Computational Chemistry

which MM is oblivious. MM cannot provide information about the shapes and
energies of molecular orbitals nor about related phenomena such as electronic
spectra.
Because of the severely empirical nature of MM, interpreting MM parameters in
terms of traditional physical concepts is dangerous; for example, the bond-stretching
and angle-bending parameters cannot rigorously be identified with spectroscopic
force constants [ 33 ]; Lipkowitz suggests that the MM proportionality constants
(Section3.2.1) be calledpotential constants. Other dangers in using MM are:

1.Using an inappropriate forcefield: a field parameterized for one class of com-
pounds is not likely to perform well for other classes.
2.Transferring parameters from one forcefield to another. This is usually not
valid.
3.Optimizing to a stationary point that may not really be a minimum(it could be a
“maximum”, a transition state), and certainly may not be aglobalminimum
(Chapter 2 ). If there is reason to be concerned that a structure is not a minimum,

0

04000 3500

3239

1540

1083

689

3000 2500 2000

FREQ_VAL

FREQ_VAL

1500 1000 500

4000 3500 3000 2500 2000 1500 1000 500

3.1e+002

37

73

1.1e+002

6.1e+002

9.2e+002

IR_INTENS

IR_INTENS

0

4000 3000 2000 1000

20

40

60

80

Benzene 3091 3050 1480

1040

674

1032

3095 1504 699

MMFF

Experimental

MP2 / 6-31G*

Fig. 3.15 Experimental (gas phase), and MM (MMFF) and ab initio (MP2(fc)/6–31G*) calcu-
lated IR spectra of benzene

3.6 Strengths and Weaknesses of Molecular Mechanics 75