Next: Project 2: Constructing Walsh
Up: No Title
Previous: Literature Review
Select any small molecule of your choice, as long as it satisfies the
following criteria:
- At least two atoms, and not H2
- No more than 6 non-hydrogen atoms, plus any number of hydrogens
- Work with the ground electronic state, which must be closed-shell
(i.e., no radicals, except by special permission)
Use three different correlation methods:
- Hartree-Fock
- Density Functional Theory (DFT): use whichever flavor of DFT you
prefer. Examples: S-VWN (Slater exchange plus VWN correlation,
often called LSDA), BLYP (Becke exchange plus LYP correlation),
B3LYP, B3PW91, EDF1.
- MP2
For each of these three correlation methods, HF, DFT, MP2, use any two basis
sets of your choice. This will give you a total of six different
levels of theory. For each level of theory, optimize the molecule to
obtain the equilibrium geometry, and at that geometry, obtain the following
equilibrium molecular properties:
- Geometry (in angstroms and degrees)
- Energy (in hartrees)
- Harmonic vibrational frequencies (in cm-1) and
infrared intensities (in km mol-1)
- Dipole moment (in Debye)
- Ho(298 K) (in kcal mol-1)
- So(298 K) (in cal mol-1 K-1)
- Go(298 K) (in kcal mol-1)
Results:
Prepare a table listing your molecular properties at each level of theory
and any available experimental results.
Questions for Discussion (VERY IMPORTANT):
- 1.
- For this question, completely ignore any experimental results you
may have obtained. This will simulate your doing computations on
systems where the experimental data isn't available (which is often
the situation when you're doing computations!)
Compare the theoretical results to each other. Do they appear to be
converging with improved basis set and treatment of correlation, or is there
a lot of scatter in the data? Would additional computations be required
to obtain reliable predictions? If so, what computations would you
recommend? Based on your analysis (and NOT by comparing
to experiment!), give a set of ``best'' theoretical predictions and
estimate the uncertainties.
- 2.
- Compare the theoretical data, including your ``best estimates'' with
the experimental data. Note any significant disagreements and explain them
if possible.
Next: Project 2: Constructing Walsh
Up: No Title
Previous: Literature Review
C. David Sherrill
2000-07-21