07.04.2013 Views

Essentials of Computational Chemistry

Essentials of Computational Chemistry

Essentials of Computational Chemistry

SHOW MORE
SHOW LESS

You also want an ePaper? Increase the reach of your titles

YUMPU automatically turns print PDFs into web optimized ePapers that Google loves.

7.3 CONFIGURATION INTERACTION 211<br />

clearly, if one is considering a reaction coordinate or a series <strong>of</strong> isomers, the active space<br />

must be balanced, so any orbital contributing significantly in one calculation should probably<br />

be used in all calculations. While methods to include dynamical correlation after an MCSCF<br />

calculation can help to make up for a less than optimal choice <strong>of</strong> active space, it is best not<br />

to rely on this phenomenon.<br />

7.2.3 Full Configuration Interaction<br />

Having discussed ways to reduce the scope <strong>of</strong> the MCSCF problem, it is appropriate to<br />

consider the other limiting case. What if we carry out a CASSCF calculation for all electrons<br />

including all orbitals in the complete active space? Such a calculation is called ‘full<br />

configuration interaction’ or ‘full CI’. Within the choice <strong>of</strong> basis set, it is the best possible<br />

calculation that can be done, because it considers the contribution <strong>of</strong> every possible CSF.<br />

Thus, a full CI with an infinite basis set is an ‘exact’ solution <strong>of</strong> the (non-relativistic,<br />

Born–Oppenheimer, time-independent) Schrödinger equation.<br />

Note that no reoptimization <strong>of</strong> HF orbitals is required, since the set <strong>of</strong> all possible CSFs<br />

is ‘complete’. However, that is not much help in a computational efficiency sense, since the<br />

number <strong>of</strong> CSFs in a full CI can be staggeringly large. The trouble is not the number <strong>of</strong><br />

electrons, which is a constant, but the number <strong>of</strong> basis functions. Returning to our methanol<br />

example above, if we were to use the 6-31G(d) basis set, the total number <strong>of</strong> basis functions<br />

would be 38. Using Eq. (7.9) to determine the number <strong>of</strong> CSFs in our (14,38) full CI we<br />

find that we must optimize 2.4 × 10 13 expansion coefficients (!), and this is really a rather<br />

small basis set for chemical purposes.<br />

Thus, full CI calculations with large basis sets are usually carried out for only the smallest<br />

<strong>of</strong> molecules (it is partly as a result <strong>of</strong> such calculations that the relative contributions to<br />

basis-set quality <strong>of</strong> polarization functions vs. decontraction <strong>of</strong> valence functions, as discussed<br />

in Chapter 6, were discovered). In larger systems, the practical restriction to smaller basis<br />

sets makes full CI calculations less chemically interesting, but such calculations remain<br />

useful to the extent that, as an optimal limit, they permit an evaluation <strong>of</strong> the quality <strong>of</strong><br />

other methodologies for including electron correlation using the same basis set. We turn<br />

now to a consideration <strong>of</strong> such other methods.<br />

7.3 Configuration Interaction<br />

7.3.1 Single-determinant Reference<br />

If we consider all possible excited configurations that can be generated from the HF determinant,<br />

we have a full CI, but such a calculation is typically too demanding to accomplish.<br />

However, just as we reduced the scope <strong>of</strong> CAS calculations by using RAS spaces, what if<br />

we were to reduce the CI problem by allowing only a limited number <strong>of</strong> excitations? How<br />

many should we include? To proceed in evaluating this question, it is helpful to rewrite<br />

Eq. (7.1) using a more descriptive notation, i.e.,<br />

occ.<br />

vir.<br />

= a0HF + a r i r i +<br />

occ.<br />

vir.<br />

a rs<br />

ij rs ij +··· (7.10)<br />

i<br />

r<br />

i

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!