A R T I C L E S
Rubin et al.
Table 2. Polarization of Triple Bond in Tolanes and Their NMR
Properties
absolute value. In contrast, the background values of shielding
density for peripheral groups are much smaller, and therefore
their input can be more easily assessed. Remarkably, deshielding
effects on â-sp-carbon atoms caused by ortho-substituents
(Figure 4, charts B, D) were greater than those on R-sp-carbons
(charts A, C) for both electron-donating (ortho-methyl) and
electron-withdrawing (ortho-trifluoromethyl) groups. On the
corresponding maps (Figure 4), this difference is indicated by
a more intense red spot reflecting the deshielding effect cast on
the â-carbon from both o-Me and o-CF3 groups (B and D). At
the same time the analogous red site on the shielding density
maps projected on the R-carbon is relatively faint (A and C).
Shielding effects of ortho-substituents on sp-carbons in
tolanes can be even better illustrated by analysis of correspond-
ing anisotropy cones,19 depicted in Scheme 1. As discussed
above, electron-withdrawing and electron-donating substituents
in the ortho-position of tolanes induce different polarization of
the triple bond, as indicated by opposite directions of dipole
moment vectors in cartoons I and II. If this were the major
factor, the pattern of substituent-induced 13C chemical shifts of
sp-carbons in ortho-tolanes would be similar to that observed
in the para-tolane series.5 Yet, the anisotropic properties of a
substituent placed in close proximity to a triple bond also
contribute to the shielding. Although the value of anisotropy
factor ∆ø may vary, orientation of anisotropy cones and their
effects are virtually the same for different groups, in accordance
with the density map analysis provided above (Figure 4). As a
result, the signal of the â-carbon atom, located closer to the
center of the deshielding lobe of the cone, is being shifted to
lower field regardless of the electronic nature of the anisotropic
group. Undoubtedly, the observed strong equalizing anisotropic
effect of ortho-substituents dominates over the differentiating
electronic polarization effect. Consequently, while 13C chemical
shifts of sp-carbons can serve well as indices of natural charges
on a triple bond in the para-tolane series, by no means can
they be used for deducing electronic polarization of a triple bond
in ortho-tolanes.
substituent
∆
NC
∆δexpt
∆δcalcd
R1
R2
NCR
−
NCâ
δR − δâ
δR − δâ
1r
1s
1d
1t
1b
1a
1i
1k
1h
1j
Me
F
H
H
H
H
H
H
Me
F
MeO
NMe2
CF3
CN
0.006
-0.007
0.002
0.85
-0.75
1.31
0.58
-0.94
-0.10
3.09
-3.95
-6.76
-5.60
-10.39
-6.64
-5.32
-10.21
-11.33
MeO
NMe2
CF3
CN
H
H
H
H
H
0.029
3.29
-0.028
-0.042
0.001
-0.022
0.008
0.006
-0.054
-0.071
-3.79
-6.04
-5.01
-11.74
-7.66
-5.81
-9.59
-10.34
1f
1e
H
positive sign for tolanes with electron-donating groups and
negative for ones bearing electron-withdrawing substituents
(Table 2) regardless of their positions. Analogously, the
difference in chemical shifts of the sp-carbons, ∆δ, is defined
as δR - δâ, and its value depends on both the nature and position
of a substituent. Thus, in para-substituted tolanes ∆δ is positive
for electron-donating and negative for electron-withdrawing
groups, whereas in the ortho-series ∆δ is always negative.
Graphical representation of these data (Figure 3) reveals a linear
correlation between ∆δexp and ∆NC in the para-tolane series,
while no trend for ortho-tolanes is obvious (Figure 3).
Having demonstrated that the values of 13C chemical shifts
of acetylenic carbons in ortho-substituted tolanes do not correlate
with the distribution of electron density on the triple bond, we
decided to investigate the influence of anisotropic properties
of ortho-substituents on chemical shifts. To this end, we
analyzed shielding density maps for ortho-methyl (1i) and ortho-
trifluoromethyl (1f) tolanes as seen from R- or â-carbon,
respectively (Figure 4).17 For each probe nucleus,18 blue and
red colors stand for shielding and deshielding effects, respec-
tively, induced by magnetic currents in the corresponding areas
(Figure 4). The intensity of color is proportional to the value
of shielding density; i.e., areas mapped with faded colors have
negligible effect on the probe. This analysis demonstrated that,
as expected, both sp-carbon atoms receive a major shielding
contribution from their own magnetic current surrounding the
triple bond. The detailed examination of this component,
however, is complicated as it would require analysis of minor
shielding density fluctuations on the background of its very high
Investigation of Factors Affecting Pd-Catalyzed Hy-
drostannation of Tolanes. Having addressed the issue of
adequate interpretation of sp-carbon chemical shifts and polar-
ization of triple bond in diarylacetylenes, we turned our attention
to the analysis of factors responsible for exclusive R-regiose-
lectivity of Pd-catalyzed hydrostannation of ortho-tolanes.9 To
verify the assignment of regiochemistry and evaluate both
electronic and steric effects, we performed analogous indepen-
dent investigation,20 in which monosubstituted tolanes 1 were
subjected to the standard conditions for palladium-catalyzed
hydrostannation.9 It was found that, in the para-substituted series
where steric effects are negligible, the regiochemistry of
hydrostannation is entirely governed by polarization of the triple
bond. All para-substituted tolanes afforded mixtures of regioi-
someric vinylstannanes; the substrates bearing electron-
withdrawing groups yielded R-vinylstannanes 2a-c as major
products (Table 3, entries 1-3), whereas those with electron-
donating groups favored predominant formation of â-vinylstan-
nanes (Table 3, entry 4).
(16) Experimental 13C NMR chemical shifts were measured for CDCl3 solutions
(6-10%) of corresponding tolanes at room temperature and were unam-
biguously assigned based on 2D correlation spectroscopy.11
(17) Shielding density maps were obtained from DFT B3LYP/6-31G(d)
computations using GIAO method to generate three cubes for anisotropic
magnetic current densities with magnetization vectors along x, y, and z
axes. Surfaces were built using a cube for isotropic magnetic current density,
which was obtained as (jx + jy + jz)/3. Analogously, shielding density
cubes were obtained for shielding tensor components Fxx, Fyy, and Fzz at
both R- and â-carbon atoms, after which isotropic shielding density cubes
were generated as a trace of the shielding tensor (Fxx + Fyy + Fzz)/3. Finally,
current density surfaces were mapped with shielding densities, using
GaussView 3.09W, Gaussian Inc, Pittsburgh, PA.
(18) Magnetic shielding phenomena may be viewed as a result of interaction
between an induced magnetic field and the permanent magnetic moment
of a given nucleus with a nonzero magnetic moment. Such a nucleus is
referred here and below as a probe nucleus. For discussion, see: Wolinsky,
K. J. Chem. Phys. 1997, 106, 6061.
(19) For example, see: Gunther, H. NMR Spectroscopy: Basic Principles,
Concepts, and Applications in Chemistry, 2nd ed.; Wiley: 1994.
(20) Tolanes 1d, h, i have been previously employed in the Pd-catalyzed
hydrostannation.9
9
10246 J. AM. CHEM. SOC. VOL. 127, NO. 29, 2005