Advances in Teaching Physical Chemistry - American Chemical Society

---

Chapter 13

Downloaded by UCSF LIB CKM RSCS MGMT on November 19, 2014 | http://pubs.acs.org Publication Date: December 18, 2007 | doi: 10.1021/bk-2008-0973.ch013

Introduction of a Computational Laboratory into the Physical Chemistry Curriculum Roseanne J. Sension Departments of Chemistry and Physics, University of Michigan, Ann Arbor, MI 48109-1055

Computational methods are of increasing importance in the chemical sciences. This paper describes a computational chemistry laboratory course that has been developed and implemented at the University of Michigan as part of the core physical chemistry curriculum. This laboratory course introduces students to the principle methods of computational chemistry and uses these methods to explore and visualize simple chemical problems.

Computational methods have long been important within the physical chemistry community. However, a quick survey of the literature in any of the major chemistry journals will highlight the impact that computational methods are now having in all areas of modern chemistry, from organic synthesis, to biochemistry, to the development of materials, in addition to the traditional areas of physical chemistry. A recent search of American Chemical Society (ACS) Journals for articles published over a four year period returned over 10,000 articles using molecular dynamics simulations. Approximately 40% of the articles were published in the Journal of Physical Chemistry A or B . The other citations were scattered throughout the remaining journals published by the A C S . Computational methods have become ubiquitous in large part because they enhance the microscopic, atomistic, and dynamic understanding of molecular systems. Scientists are able to develop a level of insight and intuition not easily obtained from experiments or equations alone.

220

© 2008 American Chemical Society

In Advances in Teaching Physical Chemistry; Ellison, M., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2007.

Downloaded by UCSF LIB CKM RSCS MGMT on November 19, 2014 | http://pubs.acs.org Publication Date: December 18, 2007 | doi: 10.1021/bk-2008-0973.ch013

221 The power of computational methods to enhance student understanding of chemical phenomena is, if not ignored, severely underutilized. Future scientists in all areas of chemistry will be expected to appreciate and evaluate the results of simulations and other calculations presented in lectures, seminars, and research articles. Many of them will be required to use sophisticated computational methods in the course of their own research. Today's students should learn how to evaluate the validity of computationally generated results and to make sound choices between the different methods available to address a given problem. Computational methods are not well integrated into most of the undergraduate curriculum, although implementation in the physical chemistry curriculum has been reported (1-5). The approach outlined here builds on many of these earlier ideas while emphasizing collaboration and exploration.

Outline of the Course The Computational Chemistry laboratory course developed at the University of Michigan is designed to educate students in computational methods within the context of the undergraduate physical chemistry curriculum. The course is a onecredit course required of all chemistry majors, accompanying two semester long lecture courses in Physical Chemistry and a three-credit hour Physical/Analytical laboratory course. The material developed in this laboratory introduces chemistry majors to methods in modern computational chemistry and molecular modeling, and uses these methods to draw connections between the equations presented in lecture, results obtained in physical measurements, and "real" chemical phenomena. Students come out of the course with an appreciation for the methods of modern molecular modeling and an improved insight and intuition for chemical phenomena. The cornerstone of this curriculum is a series of hands-on laboratories. The students learn by doing in a collaborative, exploratory environment. Thus the laboratory units are intended to meet two separate but complementary goals. 1.

Each laboratory unit introduces and develops the fundamentals of one of the important computational methods commonly used in modern chemical research. This includes an appreciation for the pitfalls that may arise in any calculation and the ways to avoid these problems. 2. Each laboratory is designed to help students investigate, visualize and explore a chemical problem, developing an insight and intuition not easily developed from equations alone. The prerequisites for the course include two years of chemistry, including organic, analytical, and an introductory inorganic chemistry course, one year of calculus-based physics, three terms of calculus, and introduction to differential equations usually taken concurrently. Most students take the computational laboratory concurrent with the physical chemistry lecture course covering

In Advances in Teaching Physical Chemistry; Ellison, M., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2007.

Downloaded by UCSF LIB CKM RSCS MGMT on November 19, 2014 | http://pubs.acs.org Publication Date: December 18, 2007 | doi: 10.1021/bk-2008-0973.ch013

222 quantum chemistry and molecular spectroscopy. The course does not require or presume any prior programming experience or any prior experience with computational chemistry software. The credit load for the computational chemistry laboratory course requires that the average student should be able to complete almost all of the work required for the course within the time constraint of one four-hour laboratory period per week. This constraint limits the material covered in the course. Four principal computational methods have been identified as being of primary importance in the practice of chemistry and thus in the education of chemistry students: (1) Monte Carlo Methods, (2) Molecular Mechanics Methods, (3) Molecular Dynamics Simulations, and (4) Quantum Chemical Calculations. Clearly, other important topics could be added when time permits. These four methods are developed as separate units, in each case beginning with the fundamental principles including simple programming and visualization, and building to the sophisticated application of the technique to a chemical problem. Each unit is introduced by a sixty to ninety-minute lecture providing an overview of the method, some necessary background information not otherwise covered in the curriculum, and an outline of the goals of the experiments and exercises. Thus the total lecture time over the course of the semester is four or five hours. The course is designed to facilitate hands-on exploration and active learning as much as possible. In this context the course cannot and does not provide comprehensive coverage of computational chemistry.

One Example: Molecular Dynamics Calculations The molecular dynamics unit provides a good example with which to outline the basic approach. One of the most powerful applications of modern computational methods arises from their usefulness in visualizing dynamic molecular processes. Small molecules, solutions, and, more importantly, macromolecules are not static entities. A protein crystal structure or a model of a D N A helix actually provides relatively little information and insight into function as function is an intrinsically dynamic property. In this unit students are led through the basics of a molecular dynamics calculation, the implementation of methods integrating Newton's equations, the visualization of atomic motion controlled by potential energy functions or molecular force fields and onto the modeling and visualization of more complex systems.

A Simple One-Dimensional Trajectory A very simple implementation of a molecular dynamics trajectory calculation is achieved by using a velocity Verlet algorithm to calculate the

In Advances in Teaching Physical Chemistry; Ellison, M., et al.; ACS Symposium Series; American Chemical Society: Washington, DC, 2007.

223 motion of a diatomic vibrator (6). In this method, given an initial position, x(0), and initial velocity, v(0) , the trajectories are calculated iteratively from:

2

x(t + St) = x(t) + a • v(t)+!