Introduction
Many important properties of molecules can now be determined through
several computer calculations, instead of by hand, allowing one to use
large basis sets. However, these calculations are not always
correct and one must be skeptical of computer output. The
electonic structure of a molecule determines much of its reactivity
using the probable locations of the electrons and their energies.
The wave functions for electrons in molecules are known as
molecular orbitals. Real wave functions always have lower energy
than any approximation. The best geometry is found by trying
different arrangements of atoms until the energy of the system is
minimized. The larger the basis set, and the more accurate the
molecular geometry, the more accurate the energy prediction will be.
The three molecules looked at during this experiment were hydrogen floride (HF), carbon dioxide (CO
2),
and Styrene. The geometries (bond lengths and angles), primary
molecular orbitals (HOMO), dipole moments, and vibrational
frequencies/modess were all evaluated using the program MacMolPlt.
Three different level of
ab initio
theory (3-21G, 6-31G, and 6-311G) were compared for consistency of
predictions. The best ab initio results were used to
calculate the vibrational frequencies of each molecule. Potential
energy plots were generated for our diatomic molecules using the best
ab initio
results as a starting point, and UVvis calculations were performed for
our aromatic molecule. The results obtained were compaired to the
experimental ones found using the NIST website (National Institute of
Standards and Technology).
Data/Results
Follow the links below to view the data and results for each molecule.
HF calculations CO2 calculations
Styrene calculationsConclusion
In conclusion, it has been determined that the computer calculated values and properties for HF, CO
2,
and Styrene are both precise and accutate compared to the experimental
values found on the NIST website. It should be true that the best
ab initio results (for the highest level of theory) be the closest in
value to the experimental values because it should have a more accurate
molecular geometry and therefore, a more accurate energy prediction.
However, this not always the case because there are
approximations made in the programs used, and they are not always 100%
accurate. Some of the calculations do not always get ran
completely and some even fail. It was also seen that the three
different level of theory were very consistent with each other,
indicating that a good prediction was made. The computational
results found are more useful when you are analyzing a smaller molecule
because less approximation are needed when performing the calculations,
and they take less time to run.
References
Lab Handout. Molecular Orbital/Quantum Calculations 2. J. Gutow 2005; revised 2009.
Atkins and De Paula. Physical Chemistry.8
th Edition. W.H. Freeman and Company. New York, 2006. Ch. 11 & 18.
Computational Chemistry Comparison and Benchmark Data Base. NIST.
<http://cccbdb.nist.gov/>.
NIST Chemistry WebBook. NIST. U.S. Secretary of Commerce. 2008.
<http://webbook.nist.gov/chemistry/>.