Quantum Calculations
by Alyssa Ladwig
and Jon Luetke

    Being able to understand the electronic structure of a molecule allows one to determine the reactivity of the molecule.  It also allows one to predict the locations of the electrons, vibrational frequencies, potential energy, and the UV-Visible transitions.  These properties are useful to determine whether a molecule will react in a reaction or not.  To find these properties, wavefunctions for the electrons are formulated.  The expectation value of energy is calculated by taking the integral of the wavefunction multiplied by the Hamiltonian operator and the conjugated wavefunction.  The calculations are complicated and require computer technology.  For the purposes of this experiment, the computer program GAMESS was used to calculate the integrals.  Two methods of calculations were used, MOPAC and ab initioAb initio is the best level of theory.  It calculates all the integrals, but the calculations can be lengthy depending upon the molecule.  From these calculations, diagrams of the highest occupied molecular orbital (HOMO) and potential energy curves were created and information about the vibrational frequencies and UV-Visible transitions were collected for the molecules fluorine (F2), carbon monoxide (CO), and para-dibromobenzene (C6H4Br2).

    The optimized geometry for each molecule was essential to determine before completing any additional calculations.  The ideal geometry was chosen by comparing the calculated values for bond lengths with the literature values.  The basis set with the closest bond length was chosen.  For fluorine, the 631Gd calculated value for the bond length of 1.412529 Aº was closest to the literature value of 1.4119 Aº.  For carbon monoxide, the 321G calculated value for the bond length of 1.128936 Aº was closest to the literature value of 1.1283 Aº. The literature values were obtained from the NIST Computational Chemistry Comparison and Benchmark Database.

                          Fluorine Optimized Geometry                                                 Carbon Monoxide Optimized Geometry
F2 Optimized GeometryCO Optimized Geometry
The calculated bonds for each of the levels of theory were compared to the literature values for para-dibromobenzene.  The double zeta calculation proved to be the closest to the literature values (Table 1). 

                                                                           Para-dibromobenzene Optimized Geometry
para-dibromobenzene optimized geometry
Table 1: Para-dibromobenzene Optimized Bond Lengths for Double Zeta Basis Set
Bond Calculated Values (Aº) Literature Values (Aº)
Br-C1
1.89344
1.906
C1-C2
1.38854
1.388
C2-C3
1.39236
1.394
C-H
1.0714
1.09
The literature values were obtained from Liu, Y., et al. J. Chemical Physics 2004, 120(14).

Using the optimized geometry, diagrams of the highest occupied molecular orbital (HOMO) were created for each molecule.

                         Fluorine HOMO (Energy Level 9)                                          Carbon monoxide HOMO (Energy Level 7)
F2 HOMOCO HOMO
                                                                          Para-dibromobenzene HOMO (Energy Level 55)
Para-dibromobenzene HOMO
The potential energy surface as a function of bond length was calculated for the diatomic molecules at each level of ab initio theory.  At higher levels of theory (more complex) the potential energy values decreased.  Comparing the two diatomic molecules, fluorine had a much more negative potential energy than carbon monoxide overall.  When comparing both diatomics it can be seen that there is a potential well in each where the minimal value is near the equilibrium bond length.

F2 Potential Energy Surface

CO Potential Energy Surface


The vibrational frequencies for the three molecules were calculated using the optimized geometry. 

Table 2: Diatomic Molecule Frequency Data
Molecule
Calculated Frequency (cm-1)
Literature Values (cm-1)
F2
1141.609985
916.6
CO
2312.689941
2169.8
The literature values were obtained from the NIST Computational Chemistry Comparison and Benchmark Database.

Table 3: Para-dibromobenzene Frequency Data (cm-1)
95.389999
479.970001
1082.750000
1213.290039
1752.900024
162.570007
602.950012
1106.609985
1314.989990
1787.920044
231.279999
696.559998
1173.420044
1325.780029
3402.360107
309.390015
763.799988
1175.619995
1446.219971
3422.179932
350.869995
867.400024
1197.609985
1533.069946
3429.979980
465.910004
1027.540039
1201.030029
1655.630005
3448.949951

A comparison to the literature values for the para-dibromobenzene is most effectively accomplished by comparing the calculated results to a literature Infared spectrum. The NIST website can be used to make such a comparison:                                http://webbook.nist.gov/cgi/cbook.cgi?ID=C106376&Units=SI&Type=IR-SPEC&Index=1#IR-SPEC

As can be seen, the calculated vibrational frequency for fluorine is well outside any range of error for the literature value.  Carbon monoxide is closer to the literature value, however still quite different.  When comparing the calculated values to the IR spectrum of para-dibromobenzene, a large number of the calculated values cannot be seen on the spectrum.  Therefore, the accuracy of any values less than 400cm-1 cannot be determined.  The calculated values over 3000cm-1 are not shown as peaks in this region represented on the spectrum.

When considering the representative motion of the significant peaks in the para-dibromobenzene spectrum the following motions are represented.

Table 4: Relative Motions for Significant Peaks in the Calculated Vibrational Frequencies for Para-dibromobenzene
Frequency (cm-1)
Description of Motion
465.910004
sliding of benzene ring between bromines
602.950012
folding of molecule around axis running through bromines
1082.750000
flapping motion of carbons with hydrogens attached
1106.609985
anti-symmetric planar stretch of carbons with hydrogens attached
1201.030029
anti-symmetric stretching between C2, C6 and C3, C4
1533.069946
anti-symmetric planar stretch between C1, C2 and C4, C5
1655.630005
symmetric planar stretches of C1 and C4 from C3
3448.949951
stretching between hydrogen and carbon on benzene ring


The values for transition energy of an electron from the ground state to an excited state were calculated for para-dibromobenzene for the 3-21G and 6-31G basis sets.  The calculated data can be seen in the following tables.  In addition to the value for transition energy, a value for oscillator strength is also given.  This oscillator strength can be used to determine the strength of the peak that would be displayed on a UV-Visible spectrum.

Table 5: Para-dibromobenzene  UV-Visible Results for 3-21G Basis Set
From Ground State to Excited State
Transition Energy (cm-1)
Oscillator Strength
1
51427.24
0.001407
2
52826.77
0.137390
3
55333.80
0.000000
4
56997.53
0.014225
5
64074.29
0.006629
6
64198.74
0.000000
7
67217.40
1.653998 (Primary)
8
68624.24
0.876194 (Secondary)
9
70255.07
0.000000
10
74088.20
0.000000

Table 6: Para-dibromobenzene UV-Visible Results for 6-31G Basis Set
From Ground State to Excited State
Transition Energy (cm-1)
Oscillator Strength
1
50824.08
0.000274
2
52116.67
0.116610
3
54799.24
0.000000
4
56446.28
0.010258
5
64607.43
0.004443
6
64721.19
0.000000
7
66374.97
1.727890 (Primary)
8
67719.53
0.902194 (Secondary)
9
68290.52
0.000000
10
73855.20
0.000000


The calculated transition energy values were compared to literature values obtained from the Journal of American Chemical Society.  In Hellmann and Bilbo's study published in 1953, they determined the value for the primary band to be 43898 cm-1 and the secondary bands to be at 36590 and 35448 cm-1.  As can be seen, the calculated values were much greater than the literature values.  Hellmann and Bilbo's results can be found at: http://pubs.acs.org/cgi-bin/archive.cgi/jacsat/1953/75/i18/pdf/ja01114a512.pdf

From the vibrational frequency and transition energy calculations, it can be seen that their ab initio calculations were not good predictors of literature values.  The bond length calculations used to determine the optimized geometry were very close to literature values.  Therefore, ab initio could be used to optimize molecules but would not be accurate in predicting vibrational frequencies and transition energies.