Full Paper
tial ratio of photooxidation products, and hence the selectivity,
can be quantified by the sum of the integrals for the respec-
tive peaks of the product pairs (2+3 and 4+5; 1.08:1.00).[28]
Other solvents such as ethyl acetate, toluene, and CH2Cl2 af-
forded the same oxidation ratio. To determine whether the oxi-
dation proceeds via a charged benzylic intermediate (path I),
several para-substituted secondary amines were investigated
(Table 1, entries 1–4).
6) and led us to investigate the influence of BDE on oxidation
selectivity.
Ortho-functionalized asymmetric dibenzylamines: a
combination of electronic and steric effects
Further investigation of substitution of the aromatic ring
revealed a good degree of selectivity for ortho-substituted
asymmetric dibenzylamines (62–69%; Table 1, entries 7–9).[31]
Moreover, the photooxidation of ortho-disubstituted aromatics
resulted in almost complete selectivity for oxidation of the un-
substituted benzyl side (90–95%; Table 1, entries 10–12).
Table 1. Correlation of A-side oxidation with the difference in %s
character.[a]
Since inductive effects are only observed over short distan-
ces and electronegative substituents increase neighboring
BDEs,[32] we hypothesized that the BDE of the CÀH bond is the
main determinant of selectivity. An NBO analysis of the sub-
strate scope was undertaken in an attempt to quantify this se-
lectivity trend.[33] A general trend was observed in which the
CÀH bonds adjacent to the nitrogen atom in the most selec-
tive substrates had higher differences in hybridization (%s
character) than less selective substrates (Table 1). Hybrid orbital
theory explains that an increase in hybridization of the carbon-
centered orbital leads to a higher BDE (sp>sp2 >sp3).[34] The
calculated differences in the %s character of the two CÀH
bonds a to the nitrogen atom are summarized in Table 1.
The %s character can also be determined experimentally
from one-bond CÀH coupling constants (1JCH =5ꢁ%s).[35,36]
Entry
R
D%s character (BÀA)
% of A side [O]
NBO
1JCH/5
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
4-OMePh
4-NO2Ph
4-MePh
3,4,5-OMePh
1-naphthyl
2-naphthyl
2-OMePh
2-ClPh
2-FPh
2,6-ClPh
2,6-FPh
2,3,4,5,6-FPh
Mes
tBu
0.01
0.01
0.08
52
55
52
57
52
À0.11
0.01
0.03
À0.19
0.34
0.26
0.36
0.41
0.70
0.66
0.75
0.36
À0.07
0.00
0.02
0.00
À0.01
0.36
0.44
0.42
0.76
0.88
1.14
0.01
À0.41
À0.34
À0.51
55
71[c]
62
62
>95
90
>95
81
43
1
Coupled 13C NMR spectra were used to determine the JCH
values of the two sets of CÀH bonds a to the nitrogen atom in
each substrate (Table 1). The differences in hybridization of the
CÀH bonds in the substrates indicate a strong inductive effect
on hybridization. Incorporation of electronegative groups at
the ortho position increases the electronegativity of the ipso
carbon atom, and thereby p character is withdrawn from the
adjacent benzylic carbon atom.[37] As a result, less p character
is available for the benzylic CÀH bonds, hence the observed in-
crease in s character. As the substrates examined only include
aromatic rings substituted with groups more electronegative
than hydrogen,[38] we did not observe any preferential oxida-
tion of the substituted side. Oxidation of the unsubstituted
benzyl side was always preferred, that is, the difference in %s
character is directly related to the rate of oxidation. Although
this difference was higher for the difluoro- than for the respec-
tive dichloro-substituted compound (Table 1, entries 11 and 10,
respectively), the observed ratio of imine products revealed
a lower selectivity. Che et al.[13] previously noted the impor-
tance of steric hindrance. In our case, the greater steric bulk of
chlorine in comparison with fluorine (A values: 0.43 versus
0.15)[39] can explain the observed selectivities.
[b]
leelamine
1-adamantyl
–
36
38
À0.34
[a] Full conversion was observed for all substrates. For full experimental
details, see the Supporting Information. [b] NBO analysis of this substrate
was not performed. [c] Average yield over four experiments with a range
of 4%.
The ratio of imine products obtained on photooxidation of
the p-nitro-monosubstituted dibenzylamine derivative (1.2:1,
Table 1, entry 1) was similar to the results obtained for the oxi-
dation of the mono p-methoxy derivative. In fact, the introduc-
tion of substituents in either the para or the meta position did
not lead to any significant differences in the oxidation selectivi-
ty, and oxidation of the unsubstituted benzyl side was slightly
favored. Based on the relative acidities of p-NO2- and p-OMe-
substituted compounds (pKa =10.8 and 19.1 for phenols; 20.9
and 27 for anilines),[29] we expected there to be a significant
difference, as well as a switch in selectivity, if the reaction
proceeded via path I. However, this was not observed, which
suggests that the acidity of this proton does not significantly
influence the outcome of the oxidation.
From what is known computationally of benzyl radicals,[30]
extension of the conjugated systems to naphthyl groups
makes radical delocalization less efficient, and the calculated
BDEs of 1-naphthyl and 2-naphthyl CÀH bonds are higher than
those of benzylic CÀH bonds by about 1 kcalmolÀ1. This is re-
flected in the poor selectivity observed (Table 1, entries 5 and
For a more detailed investigation of steric effects, a mesityl
derivative (Table 1, entry 13) was prepared. For this substrate,
each of the benzyl CÀH bonds has nearly the same %s charac-
ter and should therefore result in a poorly selective oxidation.
However, analysis of the photooxidation products revealed
that oxidation occurred predominantly (81%) at the less hin-
dered side (A value of CH3 is 1.7),[39] demonstrating that steric
effects play an important role in selectivity as well.
Chem. Eur. J. 2015, 21, 1 – 8
3
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
&
These are not the final page numbers! ÞÞ