Computational Chemistry

(Steven Felgate) #1

room-temperature, 298 K, of the dimer minus twice the enthalpy of the monomer.
Here are some calculated values of this binding enthalpy,withoutthe BSSE
correction; hartrees are converted to kJ mol#^1 by multiplying by 2,626:
CBS-Q, a high-accuracy multistep method with correlation energy correction
and large basis sets (Section 5.5.2.2b)


# 152 : 67093 ##ðÞ¼# 152 : 66546 0 : 00547 ¼# 14 :4 kJ;

MP2= 6 # 311 þþG 3dfðÞ;3pd

# 152 : 60355 ##ðÞ¼# 152 : 59780 0 : 00575 ¼# 15 :1 kJ;

MP2= 6 #31G$

# 152 : 35198 ##ðÞ¼# 152 : 34318 0 : 00880 ¼# 23 :1 kJ;

HF= 6 # 311 þþG 3dfðÞ;3pd

# 152 : 06881 ##ðÞ¼# 152 : 06510 0 : 00371 ¼# 9 :74 kJ;

HF= 6 #31G$

# 151 : 97417 ##ðÞ¼# 151 : 96798 0 : 00619 ¼# 16 :3 kJ

The correlation-correction/large basis CBS-Q calculation gives a binding
enthalpy (#14.4 kJ mol#^1 ) not too far from the experimental (#13.4 kJ mol#^1 )
and the MP2 method with the very big MP2/6–311þþG(3df,3pd) basis gives a
somewhat worse deviation, while with MP2 and a smaller basis the binding value is
still worse. This is in accord with the above assertion that accounting for BSSE will
give a smaller energy drop than without it, i.e. that non-counterpoise calculations
give the bigger energy drop. However one should add “other things being equal”:
with the Hartree–Fock method, the smaller basis (6–31G*) actually gives a
smallerenthalpy drop (9.74 kJ mol#^1 ) than the ca. 13 kJ mol#^1 decrease expected
from a good counterpoise calculation (and the binding enthalpy is, coincidentally,
slightly better estimated than with HF/6–311þþG(3df,3pd)). Presumably the
somewhat erratic results at the HF level are due to neglect of dynamic correlation
(Section 5.4.1).
The water dimer has been examined in detail with the accent on density
functional (Chapter 7) methods [ 107 ], and the use of ab initio as well as DFT in
counterpoise calculations on it [ 108 ]. Using large basis sets and high correlation
levels to get high-qualityatomization energies(which are of course not of the weak
interaction type, and which the counterpoise correction is said to worsen [ 105 b]) is
explained in the book by Foresman and Frisch [ 1 e]. Energy calculations are
discussed further inSection 5.5.2.


280 5 Ab initio Calculations

Free download pdf