Results on the frequency of the iterated processes

The results of the series of tests corresponding to the proton transfer and the carbon configuration inversion steps, when a full minimization of the environment is carried out at every TS search step in the core, are given in Tables 3.5 and 3.6, respectively. Comparing with Tables 3.3 and 3.4, it can be observed that the minimization of the environment at every TS search step in the core clearly reduces the number of core iterations, although in turn tends to augment the number of environment iterations. As a result, the difference between both strategies does not seem to be very significant, although the minimization of the environment at every TS search step rather tends to increase the total number of iterations in the two reaction steps studied here. As expected, the energy column shows that the final transition state structure for each core/environment partition is independent on the strategy used.

Table 3.5: Results testing the core TS search / environment minimization switch for the proton transfer step. At every TS search step a full minimization of the environment is performed.

core / env	(iter.core/iter.env)	energy$^a$
3 / 1295	(96/1896)	-7453.26
12 / 1286	(105/2075)	-7453.25
23 / 1275	(42/1211)	-7453.27
80 / 1218	(46/976)	-7453.29
138 / 1160	(105/1088)	-7453.27
194 / 1104	(62/752)	-7453.27
399 / 899	(80/793)	-7453.28
$^a$ In kcal/mol

Table 3.6: Results testing the core TS search / environment minimization switch for the carbon configuration inversion step. At every TS search step a full minimization of the environment is performed.

core / env	(iter.core/iter.env)	energy$^a$
7 / 1291	(88/2092)	-7443.25
17 / 1281	(143/1923)	-7442.31
23 / 1275	(115/1908)	-7442.31
34 / 1264	(66/1631)	-7442.32
80 / 1218	(69/1861)	-7442.33
138 / 1160	(52/758)	-7440.50
194 / 1104	(73/920)	-7440.50
643 / 655	(117/710)	-7440.54
$^a$ In kcal/mol

Our experience shows that since the location of TS points is complicated, the strategy of a full TS search before a next minimization is recommendable to unsure at a primary stage of the search that the final convergence will be reached. On the contrary, when the minimization is performed at every TS search step we have no information about the desired success of the search until the end of the whole process.