IDENTICAL ACYL TRANSFER REACTIONS
1611
tural effects in the reagents or, all the more, describe all
the reactions studied.
The electron affinity of a molecule or ion is usually
taken to equal [13] (Table 1) the energy of the lowest
unoccupied molecular orbital (LUMO). This EA* value
may fail to describe structural effects in cations explic-
itly because of the special features of LUMO localiza-
+
≠
Our detailed analysis of ∆G –pK correlations for
Çç
identical dimethylcarbamoyl transfer reactions [7] only
allowed us to find that the transition state of reactions (1)
becomes more “rigid” (shifts toward tetrahedral inter-
mediate products) as the basicity of the nucleophile and
leaving group grows. The factors that control the barrier
to the reaction and the reasons for barrier changes
remain unknown. It is also unclear to what extent the
conclusions drawn from our analysis are universal, that
is, transferable to other reaction series. For instance,
note that identical methyl and acyl transfer reactions (1)
are substantially different. The rate of the former drops
in the series of the strongest nucleophiles [13] and
grows in the series of nucleophiles studied in this work
+
tion. An analysis of the MOs of AcLg indeed showed
that the lowest unoccupied molecular orbital of all cat-
ions did not contain a contribution of the p orbital of
z
the carbonyl group, that is, exactly the atomic orbital
(
AO) of the reaction center that experienced the frontal
attack of a nucleophile in reaction (1) [16]. For this rea-
+
son, we selected those LUMOs of AcLg (subfrontier
orbitals with respect to that determining the EA* value)
that contained the largest contributions of the p orbital
z
of C=O. In the majority of cases (except no. 27,
Table 1), the coefficients of the p AO in the selected
z
LUMOs were of 0.4–0.6. The energies of these orbitals,
(Table 1; Table 2, Eqs. (1) and (4)).
E , are listed in Table 1.
In recent years, nucleophilic substitution reactions,
p
z
including nucleophilic substitution at the carbonyl cen-
ter [14], have more and more often been analyzed using
the model of the Shaik–Pross cross-diagrams [1, 13]. In
It follows from Table 2 that the quality of correla-
tions with Epz is much better than the quality of corre-
this approach, reactivity is written in terms of the elec- lations with EA* (Table 2, compare Eqs. (17) and (18)
+
tron affinity EA* of the electrophile (AcLg ) and the with (15) and (16) or (19) with (12)). Moreover, setting
ionization potential IP* of the nucleophile as
the energy gap of the reaction equal to IP* – Epz not
≠
only gives physically meaningful results (Table 2, com-
pare (19) and (12)) but also allows a unified correlation
to be used for processing data on all the identical acyl
transfer reactions studied (Table 2, Eq. (20)), which is,
in our view, a very interesting result. Note in conclusion
that, as distinct from the MO characteristics consid-
ered, the other calculated quantum-chemical parame-
∆
G = A(IP* – EA* ) – B,
(2)
Nu
AcLg+
where A is the value characterizing the curvature of the
intersecting potential functions (for instance, parabo-
las) and B is the resonance interaction energy of orbitals
in the transition state. We cannot use (2) for our pur-
poses without invoking the results of quantum-chemi-
cal calculations (Table 1). In addition, we must assume
that A and B are constant values.
ters, such as bond orders and charges on atoms of the
reagents of reaction (1), do not correlate with the ∆G≠
Let us consider correlations between the experimen- values; for this reason, we neither give nor discuss
tal and calculated reaction characteristics (Table 2, them.
Eqs. (11)–(20)). Processing all the reactions studied in
To summarize, the results obtained lead us to con-
the coordinates of Eq. (2) does not give satisfactory
clude that reactivity in identical acyl transfer reactions
results (Table 2, Eq. (11)). At the same time, there is a
is controlled by the interaction of frontier orbitals in the
correlation for Acyl = const (Table 2, Eq. (12)), which
transition state. The particular localization of MOs on
is, however, physically meaningless. Indeed, it predicts
electrophile fragments, especially, its reaction center
a decrease in the barrier to reaction as the energy gap
(
carbonyl group) should be taken into account in con-
(
IP* – EA*) to be overcome by the reactants in reaction (1)
structing correlations and selecting frontier LUMOs.
increases. Formally, failures of both correlations can be
related to the assumptions made (A, B = const) and the
calculated quantum-chemical parameters IP* and EA*
proper. No verified methods for the determination and
control of the A and B values have been suggested
REFERENCES
1
2
3
4
. F. A. Carroll, Perspectives on Structure and Mechanism
in Organic Chemistry (ITP, California, 1998).
[
1, 13], but the assumption that they are constant in
series of related reactions is considered a good approx-
imation [15].
. E. S. Lewis and D. D. Hu, J. Am. Chem. Soc. 106 (11),
3
292 (1984).
. R. I. Masel, Chemical Kinetics and Catalysis (Wiley,
New York, 2001).
For this reason, we concentrated on the calculated
≠
IP* and EA* values. Correlations of ∆G with IP*
. I. Lee, Chem. Soc. Rev. 19 (1), 133 (1990).
(
Table 2, Eqs. (13) and (14)) are satisfactory, but corre-
lations with EA* are not. The latter values do not reflect
5. C. F. Bernaskoni, J. A. Moreira, L. L. Huang, and
K. W. Kittredge, J. Am. Chem. Soc. 121 (8), 1674
(1999).
the expected changes in the electronic structure of
+
AcLg in reaction series with one and the same nucleo-
phile (Table 2, Eqs. (15) and (16)), that is, “take no
notice” of changes in the nature of the acyl group.
6. V. I. Rybachenko, G. Shroeder, K.Yu. Chotii, et al., Teor.
Eksp. Khim. 39 (6), 347 (2003).
RUSSIAN JOURNAL OF PHYSICAL CHEMISTRY A Vol. 81 No. 10 2007