CHEM 47

Introduction to Molecular Modelling

Much of the introduction below is borrowed from :
Dr. Jeff Gosper at Brunel university Jeffrey.Gosper@brunel.ac.uk

History of Molecular Modelling

There was an evolution from:
'flat' structures
to correct representation of stereochemistry
to consideration of molecular topology
to 3D representations of structure, surfaces, and properties
to Intermolecular interactions and accurate calculation (with solvent)

Physical Models

  1. Mostly based on molecular connectivity and stereochemistry
  2. Give some impression of molecular dimensions and shape
  3. Are inaccurate, difficult to manipulate and store
  4. Give very little information of energy and other properties
  5. Are subject to the forces of nature (gravity) and 'marauding fingers'
  6. Very difficult to construct large models

Advantages of Computer Models

  1. Geometrically accurate (only limited by knowledge)
  2. Capable of precise manipulation
  3. Conformational energies may be calculated (QM and MM)
  4. Readily superimposed (compare conformations, properties, volumes, etc)
  5. One molecule may be positioned precisely with respect to a set of spacial requirements (enzyme - substrate, activated complex, etc)
  6. Easily and conveniently stored, edited, and duplicated

What are Computer Models

They are basically mathematically representations/models of molecules and properties such as:

Why Model?

Any field where the 3D shape of molecules and their properties are important can usually benefit from modelling.

Modelling usually involve computation and graphics.

  1. Typical modelling exercises can involve:
  2. Prediction and visualization of shape/properties
  3. Comparison of shape/properties
  4. Examination/prediction of molecular interactions and reactions
  5. Investigation of unstable or excited molecules
  6. Modelling of dynamic systems (vibrations, diffusion, conformational changes, reactions)
Examples of dynamic systems / animations:

Basically modelling aids in understanding and prediction of molecular phenomena.

Applications of Modelling

  1. New drugs
  2. Chemical sensors and probes
  3. Chromatographic agents
  4. Enzymes/Catalysts
  5. Stereoselective synthesis
  6. Pesticides
  7. Microelectronics
  8. Advanced materials

Examples of (Simple) Model Types

  • Stick
  • ball and stick
  • Space filled
  • VDW surfaces
  • Dot surfaces
  • Ribbon Structures

    What does Molecular Modelling / Computer Aided Molecular Design (CAMD) Involve

    CAMD is:
    • the interplay between the modeller (Chemist) and computer
    • the interplay between display/input and calculations
    Involves:
    1. Structure input, refinement and display.
    2. Computation of properties
    3. Peparation of properties for display
    4. Display of molecular properties

    1. Structure Input, refinement and display

    Before any effective modelling can be achieved a 'good' model is required. The CAChe Model Building and Visualizing Applications that we use has Editor desaigned for model input.

    Editor lets you draw and modify molecules. It creates the atom positions, bond type, configurations, valence, and geometry used as input for all the other CAChe applications. You can design any molecule in the Editor and render it in numerous styles. Because you create your molecule sample files in the Editor, it is the logical starting point in the CAChe system if you have not used CAChe before or you need to build a new molecule, Launch the Editor from the CAChe folder on your desktop by double-clicking on the Editor icon.

    Within the Editor there is a menu option termed Beautify that is specifically defined to provide a quick means of refining the structure. Based on default valences it automatically draws in the hydrogensa and establishes reasonable bond lengths and angles. More precise structure refinement is obtained by running a geometry optimization using one of the computational options specified below.

    2. Computation of properties

    There are basically two different methods for the calculation of molecular properties:

    • Quantum Mechanics (QM)
    • Molecular Mechanics (MM)

    CAChe offers computational applications based on classical mechanical and quantum mechanical models. They differ in their approach to the prediction of the energy associated with an arrangement of atoms (i.e. a molecule).MM treats a molecule as series on balls and springs, and energies are determined in accordance with Hooke's law. QM calculations solve approximations to Schroedinger's equation in order to determine electron density, energy, and other molecular properties.

    Molecular Mechanics (MM)

    Mechanics uses classical mechanics to compute the energies of molecules. Mechanics can compute the energy of a structure, optimize a structure by minimizing its energy, and performing a variety of other complex analyses.

    Dynamics also uses a classical mechanical approach. It simulates the normal motion of atoms according to time, temperature, and the calculated forces on the atoms.

    Quantum Mechanics (QM)

    ExtHuckel is a semi-empirical quantum mechanical computation that solves the Schroedinger equation to predict the distribution of electrons in a molecule. This application is intrinsically fast and all CAChe systems can execute it for molecules of reasonable size. Results are added to the molecule file and can be converted to generate molecular orbitals, electron density, or electrostatic potential.

    MOPAC is a comprehensive semi-empirical quantum mechanical computational tool. It can search for an optimized geometry and compute molecular properties such as bond order, partial charges, orbital energies, and vibrational spectra. MOPAC also has interesting options that find and characterize transition states and reaction pathways. There are practical limitations on the molecule size and the complexity of the experiment using this application. The more computing power you have access to, the better off you are for using MOPAC.

    ZINDO is a semi-empirical tool that includes a method for computing spectroscopic properties (UV/visible spectra) and a method of computing molecular geometries. ZINDO is unique in that it includes the transition metals in its parameterized elements. Like MOPAC, the more computing power you have, the better ZINDO performs.

    How well the calculated model reflects the 'real' situation will obviously depend on the accuracy of the computtational procedure under consideration.

    3. Peparation of properties for display

    The CAChe Model Building and Visualizing Applications uses a Tabulator that converts the solved wavefunction from quantum mechanical applications into graphical representations of electron density, electrostatic potential, and molecular orbital surfaces.

    4. Display of molecular properties

    The Visualizer+ won't let you modify a molecule, but it displays complex information resulting from computational applications. You can analyze the results of energy calculations and look at electronic propenies.
    "File differences between the Editorr and the Visualizer applications are based on when you use them. For creating and modifying molecular models, use the Editor. For analyzing computational results, use the Visualizer.
    The similarities between the Editor and the Visualizer applications are in the commands they give you to represent atoms, bonds, color, 3D stereo settings, labels, and so on. After you get acquainted with the Editor, you'll already be familiar with the Visualizer because they are so similar.