions (∆Н ) and selected geometrical parameters of corresponding transition states.а ≠ Com- Con- ∆Нf, ∆Н , Bond length, Å Angle, deg. pound former kJ/mol kJ/mol Сat-Оe Сat-Оk Оk-Сat-Оe Оk-Сat-С2(3)-Оe А -524,00 205,06 2,122 2,045 139,3 147,7 В -514,76 198,61 2,139 2,042 138,2 147,6 II a С -516,43 179,79 2,065 2,039 141,8 149,1 D -519,19 182,42 2,063 2,036 141,9 149,2 А -397,98 205,89 2,116 2,052 139,2 147,4 В -393,05 199,49 2,132 2,056 139,4 146,9 II b C -375,39 199,33 2,148 2,05 139,1 -147,0 D -389,15 189,20 2,069 2,038 141,6 -150,8 А -410,74 197,53 2,112 2,064 140,0 146,7 В -409,78 209,41 2,135 2,052 139,1 147,0 II c C -401,16 181,21 2,055 2,046 142,0 -150,0 D -397,73 185,23 2,053 2,046 142,0 -149,9 a Cat – carbon atom of epoxydic ring under attack; Oe – oxygen atom of epoxydic ring; Ok – oxygen atom of carbonyl group.
O O O O O O NHR NHR + C NHR + C O C NHR C O - C C C O C O H O H H O H O - O O O
O O O O O O O O + + C NHR C NHR C NHR C NHR - C O C O C C H H O H O H O - O O O
Scheme 1
≠ Table 2. Values of activation barriers for proton-transfer (∆Н1 ) and intramolecular cyclization ≠ (∆Н2 ) of epoxyamidoacids (II a-c) and selected geometrical parameters of corresponding transition states.а ≠ Bond length, Å ≠ Bond length, Å Angle, deg. Com- ∆Н1 , ∆Н2 , pound kJ/mol kJ/mol Н-ОH Н-ОА(N) С3-Оe С3-Оk Оk-С3-Оe Оk-С3-С2-Оe Proton transfer to oxygen atom of amide group IIа 124,52 1,345 1,054 139,03 1,977 2,081 145,5 151,8 IIb 125,19 1,352 1,055 159,66 1,973 2,061 143,4 153,9 IIc 123,09 1,360 1,050 154,77 1,971 2,056 143,4 154,0 Proton transfer to nitrogen atom of amide group IIа 109,58 1,327 1,172 159,28 1,975 2,066 143,3 154,1 IIb 123,05 1,364 1,102 168,91 1,973 2,063 143,7 154,0 IIc 126,40 1,375 1,177 162,42 1,979 2,064 143,0 154,2 a OH - proton of hydroxylic group; OA – oxygen atom of amide group; Cat – carbon atom of epoxydic ring under attack; Oe – oxygen atom of epoxydic ring; Ok – oxygen atom of carbonyl group. 128 Conference on Current Trends in Computational Chemistry 2003
Modeling Through-Space Proton Magnetic Shielding Over Oxirane Ring
S.I. Okovyty, L.K. Umrikhina, J. Leszczynski
Dnepropetrovsk National University, Nauchny St. 13. 49625, Dnepropetrovsk, Ukraine
The long-held “shielding cone” model of the through-space NMR shielding effect of an oxirane ring predicts only the effect of the magnetic anisotropy; it ignores other important contributors to the overall shielding. Calculations at GIAO-PBE1PBE/6-31G## have been performed on a simple model system, methane moved sequentially above oxirane (Fig. 1). We have calculated the shielding over a box with Z= 2.0, 2.5, 3.0 and 3.5 Å and X and Y varying from –1.5 to 2 Å in 0.5 Å steps from the position of the oxygen nuclei (Fig. 2). These calculations permit the net NMR shielding surface to be mapped. Based on those results, a new different graphical model for predicting the effect of a proton’s position relative to an oxirane ring on its chemical shift is presented.
1.50 Z H C H 1.00 C H Figure 1. Orientation of axes for the model Figure 2. (De)shieldin calculations.0.50 C plane of oxirane ring at H H Y, Å 0.00 C T O H he O res -0.50 H X ults C obt -1.00 H ain Y ed -1.50 her -1.50 -1.00 -0.50 0.00 0.50 1.00 1.50 2.00 e for the anisotropy and shielding show that along the Z-axis the shielding basically is negative for Z= 2.0 Å, but for Z = 3.5 Å the methane protons were shielded by the oxirane ring for all values of Y at X > -0.7 Å. At 2.0 Å < Z < 3.5 Å the shielding is positive or negative depending on the values of the other co-ordinates. The maximum positive value of the shielding occurs at Z= 2.6 Å. This is in good agreement with the experimentally observed data. Conference on Current Trends in Computational Chemistry 2003 129
Film Growth in Reactive Aqueous Solution of Hydrophobic and Polar Components on Adsorbing Substrate: A Computer Simulation Study
R.B. Pandey1 and Marek Urban2
1 Department of Physics and Astronomy 2 Department of Polymer Science University of Southern Mississippi, Hattiesburg, MS 39406
A mixture of hydrophobic (H), polar (P), and water (W) components are considered on a discrete lattice. These components are represented by particles with appropriate molecular weights. Nearest neighbor interactions among the particles are used to incorporate their specific characteristics, i.e., attractive P-W, repulsive H-W interactions, attractive interactions with the adsorbing substrate at the bottom etc. Particles execute their stochastic motion with a periodic boundary condition along the transverse direction and open boundary along the longitudinal (vertical) directions. Impenetrable substrate prevent particles to leave the system from the bottom however, the water components can escape the system from the top (evaporation). The concentrations of H and P components are conserved. Apart from non-covalent interactions, kinetic reactions are considered for the covalent bonding among these constituents. As the simulation proceeds, the densities of these constituents grow from the substrate. Density profiles and interface are studied in detail as a function of water concentration. The interface width, a measure of the surface roughness, is found to increase with the water concentration which seems consistent with the laboratory observations. Effects of interactions and kinetics are under- investigations and results will be presented as the data will become available.
130 Conference on Current Trends in Computational Chemistry 2003
Molecular Dynamics Simulations of Polymer Nanocomposites Containing POSS
Reena R. Patel1, Rajendran Mohanraj1, and Charles U. Pittman, Jr.2
1Engineering Research Center 2Department of Chemistry Mississippi State University, Mississippi State, MS 39762
Chemical incorporation of Polyhedral Oligomeric Silsesquioxanes (POSS) into traditional polymers can impart a variety of improved properties such as increase in the glass transition temperature, resistance to oxidation, surface hardening, reduction in flammability, and improvements in mechanical properties. POSS can be added as a pendant group to the polymer backbone. These polymers find applications in photoresist coatings for electronic and optical devices, magnetic recording media, optical fiber coatings, gas separation membranes, binders for ceramics, carcinostatic drugs etc [1]. Plastics like polystyrene and polymethyl methacrylate are used below their glass transition temperature (Tg), i.e. in their glassy state, where they are hard and brittle. There is often an increase in their glass transition temperature when POSS is chemically bonded to such hard plastics thereby increasing the upper limit of the temperature range in which these plastics can be used. By retarding segmental motion, POSS units raise the heat distortion temperature. This effect is complicated to predict because it depends on the nature of the polymer packing, POSS packing and the substituents on POSS. The main aim of using molecular dynamics simulations to study polymer nanocomposites containing POSS is to identify the reason behind the changes in properties that POSS imparts to the polymer. Molecular dynamics simulations help in estimating the segmental motion of the polymer versus temperature, packing details, mechanical properties, glass transition temperature etc [2]. Molecular dynamics simulations were performed on POSS-polystyrene and POSS- polymethyl methacrylate polymers. Experimental glass transition temperatures are available for these polymers [3, 4]. By carrying out molecular dynamics calculations on these systems we aim to find the reason for enhancement of properties at the molecular level. It is also possible to determine the accuracy of the forcefield that is being used for the simulations by comparing the properties predicted by the simulations with the experimental values. The first group of systems we considered comprises of pure atactic polystyrene and atactic polystyrene randomly copolymerized with POSS containing cyclohexyl, cyclopentyl or isobutyl groups as the side groups. These functional groups are attached to the seven corner silicon atoms of the POSS cage and the eighth silicon atom is attached to polystyrene through a p-styryl function present on one corner Si atom of the POSS cage. This system is studied to analyze the effect of changing the functional groups on the seven corner silicon atoms of the POSS unit. The experimental data for the glass transition temperature is available for these systems [3]. The polymers were built as a single chain consisting of 100 monomers. This single polymer chain is packed in a cubic simulation cell at a low density. All the simulations were carried out using the Cerius2 [5] software. The COMPASS (Condensed-phase Optimized Molecular Potentials for Atomistic Simulation Studies) forcefield was selected for performing the simulations. This forcefield has been successfully used in the study of POSS compounds [2]. The structure was minimized prior to performing dynamics in order to optimize the initial geometry. Minimization was followed by equilibration at 500K for more than 1.5 nano second duration to get well equilibrated system using NPT (constant number of atoms, constant pressure, and constant temperature ensemble) dynamics. Since the rate of cooling used is a key Conference on Current Trends in Computational Chemistry 2003 131
factor that influences accuracy of the computed Tg, the step size in the annealing process was chosen sufficiently small (25 K). At each temperature the system was allowed to equilibrate for at least 200 pico second before making production runs to collect data. The effect of mole percent of POSS present in the polymer on the properties was also studied holding the seven R groups constant. For this purpose, a system with a polymethyl methacrylate (PMMA) backbone was selected. The glass transition temperature data for PMMA with varying mole percents of the POSS and with isobutyl functional group is available [4]. Syndiotactic PMMA is selected for this case.
Figure 1. Structure of POSS with seven Si atoms attached to the functional groups (R) and eighth Si atom attached to a methacrylate function through an ethyl spacer
Figure 1 shows the T8 cage structure of POSS comprised of eight Si atoms and twelve oxygen atoms. Seven corner Si atoms are attached to the R group and the eighth Si atom is attached to a methacrylate function through an ethyl spacer. This MMA-POSS unit is randomly copolymerized with the desired mole percent of pure MMA to get hybrid P (MMA-co-MMA- POSS) polymers. Enhancement of mechanical properties is expected in styryl polymer nanocomposites containing POSS as there is an increase in the glass transition temperature. POSS moieties behave as strong anchor points giving added strength to the polymer thereby causing an increase in the stiffness of the polymer. Mechanical properties can be studied by carrying out constant stress simulations in which both the shape and volume of the cell are allowed to change. This allows the internal stress of the system to match the externally applied stress. Elastic constant matrices, tensile moduli, bulk moduli, shear moduli and Poisson’s ratios can be computed from the trajectories generated by the constant stress simulations. Computation of the Radial distribution function (RDF) aids in understanding the packing details of polymer. The radial distribution function can be computed from the trajectory generated by the production runs of molecular dynamics simulations. RDF is a statistical function which describes the atomic distribution around an atom in a structure. It is possible to calculate the structure factor with the help of the radial distribution function. The structure factor aids in revealing the structure properties and it can be directly compared with X-ray diffraction data. Specific volume was plotted as a function of temperature to compute glass transition temperature. The temperature at which a discontinuity in the slope occurs is the glass transition temperature. This phenomenon is a second order transition involving just a change in heat capacity without any latent heat involved [6]. This temperature is called the glass transition temperature because this transition occurs in a very narrow range of temperature where segmental motion is suddenly activated by the available thermal energy. There is a higher heat capacity above the glass transition temperature. Figure 2 shows the specific volume versus temperature graph plotted for 100mer pure atactic polystyrene. The glass transition temperature according to this simulation is 375 K and the experimental value is 373 K. 132 Conference on Current Trends in Computational Chemistry 2003
Specific Volume Vs Temperature for atactic PS
1.04
1.03
1.02
) 1.01
1
0.99
0.98
0.97
0.96
0.95 300 350 400 450 500 T (K) Figure 2. Specific Volume versus temperature curve for pure atactic polystyrene. The temperature at which a break in the slope occurs is the Tg.
The thermal expansion coefficient can also be calculated from this graph. It is defined as the change in volume per unit volume per degree Kelvin rise in temperature. The values of volumetric thermal expansion coefficient above and below the glass transition temperature are compared with the experimental values [7]. Above the Tg, its value is (4.41-5.96)x10 −4 K −1 and below Tg it is 2.05x10 −4 K −1 from the simulations. The experimental values are (5.01-6.0) x10 −4 K −1 and (1.7-2.1) x10 −4 K −1 above and below the Tg respectively.
Figure 3. Mean Square Displacement of polystyrene backbone with and without POSS moiety (4 mole % of cyclohexyl POSS randomly copolymerized with atactic PS)
The mean square displacement of the polymer backbone can be studied from the trajectory generated in the production run. This helps in understanding the diffusive behavior of the POSS moiety. We can also compare the movement of the polystyrene backbone, both with and without the POSS moiety, thereby seeing the effect that POSS has on the polymer. For this purpose we have analyzed the trajectories generated from the simulations of pure atactic PS and Conference on Current Trends in Computational Chemistry 2003 133
another system involving atactic PS randomly copolymerized with 4 mole % cyclohexyl-POSS (cyclohexyl side groups are attached to the seven corner Si atoms of POSS and the eighth Si atom attached to the polystyrene through a p-styryl function at the eighth Si atom). This analysis helps in understanding the behavior of POSS moieties as strong anchor points in the polymer which restricts the movement of the polymer. The high mass (~1000 amu) and large volume of pendant POSS moieties greatly restrict activating its motion thermally. Thus, the POSS moieties act as “anchors”. Figure 3 shows the graph of mean square displacement, at 500 K, for pure atactic polystyrene and for a polymer nanocomposite comprised of atactic polystyrene copolymerized with 4 mole % cyclohexyl POSS. As seen from the graph, there is a decrease in the movement of the polymer backbone when POSS is added to the pure atactic polystyrene. This indicates that the addition of POSS to pure polymers restricts the movement of the polymer backbone thereby causing an increase in the glass transition temperature and improvements in the mechanical properties of the polymer it is added to.
Acknowledgement The authors would like to acknowledge NSF EPSCoR grant EPS 0132618 for financial support and the National Center for Supercomputing Applications at the University of Illinois, Urbana Champaign for providing computer time for some of the simulations in this study.
References 1. Li, Q.Z; Wang, L; Ni, H; Pittman, Jr., C.U. Journal of Inorganic and Organometallic Polymers 2002, 11, 123. 2. Bharadwaj, R. K.; Berry, R. J.; Farmer, B. L. Polymer 2000, 41, 7209. 3. Haddad, T.S.; Brent, D.V.; Shawn, H.P. Journal of Inorganic and Organometallic Polymers 2002, 11, 155. 4. Li, Q.Z.; Toghiani, H.; Mather, P.; Pittman, Jr., C. U. Manuscript in preparation. 5. Cerius2 4.7 San Diego, CA: Accelrys Inc., 2001. 6. Rodriguez, F. Principles of Polymer Systems, Second Ed., McGraw-Hill Publishers, 1982, pp 33-39 and 50-55. 7. Bandrup, J; Immergut, E.H: Polymer handbook, Wiley Interscience, Second Ed., (1989) v/77-v/86.
134 Conference on Current Trends in Computational Chemistry 2003
Simulation of Molecular Magnets within Density Functional Theory
Mark R. Pederson
Center for Computational Materials Science Naval Research Laboratory Washington DC 20375-5000
Observation of resonant tunneling of magnetization in Mn12-Acetate has focused significant attention on a novel class of spin-ordered organometallic molecules. These molecular magnets consist of approximately 70-200 atoms and are typically composed of 4-15 transition metal atoms locked in place by organic ligands and anions. In addition to very interesting chemistry and physics, such molecular systems could form the basis of future nanoscale devices. Potential applications include high-density magnetic storage devices, q-bits for quantum computing, biomedical imaging, and tunable high-frequency radiation sources. The fundamental figure of merit that determines the resonant tunneling fields is the second-order magnetic anisotropy Hamiltonian that is governed primarily by the spin-orbit interaction. The large size of these systems and the small energy scales of interest also provide an interesting challenge to quantum-mechanical-based computational methods. In this talk I will give some background on the relevant experiments and then describe our efforts at predicting properties of molecular magnets with an all-electron density-functional formalism. A brief review of the massively parallel computational method, NRLMOL, will be given followed by applications to molecules of current interest. With respect to Mn12-Acetate, I will first discuss the calculation of the second-order anisotropy Hamiltonian followed by calculations aimed at understanding the observed symmetry breaking in this molecule. To further benchmark the reliability of density- functional theory for such systems, similar results on several other molecular magnets will be presented.
Various parts of this work were performed in collaboration with S.N. Khanna, J. Kortus, S. Hellberg, T. Baruah, N. Bernstein and K. Park. Conference on Current Trends in Computational Chemistry 2003 135
Prediction of Free Energies of Aqueous Solvation
Zenaida Peralta-Inga, Ping Jin, Jane S. Murray and Peter Politzer
Department of Chemistry, University of New Orleans New Orleans, LA 70148
We have shown, in a series of studies, that a variety of condensed phase properties that depend upon noncovalent interactions can be expressed quantitatively in terms of specific features of the molecular surface electrostatic potentials. In this presentation, following upon earlier work, we extend this approach to aqueous solvation. We have developed correlations that permit the free energies of solvation to be predicted with good accuracy from two or three computed quantities that characterize the solutes’ molecular surface potentials. The physical basis for these correlations will be discussed. 136 Conference on Current Trends in Computational Chemistry 2003
Comparison ab Initio Study of the Double-Proton Transfer in Methylated and Non-Methylated DNA Base Pairs
Yevgeniy Podolyan, Leonid Gorb, Jerzy Leszczynski
Computational Center for Molecular Structure and Interactions, Jackson State University, Department of Chemistry, 1325 Lynch St., Jackson, MS 39127
Double-proton transfer in DNA pairs is a process, in which proton of one base is transferred to the other one and visa versa. As an intramolecular proton transfer in DNA bases, double-proton transfer can also lead to mutations. Since the hydrogen atoms in the position 9 of purine and 1 of pyrimidine bases are replaced with a sugar moiety in DNA, the bases methylated at corresponding positions should possess the properties more closely resembling those of the nucleotides. The processes of double proton transfers in the DNA base pairs have been studied using ab initio methods. We have optimized local minimum and transition state structures for gas- phase isolated Guanine–Cytosine (GC) and Adenine–Thymine (AT) pairs and for methylated mGmC and mAmT base pairs. The optimizations of all structures were performed at B3LYP/6- 31G(d) and MP2/6-31G(d) levels of theory. Complete potential energy surface scans have been performed for a gas-phase proton transfer in isolated base pairs at B3LYP/6-31G(d) level. The analysis of the transition state structures and the potential energy surfaces confirms that there exist only one double-proton transfer form of the CG (and mGmC) base pair, which can be obtained through a double-proton transfer. However, the proton transfer process is predicted to be synchronous in the methylated species, rather than asynchronous one-step mechanism expected in non-methylated bases. As in non-methylated base pairs, the proton transfer in methylated mAmT pair goes through a transition state that is less than 1 kcal/mol higher than the rare double-proton transfer form. Even more, the transition state disappears from the surface of the Gibbs free energies.
Conference on Current Trends in Computational Chemistry 2003 137
Conformational Properties of Anti-HIV Prodrgus: AZT Derivatived
1Jillian Pope*, 1Jesse Edwards, 2Henry Joung Lee, 2Zhengqing You
1Department of Chemistry/AHPCRC Florida A&M Tallahassee, Florida,USA 32307, 2College of Pharmacy and Pharmaceutical Sciences, Florida A&M Tallahassee, Florida,USA 32307
In an attempt to develop anti-HIV agents devoid of serious toxic effects, H.J. Lee et al. synthesized several potential anti-HIV drugs along the prodrug scheme.Using molecular mechanics techniques in the Sybyl suite of programs several low-lying conformations were calculated. The structure of these low-lying conformations will be discussed. This work will also show a strong correlation between the bioactivity, molecular volume and diffusive properties of this unique class of compounds. 138 Conference on Current Trends in Computational Chemistry 2003
Conformational Study of Cyclopentadecane
Arnaldo O. M. Porrata, Dalephine Davis, Judge Brown, Diwakar M. Pawar, Eric A. Noe
Contribution from Jackson State University, Department of Chemistry, 1400 J. R. Lynch Street, Jackson, MS 39217-0510
The 13C spectrum of a 1% solution of cyclopentadecane (1) in propane showed the presence of 8 overlapping peaks at –170°C (dial temperature), although the exact number is uncertain because of a low signal-to-noise ratio. The peaks could be consistent with a conformation of C2 symmetry, but the possibility of additional conformations cannot be excluded. The carbon NMR spectra were analyzed in terms of the conformations predicted by Allinger’s molecular mechanics program, which indicated that a conformation of C2 symmetry (1a) is of lowest strain energy and is more stable than the next (C1) conformation (1b) by 0.32 kcal/mol. The free energies for the first ten conformations were obtained at the HF/6-311G* level and compared with the MM3 results: 1b is calculated to be lower in free energy than 1b by 0.85 kcal/mol at 25°C. Carbon chemical shifts were obtained for these conformations of 1 by the Gauge-Including-Atomic-Orbital (GIAO) method and compared with experimental values. This work was supported by NSF-CREST Grant No. HRD – 980 5465.
Conference on Current Trends in Computational Chemistry 2003 139
Stereoelectronic Interactions and Chemical Reactivity of a Masked Electrophile
Harry L. Price
Department of Chemistry, Stetson University, DeLand, Florida 32724
In biological systems electrophilic centers are often stabilized or masked via structural modifications. Masking serves to control chemical reactivity and in some cases, may assist in steering a reaction to a specific region or face of a molecule. In the work presented here, calculations were performed to investigate the chemical reactivity of the cyclopropylpyrrololindole (CPI) alkylating center in the microbial antibiotic CC-1065.
CH3 HN NH2 N O O N OH O N N OMe N H H O OH MeO
CC-1065 The biological activity of CC-1065 is linked to the covalent modification of DNA via a cyclopropyl group1,2. The reactive center (I) of this molecule represents a complex system of atoms whose reactivity and hence biological activity, is effectively switched on and off by the simple process of protonation (II). The CPI moiety is interesting because upon activation (i.e., unmasking) it reacts with nucleophiles to yield a distribution of the two products shown below, instead of a 50:50 product mixture, or even a reversed distribution in terms of the major and minor products.
Nu Nu CH 3 CH3 CH H C N CH3 3 3 H3C N H3C N H3C N O O N N O + O H N N O H H H OH+ OH OH I II 80% 20%
The primary goal of this study was to catalog stereoelectronic changes in the molecular framework of I brought about by protonation of the quinone oxygen. Unprotonated (I) and protonated (II) CPI fragments were studied using the B3LYP DFT method and the 6-31G* basis set. Results were used to better understand the observed product distribution in terms the 3-5 structural and electronic descriptors bond length (r), charge density (ρb), the Laplacian of the 2 charge density (∇ ρb), and bond ellipticity (εb). The numbering scheme of backbone atoms is shown below. 140 Conference on Current Trends in Computational Chemistry 2003
10C 1C 13 2 C15 C24 C28 N C C19 C7 C3 C8 20 6 4 O C C C21 C5 N22 14O
How does protonation effect the molecular structure of the CPI fragment? The answer to this question is, subtly yet significantly. A root-mean-square overlay of the optimized geometries of structures I and II reveal interesting deformations distributed throughout the molecular fragment. Most notably, both the C1-C2 and C2-C15 cyclopropyl bonds become elongated, with the bonds increasing by 0.014 and 0.024Å, respectively. Notably, bond C1-C15 underwent a 0.016Å contraction. Protonation increased the C5-O14 bond length by 0.099Å. Bonds containing nitrogen that exhibit pronounced changes are C19-N13, and C7-N13, with these bonds increasing and decreasing in length in a complimentary fashion by 0.030Å and 0.035Å, respectively. Of the bonds making up the quinone ring, bonds C3-C4 and C6-C7 increased in length by 0.021 and 0.027 Å, respectively, whereas the remaining bonds of the quinone ring underwent contraction, with bonds C4-C5, C5-C6 exhibiting decreases of 0.054 and 0.060Å, respectively, and bonds C2-C7, and C2-C3 exhibiting contractions of 0.017 Å and 0.009 Å, respectively. It is interesting that the non-bonded distance between N13 and C15 increased more than 0.115 Å upon protonation. Movement away from N13 could reduce thru-space transfer of electron density to C15, which in turn could change the electrophilic character of this key atom. Are structural changes brought on by protonation associated with electronic changes? If 2 so, are these changes significant? Comparision of ρ(r),∇ ρ(r), and εb in I and II reveal a great deal of useful information that may be used to answer these questions. The results clearly indicate that the effects of protonation are distributed throughout the CPI fragment and not localized to the cyclopropyl and quinone groups. Protonation produces decreases in ρ(r), for both C1-C2 and C2-C15 bonds, with the C2-C15 bond losing more than 40% of its charge density compared to C1-C2. Bond C5-O14 loses more than 20% the original ρ(r) upon protonation. The bonds attached directly to the quinone group, bonds C4-C5 and C5-C6 exhibit an increase in ρ(r) of 8.3 and 9.4%, respectively. Bonds C19-N13 and C7-N13 parallel changes in bond length, with bond C19-N13 exhibiting a decrease of 0.019 a.u. in ρ(r) while C7-N13 exhibited an increase of 0.013 a.u. Overall, changes in ρ(r) parallel changes in bond length. That is, increases in bond length result in decreased ρ(r), and vice versa. Protonation of I results in significant changes in ∇ 2ρ(r). This descriptor is an indicator of the amount of charge located in the internuclear region between bonded atoms. When ∇ 2ρ(r) < 0, charge is concentrated in the internuclear region between atoms, and is indicative of a shared interaction. Conversely, a closed-shell interaction is characterized by a∇ 2ρ(r) > 0, and a depletion of charge in the internuclear region. Results of this study indicate that bonds C1-C2 and C2-C15 exhibit a 13 and 23% increase in ∇ 2ρ(r) respectively, in II compared to I. Bonds C2-C3, C2-C7, C4-C5, C5-C6, and C5-O14 all exhibit increased negative values of ∇ 2ρ(r), while bond C3-C4 and C6-C7 exhibit positive increases in ∇ 2ρ(r). This trend is clearly consistent with increased delocalization. Protonation influences nitrogen containing bonds, with bonds C7-N13, C19-N13, and C4-N22 each exhibiting marked decreases in ∇ 2ρ(r). Most notable is the reduction of ∇ 2ρ(r) for bond C4-N22. Protonation resulted in a 57% decrease in ∇ 2ρ(r) for this bond. The apparent contribution of charge to the quinone ring by these nitrogens is complimented by a increase in ∇ 2ρ(r) for bonds C19-O20 and C21-N22. Conference on Current Trends in Computational Chemistry 2003 141
Changes in εb, an indicator of how much a bond is elongated in the plane perpendicular to the bond path reveals the general shape and character of the bond (i.e., the degree of π-character, etc.), also reveal sensitivity towards protonation. In keeping with the observed trend, bond C2- C15 reveals a greater change in εb upon protonation than bond C1-C2 (0.196 versus 0.084), changes consistent with bond distortion and alteration of charge density along the bond path. Changes in εb of the quinone ring reflect increases in εb for bonds C2-C3, C2-C7, C4-C5, and C5-C6. Decreases in εb are observed for bonds C3-C4, C6-C7, and the quinone group C5-O14. These changes support increased delocalization in the direction of bond C5-O14. Further evidence for the donation of charge by N13 and N22 is provided in the values of εb for bonds C7- N13 and C4-N22. Bond C7-N13 undergoes a loss of ellipiticity from 0.160 in I to 0.0501 in II, whereas bond C4-N22 exhibits an increase in εb of from 1.474 in I to 1.591 in II. These changes are consistent with changing degrees of conjugation, with C7-N13 exhibiting more single-bond character than C4-N22, which is assuming greater π-character. Perhaps the most intriguing change in εb is observed in bond C5-O14, where protonation causes a drastic reduction in εb from a value of 0.268 in I to 0.086 in II, a 68% reduction, again a reduction which suggests a loss of π-character and increased single bond character. In summary, the results of this study reveal a great deal about the effect that protonation has on the structure and electronic features of the CPI fragment. Protonation clearly perturbs the molecular scaffold of the fragment in such a way that C15 of the cyclopropyl ring becomes more reactive towards nucleophiles than C1. Moreover, activation is clearly linked to long-range electronic interactions involving both conjugation and hyperconjugation as evidenced by changes 2 in ρ(r),∇ ρ(r), and εb. The results of this study also revealed a possible stabilizing interaction between the bridgehead carbon C15 and N13, whose close proximity may provide supplemental electron density and hence, suppress the electrophilic character of C15 and possibly even the entire cyclopropyl group in the unprotonated state. This study also revealed that protonation diminishes this potential stabilizing interaction between N13 and C15 via an increase in the distance between these atoms. The decrease in non-bonded interactions between C15 and N13 together with conjugative effects distributed throughout the CPI fragment result in polarization of the molecule in such a way that the electrophilicity of C15 is selectively enhanced relative to C1. These stereoelectronic changes induced by protonation may explain, at least in part, the observed product distribution when II undergoes nucleophilic substitution.
References 1. Warpehoski, M.A., and Harper, D.E. (1994) J. Am. Chem. Soc. 116, 7573-7580. 2. Boger, D.L., and Johnson, D.S. (1995) Proc. Natl. Acad. Sci. USA 92, 3642-3649.Bader, R.F.W., and Chang, C. (1989) J. Am. Chem. Soc. 93, 5095-5107. 4. Carroll, M.T., Cheeseman, J.R., Osman, R., and Weinstein, H. (1989) J. Phys. Chem. 93, 5120-5123. 5. Shi, Z., and Boyd, R.J. (1993) J. Am. Chem. Soc. 115, 9614-9619. 142 Conference on Current Trends in Computational Chemistry 2003
Theoretical Predictions as to Initial/Intermediate Steps in Chemical Transformation of Cyclic and Cage Cyclic Nitramines Are Supported by UV/VIS/FTIR Spectrophotometry, Using CL-20 as the Model
Mohammad (Mo) Qasim, Herbert L. Fredrickson, John Furey, Chris McGrath
ERDC-EL, MS, Vicksburg, MS 39180
Establishing feasibilities as to energetic compound degradation through structural comparison of these compounds and their initial/intermediate transformation products is useful to present DoD cleanup efforts and to providing a basis for extrapolation of predictions to new energetic materials. The resultant database, identifying molecular characteristics that can be predicted to have greatest effect on energetic environmental fate relationships, helps to indicate these feasibilities or lack of them as new degradation methodologies are developed. Combining of computer chemistry calculation with experimental verification is useful in proving concepts and in proving what is chemically possible. UV/VIS/FTIR and a monochromatic irradiation system (to photo-induce free radical reactions) offers a clean, flexible, inexpensive and fast way to track subtle changes in spectral shape and follow rates of fast-forming intermediates. The hypothesis: molecular structure, under homologous conditions, determines preferred degradation pathways that can be theoretically predicted was first examined in regard to cyclic and cage cyclic nitramines through semi-empirical methods. Then, UV/VIS spectrophotometry and FTIR were used to verify both hypothesis and theoretical study, using 2,4,6,8, 10,12-hexanitrohexaazo-iso-wurtizane (CL-20) as the experimental model and following its course of degradation to fast-forming transition states and initial/intermediate products. Thus far, the results of our spectrophotometric study do support our structural computational chemistry study and predictions. Our Approach included a theoretical study which supported the hypothesis plus verification of first tier transformation products of CL-20 due to reactions with hydroxide ions and with free radical producing agents through UV/VIS spectrophotometry to follow the course of degradation. FTIR was used to study the functional groups—nitro, amino, carbon-carbon (C- C) and carbon-nitrogen (C-N) bonds—indicating breaking and formation of bonds.
The poster presents both computational predictions and supporting UV/VIS/FTIR spectral data: Computational predictions: Most likely first tier intermediates of cyclic and cage cyclic nitramines were compared as to bond lengths and angles, heats of formation, steric energy, dipole moments, solvent accessibility and electrostatic potential surfaces, partial charges, and HOMO/ LUMO energies through MOPAC quantum mechanical and classical force field mechanics. UV/VIS/FTIR spectral data: Two competing modes of degradation are summarized: through addition of hydroxide ions and through addition of photo induced free radicals. A monochromatic irradiation system photo- induced free radical reactions through irradiation at targeted wavelengths of compound absorbance. UV/VIS measured concentration and followed the course of reactions. FTIR followed CL-20 degradation through alkali hydrolysis, where FTIR measurements verified theoretical predictions.
Conference on Current Trends in Computational Chemistry 2003 143
Conclusions: Calculations indicate that the free radical mechanism (symmetrical bond-breaking) is more apt to occur upon increase in number of symmetrical C-C (preferred) then N-N bonds contained within the molecule; The less polar bonds have a greater tendency toward free radical bond-breaking. Therefore, the higher the ratio of C-C bonds (CL-20 has three cage C-C bonds), the more likely that degradation will proceed via free radical oxidation, as reflected by lower HOMO/LUMO energies. In other words, under similar conditions, the higher the ratio of non-polar bonds, the more likely that degradation will proceed via free radical oxidation. Also, the more strained the molecular structure, the greater is the tendency toward free radical bond-breaking via intersystem crossing to the triplet state—where changes in molecular geometry take place. FTIR measurements of changes in functional groups (nitro, amino, C-C and C-N bonds) during appearances/ disappearances of intermediate species verified the hypothesis.
144 Conference on Current Trends in Computational Chemistry 2003
Quantitative Structure-Activity Relationship (QSAR) Study by GA- MLR Analysis of the Antioxidant Activity of Flavonoids
B.F. Rasuleva,b, N.D. Abdullaeva, V.N. Sirova, J. Leszczynskib
aInstitute of Chemistry of Plant substances AS RUz, Kh. Abdullaev Str., 77, Tashkent, 700170, Uzbekistan bComputational Center for Molecular Structure and Interactions, Jackson State University, 1325 J.R. Lynch Street, P.O. Box 17910, Jackson, Mississippi 39217-0510
Flavonoids occur in most plant species. Flavonoids have been shown to have anti- inflammatory, antiallergic, antibacterial, antimutagenic, antiviral, antineoplastic, anti-thrombotic, and vasodilatory activity. The potent antioxidant activity of flavonoids—their ability to scavenge hydroxyl radicals, superoxide anions, and lipid peroxy radicals—may be the most important function of flavonoids, and underlies many of the above actions in the body. Oxidative damage is implicated in most disease processes, and epidemiological, clinical, and laboratory research on flavonoids and other antioxidants suggest their use in the prevention and treatment of a number of these. Catechin and its derivatives, oligomeric proanthocyanidins, quercetin and quercetin chalcone, Ginkgo flavone glycosides, silymarin, and others can be utilized in preventative and treatment protocols for cardiovascular disease, cancer, inflammatory conditions, asthma, periodontal disease, liver disease, cataracts and macular degeneration. The structural components common to these molecules include two benzene rings on either side of a 3-carbon ring. Multiple combinations of hydroxyl groups, sugars, oxygens, and methyl groups attached to these structures create the various classes of flavonoids: flavanols, flavanones, flavones, flavan-3-ols (catechins), anthocyanins, and isoflavones. In this report, the quantitative structure-activity relationship (QSAR) analyses and ab initio calculations were performed on a series of flavonoids. Molecular descriptors were computed mainly using the software DRAGON of Todeschini and Consonni for a set of 27 flavonoid compounds with antioxidant activity. The input files for descriptor calculation, containing information on atom and bond types, connectivity, partial charges and atomic spatial coordinates, relative to the minimum energy conformation of the molecule, were obtained from MDL Mol file after optimization by the DFT quantum-chemical method B3LYP/6-31G(d, p). 1150 molecular descriptors of different kinds were used to describe compound chemical diversity from DRAGON software. In addition, several quantum-chemical parameters, calculated by DFT B3LYP/6-31G(d, p) method were added as molecular descriptors. The correlation between biological activity and structural properties was obtained by using the variable selection Genetic Algorithm (GA) and multiple linear regression analysis (MLR) methods. The model obtained showed not only statistical significance but also predictive ability. The significant molecular descriptors related to the compounds with antioxidant activity were: Molecular shape descriptor, Dipole moment, indicator variable of the presence of hydroxyl groups (OH) in relevant positions and some topological descriptors. These variables led to a physical explanation of electronic molecular property contributions to potency of lipids peroxidation inhibition. Two models have been derived using calculated. These models can be used to estimate the antioxidant activities of new substituted phenolic compounds or flavonoids derivatives. The QSAR model established here may reasonably predict the antioxidant activity using simple calculated parameters. Conference on Current Trends in Computational Chemistry 2003 145
Ab Initio Study of the Rotation around the C=C Bond in Push-Pull Systems
Pornpun Rattananakin,1 Yevgeniy Podolyan,2 Svein Saebø,1 and Charles U. Pittman, Jr.1
1Department of Chemistry, Mississippi State University, MS 39762 2Department of Chemistry, Jackson State University, MS 39217
Push-pull electron systems consist of an electron donor and an electron acceptor coupled through a conjugated system. Two simple examples are cis-3-amino-2-propenenitrile 1a and trans-3-amino-2-propenenitrile 1b where NH2 is the electron donor (push) and C≡N is the electron acceptor (pull). The purpose of this study is to examine the push-pull effect across C=C (+) system while varying the electron acceptor, e.g. C≡N (1), NO2 (2), BH2 (3), and CH2 (4). We chose the rotational barrier for the internal rotation around the C=C bond as an indicator of the push-pull effects. Our studies included optimization of the local minima (cis- and trans- forms) as well as the transition state structures at B3LYP/6-31++g(d,p) and MP2/6-311++G(d,p) ab initio methods. The Synchronous Transit-Guided Quasi-Newton (STQN) method was used to search for the transition state structures. Our transition states all have closed shells unlike ethylene’s transition state which exists as a diradical. The results from our calculations indicate (+) that the strongest push-pull effect is detected in the case of the CH2 acceptor which gives a barrier (8.2 kcal/mol) which is about 7 times smaller than ethylene (58.0 kcal/mol). On the other hand the weakest push-pull effect is observed for the C≡N acceptor for which the barrier (51.2 kcal/mol) is only slightly lower than in ethylene. The analysis of the electron density distribution has been performed by mapping electron density in Molden program and by calculation of the electrostatic potential-derived charges on atoms according to Merz-Singh-Kollman scheme.
N N H C H C C C C C H NH2 H2N H 1a 1b
+ H NO2 H BH2 H CH2 C C C C C C H NH2 H NH2 H NH2 2 3 4
146 Conference on Current Trends in Computational Chemistry 2003
Electron Density Distribution
(+) cis- CH2CH=CHNH2
cis-CH3CH=CHNH2 Conference on Current Trends in Computational Chemistry 2003 147
Molecular Design for Ionic Nonlinear Optical Systems
Paresh Chandra Ray
Department of Chemistry, Jackson State University, Jackson, MS, USA, 39217.
In this presentation I want to emphasis mainly on the design strategy to maximize the Nonlinear Optical (NLO) properties of ionic molecules. Two series of push-pull ionic molecules are investigated to establish the structure –NLO- property relationships by carrying out the ab initio (HF/6-31G) calculations of the molecular polarizability (α), the first hyperpolarizability (β) and second hyperpolarizability (γ). It is found that the absolute Bond Length Alternation (BLA) values decreases and α, β and γ increases with introducing strong donors or acceptors. The solvent effects on the NLO properties are studied by using the self-consistent reaction field (SCRF) method. As the solvent polarity increases α, β and γ increases monotonically. 148 Conference on Current Trends in Computational Chemistry 2003
Molecular Dynamics of 5S rRNA Loop E
Kamila Reblova,# Nada Spackova,§ Richard Stefl,# Kristina Csaszar,¶ Jaroslav Koca,# Neocles B. Leontis¶* and Jiri Sponer§*
#National Center for Biomolecular Research, Kotlarská 2, 611 37 Brno, Czech Republic §Institute of Biophysics, Academy of Sciences of the Czech Republic and National Center for Biomolecular Research, Královopolska 135, 612 65, Brno, Czech Republic ¶Chemistry Department and Center for Biomolecular Sciences, Bowling Green State University, Bowling Green, OH 43403
Explicit solvent and counterion molecular dynamics simulations have been carried out for a total of > 80 ns on the bacterial and spinach chloroplast 5S rRNA Loop E motifs. The Loop E sequences form unique duplex architectures composed of seven consecutive non-Watson-Crick base pairs. The starting structure of spinach chloroplast Loop E was modeled using isostericity principles and the simulations refined the geometries of the three non-Watson-Crick base pairs that differ from the consensus bacterial sequence. The deep groove of loop E motifs provides unique sites for cation binding. Binding of Mg2+ rigidifies Loop E and stabilizes its major groove at an intermediate width. In the absence of Mg2+, the Loop E motifs show an unprecedented degree of inner-shell binding of monovalent cations that, in contrast to Mg2+, penetrate into the most negative regions inside the deep groove. The spinach chloroplast Loop E shows a marked tendency to compress its deep groove compared with the bacterial consensus. Structures with narrow deep groove essentially collapse around a string of Na+ cations with long coordination times. The Loop E non-Watson-Crick base pairing is complemented by highly specific hydration sites ranging from water bridges to hydration pockets hosting 2 to 3 long-residing waters. The ordered hydration is intimately connected with RNA local conformational variations.
K. Reblova, N. Spackova, R. Stefl, K. Csaszar, J. Koca, N.B. Leontis and J. Sponer: Non- Watson-Crick base pairing and hydration in RNA motifs: Molecular dynamics of 5S rRNA Loop E. Biophys. J. 84, 2003, 3564-3582. Conference on Current Trends in Computational Chemistry 2003 149
Theoretical Study on Thermal Conversion of Benzopyran: Substituent and Solvent Effects
Demarcio Reed, Yinghong Sheng, and Jerzy Leszczynski
The Computational Center for Molecular Structure and Interactions, Department of Chemistry, Jackson State University, P.O. Box 17910, 1400 J.R. Lynch Street, Jackson, Mississippi 39217
A comprehensive theoretical study has been conducted to study the possible reaction mechanisms for the conversion between the closed-form and open-form of substituted pyrans and benzopyrans. The reaction mechanisms for the thermal conversion between closed-form and open-form of substituted pyrans and benzopyrans, as well as that on the triplet potential energy surface were studied; several possible reaction mechanisms were proposed. A comprehensive mechanistic view of the ultrafast photochemistry of substituted pyrans and benzopyrans were obtained and interpreted in terms of the strengths of substituents. 150 Conference on Current Trends in Computational Chemistry 2003
gem-Dimethyl Substituents of Cyclopropane and Cyclobutane
Ashley L. Ringer and David H. Magers
Mississippi College, Department of Chemistry and Biochemistry Clinton, Mississippi
The gem-dimethyl effect is the acceleration of cyclization by substituents in the chain and is often used in organic synthesis as a ring-closing effect. In the study, calculations on methylcyclobutane (Figure 1), 1,1-dimethylcyclobutane (Figure 2), 1,2-dimethylcyclo-butane (Figure 3), and 1,3 dimethylcyclobutane (Figure 4) are performed to determine if this effect is a thermodynamic effect caused by lower strain energy or a kinetic effect that simply lowers the activation barrier for the cyclization reaction. 1,1-dimethylcyclobutane is a four-membered carbon ring with gem-dimethyl substituents.
Figure 1. methylcyclobutane Figure 2. 1,1-dimethylcyclobutane
Figure 3. cis-1,2-dimethylcyclobutane Figure 4. trans-1,3-dimethylcyclobutane
One hypothesis suggests that the strain energy is lower because the methyl groups have more freedom to move in the ring than in the open chain. Optimum equilibrium geometries, harmonic vibrational frequencies, and corresponding electronic energies are computed for all pertinent molecular systems using SCF theory, density functional theory (DFT), and second- order perturbation theory (MP2). Comparisons are made between the conventional strain energy of cyclopropane and cyclobutane, methylcyclobutane, and the dimethyl-substituted cyclobutanes. SCF and DFT calculations indicate that 1,1-dimethylcyclobutane is approximately 3 to 3.5 kcals/mol less strained than methylcyclobutane, which in turn is 3 to 3.5 kcals/mol less strained than cyclobutane; that is, there is at least some thermodynamic component to the gem- Conference on Current Trends in Computational Chemistry 2003 151
dimethyl effect. The study continues by calculating conventional strain energies for the 1,2 and the 1,3 dimethyl substituted cyclobutanes. Additionally, the optimum geometries of all relevant systems are examined, particularly noting the bonding angles in the rings, to study the relevance of geometry to the gem-dimethyl effect.
We gratefully acknowledge the support of NSF EPSCoR (EPS-0132618). 152 Conference on Current Trends in Computational Chemistry 2003
Time-Dependent Non-Wavepacket Theory of Electron Scattering
Burke Ritchie# and Charles A. Weatherford*
#University of California, Lawrence Livermore National Laboratory, Livermore, CA 94550 *Physics Department, Florida A&M University, Tallahassee, FL 32307
We are motivated to overcome the limitation of time-dependent wavepacket theory due to wavepacket spreading, which effectively limits scattering-electron momenta (k) to values greater than 1 a.u. We present in the paper, results for k2=0.1 and k2=1.0 a.u. We solve the time- dependent Schrödinger equation
∂ 1 [i + ∇ 2 ]ψ(r ,t) = V (r ,t)ψ(r ,t) ∂t 2
by direct solution of the integral equation version of the Schrödinger equation
r k 2 1 t i ψ(r ,t) = exp[ik • r − i t]+ dt'exp[ ∇ 2 (t − t')]V (r )ψ(r ,t') 2 i ∫ 2 0
*Supported by FAMU-Army High Performance Computing Research Center (FAMU-AHPCRC)
Conference on Current Trends in Computational Chemistry 2003 153
Bond Dissociation of Novel AZT Derivative Prodrugs
1Jamar Robinson*, 1Jesse Edwards, 2Henry Joung Lee, 2Zhengqing You
1Department of Chemistry/AHPCRC Florida A&M Tallahassee, Florida,USA 3230 2College of Pharmacy and Pharmaceutical Sciences, Florida A&M Tallahassee, Florida,USA 32307
In an attempt to develop anti-HIV agents devoid of serious toxic effects, H.J. Lee et al. synthesized 11 potential anti-HIV drugs along the prodrug scheme.The purpose of this work is to determine bond dissociation energies of the 11 prodrugs. This work will produce the bond dissociation energies of these 11 prodrugs using quantum mechanical techniques and discuss the implications to the mechanism of action of these prodrugs. We will also present correlation between the bond dissociation energy and activity.
154 Conference on Current Trends in Computational Chemistry 2003
An ab Initio Study of C-H···O Hydrogen Bonding in Peptide Models
Amy Rowe
Department of Physical Sciences, Delta State University, Cleveland, Mississippi
Hydrogen bonding involving the alpha carbon of a peptide backbone occurs frequently in proteins. Due to the lack of knowledge about the strength of these weak hydrogen bonds, however, the importance and function they serve in proteins is ambiguous. High level ab initio calculations performed on n-formyl-alanine-amide and acetamide offer evidence of a hydrogen interaction between the alpha carbon of alanine with the carbonyl oxygen of acetamide. The geometries for n-formyl-alanine-amide and acetamide were optimized using density functional theory (DFT) at the BLYP level and 3-21G, 6-31G*, and 6-311G** basis sets. Scans of the C- H…O hydrogen bond distance at varying C-H…O bond angles were computed at the 6- 311G**/MP2 level. Results show that the binding energies of some of the C-H…O hydrogen interactions have sufficient strength. Therefore, C-H…O hydrogen bonding and may be considered a factor in protein structure and function.
Conference on Current Trends in Computational Chemistry 2003 155
Hydrogen Bond Effect on Electronic and Vibrational (Hyper)Polarizability of The Dimmers (HF)2, (H2O)2, and (NH3)2: Theoretical Study
Amar Saal, Yahia Moussaoui, and Ourida Ouamerali
Laboratory of Physico-Chimie Théorique et Chimie Informatique, Faculty of Chemistry, USTHB- University, Algiers Algeria
Nonlinear optical responses such as generation of harmonics, Kerr and Pockels effects and wave mixing are governed by hyperpolarizabilities. It is now widely recognized that these quantities may be written as the sum of electronic and vibrational contributions[1], and references therein. The molecules HF, H2O, and NH3 are largely investigated. Their electrical properties, particularly vibrational (hyper)polarizability are calculated experimentally and at different level of theory. We are interested in this study by the evaluation of these quantities, as well as their electronic counterparts, of the dimmers (HF)2, (H2O)2, and (NH3)2. Our goal is to highlight the effect of the hydrogen (X······H) bond on the two contributions electronic and vibrational to the (hyper)polarizability. Note that the dimmer (HF)2 has been the subject of a previous investigation made by Bishop et al. [2]. Electronic and vibrational polarizability and first hypolarizability of the molecules HF, H2O, and NH3 and the dimmers (HF)2, (H2O)2, and (NH3)2 have been calculated at CPHF/aug- v e cc-pvtz level of theory. The results show that, the ratio β L / β L decreases in the order (NH3)2 >> (H2O)2 >> (HF)2, i.e. when the electronegativity of the element X (X······H) increases. Individual normal mode contribution to the first hyperpolarizability are also determined.
[1] Bishop D.M.; Norman P. Handbook of advanced Electronic and photonic Materials and devices, Volume9: Nonlinear optical, H. S. Nalwa, ( 2001 ) Chap. I.;Champagne B.; Kirtman B. Handbook of advanced Electronic and photonic Materials and devices, Volume9: Nonlinear optical, H. S. Nalwa, ( 2001 ) Chap. II. [2] Bishop D.M.; Pipin J. and Kirtman B. J. Chem. Phys. 102 (1995) 6778. 156 Conference on Current Trends in Computational Chemistry 2003
Efficient AO-Formulation of MP2-Gradients
Svein Saeboa, Krzysztof Wolinskib,d, Peter Pulayb,c, and Jon Bakerb,c,
aDepartment of Chemistry, Mississippi State University, Mississippi State, MS 39762 bParalell Quantum Solutions, LLC, 2013 Green Acres, Suite A, Fayetteville, AR 72703 cDepartment of Chemistry and Biochemistry, University of Arkansas, Faytetteville, AR 7270 dDepartment of Chemistry, University of Lublin, Lublin, Poland
We have recently implemented an efficient AO-formulation of MP2-gradients. The formulation can be used for any set of internal orbitals (e. g. localized orbitals), but currently it has only been mplemented for Canonical internal orbitals.. The orbital-invariant form of second- order Møller-Plesset perturbation theory can be derived from the Hylleraas functional form of the second-order energy. The functional form is very convenient for the formulation of the gradient, and the present derivation essentially follows our formulation from 1986 [1]. The formalism as well as its computer implementation will be described in detail. The program is currently only running on a single CPU. However, we will provide examples of MP2-gradient calculations on systems with ~100 atoms and about ~1000 contracted basis functions carried out on a PC, and timings and test-results will be provided. Parallelization of the program is in progress, and when this is completed, we expect that geometry optimization of systems with several hundred atoms can be routinely carried out on small PC-clusters at the MP2 level using suitable (~TZ+P or better) atomic basis sets.
1. Pulay, P and Saebo, S. “Orbital-invariant formulation and second-order gradient evaluation in Moller-Plesset perturbation theory” Theor. Chim. Acta 69, 357 (1986)
Conference on Current Trends in Computational Chemistry 2003 157
Theoretical Studies of Neutral and Ionized Lanthanide Halides
Julia Saloni1,2, Szczepan Roszak1,2, and Jerzy Leszczynski1
1The Computational Center for Molecular Structure and Interactions, Department of Chemistry, Jackson State University, 1400 J.R. Lynch Street, Jackson, Mississippi 39217 USA 2Institute of Physical and Theoretical Chemistry, Wroclaw University of Technology, Wyb. Wyspianskiego 27, 50-370 Wroclaw, Poland
Lanthanide halides LnX3, with Ln = Ce, Nd, Gd, Tb, Dy, Ho, Er, Tm and X= F, Cl, Br, I were investigated experimentally by Knudsen effusion mass spectrometry [1-2]. In high temperatures gaseous molecules of LnX3 follows processes of ionization and fragmentation. The mechanisms of those processes are poorly understood. Molecular structures and the nature of bonding of species appeared in such processes are not directly available from experimental data. The theoretically determined data can be very helpful in exploring above mentioned chemistry. + + + + In this project theoretical studies on LnX3, LnX3 , LnX2, LnX2 , LnX, LnX , and Ln were performed [3]. The quantum-chemical calculations were carried out using the density functional theory (DFT). All the calculations were obtained using relativistic effective core potentials (RECPs), developed by Stuttgart/Dresden groups. The basis sets of the halogens were additionally supplemented by two d polarization and a set of sp diffuse functions proposed as a part of SDB-aug-cc-PVTZ and aug-cc-PVTZ basis sets. Ground state molecular geometries were obtained for all studied species. The location of true minima was confirmed in calculations of vibrational frequencies. The electronic density distributions were studied applying the Mulliken population analysis. An example of obtained results dysprosium is discussed.
References: 1. J. Kapala, S. Roszak, S. Nunziante-Cesaro, and M. Miller J. Alloys and Compounds, 345 (2002) 90-99 2. K. Hilpert, U. Niemann, Therochim. Acta 299 (1997) 49 3. M. Miller Scientific Papers of the Institute of Inorganic Chemistry of The Wroclaw University of Technology, Oficyna Wydawnicza Politechniki Wroclawskiej, Wroclaw, 1997, No.66, Ser. Monographs. No 33.
158 Conference on Current Trends in Computational Chemistry 2003
Theoretical Study of Gas Phase Tautomerization Reactions for the Ground and First Excited Electronic States of Adenine
Latasha M. Salter*1,2 , Galina M. Chaban3 , and Jerzy Leszczynski2
1Tougaloo College, 500 West County Line Road, Tougaloo, Mississippi 39174 2Computational Center for Molecular Structure and Interactions, Jackson State University, Jackson, Mississippi 39217 3NASA Ames Research Center, Mail Stop T27B-1, Moffett Field, California 94035-1000
Geometrical structures and energetic properties for different tautomers of adenine are calculated in this study, using multi-configurational wave functions. Both the ground and the lowest singlet excited-state potential energy surface are studied. Four tautomeric forms are considered, and their energetic order is found to be different on the ground and the excited state potential energy surfaces. Minimum energy reaction paths are obtained for hydrogen atom transfer (tautomerization) reactions in the ground and the lowest excited electronic state. It is found that the barrier heights and the shapes of the reaction paths are different for the ground and the excited electronic state, suggesting that the probability of such tautomerization reaction is higher on the excited state potential energy surface. This tautomerization process should become possible in the presence of water or other polar solvent molecules and may play an important role in the photochemistry of adenine.
Figure 1. Schematic structure and atom numbering for adenine tautomers. Conference on Current Trends in Computational Chemistry 2003 159
Calculations on Transition Metal Oxides – From Gas Phase Clusters to Solid Catalysts
Joachim Sauer
Institut für Chemie, Humboldt-Universität, 10099 Berlin, Germany
Vanadium oxides are selective catalysts for partial oxidations. There is particular interest in the effect that the aggregation level of the material has on its reactivity. The materials considered include the ideal single crystal V2O5 (001) surface, thin films of V2O5 on α-Al2O3, isolated V2O5 sites supported on silica and both neutral and charged V2O5 gas phase clusters. Periodic slabs and large finite clusters are used as models. As test reaction oxygen removal from the surface is examined and the differences found are related to different structural and electronic features [1,2]. In addition the energy profile for the selective oxidation of methanol to formaldehyd is calculated. This reaction has been experimentally studied both as a gas phase reaction for a related molecule cation [3] and as a catalytic process on solid catalysts. It is frequently claimed that gas phase studies on cluster cations in the gas phase (mass spectrometer) can provide information on particle size effects on the reactivity of supported catalysts. The results obtained here reveal large reactivity differences between cationic gas phase species and neutral clusters on supports. The studies are made by DFT. The accuracy of gradient corrected functionals (PW91) and the hybrid B3LYP functional for the oxygen-metal bond and the activation barrier is discussed. The methanol oxidation is one of the less common examples that DFT may yield too high barriers [4]. The reason is the biradicaloid nature of the transition structure. Improved results are obtained by the broken symmetry approach. The reliability of this approach for the general case of weakly coupled d-electrons on neighboring transition metal sites is examined by comparison with MR-CI calculations [5,6].
1. V. Ganduglia-Pirovano, J. Sauer (2003). 2. V. Brázdová, V. Ganduglia-Pirovano, J. Sauer, Phys. Rev. B, in preparation. 3. D. Schröder, J. Loos, M. Engeser, H. Schwarz, H.-C. Jankowiak, R. Berger, R. Thissen, O. Dutuit, J. Döbler, J. Sauer, Inorg. Chem., submitted. 4. J. Döbler, M. Pritzsche, J. Sauer (2003). 5. M. Pykavy, C.v. Wüllen, J. Sauer, in preparation. 6. O. Hübner, J. Sauer, Phys. Chem. Chem. Phys. 4 (2002) 5234 160 Conference on Current Trends in Computational Chemistry 2003
“Aromatic” 4n π-system Diazapentalenes
Yinghong Sheng, 1 Ray Butcher, 2 Bakhtiyor Rasulev,3 Jerome Karle 2 and Jerzy Leszczynski 1
1 The Computational Center for Molecular Structure and Interactions, Department of Chemistry, Jackson State University, P. O. Box 17910, 1400 J. R. Lynch Street, Jackson, Mississippi 3921 2 Laboratory for the Structure of Matter, Code 6030, Naval Research Laboratory, Washington, DC 20375-5341 3 Institute of Chemistry of Plant substances, AS RUz, Kh.Abdullaev str., 77Tashkent, 700170, Uzbekistan
Conjugated hydrocarbons offer a vast resource of structural diversities. The conjugated cyclic hydrocarbons, such as pentalene, are of particular interests in the modern organic, as well as physical and theoretical chemistry. Pentalene is a thermally unstable compound belonging to the class of destabilized, anti-aromatic π-systems. Stabilization of the pentalene system can be achieved by steric shielding (e.g., by tert-butyl groups) or by the introduction of donor groups in the 1(3,4,6) position and acceptor groups in the 2(5) position [1]. 1,3,4,6-tetradonor-2,5- diacceptor-substituted pentalenes has been reported to exhibit aromatic stabilization and a delocalized π-bonding system [2]. Replacement of two carbon atoms on the pentalene rings leads to the diazapentalenes. Among them, 2,5-diazapentalene and 7,8-diazapentalen are of particular interests. So far 7,8- diazapentalenes which contain diones have been reported. Recently successfully synthesizes of the nitro-substituted 7,8-diazapentalenes, such as 1-nitro-7,8-diazapentalene, 1,3,4-tris(nitro)- 7,8-diazapentalene and 1,3,4,6-tetrakis(nitro)-7,8-diazapentalene, has been reported. It is expected that 7,8-diazapentalene should show the similar features as the pentalene, thus electron- donating groups, such as amino group, would stabilize the 7,8-diazapentalene, while electron- withdrawing groups destabilize the system. However, no 7,8-diazapentalene containing electron- donating groups has been reported. Therefore it is of interest to assess how the nitro groups stabilize the 7,8-diazapentalene. Its analogue, 2,5-diazapentalene, also called pyrrolo[3,4-c]pyrrole, is expected to be nonaromatic. Recently the X-ray structure of the potassium and ammonium salts of the dianion + 2- + 2- of 1,3,4,6-tetrakis(nitro)-2,5-diazapentalene, 2K .C6N6O8 and 2NH4 .C6N6O8 have been reported [3]. This is the first time that 2,5-diazapentalene form anions, whereas in the literature 2,5-diazapentalenes are bases and forms cations. The mechanism that allows 2,5-diazapentalene to form the dianion or dication is still unknown. In this paper, the stablibilty of pentalene, and 7,8-diazapentalene, 2,5-diazapentalene as well as their substituted derivatives as predicted at the ab initio HF/6-31G(d) and B3LYP/6- 31G(d) levels are discussed. In addition, the cationic and anionic forms of 2,5-diazapantalenes are also investigated.
Conference on Current Trends in Computational Chemistry 2003 161
pentalene 7,8-diazapentalene 2,5-diazapentalene
K K
References: 1. Gimarc, B. M. J. Am. Chem. Soc. 1983, 105, 1979. 2. Kataoka, M.; Ohmae, T.; Nakajima, T. J. Org. Chem. 1986, 51, 358. 3. Butcher, R. J.; Bottaro, J. C.; Gilardi, R. Acta Cryst, 2003, 59, 591.
162 Conference on Current Trends in Computational Chemistry 2003
MCSCF Study on Excited State Properties of Thiouracil and Its Comparison with Uracil
M.K. Shukla and Jerzy Leszczynski
Computational Center for Molecular Structure and Interactions Department of Chemistry, Jackson State University 1400 J.R. Lynch Street, Jackson, MS 39217 USA
The thio analogs of nucleic acid bases have distinctive biological and pharmacological activities. Several thio derivatives have been used as drugs against different diseases. Thiouracil has been shown as a minor constituent of t-RNA and can be used as anticancer and antithyroid drugs. A comprehensive theoretical study of singlet electronic transitions of 2-thiouracil, 4- thiouracil and 2,4-dithiouracil was performed at the Multi-Configurational Self-Consistent Field (MCSCF) level of the theory and their results were compared with those of the uracil at the same level of the theory. The ground state geometry of the molecule was optimized at the MP2/6- 311G(d,p) level. The MCSCF calculations were performed using the 6-311+G(d) basis set. The active space for the MCSCF calculation was consisted of 12 orbitals (11 orbitals in case of uracil) in which 6 orbitals (5 orbitals in uracil) were occupied π type while remaining 6 orbirals were virtual π* type. In order to compute nπ* transitions, the two occupied orbitals were replaced with two σ orbitals localized at the thio (oxo in uracil) groups. The effects of dynamic electron correlation on the MCSCF energies were considered at the second-order multiconfigurational quasi-degenerate perturbation (MCQDPT2) theory. The computed transition energies after the MCQDPT2 correlation were found to be in agreement with the experimental data. The ground and excited state charge distributions, dipole moments and molecular electrostatic potentials maps are also discussed in the present work.
Conference on Current Trends in Computational Chemistry 2003 163
X2=y4=o : uracil Y4 x2=s,y4=o : 2-thiouracil x2=o,y4=s : 4-thiouracil x2=y4=s : 2,4-dithiouracil C4 H5 C5 H3 N3
C6 C2 N1 X2 H6
H1
S0 S0
S1(ππ*) S2(ππ*)
S3(ππ*) S1(nπ*) Figure. Molecular Electrostatic Potential Maps of 2,4-dithiouracil in the ground and vertical singlet excited states
164 Conference on Current Trends in Computational Chemistry 2003
MCSCF Investigation of Singlet Excited State Geometries of Guanine Tautomers
M.K. Shukla and Jerzy Leszczynski
Computational Center for Molecular Structure and Interactions Department of Chemistry, Jackson State University 1400 J.R. Lynch Street, Jackson, MS 39217 USA
Guanine is an important brick of building blocks of nucleic acids. It shows the maximum tautomeric activity among the natural nucleic acid bases. Recent experimental investigations in supersonic beam show that guanine can have four tautomeric forms (keto-N9H, keto-N7H, enol- N9H, enol-N7H). There are two experimental studies in this regard. However, they differ with respect to the spectral origin (0-0 transition) of these tautomers. Resolving these ambiguities demand high level of theoretical calculations on guanine tautomers in the ground and lowest singlet excited state. We have used Multi-Configurational Self-Consistent Field (MCSCF) level of theory in this work. Ground and lowest singlet excited state geometries were optimized at the MCSCF level employing the 6-31G basis set with d-type of polarization function on amino nitrogen (the basis set hereafter called as 6-31G(NH2*). The active space in this calculation consisted of 10 electrons distributed in the 9 π orbitals. The MCSCF/6-31G(NH2*) optimized geometries were used to performed single point calculation employing the 6-311++G(d,p) basis set at the MCSCF level. In this case active space was consisted of 10 electrons distributed in the 10 π orbitals. The effects of dynamic electron correlation on the MCSCF/6-311++G(d,p) energies were considered at the second-order multiconfigurational quasi-degenerate perturbation (MCQDPT2) theory. The spectral origins of tautomers were computed as the energy difference between the optimized ground and singlet excited state of the corresponding tautomer. The electronic charge distributions, dipole moments, molecular electrostatic potentials of tautomers in the ground and excited states are also discussed. Conference on Current Trends in Computational Chemistry 2003 165
Figure. Geometries of guanine tautomers in the ground and lowest singlet ππ* excited state at the MCSCF(10,9)/6-31G(NH2*) basis set. 166 Conference on Current Trends in Computational Chemistry 2003
Computational Study of a Fluorescent Photoinduced Electron Transfer (PET) Sensor for Cations
S. A. de Silva, M. L. Kasner, M. A. Whitener and S. L. Pathirana
Department of Chemistry & Biochemistry, Montclair State University, Upper Montclair, NJ 07043
The design of photoresponsive molecules that can function as sensors for various cations is a rapidly developing frontier of supramolecular chemistry. Some of these sensors utilize photoinduced electron transfer (PET) between a covalently linked chromophore and a receptor to generate or quench fluorescence as a means of translating a molecular recognition event into an optical signal.1 These PET sensors have evolved from simple first generation systems2 that utilize one PET process to indicate a single cation binding event, to third generation systems3 that utilize three PET processes to indicate three cation binding events. Although fluorescent PET sensors have been investigated for nearly two decades, only a few sensors and sensor-cation complexes have been isolated and characterized by X-ray crystallography. Computational studies of fluorescent PET sensors are even more rare, and to the best of our knowledge, the first computational study of pyrazoline PET sensors was reported recently.4 In this poster, we present computational studies of a fluorescent PET sensor based on a anthracene chromophore (L) and its Zn2+ complex (L-Zn).5
N N ZnCl2
N N ZnCl2 N N
L L-Zn Non-Fluorescent Fluorescent
The sensor L was designed to function as an “off-on” fluorescent switch for post transition metal ions such as Zn2+, and as a fluorescent “off-on-off” switch for protons.4 Intensity of fluorescence of L increases with the chelation of Zn2+ leading to the formation of a 1:1 complex, L-Zn. The fluorescence of L is quenched due to the thermodynamically favored PET between the tertiary aliphatic nitrogen (tertiary amine) and the excited chromophore (*Anth). The driving force for these electron transfer processes (∆Get) can be expressed by a modified Weller equation as follows: ∆Get = - Es - Ered.chrom + Eox.receptor where Es, Ered.chrom and Eox.receptor are the singlet energy, reduction potential of the chromophore and the oxidation potential of the receptor, respectively. An increase in the oxidation potential of 2+ the receptor (caused by the binding of Zn ) will increase the ∆Get, which will decrease and eventually quench the PET process. This in turn will allow the excited chromophore to relax by Conference on Current Trends in Computational Chemistry 2003 167
fluorescence, and will result in the generation of a fluorescent signal due to the binding of Zn2+. The process described above can be shown schematically by a simplified MO diagram as follows:
E
LUMO LUMO
ET
HOMO Fluorescence ET HOMO X HOMO
HOMO
Cation free Excited Cation bound Excited receptor chromophore receptor chromophore (Non-Fluorescent) (Fluorescent) Figure 1. Energy levels L and L-Zn, that shows “Off-On” fluorescence switching with Zn2+ binding.
Our computational studies of this sensor are focused on modeling the energetics of the fluorescent “off-on” switching of L with Zn2+ binding, and a comparison of the X-ray crystal structures of L and L-Zn with optimized geometries. The Hartree-Fock and Density Functional Methods: HF/6-31G(d,p)//HF/6-31G(d,p), B3LYP/6-31G(d,p)//B3LYP/6-31G(d,p), were used to determine the optimized geometries and molecular orbital energies.
References
1. de Silva, A. P.; Fox, D. B.; Moody, T. S.; Weir, S. M. Trends in Biotech. 2001, 19, 29. 2. de Silva, A. P.; de Silva, S. A. Chem. Commun. 1986, 1709. 3. de Silva, S. A.; Amorelli, B.; Isidor , D. C.; Loo, K. C.; Crooker, K. E.; Pena, Y. E. Chem. Commun. 2002, 1360. 4. Fahrni, C. J.; Yang, L.; VanDerveer, D. G. J. Am. Chem. Soc. 2003, 125, 3799. 5. de Silva, S. A.; Zavaleta, A.; Baron, D. E.; Allam, O.; Isidor, E. V.; Kashimura, N.; Percarpio, J. M. Tetrahedron Lett. 1997, 38, 2237.
Acknowledgements Financial support was provided by Montclair State University and the Petroleum Research Fund of the American Chemical Society.
168 Conference on Current Trends in Computational Chemistry 2003
Theoretical Study of Oxygen Absorption on Single-Walled Carbon Nanotubes
Tomekia M. Simeon, Chi-Cobi Speaks, Glake Hill, and Jerzy Leszczynski
The Computational Center for Molecular Structure and Interactions Department of Chemistry, Jackson State University 1400 J. R. Lynch Street, Jackson, MS 39217
Classifying new mechanical, chemical, and electrical properties of carbon nanotubes poses an interesting challenge. Recent studies show that electronic properties of carbon nanotubes can be appreciably altered by the presence of absorbed molecules [1]. This has important ramifications for device applications because individual nanotube sensors can be used to detect different type of molecules [1]. Oxygen, in particular, has been found to influence electronic properties of SWNTs [1, 7]. This project will discuss the unique characteristics of oxygen absorption on SWNTs and present mechanisms for binding and interactions. Calculations of these properties are interesting because of the possible technological applications in the foreseeable future.
References: [1] D. C. Sorescu, K. D. Jordan, P. Avouris, J. Phys. Chem. B, 105, 11227, 2001. [2] O. Gulseren, T. Yildirim, S. Ciraci, Physical Review Letters, Vol. 87, 11, 2002. [3] S. Jhi, S. G. Louie, M. L. Cohen, Physical Review Letters, Vol. 85, 8, 2000. [4] J. Kong, N. R. Franklin, C. Zhou, Science, 285, 622, 2000. [5] P. G. Collins, K. Bradley, M. Ishigami, A. Zetta, Science, 287, 1801, 2000. [6] J. Kong, J. Cao, H. Dai, Applied Physics Letters, 80, 73, 2002. [7] K. Wong, G. Tikhonov, V. Kresin, Physical Review B, 66, 125401, 2002.
Conference on Current Trends in Computational Chemistry 2003 169
The Density Functional Theory Study of the Superacid Acidity Scale
Vitaly Solkan1 and Jerzy Leszczynski2
1Zelinsky Institute of Organic Chemistry, RAS, 119991 Moscow, Leninskii pr.47, Russian Federation 2Computational Center for Molecular Structure and Interactions, Department of Chemistry, Jackson State University, 1400 J. R. Lynch Street, Jackson, MS 39217, USA
The change in the superacidic’s acidity scale when passing from the gas phase to solution were studied. Gas-phase energies for FSO3H, ClSO3H, CF3SO3H, H2SO4, HF, and HClO4 were calculated by the density functional theory ( at B3LYP level) with the 6-31++G(d,p) basis set. In order to compute deprotonation enthalpies, zero-point corrections have been obtained from unscaled frequency calculations on minima at the B3LYP level. The entropic contributions to the gas-phase free energies have also been obtained at the B3LYP level. The solute-solvent interaction term has been broken down into the electrostatic and nonelectrostatic contributions. We have obtained the electrostatic term using the polarizable continuum model (PCM), where the solute is inscribed in a cavity surrounded by a continous dielectric that represents the superacid. Calculated deprotonation free energies are in good agreement with the experimental values. We have also carried out MP2 calculations with larger basis sets, but they do not introduce any significant improvement over the level of calculation we have chosen. The main contribution to the change in the acidity scale in liquid superacids is the electrostatic component of the solvation energies of the basic forms, and the charge delocalization. The electrostatic stabilization is lower when the size of acid increases, favoring the displacement of the equilibrium to the left. The advantages of the hybrid solvation model, which includes both the explicit contribution of the first solvate shell and the continuum model PCM, is shown for calculation of deprotonation free energy for fluorosulfonic, chlorosulfonic, and trifluoromethane- sulfonic acids. Based on the results of the calculations, a quantitative estimation of the catalytic activity of protons is given with an explicit consideration of solvation with acidic molecules.
170 Conference on Current Trends in Computational Chemistry 2003
The Nature of Superacid Electrophilic Species in HF/SbF5 and FSO3H/SbF5: The Ab Initio and Density Functional Theory Studies Taking Into Account the Solvation Effects
Vitaly Solkan1 and Jerzy Leszczynski2
1Zelinsky Institute of Organic Chemistry, RAS, 119991 Moscow, Leninskii pr.47, Russian Federation 2Computational Center for Molecular Structure and Interactions, Department of Chemistry, Jackson State University, 1400 J. R. Lynch Street, Jackson, MS 39217, USA
A liquid superacid media allow the formation and stabilization of energetic species like carbocations and carbenium ions, and activation of molecules of low reactivity, such as alkanes under mild conditions. A special category of superacids is generally obtained by the combination of a strong Bronsted acid with a strong Lewis acid, such as HF/SbF5 and FSO3H/SbF5. The former system is believed to possess the strongest acidity in the condensed phase. Spectroscopic studies of HF/SbF5 and FSO3H/SbF5 with different molar ratios indicate that species such as + + + - - H2F , H3F2 , H2SO3F , and more solvated cations exist these media, as well as SbF6 , Sb2F11 , - - [SbF5 SO3F] , [Sb2F10 SO3F] anions. The ab initio at MP2/6-31++G(d,p) +RESP(Sb) level and density functional theory at B3LYP/6-31++G(d,p) +RESP(Sb) level studies of HF/SbF5 and + - FSO3H/SbF5 superacid systems were carried out. The geometries of possible species H SbF6 , + - + - + - + - + - + - + H2F SbF6 , H3F2 SbF6 , H Sb2F11 , H2F Sb2F11 , H3F2 Sb2F11 , H [SbF5 SO3F] , H [Sb2F10 - SO3F] were calculated and correspond with available experimental data. The contribution of superacidic environment to a electrostatic solvation free energy for these clusters are included at MP2/6-31++G(d,p) and B3LYP/6-31++G(d,p) levels by means of a polarizable continuum model (PCM). Calculations of the acid strength of the electrophilic species were also performed and indicated that the acid strength increases with the solvation degree. In accordance with the experiment, larger anions provide more acidic species than do smaller ones. The more solvated ions afford a more disorganized species upon deprotonation. In FSO3H/SbF5 solutions by coordination with fluorosulfonate SbF5 increases the proton acidity of the medium which behaves then only as an acid in the Bronsted sense. When the FSO3H/SbF5 system reacts only as - a proton acid, the negative charge is distributed over the complex anion [Sb2F10 SO3F] . Based on the results of the calculations, a quantitative estimation of the catalytic activity of protons is given with an explicit consideration of solvation .
Conference on Current Trends in Computational Chemistry 2003 171
A Theoretical Study of Carbon Nanotube Interactions
Chi-Cobi Speaks, Tomekia Simeon, Glake Hill, Jerzy Leszczynski
Computational Center for Molecular Structure and Interactions Department of Chemistry, Jackson State University 1400 J. R. Lynch Street, Jackson, MS 39217
The discovery of carbon nanotubes, long, thin cylinders presents promising applications in the new possible carbon science. These large macromolecules are unique for their size, shape, and remarkable physical properties. Nanotubes have a very broad range of electronic, thermal, and structural properties that change depending on the different kinds of nanotubes. For this project, we studied the interactions of single-walled nanotubes (SWNTs) at different bond distances and angles with various dimers. A detailed study of these interactions were performed at the Density Functional Theory level employing B3LYP functional. 172 Conference on Current Trends in Computational Chemistry 2003
The Effects of Metal Cations on Nucleic Acid Bases
J. Sponer,1,2,3 J. E. Sponer,1 J.V. Burda,4 B. Lippert,5 and J. Leszczynski3
1Institute of Biophysics, Academy of Sciences of the Czech Republic and National Center for Biomolecular Research, Kralovopolska 135, 612 65 Brno, Czech Republic. 2 J. Heyrovský Institute of Physical Chemistry, Academy of Sciences of the Czech Republic, Dolejškova 3, 182 23 Prague, Czech Republic 3 Department of Chemistry, Computational Center for Molecular Structure and Interactions, Jackson State University, Jackson, MS 39217 4Department of Chemical Physics and Optics, Faculty of Mathematics and Physics, Charles University, Ke Karlovu 3, 121 16 Prague 2, Czech Republic 5 Department of Chemistry, University of Dortmund, 44221 Dortmund, Germany
Interactions of metal cations with DNA bases can significantly influence the electronic structure of nucleobases. We have carried out a systematic computational analysis of these effects complemented by a careful evaluation of available experimental data. Our results show that the intrinsic gas phase trends, as revealed by modern quantum chemical calculations, are highly relevant to the situation encountered in the solution and solid state experiments. Understanding of the correlation between intrinsic gas phase trends and solution studies is of primary importance for a correct interpretation of both computed and experimental results with respect to the role of metal cations in nucleic acids. Selected results will be reviewed.
Selected recent studies. N. Gresh, J. E. Sponer, N. Spackova, J. Leszczynski, J. Sponer: Theoretical study of binding of hydrated cations Zn(II) and Mg(II) to guanosine 5' monophosphate. Towards polarizable molecular mechanics for DNA and RNA. J. Phys. Chem. 107, 2003, 8669.
J.V. Burda, J. Sponer, J. Hrabakova, M. Zeizinger and J. Leszczynski The Influence of N7 Guanine Modifications on the Strength of Watson-Crick Base Pairing and Guanine N1 acidity: comparison of gas phase and condensed phase trends. J. Phys. Chem. B 107, 2003, 5349. L. Rulisek and J. Sponer: Outer-shell and inner-shell coordination of phosphate group to hydrated metal ions (Mg2+, Cu2+, Zn2+, Cd2+) in presence and absence of nucleobase. The role of non-electrostatic effects. Journal of Physical Chemistry B 106, 2003, 1913. K.S. Schmidt, J. Reedijk, K. Weisz, E.M.B. Janke, J.E. Sponer, J. Sponer, B. Lippert: Loss of Hoogsteen pairing ability upon N1 adenine platinum binding. Inorg. Chem., 2002, 2855. J. E. Sponer, J. Leszczynski, F. Glahe, B. Lippert, J. Sponer: Protonation of platinated adenine nucleobases. Gas phase vs condensed phase picture. Inorg. Chem. 40, 2001, 3269. J.V. Burda, J. Sponer, J. Leszczynski: The influence of square planar platinum complexes on DNA base pairing. An ab initio DFT study. Phys. Chem. Chem. Phys. 3, 2001, 4404. J. Sponer, J. E. Sponer, L. Gorb, J. Leszczynski, B. Lippert: Metal-stabilized rare tautomers and mispairs of DNA bases: N6-metalated adenine and N4-metalated cytosine. Theoretical and experimental views. J. Phys. Chem. A 103, 1999, 11406. J.V.Burda, J. Sponer, J. Leszczynski: The interactions of square platinum(II) complexes with guanine and adenine: a quantum-chemical ab initio study of metalated tautomeric forms. J. Biol. Inorganic Chem. 5, 2000, 178. J. Sponer, J.E. Sponer, J. Leszczynski: Cation – pi and amino-acceptor interactions between hydrated metal cations and DNA bases. Quantum-chemical view. J. Biomol. Struct. Dyn. 17, 2000, 1087. Conference on Current Trends in Computational Chemistry 2003 173
Interaction of Model α-Helical Peptides with Lipid Bilayers
E. Štefáneková, J. Urban, P. Mach and T. Hianik
Department of Biophysics and Chemical Physics, Comenius University, Mlynská dolina F1, 842 48 Bratislava, Slovakia
Lipid-protein interactions are of fundamental importance for understanding both the structural integrity and function of biological membranes. To overcome the problem of the complicated structure of integral proteins and their isolation and purification, the synthesized peptide models of specific regions of natural membrane proteins could be used in biophysical studies of the mechanisms of protein-lipid interactions. Among others the α-helical peptide acetyl-Lys2-Gly-Leu24-Lys2-Ala- amide (P24), has been successfully utilized as a model of the hydrophobic transmembrane α-helical segments of integral proteins. This peptide contains a long sequence of hydrophobic leucine residues capped at both the N- and C-termini with two positively charged, relatively polar lysine residues. The central polyleucine region of these peptides was designed to form a maximally stable α-helix which will partition strongly into the hydrophobic environment of the lipid core, while the dilysine caps were designed to anchor the ends of these peptides to the polar surface of the BLM and to inhibit the lateral aggregation of these peptides. This contribution reports on the study of the process of the interaction of P24 peptide with fully hydrated dipalmitoylphosphatidylcholine (DPPC) and dimyristoylphosphatidylcholine (DMPC) bilayers using the molecular dynamics simulations on this molecular system. Both systems were studied under constant pressure with boundary conditions at various temperatures using the GROMACS package [1]. The results obtained from the simulations are discussed and compared with the experimental results obtained from the study of the similar L24 peptide (acetyl-Lys2-Leu24-Lys2-amide). This work was supported by NATO Collaborative linkage grant (LST.CLG.978567), INTAS (Project 01-0224), by the Slovak Grant Agency (Project No. 1/8310/01 to T.H.) a grant from the Protein Engineering Network of Centers of Excellence (to R.N.M.)
[1] E. Lindahl, B. Hess and D. van der Spoel GROMACS 3.0 A package for molecular simulation and trajectory analysis, J. Mol. Mod. 7 (2001) 306, Internet: http://www.gromacs.org
174 Conference on Current Trends in Computational Chemistry 2003
Effects of the Binding of Molybdenum on the Electrostatic Charge, Solvation and Solute energy of 4-hydroxyl-2, 6-dicarboxylate- pyridine Anion
AlTrev Sykes and Suely M. Black
Center for Materials Research, Norfolk State University
This study reports the binding energy, and structural and electrostatic potential charge changes upon the binding of 4-hydroxyl-2, 6-dicarboxylate-pyridine anion and the Molybdenum MoO3, forming coordinate compound. The optimal constrained and unconstrained geometries of the ligand, MoO3, and complex were obtained using density functional theory, with B3LYP exchange-correlation functional, and LACVP* and 6-31G** basis sets. The results were obtained for water solvated systems, where the solvation effects are taking into account via the dielectric continuum model. Unconstrained and C2 constrained geometry optimizations yield slightly different structures. The binding energies for the two structures were calculated from the individual energies. The structural parameters for each species, and the re-distribution of electrostatic potential charges due to binding were also studied. This investigation was conducted using Jaguar 4.0 computational chemistry software package. The theoretically calculated bond lengths were compared to that of the experimental values for the crystal reported by Hicks and colleagues, and are in good agreement. The unconstrained geometry optimization for the complex yields a symmetrically distorted octahedral geometry. It was found that the binding energy for the Molybdenum oxo complexed system was 0.227 hartrees and that there was a -0.67 a.u. net charge transfer from the ligand to the MoO3 upon formation of the complex. A greater change in the partial charge of N (+0.60 a.u.) than in the O (+0.17 a.u.) upon formation of the complex reflects the steric hindrance to the formation of a bond between the oxygen atoms in the carboxylate groups and the Molybdenum atom. This may also explain the formation of the distorted octahedral geometry. Conference on Current Trends in Computational Chemistry 2003 175
+ Molecular Structures and Nature of Interactions in OCH (Ne)n (n=1-17) Complexes
Jaroslaw J. Szymczak,a,b Szczepan Roszak,a,b and Jerzy Leszczynskib
aInstitiute of Physical and Theoretical Chemistry, Wroclaw University of Technology, Wyb. Wyspianskiego 27, 50-370 Wroclaw, Poland bThe Computational Centre for Molecular Structure and Interactions, Department of Chemistry, Jackson State University, Jackson, MS 39217 USA
The results of ab-initio studies of structural and thermodynamical aspects of the + + OCH (Ne)n + Ne = OCH (Ne)n+1 (n=0-16) gas-phase clustering reaction are reported. The formation of clusters follows the pattern of the consecutive filling of four distinct shells (three five-member rings perpendicular to the OCH+Ne core cation). Detailed analysis of the interactions in consecutive complexes is presented.
+ Figure 1. The OCH (Ne)17 cluster. 176 Conference on Current Trends in Computational Chemistry 2003
The Computational Studies of the Intermolecular Interactions in Donor-Acceptor XL3NH3 (X=B, Al, Ga ; L=H, Cl) Complexes
Jaroslaw J. Szymczak,a,b Szczepan Roszak,a,b and Jerzy Leszczynskib
aInstitiute of Physical and Theoretical Chemistry, Wroclaw University of Technology, Wyb. Wyspianskiego 27, 50-370 Wroclaw, Poland bThe Computational Centre for Molecular Structure and Interactions, Department of Chemistry, Jackson State University, Jackson, MS 39217 USA
The systematic studies of the intermolecular interaction energies are presented for donor- acceptor complexes BH3NH3, BCl3NH3, AlH3NH3, AlCl3NH3, GaH3NH3, and GaCl3NH3. The total interaction energy calculated at the MP2 level EMP2 = EAB - EA- EB has been decomposed (2) EMP2 = EHF + MP into Hartree-Fock and correlation components. The applied SCF energy decomposition was performed within the variational-perturbational scheme corrected for the basis set superposition error.
X=B . X=Al . X=Ga . 10
5
0 1.2 1.4 1.6 1.8 2 2.2 2.4 2.6 2.8 3 3.2 3.4 3.6 3.8 4 4.2 4.4 4.6 4.8 5
-5 ]
ol -10
-15
-20
-25 er Interaction Energy [kcal/m -30 im D
-35
-40
-45
-50 X-N distance [Angst.] Figure 1. Energy curves for X Cl3NH3,( X=B, Al, and Ga) for methods with and without correlation. Conference on Current Trends in Computational Chemistry 2003 177
The Structure and Thermodynamics of Solvated Monatomic Ions Using a Hybrid Model of Solvation
Gregory J. Tawa1, Igor A. Topol2, Stanley K. Burt2, Richard. A. Caldwell3, A. Rashin4
1Wyeth Research, CN80000, Princeton NJ 08543 2Advanced Biomedical Computing Center, SAIC Frederick, NCI Frederick Cancer Research And Development Center, P.O. Box B, Frederick Maryland 21702-1201 3Department of Chemistry, University of Texas at Dallas, P.O. Box 830688, Richardson TX 75083-0688 4BioChemComp, Inc., 543 Sagamore Avenue, Teaneck, New Jersey 07666
An understanding of some key biological processes such as thermodynamics of enzyme- ligand binding or the selectivity of ion-channels is ultimately dependent on an understanding of ion hydration. Therefore, a model for calculating the hydration free energy of ions in aqueous solvent is presented. The model is used to first calculate the proton hydration free energy in an + effort to resolve uncertainty concerning its exact value. In the model we define ∆Ghyd(H ) as the free energy change associated with the following process:
+ + ∆G{H (gas) + [H2O]n(aq) → H [H2O]n(aq)},
where the solvent is represented by a neutral n-water cluster embedded in a dielectric continuum and the solvated proton is represented by a protonated n-water cluster also in the continuum. All solvated species are treated as quantum mechanical solutes (B3LYP, MP2, MP4, CCSD(T)) coupled to a dielectric continuum using a self-consistent reaction field (SCRF) cycle. An + investigation of the behavior of ∆Ghyd(H ) as the number of explicit waters of hydration is increased reveals convergence by n = 4. The converged value of –262.23 kcal/mol. This result strongly suggests that the proton hydration free energy is at the far lower end of the range of values obtained from the literature. The methodology is then used to calculate the hydration free energies of other monatomic ions relative to that of the proton. These include cationic forms of the alkali earth elements Li, Na, and K, and anionic forms of the halogens F, Cl, and Br. The relative in hydration free energy in this case is defined as
± ± + ± + ∆[∆Ghyd(Z )] = G(Z [H2O]n(aq)) – G(H [H2O]n(aq)) – G(Z (gas))-G(H (gas))
where again the solvated ions are represented by ion-water clusters coupled to a dielectric ± continuum using an SCRF cycle. An investigation of the behavior of ∆[∆Ghyd(Z )] as the number of explicit waters of hydration is increased reveals convergence by n = 4. This convergence indicates that the free energy change for addition of water to a solvated proton-water complex is the same as the free energy change associated with the addition of water to a solvated Z± water complex. This is true as long as there are four explicitly solvating waters associated with the ion. This convergence is independent of the type of monatomic ion studied and it occurs before the first hydration shell of the ions (typically ≥ 6) is satisfied. Structural analysis of the ion-water clusters reveals that waters within the cluster are more likely to form hydrogen bonds with themselves when clustering around anions, than when clustering around cations. This suggests that for small ion-water clusters anions are more likely to be externally solvated than cations. 178 Conference on Current Trends in Computational Chemistry 2003
Non-Linear Optical Properties of Novel Fluorenyl Derivatives: Ab Initio Quantum Chemical Calculations
Kanchana S. Thanthiriwattea,b and K.M. Nalin de Silvab
a Department of Chemistry, Mississippi State University, Mississippi State, MS 39762 b Department of Chemistry, University of Colombo, Sri Lanka.
The field of the non-linear optics (NLO) has developed dramatically throughout the last two decades. The NLO properties of molecules have become a focus of current research, in view of their potential applications in various photonic technologies, including all optical switching and data processing. We report ab initio studies of the first static hyperpolarizabilities (β) of fluorenyl derivatives, in which electron donating (D) and electron accepting (A) groups were introduced on either side of the fluorenyl ring system. For designing systems with high β, intra molecular charge transfer (CT) is a key between donor and acceptor and leads to a very large value for the hyperpolarizability (β). Geometries of all molecules were optimized at the Hartree–Fock level in a series of steps, first with the STO-3G minimal basis set, then with the 3-21G split valence basis set and finally with the 6-31G basis set. The first static hyperpolarizabilities of these molecules were calculated using Hartree–Fock and the 6-31G** basis set. The calculated hyperpolarizabilities of these molecules were compared with biphenyl derivatives and other available data in the literature. To understand this phenomenon in the context of a molecular orbital picture, we examined the HOMO and LUMO of all molecules, and it clearly shows the inverse relationship with HOMO–LUMO energy gap. The study reveals that the fluorenyl derivatives have large values, hence in general may have potential applications in the development of non-linear optical materials.
Conference on Current Trends in Computational Chemistry 2003 179
Debye Temperature of Ionic Solids Revisited
S. B. Tripathi
Department of RF Planning & Optimization Telephone Systems International, Inc. 400 Kelby Street - 16th Floor Fort Lee, NJ 07024
The Debye temperature of ionic solids is reexamined by a new formalism. With the introduction of an alternative approach for the evaluation of potential parameters, a new method is developed systematically which derives excellent results. Thus, it establishes the alternative method, and hence whenever the compressibility data are not available, molecular data known with a high degree of accuracy, can be used to evaluate the parameters of the interaction potentials. Various attempts [1-10] have been made earlier to understand the nature of ion-ion interaction and hence to decide the form of interaction potential of a diatomic ionic crystal. Once the reliable form of interaction potential is known, it is possible to compute various macroscopic properties of the crystals. It becomes of more practical importance where the direct measurements of these properties have not yet been carried out. The anharmonic interactions in solids have become the subject of a number of recent investigations [11-14], as they play very important role in explaining the various thermodynamic properties. Debye temperature is one of such property. The calculation of Debye temperature qD) of solids is a problem which has a long history and has generated an enormous amount of literature [15-18]. The evaluation of adjustable parameters of the interaction potential is generally done with the help of compressibility data. Another method for the evaluation of these parameters has been suggested by Kachhawa and Saxena [19], which makes use of only molecular data. This method is better than the first one because the compressibility values required for the calculation of Debye temperature in the later one is not always available and also their values at 0 0 K involve a large amount of uncertainty. Moreover, the experimental values of molecular constants employed for the calculation of Debye temperature in the Kachhawa and Saxena [19] method are known with high degree of accuracy from spectroscopic measurements. The compressibility method requires the knowledge of crystal structure, on the other hand, this method is completely independent of crystal structure. In the present communication, attention has been focused on the alkali halogenides and an attempt is made to evaluate their Debye temperature from the knowledge of their molecular constants by using more realistic Woodcock potential. We have made entirely new calculations for the alkali halide crystals. The result of present calculations are compared with the experimental Debye temperature values and the universality of the alternative procedure has also been tested.
References [1] K P Thakur and J D Pandey, J. Inorg. Nucl. Chem., 37, 645 (1975). [2] K P Thakur, J. Inorg. Nucl. Chem., 36, 2171 (1974). [3] K S Krishnan and S K Roy, Proc. Roy. Soc., London, 207-A, 447 (1951). [4] S P Srivastava, M N Sharma and M P Madan, J. Phys. Soc., Japan, 25, 212 (1968). [5] S S Mitra and S K Joshi, Physica, 26, 284 (1960). [6] K P Thakur, Indian J. Pure & Appl. Phys., 11, 549 (1973). [7] K P Thakur and J D Pandey, J. Chem. Phys., 71, 850 (1974). [8] K P Thakur, Indian J. Chem., 12, 376 (1974). 180 Conference on Current Trends in Computational Chemistry 2003
[9] L L Kazmersk, Nuovo Cemento, 20, 2013 (1983). [10] J D Pandey, A K Singh and S B Tripathi, J. International Academy of Phys. Sci., 2, 87(1998). [11] H Siethoff and K Ahlborn, Phys. Status Solidi (b), 179,190 (1995). [12] V Kumar, G M Prasad and D Chandra, Phys. Status Solidi (b), K45,186 (1994). [13] T S Moss, Phys. Status Solidi (b), 415,131 (1985). [14] R R Reddy and Y Nazeer Ahammed, Infrared Phys. Technol, 36, 825 (1995). [15] M Blackmann, Handbuch der Physik VII/I, Springer Verlag, Berlin, 325 (1955). [16] L Dass and S C Saxena, J. Chem. Phys., 43, 1747 (1965). [17] J D Pandey and S P Gupta, Indian J. Chem., 7, 331 (1969). [18] J D Pandey, Z. Phys. Chemie, Leipzig, 243, 221 (1970). [19] C M Kachhava and S C Saxena, Phil. Mag., 8, 1429 (1963). [20] L V Woodcock, J. Chem. Soc., Faraday Trans. 2, 1405 (1974). [21] M P Toshi, Solid State Phys., 61,1 (1964). [22] Y P Varshani and R C Shukla, J. Mol. Spectro, 16, 63 (1965). [23] K S Krishnan and S K Roy, Proc. Roy. Soc., London, 207-A, 447 (1951). [24] R P Jain, S Saxena and J D Pandey, J. Inorg. Nucl. Chem., 36, 203 (1974). [25] M Börn and K Huang, " Dynamical Theory of Crystal Lattice", Clanendon Press, Oxford,(1956). [26] R P Pandey and J D Pandey, Indian J. Chem., 19, 366 (1980). Conference on Current Trends in Computational Chemistry 2003 181
Quantum-Chemical and Experimental Study of Cyano-(2-3H- quinazolin-4-ylidene)-acetic Acid Ethyl Ester
M. Tulyasheva 1, B. F. Rasulev 1,2, N.D. Abdullaev 1, K.K. Turgunov1, Kh. M. Shakhidoyatov 1 and J. Leszczynski 2
1Institute of Chemistry of Plant substances AS RUz, Kh.Abdullaev str., 77, Tashkent, 700170, Uzbekistan 2Computational Center for Molecular Structure and Interactions, Jackson State University, Jackson, MS, 39217
A new compound - Cyano-(2-3H-quinazolin-4-ylidene)-acetic acid ethyl ester have been synthesized and studied by experimental and theoretical methods. We present the NMR and X-Ray study combined with theoretical Density Functional Theory (DFT) calculations of this quinazoline-4 derivative. The relative stabilities of cyano-(2- 3H-quinazolin-4-ylidene)acetic acid ethyl ester are calculated and compared with experimental data. DFT B3LYP/6-311G(d,p) calculations are used to investigate the structure and stability tautomeric forms of the quinazoline-4 derivative. B3LYP/6-311G(d,p) have shown the tautomer B to be the most stable, in agreement with the experimental results. Most stable conformation B show presence an internal C=O⋅⋅⋅HN hydrogen bond between side chain and main skeleton. The 1H-NMR spectrums also show that assignment of the tautomeric form B as the dominant one at room temperature.
Figure 1. X-Ray structure and calculated by DFT tautomer B. 182 Conference on Current Trends in Computational Chemistry 2003
Object Oriented Approaches for Parallelizing Electronic and Structure Calculations of Nano & Biomaterials†
Ravi K. Vadapalli1, John W. Mintmire1 and Brett I. Dunlap2
1Department of Physics, Oklahoma State University, Stillwater OK 74078 2Code 6189, US Naval Research Laboratory, Washington D.C.
Simulation of single and multi-wall carbon nanotubes, nanowires and biopolymers with helical symmetry represents a computationally complex and challenging set of applications. The availability of high performance computing paradigms and object oriented programming (OOP) techniques offer promising new avenues for efficient simulation. Our work is focused on developing a scalable and portable software suite of programs in support of the National Nanotechnology Initiative (NNI) of the Department of Defense (DoD) and to extend future DoD capabilities to develop miniaturized detectors and intelligent sensors for defense applications. These codes are based on first-principles calculations using local density functional theory (DFT). The software suite will be implemented for NUMA architectures using an object oriented Fortran (OOF) programming approach. The software suite primarily consists of an interoperable set of computational tools for calculating electronic structures of nanowires, biopolymers with helical symmetry-allowing us to treat carbon nanotubes, silicon, and silica nanowires, and a range of biomolecular materials such as DNA, etc., and extremely large finite clusters-allowing us to treat end-caps and junctions of carbon nanotubes, etc. This effort should help develop new avenues to better understand exceptional properties of nanostructures such as strength, electrical resistivity and conductivity, and optical absorption that are significantly different from the same matter at either the molecular or bulk scale. For example, carbon nanotubes with diameters on the nanometer scale exhibit anomalous electronic properties relative to those of larger diameter graphitic fibers. DFT is an extremely successful approach for the simulation of complex materials such as large molecular clusters, carbon nanotubes, biomolecules such as proteins, etc. The efficient computation of two-electron Coulomb repulsion integrals plays a vital role in any Gaussian basis set electronic structure calculation, whether it be DFT or more traditional Hartree-Fock-type methods. Dunlap [1] has pioneered the development of Gaussian-based methods for DFT, and has carried out extensive development subsequently in the use of solid-harmonic Gaussian basis sets for electronic structure simulations. This approach is implemented in his SHGDFT serial code, which is one of the legacy codes used as a basis for our current project. Because helical symmetry can be used to reduce substantially the computational costs of simulating the electronic and structural properties of nanowires and related extended quasi 1-D structures such as chain polymers and extended biological molecules such as DNA, Mintmire [3] developed and tested the helical nanostructures (HENS) code for serial computing environment that explicitly incorporates helical symmetry based on first-principles LDF. Mintmire [3] used Cartesian- Gaussian basis sets in his work. Our goal is to parallelize the serial codes HENS and SHGDFT for high performance parallel architectures by employing cutting edge technologies such as Global Arrays [4], a NUMA model which provides a portable interface for a parallel program to independently,
† This work is supported by the DoD HPCMO CHSSI program through the Naval Research level as part of CHSSI project MBD-05. Any opinions, findings, conclusions, or recommendations expressed in this work are those of the author(s) and do not necessarily reflect the view of the Naval Research Laboratory Conference on Current Trends in Computational Chemistry 2003 183
asynchronously, and efficiently access logical blocks. We will use NWChem [5] as a benchmark and prototype for our work. Figure 1 depicts our approach for the design and implementation of parallel software suite of programs SHGDFT-HENS. In this approach, each layer is self- contained as much as possible, and offers sufficiently exclusive tasks and support utilities. For example, the Application Program Interface (API) for User Interface Layer should support extraction of data from an input file from the web or any other remote location and translate according to a set of predefined syntactical rules in Fortran 90/95. This could be accomplished by providing support through XML or similar languages on the remote side and scripting languages such as awk, perl, and python to transmogrify the input data so as to upload data structures that will be used by lower layers to execute tasks appropriately. Similar procedures could be followed to return output files to the remote side. As noted earlier, the backbone for this design is implementation of methods and behaviors of the members of the Core Data Objects (CDO) and the Integral Object in particular. Communications between processors is exclusively handled by the Node Set Object. Services provided by the Node Set Object include, interprocess communication (IPC), parallel file management, processor, and data management for distributed computing environment, barrier synchronizations and other semantics for efficient implementation of parallel algorithms. While the purpose and functionality of both the Software Tools Layer, and Utility Support Layer is self- explanatory, the Quantum Chemistry Models Layer is the trivial candidate for future expansions. Hence, methods and behaviors of CDOs should be developed and implemented with this in mind. The salient features of the design and implementation of the present software suite include: • Multilayered architecture with each layer to offer exclusive services • To provide comprehensive, but simplified user and Application Program Interface • Efficient memory management • To exploit OOP paradigm for abstraction, encapsulation and data security • OOF for software reusability and backward compatibility as far as possible • Support through scripting languages such as awk, perl, python for input parsing, dependency graph, and data packing/unpacking etc. • Support for C/C++ wrappers wherever necessary • Data structures and templates in Fortran, to emulate C++ STL, and to exploit inheritance, function overloading, and polymorphism • Support for future expansion
184 Conference on Current Trends in Computational Chemistry 2003
User Interface Layer Input Parser/ Web interfacing
S Quant. Chem. Models SHGDFT/HENS/SCF etc. I M U Core Data Objects Atom; Basis; Geometry; Integral L A T MPI, Inter Process Comm., Node Set Object I Parallel File Mgmt. etc. O N Linear Algebra, C/C++ Wrappers, Software Tools Layer Runtime Database
I/O Management, Error Handling, Utility Support Layer Check Sums, Data (UN) packing, Time Stamp, CVS, etc.
Figure 1. Preliminary Design of Multilayered Architecture for Parallelization of SHGDFT- HENS.
References
1. B. I. Dunlap, “Three-center Gaussian-type-orbital integral evaluation using solid spherical harmonics”, Phys Rev A 42, 1127 (1990) 2. B. I. Dunlap, “Direct quantum chemical integral evaluation”, Int J Quant Chem 81, 373 (2001) 3. J. W. Mintmire, “Local-Density Functional electronic structure of helical chain polymers”, Density Functional Theory Approaches to Chemistry, J. K. Labanowski and J. W. Andzelm, Eds. (Springer-Verlag, New York, 1991), 125-137 4. J. Nieplocha, R. J. Harrison, R. J. Littlefield, "The Global Array Programming Model for High Performance Scientific Computing ", SIAM News, August 1995. http://citeseer.nj.nec.com/nieplocha95global.html 5. NWChem Programmers Manual http://www.emsl.pnl.gov/docs/nwchem/nwchem.html Conference on Current Trends in Computational Chemistry 2003 185
Modeling of Electroreduction Reaction of Some Zn 2+ and Mn2+ Heterocomplexes
V.F. Vargalyuk and V.A. Seredyuk
Dnepropetrovsk National University Department of Chemistry, Dnepropetrovsk, Ukraine
2+ - Recently we have performed quantum-chemical calculations of reaction Me aq + 2e → 0 2+ 2+ 2+ Me aq with participation of aquacomplexes Me (H2O)32 for Zn and Mn ions. From the practical point of view it is of interest to calculate structures of complexes where internal coordination sphere includes not only water molecules but also molecules of organic compounds (L) used in galvanotechnics as effective regulators of process of electrosedimentation of metals. In this study instead of big structures Me(H2O)32 we have calculated compact Me (L) (H2O)x structures where х does not exceed 5, which allows us considerably save computer time without damage to revealing the basic tendencies. Calculations have been carried out by computer modeling using GAMESS package of program at unrestricted Hartree-Fock level with 3-21G and TZV basis sets. During calculation in preliminary optimized structure of aquacomplexes with an ion of metal in a degree of oxidation +2 one of water molecules was replaced by organic ligand then the system was optimized. The received coordinates of the optimized systems used as initial approach for calculations of the following stage of electroreduction. Simultaneously energies of transition complexes for geometry with of precursors have been calculated (marked by an asterisk). These structures form at the moment of carrying out of electron when nucleus of atoms had not time to change the position yet. As organic ligands widely used in the industry in sour electrolits zincing additives have been examined: thiourea (Tm), acrylamide (Aa) and acrylnitril (An). Obtained values have been compared with parameters of aquacomplexes (Aq). Fig. 1 shown the relative energies of calculated systems (∆E=E-E0), where E0 is energy of a metal complex in oxidations degree +2. The changes of the central atom charge (Z (Me)) during electroreduction process Me2+ + e-→ Me+* → → Me+ +e-→ Me0* → Me0are shown on Fig. 2. From Fig. 1a one can see that organic ligands significantly stabilize the mixed complex of Zn+ ions, thus, creating favorable conditions for transfer of the second electron. In case of reaction Mn2+ →Mn0 increasing of energy of transition complexes is not as remarkable as for reduction of Zn2+. The general unfavorable energetic of electroreduction process obtained here is the same as in case of aquacomplexes Mn2+ reduction reaction: energy of transition complexes (Mn+)* and (Mn0)* significantly higher, if compare to energy of initial particle Mn2+. 186 Conference on Current Trends in Computational Chemistry 2003
Fig.1 Energy diagrams of electroreduction of heterocomplexes of the Zn (a) and Mn (b).
Fig.2 Diagrams of central atom charge changing of the heterocomplexes of the Zn (a) and Mn(b).
Consideration of the character of influence of the investigated organic substances on the distribution of electronic density to complexes has shown (fig. 2), that in general reactions Zn2+ → Zn0 and Mn2+ → Mn0 the high degree of smoothing of central atom charge takes place during reaction. Taking into account heterogeneity of reaction Me2+ → Me0 the geometry of complex structures has been investigated since it could plays significant role on adsorption of these particles on surfaces of metal. It has been shown, that thiourea connects to the central atom through sulfur, acrylamid – through oxygen atom of carbonyl groups, and acrylonitril – through nitrogen atom of nitril groups. It allows the acrylic compounds to interact with superficial metal atoms of the vinyl group providing easier electron transfer along π-conjugated chain. Conference on Current Trends in Computational Chemistry 2003 187
Self-Assembly Based on the Levulinic Acid-Melamine Lattice
Ramaiyer Venkatraman and Paresh Chandra Ray
Department of Chemistry, Jackson State University, Jackson, MS, USA
Equimolar amounts of levulinic acid and melamine formed a self-assembled unit through hydrogen bonding net work at room temperature. The single crystal structure diffraction studies and FTIR spectroscopy of the system indicated a 1:1 adduct formation between melamine and levulinic acid. Theoretical calculations on this system were carried out using ab-initio method for optimized geometries at HF/3-21G level theory. We also computed the IR frequency for the adduct and compared the result with the FTIR spectral data. A detailed discussion of the experimental and theoretical investigations will be presented.
188 Conference on Current Trends in Computational Chemistry 2003
Weak Forces in Building the Molecular Nanostructures
Chen Wang and Chunli Bai*
Center of Molecular Sciences, Institute of Chemistry, Chinese Academy of Sciences, Beijing 100080, China
The intermolecular weak forces play a critical role in nature, from a wide range of biological phenomena such as DNA helix and protein, to molecular crystal, physisorption and solvation phenomena. According to bottom-up technique the molecules and atoms are organized within an aggregation to function as special devices by utilizing weak forces, including hydrogen bond, electrostatic interaction, ππ stacking interactions, hydrophobic interaction etc.. The realization of these aims depends on the depth of the understanding and control of the weak forces. In the past decades a great deal of efforts have been made in the design and development of novel functional nanostructures, which have been characterized using all kinds of analytical techniques. Some of the recent examples could be found in the 2D moleular crystals and ordered self-assembled structures using scanning tunneling microscopy combined with theoretical simulations. At atomic level the 2D molecular crystals and ordered self-assembled structures have been achieved including molecules of phthalocyanine, porphrin, aromatic molecules and alkane derivatives and binary systems. The theoretical methods ranging from density functional theory to molecular mechanics and molecular dynamics simulations were employed to elucidate the weak force mechanisms in building the self-assembled nanosystems. The competitions of intermolecular interactions of adsorbate-substrate and adsorbate-adsorbate have been discussed. The hydrogen bonding and ππ stacking interaction are especially argued in such physisorption systems. The simulations lead to insightful information on the strength and relationship of functional group interactions.
Conference on Current Trends in Computational Chemistry 2003 189
Modeling Interactions of Fasciculin 2 with AChE: A DFT-based Molecular Dynamics and Quantum Chemical Study
Jing Wang, Jiande Gu, Jerzy Leszczynski*
Computational Center for Molecular Structure and Interactions, Department of Chemistry, Jackson State University, Jackson MS 39217, USA
Acetylcholinesterase (AChE) is an especially efficient serine hydrolase to terminate synaptic transmission at cholinergic synapses through catalyzing the breakdown of the neurotransmitter acetylcholine (ACh). It has an active site gorge. Two separate ligand binding sites has been located for the gorge: the acylation site and the peripheral site. The acylation site lies at the bottom of the gorge. The peripheral site locates at the entrance of the gorge, including residue W279. Peptidic three-finger snake toxin, fasciculin 2 (Fas2), from green mamba venom, is found to be a neurotoxin inhibitor of AChE. The crystal structure of AChE binding with Fas2 reveals that Fas2 caps the entrance of the gorge with excellent complementary. Site-specific mutant studies indicate that W279 on the rim of the active site gorge is a key component of the peripheral site and is essential for high-affinity Fas2 binding. In order to understand the influence of Fas2 on the W279 residue of AChE, the DFT-based molecular dynamics approach and quantum chemistry calculation methods have been employed in the present study to investigate the specific interactions at the binding site of Fas2 and AChE. Two models were selected from the crystal structure (with code 1FSS in the Protein Data Bank), for AChE and for Fas2 bounded AChE, respectively. A Model contains only the selected fragment of AChE, including W279 residue with its neighbouring negative charged residue E278. B Model includes P31 residue of Fas2 and W279-E278 of AChE. (Fig. 1) Applying DFT-based molecular dynamics approaches, the two models are investigated at the simulated temperature of 300K. During the simulation processes, A Model was noticed to poorly retain its initial structure. The peptide chain rotated easily, resulting in the considerable change of the relative orientation for W279 and E278 residues. On the other hand, with the Fas2’s binding, the structure of the AChE moiety appeared to be less flexible in B model. The relative orientation of W279 and E278 remained almost the same as the initial structure while a hydrogen bonding interaction which stabilized the AChE moiety was revealed between the H41 of W279 and the O30 of E278.
P31 W279
E278
A Model (AChE) B Model (Fas2-AChE) Figure 1. Models selected for simulation 190 Conference on Current Trends in Computational Chemistry 2003
P31 W279-E278 Figure 2. Optimization structures of P31 and W279-E278 and their electrostatic potential mapped onto the electron density. The isosurface value is 0.002 with a range for the MEP of 0.00 to 0.20 au and -0.20 to 0.00 au for P31 and W279-E278, respectively.
Based upon the results of the fluctuation trajectories for the two models, quantum chemistry theory is applied to further detail the interactions between the two moieties. P31 and W279- E278 complex have been fully optimized at the B3LYP/6-311G(d,p) level of theory. Vibrational frequency analyses reveal no imaginary frequency and suggest that both species are energy minima on the potential energy surface. For the AChE moiety, the atomic distance between H41 and O3 is predicted to be 1.55 Å, suggesting a strong H-bonding interaction between W279 and E278. The electrostatic potential map (MEP) shows that the electrostatic potential near the binding position of the Fas2 moiety appears to be positive, while the electrostatic potential above the aromatic ring of W279 in the AChE moiety is predicteded to be negative (Fig. 2). This suggests existence of electrostatic interactions between Fas2 and AChE when Fas2 approaches AChE, which might induce other weak polar interactions. Conference on Current Trends in Computational Chemistry 2003 191
1.08 1.54
Figure 3. Structure of complex composed of P31-W279-E278
The complex composed of P31, W279, and E278 has also been optimized at the B3LYP/6- 311G(d,p) level of theory. It is proved to be a local energy minimum on the potential energy surface by the vibrational frequency analysis showing no imaginary frequencies. For this complex, the atomic distance between atom H41 from W279 residue and atom O30 from E278 residue is predicted to be 1.54 Å, suggesting a strong hydrogen bonding between these two neighboring residues. The binding energy between P31 and W279 is evaluated to be 1.6 kcal/mol at the B3LYP/6-311G(d,p) level of theory. The single point energies are estimated at the higher calculation level (MP2/6-311G(d,p)), which predicts the interaction energy to be 5.4 kcal/mol. This small interaction energy suggests that the electrostatic interaction between the two moieties is so weak that it couldn’t be of importance. The rigidity of B model during the MD simulation suggests that the influence of Fas2 moiety on W279 might change its dynamic properties. 192 Conference on Current Trends in Computational Chemistry 2003
Molecular Dynamics Simulations of Crystal-Induced Membranolysis
Andrzej Wierzbicki1, Pranav Dalal2, Jeffry D. Madura2, and Herman S. Cheung3
1Department of Chemistry, University of South Alabama, Mobile, Alabama 36688 2Department of Chemistry and Biochemistry, Duquesne University, Pittsburgh, PA 15282 3Department of Biomedical Engineering, University of Miami, and Geriatric Research, Education, and Clinical Center, V.A. Medical Center, Miami, Florida 33135
Experimental evidence has already shown that calcium-containing crystals bind to several types of cells. This binding often leads to severe pathological conditions, frequently resulting in cell damage. The nature of the crystal-cell interaction is of considerable interest since the understanding of the mechanism of the crystal-cell sticking can be used in developing efficient inhibitors that would diminish the adhesion of calcium-containing crystals to cells. It has been proposed that cell adhesion occurs due to the interactions between calcium-containing crystals and phospholipid membranes. In this study using the methods of CHARMM molecular dynamics, we investigate the molecular mechanism of crystal-induced membranolysis. We show that the interactions between the surface of Calcium Pyrophosphate Dihydrate Crystal (CPPD) and the extracellular layer of the hydrated dimyristoyl phosphatidylcholine (DMPC) phospholipid bilayer may lead to the decoupling of this external layer from the intracellular side of the membrane. We demonstrate that this decoupling is primarily due to the immobilization of the external layer induced by the crystal-bilayer interactions. In turn, a local thinning of the layer on the intracellular side of the membrane occurs which favors water penetration of this layer, leading to membranolysis.
Conference on Current Trends in Computational Chemistry 2003 193
Substitutional and Endohedral Structures Based on Polyhedral Oligomeric Silsesquioxane (POSS) Molecules
Chuanyun Xiao and Frank Hagelberg
Computational Center for Molecular Structures and Interactions Department of Physics, Atmospheric Sciences, and General Science Jackson State University, Jackson, MS 39217
Polyhedral Oligomeric Silsesquioxane (POSS) units of the form Si8O12R8, where R denotes an organic ligand, attract much interest from the sides of both experimental and computational research. This attention is largely motivated by the numerous technological applications of POSS units and species derived from them that have emerged in the recent past. Thus, it has been shown that POSS molecules bond to organic polymers, forming long chains that extend through the polymer and resulting in a nanostructured organic-inorganic hybrid. In these units, the POSS chains act like nanoscale reinforcing fibers, producing extraordinary gains in heat resistance. The use of POSS segments in plastics results in improved physical properties of these materials, specifically in the enhancement of fire retardation and of usage temperatures as well as in increased hardness of the plastics. Further, various research projects have focused on metal-substituted POSS species which have been demonstrated to exhibit catalytic activity [1]. Particularly, titanosilsesquioxanes (Ti-POSS) have turned out to be effective as catalysts for alkene epoxidation [2]. In this contribution, we present results of a recent computational project, involving POSS monomers with various ligands as well as foreign atom impurities. Specifically, geometry optimizations were carried out using a variety of quantum-chemical procedures for Si8O12R8 with R = H (‘Spherosiloxane’) and R = CH3. Very little sensitivity of the Si8O12 core is observed with respect to the exchange of ligands. Further, a diversity of endohedrally doped POSS monomers was investigated in terms of their geometric, electronic, bonding and energetic characteristics. In particular, we present results obtained for POSS in combination with alkali metal atom + impurities, namely the clusters (M@ Si8O12H8) with M = Li, Na as well as with a halogen - impurity, (F@ Si8O12H8) . In response to numerous experiments involving Ti-POSS [3], we analyzed in addition the unit Si7O12H8TiOiPr with iPr = Isopropyl, acknowledging that TiOiPr substituted POSS monomers have been generated in the laboratory and have shown extraordinary catalytic activity.
[1] T. Kudo, M.S. Gordon, J.Phys. Chem. A 105,11276 (2001) [2] S. Krijnen, B. Mojet, H. C. L. Abbenhuis, R. A. van Santen, Phys. Chem. Chem. Phys. 361, 1 (1999). [3] R. Dagani, Chemical &Engineering News 79, 35, 59 (2001).
194 Conference on Current Trends in Computational Chemistry 2003
Quantum Transport in Biosystems
Ilya Yanov and Jerzy Leszczynski
Computational Center for Molecular Structure and Interactions (CCMSI) Department of Chemistry, Jackson State University
The main feature of electronic devices is that they are open systems with respect to electron flow. A theoretical consideration of such devices should be done in terms of statistically mixed states which address the problem to quantum kinetic theory. We present here the results of ab initio non- equilibrium Green’s function study of the electron transport properties of the adenine-thymine DNA base pair. Current-voltage dependence I(V), conductance spectrum G(V), and transmission coefficients are calculated. The influence of the hydrogen bonds on the resistance of base pair is discussed. The obtained results are compared with the data for the benzene-1,4-dithiolate molecule. It is shown that although the overall conductance of DNA base pair is much smaller than in the case of benzene-1,4-dithiolate it exhibits significant conductance at small voltage biases that gives the natural way to construct wires, switches, and diodes using the single DNA molecule. Conference on Current Trends in Computational Chemistry 2003 195
Ab Initio and DFT Study of Intrinsic Band Gap in Semiconductor Oxides and Ti-silicalite
N.U. Zhanpeisov,a J. Leszczynski,b M. Anpoa
aDepartment of Applied Chemistry, Osaka Prefecture University, 1-1 Gakuen-cho, Sakai, Osaka 599-8531, JAPAN bDepartment of Chemistry, Jackson State University Jackson, MS 39217, USA
Here we present the results of an ab initio and DFT calculations mimicking the effect of transition metal ion on electronic and structural properties of Ti-silicalite as well as of Na and N on rutile structures. Main discussion concerns to the excitation shifts into a longer wavelengths region and their relation to the frontier orbitals that are expected to be responsible for the observed band gap changes. The results obtained allow us to give some insight in understanding the target phenomenon as well as point out on the importance of boundary frontier orbitals. These and other related findings are presented. Ab initio molecular orbital and density functional theory calculations were performed using the Gaussian 94 and Gaussian 98 program packages. Precursor silicalite were simulated by 5T cluster model that consist of a central SiO4 tetrahedron sharing its corners with four other SiO4 tetrahedra. This model is further used to obtain the terminal and/or vicinal silanol groups of silicalite via replacing of one or two SiO4 tetrahedra around the central Si atom as well as to mimic the respective Ti-silicalite structures promoted and/or unpromoted by vanadium impurities. The titanium dioxide as a rutile were modeled either by the edge-shared dititanium containing dioctahedra surrounded by 5-fold coordinated titanium oxide or by an extended cluster model. The most optimal active sites to stabilize an impurity Na atom within rutile modified with a Na metal is considered on the latter extended model while the former one is used to mimic effects of transition metal ions as well as to model the N impurities within rutile structures. Geometry optimizations were carried out at the Hartree-Fock as well as DFT levels of theory using the standard 6-31G* and Lanl2dz basis sets. The latter calculations were performed with the use of Becke’s hybrid method with the Lee, Yang and Parr (B3LYP) gradient-corrected correlation functional. Characteristics of different minima were verified by analyzing the Hessian matrices of the energy second derivatives. The vertical excitation energies were estimated using a single excitation CI (configuration interactions) method.
196 Conference on Current Trends in Computational Chemistry 2003
A Density Functional Theory Study of the Oxidation of Methanol to Formaldehyde over Vanadia Supported on Titania
N.U. Zhanpeisov
Department of Applied Chemistry, Osaka Prefecture University, Osaka 599-8531, Japan
The mechanism of the selective oxidation of methanol to formaldehyde over vanadia supported on silica, titania and zirconia suggested recently by Khaliullin and Bell1 have been critically reconsidered within the same level of the theory. It was shown that an improper use of cluster models mimicking an intrinsic support structure may result in the failure to explain the observed experimental findings like those found in the above paper,1 i.e., when considering the activation energies and TOF between those three different supports as well as the next-nearest- neighbor V environment geometry for vanadia supported on silica catalyst. DFT molecular orbital calculations were performed using the Gaussian98 program package. These calculations are based on the use of Becke’s three parameter hybrid method with the Lee, Yang and Parr (B3LYP) gradient-corrected correlation functional. Characteristics of the different minima were verified by analyzing the Hessian matrices of the energy second derivatives. Geometries were optimized using the standard 6-31G* basis set.
1. R.Z. Khaliullin, A.T. Bell. J. Phys. Chem. B. 2002, 106, 7832 Conference on Current Trends in Computational Chemistry 2003 197
IR Bands Originated from the Adsorption of CO on Platinum Supported Zeolites: A Density Functional Theory Study
Nurbosyn U. Zhanpeisov, Hiromasa Ohmukai, Masakasu Anpo
Department of Applied Chemistry, Osaka Prefecture University Osaka 599-8531, Japan
CO adsorption on platinum supported on zeolites and other metal oxides have attracted a great attention of scientists primarily because of the enormous importance of these catalysts in industry. Several catalytic processes such as selective reduction of nitrogen oxides, production of synthetic hydrocarbons and alcohol fuels, etc. are accompanied by partial oxidation of the platinum. Nevertheless, the data concerning CO adsorption on oxidized platinum are rather scarce as compared to the increased number of papers dealing with surface carbonyls on the reduced Pt sites.1 IR spectroscopy has been widely used to characterize the nature of active Pt sites in supported oxides.2,3 Although some contrasting attribution of the observed bands in the C-O stretching vibrations made by different group of scientists,1 it is certain that high-frequency bands characterize higher oxidation states of platinum, while unambiguous determination of the oxidation state of the cation in the complex is still uncertain, and additional experiments along with theoretical investigations are needed. In present paper the results of density functional theory quantum chemical calculations have been presented to characterize the interaction of CO with both structurally and electronically different Pt sites on zeolite surfaces. Main attention is paid to the clarification of the origin of bands that appeared during the adsorption of carbon monoxide on the target catalyst surface pretreated at different temperatures and at different oxidation or reduction atmospheres. Calculations were performed using the Gaussian98 program packages. The precursor zeolites with the Si/Al ratio equal to 1 or 2 as well as purely silicalite structure was simulated by a six-member ring containing cluster models (6T) in addition to 0T, 1T, 3T and 4T cluster models. They have been used then to obtain the respective Pt containing derivatives either via an ion exchange (for oxidized Pt state) or adding metal Pt (for reduced Pt state) to the neutral silicalite. Also, for the latter reduced state we have considered pure Pt metal cluster consisting of 10 atoms in double layer structure. Geometry optimizations were carried out at the DFT level using the standard 6-31G* basis set for the zeolite framework and adsorbate atoms except the Pt sites for which Lanl2dz basis set were applied. DFT calculations were performed with the use of Becke’s three- parameter hybrid method with the Lee, Yang and Parr (B3LYP) gradient- corrected correlation functional.
1. K.I. Hadjiivanov, G.N. Vayssilov. Adv. Catal. 47 (2002) 307. 2. A. Solomennikov, A. Davydov. Kinet. Katal. 25 (1984) 403. 3. A.A. Davydov. IR spectroscopy in chemistry of oxide surfaces. Novosibirsk.: Nauka, 1984. 245 p. 198 Conference on Current Trends in Computational Chemistry 2003
Structure, Chemisorption and Catalytic Activity of Multivalent Metal Oxide Species in High Silica Zeolites
G.M. Zhidomirov
Boreskov Institute of Catalysis, Siberian Branch of the Russian Academy of Sciences, Pr. Akad. Lavrentieva 5, Novosibirsk 630090, Russia
One of the modern trends in the heterogeneous catalysts development is connected with zeolites connecting metal oxide species. Metal cations can be put into zeolites at the synthesis stage as well as via various methods of zeolite modification (like the ion exchange, metal compound implantation, etc.). By now there has been obtained a number of unique catalysts revealed high selectivity in diverse processes. To undestand how these catalysts work it is necessary to determine structure of the stabilized cation species. The latter depends in general on the lattice type, the module of zeolite (Si/Al ratio), the method of cation introduction, the redox treatments.One can consider few basic structural types of the cation localization within the zeolite lattice: lattice positions, cation positions, oxo ions containing extra lattice oxygen, channel and cavity stabilized small neutral oxo clusters. This talk is devoted to the theoretical consideration of the relative stability of various cation spesies in high silica theolites; special attention is paid to ZnZSM5 and FeZSM5 zeolites.Cluster model calculations on the DFT level were performed taking into account the geometrical peculiarities of the zeolite structure. Zn containing ZSM5 zeolites are active in dehydrogenation and aromatization of alkanes. The activity of various zinc oxide species in adsorption and activation of dihydrogen and methane molecules were compared for zinc cations in different cation positions and for oxygen bridged zinc cation pairs in ZSM5 zeolites as well as for small neutral zinc oxide clusters.The heterolytic dissociation of ethane molecule was also studied. Fe containing ZSM5 are active in full decomposition of nitrogen oxide; these systems are also unique catalysts for benzene to phenol oxidation by nitrous oxide.These activities are discussed on the base of DFT cluster model study.
Author is grateful to NWO for financial support of this work. Conference on Current Trends in Computational Chemistry 2003 199
Analysis of Spin-Unrestricted Wave Function by Means of the Expansion of β Orbitals in the Basis of α Orbitals
Igor Zilberberg∗, Sergey F. Ruzankin
Boreskov Institute of Catalysis, Novosibirsk 630090, Rissian Federation
An analysis of spin-unrestricted solutions in terms of α-spin spatial orbital set {ai} is developed. It is shown that spin-unrestricted Hartree-Fock (HF) or Kohn-Sham (KS) determinant is expressed as a linear combination of restricted determinants constructed from the ai orbitals when the β-spin spatial orbital set {bi} is expanded in the basis of occupied and virtual ai orbitals. The mean value of the S2 operator for such restricted determinants is shown to exceed its eigenvalue S(S+1) by integer number M. The latter appears to be equal to the number of 2 unpaired β electrons occupying virtual aj orbitals. Therefore, the value can be considered as an indicator of the configurations with spatially separated α and β electrons. For example, the unrestricted density functional studies of diiron-oxo proteins system consisting of two paramagnetic Fe+3(d5) centers bridged by nominally diamagnetic oxo center reveal the
∗ Corresponding author E-mail: [email protected] 200 Conference on Current Trends in Computational Chemistry 2003
Activation and Binding of Qinghaosu and Its Derivatives
Mark A. Zottola and Jean Karle
504 Eastview Terrace, Abingdon, MD 21009 USA
Qinghaosu-based anti-malarial drugs (1) have shown an effective broad spectrum response to all forms of malaria. However, despite two centuries of use, there is still significant disagreement on the mechanism of of action for Qinghaosu. This work explores both the mechanism of activation and potential mechanism of action for Qinghaosu. Based on an acid- catalyzed mechanism of activation, a tertiary di-oxonium ion is postulated as a key intermediate. Docking studies of this with falcipain, a cysteine protease, have revealed some rather interesting binding modes. THis poster will show the mechanism of activation and some of the proposed binding modes of qinghaosu to falcipain.
1 Conference on Current Trends in Computational Chemistry 2003 201
Investigation of Adenine Tautomer Complexes with Metal Ions: Stability and Molecular Structure
A.Yu. Rubina1, Yu. V. Rubin1, M. K. Shukla2, J. Leszczynski2
1Molecular Biophysics Department, Institute for Low Temperature Physics and Engineering, , Kharkov, 61103, Ukraine, [email protected] 2Computational Center for Molecular Structure and Interaction, Jackson State University, 1400 Lynch Street, Jackson, Mississippi, 39217, USA
Adenine (Ade) is one of the fundamental "bricks" composing the DNA helix . It is known that heavy metal ions have a mutagenic action on DNA. It is interesting to carry out studies of Cu(I), Ag(I), Zn(II) ion complexes with adenine tautomers. It seems interesting, in particular, to calculate the stability, molecular structure and IR spectra for adenine complexes with ions by post-Hatree-Fock ab initio methods. The geometry optimization, total energies and IR spectral calculations were performed for two tautomers of adenine and their complexes with ions of some transition metals (Cu(I), Zn(II), Ag(I)) using Gaussian 98 program. The ground state geometry optimizations and IR frequencies calculations were performed at the MP2/6-31+G(d,p) level of theory. Single point energies were also computed at the MP2/6-311++G(d,p)// MP2/6-31+G(d,p) level for Cu(I) and Zn(II) adenine complexes. Ag(I)-adenine complexes were calculated by using pseudopotential basis sets. The Zn (II)-adenine tautomer complexes have following stability order: N9H,N7Zn > N7H,N9Zn > N9H,N1Zn > N7H,N1Zn > N9H,N3Zn; for Cu(I)-adenine complexes the stability order follows: N9H,N7Cu > N7H,N3Cu > N7H,N9Cu > N9H,N1Cu > N9H,N3Cu > N7HN1Cu. Analysis of molecular structure showed that amino group is perpendicular to the adenine moleculal plane in the case Cu (I), Zn(II), Ag(I) adenine complexes with a metal ion placed at N7. Formation of complexes is accompanied with changes of bond lengths mainly at C6-N1-C2 and C6-N10 fragments, moreover the most alterations occurs in Zn(II)-adenine complexes. The sequence of high frequency vibrations of complexes is changed in their IR spectra owing to the rotation of amino group.