In the Laboratory www.JCE.DivCHED.org • Vol. 84 No. 2 February 2007 • Journal of Chemical Education 329 Characterization of High Explosives and Other Energetic Compounds by Computational Chemistry and Molecular Modeling Experiments for the Undergraduate Curriculum John A. Bumpus,* Anne Lewis, and Corey Stotts Department of Chemistry, University of Northern Iowa, Cedar Falls, IA 50614; *john.bumpus@uni.edu Christopher J. Cramer Department of Chemistry and Supercomputing Institute, University of Minnesota, Minneapolis, MN 55455-0431 Heat of formation is one of a variety of properties used to assess and compare the energy content of high explosives 1 (1). We present a series of introductory and advanced experi- ments in which heats of formation are predicted from calcu- lation. These experiments can be used to introduce students to commercially available computational chemistry and mo- lecular modeling programs. They also help reinforce several concepts in thermodynamics and computational chemistry. In the initial experiment several compounds are arranged such that students are provided structures with increasing levels of complexity so that they may master the drawing programs in a stepwise manner. Ab initio quantum mechanical (QM) cal- culations are used to model simple compounds. However, as structural complexity increases, students quickly learn time and expense are considerations in the use of such calculations. In the second experiment students are shown how to set up and use semiempirical calculations, the Austin Model 1 (AM1), the Modified Neglect of Differential Overlap (MNDO) Model, and the Parametric Model 3 (PM3), to es- timate the heats of formation for a series of energetic com- pounds, including some that are used as high explosives. In the third experiment, selected results from computational stud- ies are compared with those predicted using an older, but still useful, group additivity method (2, 3). Both approaches are compared to experimental values. Explosive compounds se- lected for study range from the very simple (e.g., methane) to the moderately complex (HMX and octanitrocubane). Al- though the PM3 model proves reasonably accurate for most of the compounds studied, it is not suitable for highly strained compounds such as cubane and octanitrocubane. This defi- ciency is used in the fourth experiment to introduce a more rigorous theoretical treatment in which density functional theory (DFT) is coupled with an isodesmic reaction approach to calculate the heat of formation of octanitrocubane. The structures of several explosives are presented in Figure 1. Experimental Procedures The following series of in silico experiments have been developed to introduce students to molecular modeling and computational chemistry of high explosives and other selected energetic compounds. The first experiment is introductory in nature and serves the dual purpose of teaching students how to use available software and familiarizing them with the structures of the explosives and other compounds used in these experiments. The remaining three experiments in- troduce students to new and old computational methods used to calculate heats of formation. Experiment 1: Molecular Model Construction Students are provided the structures of a number of high explosives and other selected energetic compounds and are asked first to draw them in any one of several molecular mod- eling programs that are commercially available. For example, we have available PCSpartan (Wavefunction, Inc.), HyperChem 7.0 (Hypercube, Inc.), and ChemDraw Stan- dard 7.0, which is used in conjunction with Chem 3D Ultra 7.0 (CambridgeSoft Corp). Manufacturers of these software programs provide relatively straightforward instructions for W Figure 1. Structures of several highly energetic compounds.