Coupling of azomethine ylides with nitrilium derivatives of closo-decaborate clusters: A synthetic and theoretical study

Article abstract of DOI:10.1002/cplu.201200257

The azomethine ylides p-R₃C₅H₄N+CHCOC ⁶H⁴R²-p (3a: R3=H, R2=H, X=Br; 3b: R3=H, R2=Me, X=I; 3c: R3=H, R2=OMe, X=I; 3d: R3=H, R2=F, X=I; 3e: R3=Me, R2=Me, X=Br) react with the nitrile functio

Full text of DOI:10.1002/cplu.201200257

DOI: 10.1002/cplu.201200257
Coupling of Azomethine Ylides with Nitrilium Derivatives
of closo-Decaborate Clusters: A Synthetic and Theoretical
Study
[
a]
[a]
[b]
[c]
Aleksey L. Mindich, Nadezhda A. Bokach,* Maxim L. Kuznetsov, Matti Haukka,
[
d]
[d]
[a]
[d]
Andrey P. Zhdanov, Konstantin Yu. Zhizhin, Serguei A. Miltsov, Nikolay T. Kuznetsov,
[a, e]
and Vadim Yu. Kukushkin*
3
+
ꢁ
2
3
The azomethine ylides p-R C H N CH COC H R -p (3a: R =H,
bond formation in the reaction of nitrilium derivatives of
boron clusters with any nucleophiles so far tested. Compound-
s 4a–h,k–n were characterized by ICP-MS, high resolution ESI-
5
4
6
4
2
3
2
3
2
R =H, X=Br; 3b: R =H, R =Me, X=I; 3c: R =H, R =OMe,
3
2
3
2
X=I; 3d: R =H, R =F, X =I; 3e: R =Me, R =Me, X=Br) react
with the nitrile functionality of the closo-decaborate clusters
1
13
1
11
1
MS, molar conductivity, and IR, H, C{ H}, and B{ H} NMR
spectroscopy. The structures of 4a and 4b were determined
by a single-crystal X-ray analysis. Theoretical calculations at the
DFT level allowed the interpretation of the observed reaction
selectivity, which is driven mostly by thermodynamic rather
than kinetic factors, and steric repulsion between the boron
cluster and the azomethine ylide molecule plays a crucial role.
The activation of the nitrile group upon binding to the boron
cluster is explained in terms of the electrostatic arguments.
The mechanisms of the nucleophilic addition and the hypo-
thetical cycloaddition between the azomethine ylide and ni-
triles were investigated in detail.
n
1
1
1
1
[
Bu N][B H (NCR )] (1a: R =Me; 1b: R =Et; 1c: R =Ph) in
4 10 9
a CH NO solution under mild conditions (20–258C, 2 min) to
afford selectively products of the nucleophilic addition (ca.
3
2
quantitative yields based on NMR analysis in [D ]DMSO, 71–
6
8
7% yield of isolated products). These products are the bory-
lated enamino ketones as the salts bearing exclusively a tetra-
n
1
+
3
butylammonium
cation
[Bu N][B H {NCR =C(N C H R -
4 10 9 5 4
2
n
4
p)COC H R -p}] (4a–h,k–n) or the mixed salts [Bu N] [p-
6
4
1-x
3
+
2
1
+
3
2
R C H N CH COC H R -p] [B H {NCR =C(N C H R -p)COC H R -
5
4
2
6
4
x
10
9
5
4
6
4
p}] (4i, 4j, and 4o). This reaction represents the first example
of the selective nucleophilic addition of pyridinium azomethine
ylides to the nitrile group and the first example of new CꢁC
Introduction
In the past decade, various polyhedral boranes were the sub-
ject of significant attention, first of all, because of their various
useful properties and important applications such as, for exam-
zolinium systems that functionalizes the closo-decaborate clus-
ter unit.
[
1]
ple in boron neutron-capture cancer therapy. Although some
reactions leading to functionalization of boron clusters are
[
a] A. L. Mindich, N. A. Bokach, S. A. Miltsov, V. Y. Kukushkin
Department of Chemistry, Saint Petersburg State University
Universitetsky Pr. 26, 198504 Stary Petergof (Russian Federation)
Fax: (+7)812-4286939
[
2,3]
known,
the reactivity of the borylated CꢀN functionality still
remains scarcely investigated. However, the data for various
borylated nitrilium species gradually emerging in the literature
disclose a high reactivity of the CꢀN group that is subject to
E-mail: bokach@nb17701.spb.edu
kukushkin@vk2100.spb.edu
[
4–6]
[7–9]
[7,8,10]
facile additions of H O,
R’OH,
or amines
thus provid-
2
[
b] M. L. Kuznetsov
ing an easy route for the functionalization.
Centro de Quꢀmica Estrutural, Complexo I, Instituto Superior Tꢁcnico
Universidade Tꢁcnica de Lisboa
Avenida Rovisco Pais, 1049-001 Lisbon (Portugal)
[
11]
Previously
our research group observed that the CꢀN
moiety bound to the closo-decaborate cluster unit is so reac-
tive toward 1,3-dipolar cycloaddition (DCA) of acyclic Z-nitro-
nes, that the cycloaddition (CA) of these dipoles proceeds rap-
idly and under mild conditions furnishing substituted 2,3-dihy-
[
c] M. Haukka
Department of Chemistry, University of Jyvꢂskylꢂ
P.O. Box 35, 40014 University of Jyvꢂskylꢂ (Finland)
[
d] A. P. Zhdanov, K. Y. Zhizhin, N. T. Kuznetsov
N. S. Kurnakov Institute of General and Inorganic Chemistry
Russian Academy of Sciences
[
11]
dro-1,2,4-oxadiazoles. Following this study, and in the frame-
work of our ongoing project on 1,3-dipolar cycloaddition to
the CꢀN group in nitriles (for reviews see Ref. [12–15] and for
recent studies see Ref. [16–20]) and isocyanides (for a review
see Ref. [12] and for a recent study see Ref. [21]) we attempted
CA between nitrilium decaborates and azomethine ylides (see
Scheme 3 later) having an initial aim to generate novel imida-
3
1 Leninsky Pr., 119991 Moscow (Russian Federation)
[
e] V. Y. Kukushkin
Saint Petersburg State Forest Technical University
Institutskii per. 5, 194021 Saint Petersburg (Russian Federation)
Supporting information, including synthetic details, for this article is
available on the WWW under http://dx.doi.org/10.1002/cplu.201200257.
ChemPlusChem 2012, 77, 1075 – 1086
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
1075

V. Yu. Kukushkin, N. A. Bokach et al.
Our hypothesis on occurrence of DCA (Scheme 1) was based
on the previous reports indicating that pyridinium azomethine
ylides in the vast majority of cases (exceptions are known al-
ing the first example of addition of any C-nucleophiles to nitri-
lium derivatives of boron clusters—is described herein. We
also report our theoretical considerations of plausible mecha-
nisms and driving forces of the coupling between free and
borylated nitriles with azomethine ylides.
[
22–24]
though scarce
) act as rather reactive 1,3-dipoles toward al-
[
25–29]
kenes with electron-withdrawing groups
or al-
(route A) accomplishing, after the oxidative aro-
[
27,28,30–32]
kynes
matization, the indolizine derivatives.
Results and Discussion
Reactivity of 1a–c toward 3a–e
For this study we employed three closo-decaborates [nBu N]-
4
1
1
1
1
[
B H (NCR )] (1a: R =Me; 1b: R =Et; 1c: R =Ph) and five
10 9
3
+
ꢁ
2
3
azomethine ylides, that is, p-R C H N CH COC H R -p (2a: R =
H, R =H, X=Br; 2b: R =H, R =Me, X=I; 2c: R =H, R =
OMe, X=I; 2d: R =H, R =F, X =I; 2e: R =Me, R =Me, X=
5
4
6
4
2
3
2
3
2
3
2
3
2
Br). The latter species were generated in situ by the treatment
3
+
of the corresponding pyridinium salts [p-R C H N
5
4
2
CH COC H R -p](X) with Et N in accord with the known proto-
2
6
4
3
[22]
col.
The reaction between the nitrile functionality in 1a–
c and azomethine ylides 3a–e (in all possible combinations)
Scheme 1. Cycloaddition of the pyridinium-based azomethine ylides to al-
kenes, alkynes (A), and nitriles (B) followed by the oxidative aromatization of
the ring systems.
proceeded rapidly in a solution of CH NO under mild condi-
3
2
tions (RT, 2 min) to afford 4a–o (ca. quantitative yields based
on NMR analysis in [D ]DMSO, 71–87% yield of isolated prod-
6
ucts for 4a–h,k–n; Scheme 3. Compound numbering is given
in Table 1).
Moreover, few examples of DCA of pyridinium azomethine
f
ylides to reactive R CN species were also reported (route B in
Anionic species 4i, 4j, and 4o, upon removal of the solvent
and treatment with methanol, formed the mixed salts
[33–36]
Scheme 1; R=Ar, OAlk; R =CF , C F , C F in Scheme 2).
f
3
3
7
7 15
3
+
2
[
{
nBu N] [p-R C H N CH COC H R -p] [B H -
4 1-x 5 4 2 6 4 x 10 9
1
+
3
2
NCR =C(N C H R -p)COC H R -p}] with varied ratios
5
4
6
4
between the cations depending on the insignificant
differences in the isolation conditions.
The reaction rate is almost independent on the
1
2
3
nature of the substituents R , R , R and it is mostly
determined by the rate of dissolution of the borylat-
ed nitriles. All isolated compounds 4a–h,k–n are
stable at room temperature both in the solid state
Scheme 2. Presumable mechanism of pyridinium ylide/perfluoronitrile interaction.
and in a solution of [D ]DMSO for at least one
6
The cycloaddition to nitriles is highly unselective and the het-
erocycles formed were separated in low yields from a broad
mixture of CA products and unidentified compounds. In only
month.
Monitoring the reactions with ESI-MS and NMR techniques
indicate that under similar reaction conditions (RT, CH NO ) no
3
2
[
34,36]
two studies,
the products derived from the nucleophilic
coupling occurs upon treatment of the electron-deficient ni-
triles CCl CN, MeC(O)CN, and CHCl CN with the most reactive
[
33–36]
addition were identified, whereas in other cases
CA prod-
3
2
ucts were isolated. Ten reactions between pyridinium azome-
thine ylides and the nitrile functionality were reported and
they all gave DCA products. However, in four cases, some by-
products derived from the nucleophilic addition were isolated
along with DCA adducts; no single example of the selective
generation of compounds originating from the nucleophilic
azomethine ylide (2c) in the presence of Et N over a period of
3
three weeks. These data suggest that the CꢀN group in the
boron clusters is substantially better activated toward the nu-
cleophilic addition as compared to the CꢀN group even in
f
electron-deficient nitriles such as R CN.
f
Unlike the reaction of R CN, where DCA was observed (see
[
36]
addition was described.
Based on these observations, the
Introduction), pyridinium azomethine ylides undergo nucleo-
stepwise mechanism that includes the nucleophilic addition
followed by the cyclization (but not vice versa) was suggested
philic addition without cyclization, (Scheme 4, compound a:
[36]
Y=CF , C F ; A=NH), to such C-electrophiles as CS in pres-
3
3
7
2
[
36]
(Scheme 2).
ence of alkylating agent (Scheme 4, compound b: Y=S-Alk;
[
37]
In contrast to our expectations, we observed that the treat-
A=S),
acyl chlorides (Scheme 4, compound c: Y=Ar; A=
isocyanates (Scheme 4, compound d: Y=NHR; A=
[
38,39]
39]
ment of the nitrilium borates with azomethine ylides proceeds
highly selectively as a nucleophilic addition, rather than DCA,
leading to formation of a new carbon–carbon bond and fur-
nishing enamino ketones. The observed reaction—represent-
O),
O),
[
isothiocyanates (Scheme 4, compound e: Y=NHR; A=
[
40]
[41,42]
S),
aldehydes (Scheme 4, compound f),
quinolone
[43]
N-oxide (Scheme 4, compound g),
2,4,6-trinitrochloroben-
1
076
www.chempluschem.org
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemPlusChem 2012, 77, 1075 – 1086

Reaction Mechanisms
Characterization of the bory-
lated enamino ketones
All isolated species 4a–h,k–n
were characterized by ICP and
+
/ꢁ
high-resolution ESI
spectrometry, molar conductivi-
mass
1
13
1
11
1
ty, IR, H, C{ H}, and B{ H}
NMR spectroscopy, and X-ray
crystallographic analysis (for 4a
and 4b). All physicochemical
data are in agreement with the
proposed
formulae.
Com-
pounds 4i,j,o were character-
ized by high-resolution ESI-MS
1
11
1
and H and B{ H} NMR meth-
ods.
The closo-decaborate deriva-
tives are hardly combustible to
achieve reproducible CHN anal-
yses, therefore all synthesized
compounds 4a–h,k–n
were
Scheme 3. Synthetic transformations.
characterized by ICP-mass spec-
trometry (B%). The obtained
compounds give satisfactory
ICP-MS-based elemental analysis (see the Supporting Informa-
ꢁ
Table 1. Compound numbering for the borylated enamino ketones.
tion). In the ESI mass spectra of 4a–o, the most intensive sig-
ꢁ
nals correspond to [A] and the less intensive ones correspond
1
2
3
1
2
3
Compd
R
R
R
Compd
R
R
R
ꢁ
ꢁ
+
to [2A+Q] (where A is the anion and Q is the pyridinium
4
4
4
4
4
4
4
4
a
b
c
d
e
f
Me
Me
Me
Me
Me
Et
H
Me
F
OMe
Me
H
Me
F
H
H
H
H
Me
H
H
4i
4j
Et
Et
OMe
Me
H
Me
F
H
Me
H
H
H
+
counterion), whereas the ESI mass spectra of 4a–h,k–n
exhibit intensive signals corresponding to [Q] and less inten-
+
4k
4l
4m
4n
4o
Ph
Ph
Ph
Ph
Ph
+
sive ones corresponding to [2Q+A] . For all spectra, the iso-
topic patterns agree well with the calculated ones. The
OMe
Me
H
Me
L values of 4a–h,k–n measured solutions of acetonitrile
with exact concentration in the range of 2–6ꢁ10 molL are
M
g
h
Et
Et
ꢁ
4
ꢁ1
H
ꢁ
+
characteristic for ionic species of the [A] [Q] type in this sol-
ꢁ
1
2
ꢁ1
vent (measured values: L =119–127 W cm mol , typical
M
[48]
ꢁ1
2
ꢁ1
range: L =120–160 W cm mol ).
M
In the IR spectra (see the Supporting Information) of 4a–
h,k–n, the most intensive signals are strong bands correspond-
ing to n(BꢁH) vibrations observed in the range of 2451–
ꢁ
1
2
515 cm , meanwhile strong bands corresponding to n(C=O)
of the cationic part of the molecule were observed in the
ꢁ
1
region of 1681–1703 cm . The most characteristic band of
ꢁ
1
each 4a–h,k–n falls in the range 1335–1340 cm and it corre-
sponds to the vibrations of the conjugated system NH-C=C-
Scheme 4. Products derived from the nucleophilic addition of pyridinium-
based azomethine ylides to C-electrophiles.
ꢁ
1
C=O. No n(CꢀN) bands at approximately 2300 cm , specific
[4]
for the starting nitrilium salts, were observed.
1
The specific feature of the H NMR spectra (see the Support-
ing Information) of 4a–o is the presence of a singlet at
[
44]
13
1
zene (Scheme 4, compound h), and several alkenes (substitu-
d=11.42–11.13 ppm from the NH···O group. In the C{ H} NMR
spectra of 4a–h,k–n, the most characteristic signals are the
C=O resonance of the anion that fall in the interval of d=
180.6–182.4 ppm and C=C-N peak in the range d=165.0–
[
45–47]
tion at the Csp
2
atom).
Hence, the reaction reported here
represents the first example of the selective nucleophilic addi-
tion of pyridinium azomethine ylides to the nitrile group and
the first example of new carbon–carbon bond formation in the
reaction of nitrilium derivatives of boron clusters with any nu-
cleophiles so far tested.
11
1
171.8 ppm. The B{ H} NMR confirmed the closo-decaborane
structure of 4a–o. Two singlets from the apical boron atoms
10
1
B
and B were observed in the range d=0.5–1.0 ppm and
ChemPlusChem 2012, 77, 1075 – 1086
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chempluschem.org
1077

V. Yu. Kukushkin, N. A. Bokach et al.
from d=ꢁ0.6 to ꢁ1.0 ppm, respectively, the weak broad sin-
lengths and angles are very close to similar systems observed
2
[52–54]
glet of the substituted boron atom B was found between d=
previously.
Owing to the hydrogen bonding N1ꢁH···O1
ꢁ
12.6 and ꢁ13.7 ppm, and the other equatorial atoms
and the planarity of the H-N1-C1-C3-C9-O1 system, the aromat-
ic rings are sterically hindered. The torsion angle O1-C9-C10-
C11 is 58.88 for 4a and 51.48 for 4b and, consequently, the
carbonyl group is extensively turned away from conjugation
with the phenyl ring. This observation obviously explains the
emerged as a multiple signal in the intervals between d=
3
5
6
9
ꢁ
23.7 and ꢁ24.2 ppm (B , B , B , B ), and d=ꢁ26.0 and
27.5 ppm (B , B , B ). All signals of the unsubstituted boron
4
7
8
ꢁ
1
atoms split into doublets without the broadband H decou-
2
2
1
pling, whereas the substituted B atom still gives the singlet
unconventional chemical shifts of CO-C H R -p in the H NMR
6 4
signal.
spectra of the anionic part of 4a–o. The pyridinium ring is
turned at 69.98 for 4a and 69.28 for 4b with respect to the
plane C1-C3-C9 and does not conjugate with the system H-N1-
C1-C3-C9-O1. All other bonds and angles have expected
values.
The single-crystal X-ray analysis of 4a and 4b (Figure 1 and
Figure S1 in the Supporting Information) disclose the presence
of two independent ionic parts. All bond lengths and angles of
Theoretical considerations and reaction mechanism
With an aim to interpret and explain the experimental observa-
tions, quantum chemical calculations of the plausible reactions
mechanisms of both cycloaddition and nucleophilic addition of
ꢁ
free and borylated nitriles [NꢀCMe and [B H (NꢀCMe)] (I)]
10
9
+
ꢁ
with the azomethine ylide C H N CH C(O)Me (II) have been
5
5
performed at the DFT level. Two main questions were raised,
that is, 1) why are the borylated nitriles much more active
toward the azomethine ylides than free NꢀCR and 2) why
does the reaction between borylated nitriles and azomethine
ylides affords the nucleophilic addition product instead of a cy-
cloadduct?
Figure 1. ORTEP view of 4a with the atom numbering. The thermal ellip-
soids are drawn at the 30% probability level.
the cationic part are the same, within 3s, as those described in
[
49]
the literature. The anionic parts of 4a and 4b consist of the
closo-decaborate cluster bound to the organic fragment
through the N1 atom. In the cluster part, the BꢁB bond distan-
ces and angles are typical for 2-substituted nonahydro-closo-
Cycloaddition of II to nitriles
Reaction mechanism
For this reaction, there are two regioisomeric pathways leading
to 5-acyl (III1, IV1) and 2-acyl (III2, IV2) 2,5-dihydro-1H-imida-
zoles (Scheme 5). Within each regioisomeric pathway, two ste-
reoisomers may be formed [III1a(b), III2a(b), IV1a(b), and
IV2a(b). Finally, the formation of two conformers with different
position of the C(O)Me substituent relative to the heterocycle
is possible for each stereoisomer except III2a for which only
one conformer III2a2 was found.
For CA of II to NꢀCMe or I, transition states of a concerted
mechanism were located for each reaction pathway except for
that leading to IV1a2 (Figure 2 and Figure S2). At the same
time, isomer IV1a2 can be easily formed from IV1a1 upon ro-
tation around the carbon–carbon bond (Scheme 5). Besides,
two stepwise pathways were found for CA of II to I involving
the generation of acyclic intermediates INTIV1a1 or INTIV2a2
and leading to the CA products IV1a1 and IV2a2, respectively
[
4,50]
decaboron clusters.
The bond lengths N1ꢁB2 are equal,
[50]
within 3s, to those in the starting borylated nitrile. The ge-
ometry of the anionic part of the molecule is determined by
hydrogen bonding, N1ꢁH···O1, giving a slightly distorted
planar six-membered system H-N1-C1-C3-C9-O1. The bond
lengths N1ꢁC1, C1ꢁC3, C3ꢁC9, and C9ꢁO1 (see Table 2) of this
system are intermediate between the reported values for the
[
51]
corresponding standard single and double bonds. The bond
Table 2. Selected bond lengths [ꢂ] and dihedral angles [8] for 4a and
4
b.
Bond lengths
Compounds
Angles
Compounds
4
a
4b
4a
4b
(
Scheme 6). The second steps (ring closure) are rate determin-
N1ꢁB2
N1ꢁC1
C1ꢁC3
C3ꢁC9
O1ꢁC9
C1ꢁC2
N2ꢁC3
C9ꢁC10
1.522(2) 1.523(4) C1-N1-B2
1.311(2) 1.315(3) N1-C1-C3
1.420(2) 1.412(4) C1-C3-C9
1.408(3) 1.406(4) C3-C9-O1
1.256(2) 1.262(3) N1-C1-C2
1.492(3) 1.494(4) C3-C1-C2
1.452(2) 1.463(3) C1-C3-N2
1.502(2) 1.492(4) C9-C3-N2
130.60(16) 131.5(2)
119.96(16) 120.9(2)
124.91(16) 125.6(2)
122.36(16) 121.0(3)
118.63(16) 117.4(3)
121.37(16) 121.6(2)
117.33(15) 117.1(2)
117.73(15) 117.2(2)
ing. No stepwise route exists for the reaction of II with
NꢀCMe.
Kinetic factors
¼6
The calculated activation energy (in terms of DGs ) is lowest
for the pathway that affords isomers III1b1 and IV1b1
(
C3-C9-C10 120.72(16) 121.3(2)
O1-C9-C10 116.92(16) 117.7(3)
Table 3). For the reaction between NꢀCMe and II, the corre-
1
078
www.chempluschem.org
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim ChemPlusChem 2012, 77, 1075 – 1086

Reaction Mechanisms
Scheme 5. 1,3-Dipolar cycloaddition of azomethine ylide II to NꢀCMe and [B10
H
9
(NꢀCMe)]ꢁ (I; transition states of the concerted mechanisms are indicated).
¼
6
ꢁ1
sponding DG value is 33.9 kcalmol . This value is higher
Thermodynamic factors
s
¼6
than DGs for the reactions of NꢀCMe with the simple acyclic
azomethine ylide CH =N (Me)CH (19.0 kcalmol ) and with
+
ꢁ
2
ꢁ1
The calculations demonstrate that isomers III1b2 and IV1b2
are the most thermodynamically stable among the others.
They can be easily formed from the most kinetically accessible
isomers III1b1 and IV1b1 upon rotation around the carbon–
carbon bond. At the same time, the energies of all other iso-
2
+
ꢁ
ꢁ1
the nitrone CH =N (Me)O (31.5 kcalmol ) in a solution of
2
[
55–58]
CH Cl .
Taking into account that NꢀCMe does not react
2
2
[
59,60]
with acyclic nitrones, its reaction with II is also not expect-
ed, in accordance with the experimental observations (see
above).
ꢁ
1
mers are higher only by 0.7–4.3 kcalmol (Table 3). The Gibbs
free energies of the formation of all isomers III and IV in
a CH NO solution (DG ) are significantly positive (9.6–15.2 kcal
The lowest activation barrier to CA of II to I is 21.1 kcal
mol . This is 12.8 kcalmol lower than the DGs value of the
ꢁ
1
ꢁ1
¼6
3
2
s
ꢁ
1
boron-free reaction and corresponds to an enhancement of
mol ). Thus, CA of II to both NꢀCMe and I is not favorable
thermodynamically, and generation of CA products upon the
reaction of II with I is not expected despite the comparatively
low activation barrier of this process, in agreement with the
experimental data (see above).
9
the reaction rate by a factor of 2.5ꢁ10 ! Thus, the binding of
the NꢀC group to the B atom results in a dramatic activation
of nitriles toward CA with azomethine ylides, and the reaction
II+I is quite favorable kinetically. The second regioisomeric
pathway (i.e. that yielding isomers III2 and IV2) is clearly less
favorable.
Nucleophilic addition of II to nitriles
Synchronicity
Concerted mechanism
Another important characteristic of the concerted mechanism
is the degree of its synchronicity. The most reliable criterion of
the synchronicity is based on the estimate of parameter Sy,
which may vary from 0 (stepwise mechanism) to 1 (completely
synchronous CA; see computational details in the Experimental
For the nucleophilic addition reactions (NAs) to nitriles, three
general types of mechanism may be considered, that is, con-
certed, associative, and dissociative. The first one (Scheme 7)
includes the formation of a new carbon–carbon bond and the
hydrogen transfer from the azomethine carbon atom to the ni-
trile nitrogen in a single step via one cyclic transition state.
This transition state may be either four-membered (type a) or
six-membered (type b). In the latter case, a solvent molecule
participates in the formation of TS and plays the role of the hy-
drogen-transfer promoter. Taking into account that the solvent
used in the experimental part (nitromethane) was not dried
Section). The S value of CA II+NꢀCMe is 0.70 (the III1b1
y
pathway) indicating that the mechanism is asynchronous by
ꢁ
3
0%. The coordination of NꢀCMe to [B H ] leads to a signifi-
1
0
9
cant increase of the asynchronicity to S =0.54 with the forma-
y
tion of the carbon–carbon bond preceding the carbon–nitro-
gen bond making. Such an increase of the asynchronicity is ac-
counted for by significant steric repulsions between the pyri-
dine ring and the boron cluster in TSIV1b1 (Figure 2).
and, hence, contains some water, namely H O was considered
2
as the hydrogen-transfer promoter in TSs of type b (however,
the bulky solvent effect was calculated at the CPCM level for
CH NO as solvent, see computational details in the Experimen-
3
2
tal Section).
ChemPlusChem 2012, 77, 1075 – 1086
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chempluschem.org
1079

V. Yu. Kukushkin, N. A. Bokach et al.
via the four-membered transi-
tion states are significantly
higher than those via the six-
membered TSs (Table 3). This
finding is not surprising consid-
ering the fact that six-mem-
bered structures are usually
more stable than four-mem-
bered ones. The barrier of the
nucleophilic addition II+Nꢀ
CMe via TSV6mem is 45.9 kcal
ꢁ
1
mol , which is too high for an
effective realization of this
route. The activation energy of
the reaction II+I via TSVI6mem
ꢁ
1
is 24.3 kcalmol , which is quite
acceptable for occurrence of
the reaction. However, this
value is higher than the barrier
found for the CA route between
ꢁ
1
II and I (21.1 kcalmol ). Hence,
NA via a concerted mechanism
should be less favorable than
CA, thus contradicting the ex-
perimental findings.
Associative mechanism
The associative mechanism of
NA to nitriles (Scheme 8) in-
volves the addition of the nu-
cleophile to the nitrile carbon
atom and subsequent two-step
proton transfer via either
route A or route B. As was men-
tioned above, no acyclic inter-
mediate was found for the reac-
tion of II with NꢀCMe. Howev-
er, in the case of NA of II to I,
Figure 2. Equilibrium geometries of selected transition states and intermediates.
two intermediates INTIV1a1
and INTIV2a2 were located. In
fact, these species represent the acyclic intermediates of the
stepwise cycloaddition discussed above (Scheme 6). The activa-
Transition states of both types a and
b (TSV4mem,
TSV6mem, TSVI4mem, and TSVI6mem) were found for con-
certed NAs (Figure 2). The activation barriers of the reactions
ꢁ
1
tion energy for the formation of INTIV1a1 (19.6 kcalmol ) is
significantly lower than that of INTIV2a2 (Table 3).
Scheme 6. Stepwise mechanisms of cycloaddition of II to I.
1
080
www.chempluschem.org
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemPlusChem 2012, 77, 1075 – 1086

Reaction Mechanisms
tion of INTIV1a1 should occur first, and the calculations dem-
Table 3. Calculated Gibbs free energies of activation and reaction in a so-
+
+
onstrate that the process INTIV1a1+Me NH !INTIV1a1H +
3
lution of CH
3 2
NO .
ꢁ
1
+
Me N is exoergonic by 14.0 kcalmol (Me NH is formed
3
3
s6
¼
Reaction
DG
[
DG
s
[kcalmol ]
upon the in situ generation of the azomethine ylide; see the
Experimetal Section). Thus, the protonation of INTIV1a1 is
much more favorable compared to its deprotonation.
ꢁ
1
ꢁ1
kcalmol
]
NꢀCMe+II!III1a1 via TSIII1a1
NꢀCMe+II!III1a2 via TSIII1a2
NꢀCMe+II!III1b1 via TSIII1b1
NꢀCMe+II!III1b2 via TSIII1b2
NꢀCMe+II!III2a2 via TSIII2a2
NꢀCMe+II!III2b1 via TSIII2b1
NꢀCMe+II!III2b2 via TSIII2b2
I+II!IV1a1 via TSIV1a1
36.1
40.6
33.9
36.3
45.8
41.3
46.7
21.5
–
21.1
24.9
39.0
47.0
38.3
52.1
19.6
16.1
45.5
3.1
68.0
45.9
44.3
24.3
–
–
–
–
–
15.2
14.3
14.3
12.8
15.1
13.8
14.3
13.8
12.4
13.4
9.6
13.7
11.4
10.5
10.2
11.6
2.3
46.6
ꢁ35.3
14.4
14.4
ꢁ3.6
ꢁ3.6
54.9
29.3
15.9
ꢁ14.0
ꢁ1.2
The proton elimination from the former ylide fragment in
+
INTIV1a1H leads to the final experimentally isolated NA
ꢁ
1
product V and this step is also exoergonic by 1.2 kcalmol .
Note that the ring closure in INTIV1a1 to give IV1a1
(
Scheme 6) requires overcoming an activation barrier of
ꢁ
1
1
6.1 kcalmol .
I+II!IV1a2
I+II!IV1b1 via TSIV1b1
The overall activation barrier of NA I+II!V via the associa-
ꢁ1
I+II!IV1b2 via TSIV1b2
tive mechanism (route A) is 19.6 kcalmol , the formation of
INTIV1a1 is the rate limiting step (Figure 3). This value is lower
than the activation energies of both concerted NA (24.3 kcal
I+II!IV2a1 via TSIV2a1
I+II!IV2a2 via TSIV2a2
I+II!IV2b1 via TSIV2b1
ꢁ
1
ꢁ1
I+II!IV2b2 via TSIV2b2
mol ) and CA (21.1 kcalmol ), thus, correlating with the ex-
perimental formation of the NA product instead of the CA one.
However, the main driving force determining the direction of
the reaction of II with I has a thermodynamic rather than a ki-
I+II!INTIV1a1 via TS1IV1a1
INTIV1a1!IV1a1 via TS2IV1a1
I+II!INTIV2a2 via TS1IV2a2
INTIV2a2!IV2a2 via TS2IV2a2
NꢀCMe+II!V via TSV4mem
NꢀCMe+II!V via TSV6mem
I+II!V via TSVI4mem
netic nature. Indeed, the DG value of NA II+I!V is negative
s
ꢁ
1
(
ꢁ3.6 kcalmol ), whereas the reaction energy of CA II+I!
ꢁ1
IV1b2 is clearly positive (9.6 kcalmol ).
I+II!V via TSVI6mem
ꢁ
+
Besides, the dissociative mechanism of NA of II to nitriles
was also considered but found to be highly unfavorable (see
the Supporting Information for details).
INTIV1a1+H
2
O!INTIV1a1-H +H
3
O
ꢁ
+
INTIV1a1+Me
3
N!INTIV1a1-H +Me
3
NH
ꢁ
+
INTIV1a1+II!INTIV1a1-H +II
+
+
INTIV1a1+Me
3
NH !INTIV1a1H +Me
3
N
+
+
INTIV1a1H +Me
3
N!V+Me
3
NH
Factors controlling the reactivity of the borylated and the
boron-free nitriles toward II
In accord with route A (Scheme 8), intermediate INTIV1a1
should undergo deprotonation. The reactions INTIV1a1+B!
Steric factor
ꢁ
+
ꢁ
INTIV1a1-H +BH are still endoergonic by 54.9, 29.3, and
The presence of a bulky substituent such as [B H ] at the ni-
10
9
ꢁ
1
1
5.9 kcalmol , for B=H O, Me N, and II, respectively. However,
trile nitrogen atom provides a strong steric repulsion between
the boron cluster and the approaching ylide upon the reaction
of I with II. This steric effect dramatically increases the asyn-
2
3
the azomethine CꢁH bond in INTIV1a1 is significantly more
acidic than that in II. Within route B (Scheme 8), the protona-
Scheme 7. Concerted mechanism of nucleophilic addition of II to nitriles.
ChemPlusChem 2012, 77, 1075 – 1086
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chempluschem.org
1081

V. Yu. Kukushkin, N. A. Bokach et al.
Scheme 8. Associative mechanism of nucleophilic addition of II to I.
electron demand processes with the predominant
HOMOdipole–LUMOdipolarophile interaction), group III (in-
verse electron demand reactions with the prevailing
HOMOdipolarophile–LUMOdipole interaction), and group II
(
“neutral” processes with both HOMO–LUMO interac-
tions efficiently operating).
The analysis of the FMO energies and composition
indicates that CA of II to NꢀCMe is clearly a normal
electron demand process (Figure 4). Thus, for
a given dipole, the reactivity of nitrile should
depend on its LUMO energy.
Several highest occupied MOs of I are localized on
the B H fragment while the first occupied MO cen-
10
9
tered on the CꢀN bond [p(CꢀN)] is HOMO-16. The
first unoccupied MO corresponding to p*(CꢀN) is
the LUMO (Figure 5). The binding of NꢀCMe with
Figure 3. Energy profiles of some reaction pathways: stepwise nucleophilic additions in
green (the most favorable route) and magenta (less favorable route), concerted nucleo-
philic addition in red, and cycloadditions in blue].
ꢁ
the B atom at [B H ] results in an increase of the
10
9
energies of both p(CꢀN) and p*(CꢀN) because of ac-
quired overall negative charge of I. As a result, both
HOMOylide–LUMOI and HOMO-16 –LUMO
gaps
I
ylide
chronicity of the process, thus hampering the carbon–nitrogen
bond formation and facilitating creation of the carbon–carbon
bond. As a result, the generation of the acyclic intermediate
INTIV1a1 further converting to NA product V becomes possi-
ble and favorable for the borylated nitrile; although no species
of such type can be formed in the case of acetonitrile. More-
over, the steric repulsion upon the formation of the carbon–ni-
trogen bond makes CA products IV thermodynamically unsta-
ble compared to NA product V.
become similar (6.85 and 6.58 eV, respectively) and CA of II to I
belongs to group II. It is interesting that the smallest
HOMO(CꢀN)–LUMO gap increases ongoing from NꢀCMe
(5.84 eV) to I (6.58 eV; Figure 4), and a lower reactivity of I is
expected compared to NꢀCMe. This finding contradicts the ex-
perimental data and the computational results discussed
above and, thus, demonstrates that the FMO theory fails to
predict correctly the reactivity of the systems under study.
Effective atomic charges
Molecular orbital composition and energies
Another factor which controls the reactivity of the nitriles
toward II is the effective atomic charge on the reacting atoms.
Taking into account that CA of II to I is asynchronous with the
prior formation of the carbon–carbon bond, namely the
charge on the nitrile C atom is crucial for both CA and NA
pathways. The calculations indicate that the NBO atomic
In terms of the frontier molecular orbital (FMO) theory, the NA
reactions are controlled by the interaction of the HOMO of the
nucleophile and of the LUMO of the substrate. For the CA reac-
tions, three main groups may be distinguished in accord with
[
61,62]
the classification by Sustmann,
that is, group I (normal
1
082
www.chempluschem.org
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemPlusChem 2012, 77, 1075 – 1086

Reaction Mechanisms
[1,2,63]
neutron capture therapy.
This reaction also represents the
first example of the nucleophilic addition of phenacyl pyridini-
um-based azomethine ylides to the nitrile functionality.
To explain the selectivity of the reaction (nucleophilic addi-
tion vs. cycloaddition), we studied the reactivity of the free
ꢁ
and the borylated nitriles, NꢀCMe and [B H (NꢀCMe)] (I),
10
9
+
ꢁ
toward the azomethine ylide C H N CH C(O)Me (II) by theo-
5
5
retical (DFT) methods. The calculations indicate that DCA of II
either to NꢀCMe or to I is clearly thermodynamically unfavor-
ꢁ
1
able (DG = +12.8 and +9.6 kcalmol , respectively). However,
s
ꢁ
1
different activation barriers (33.9 vs. 21.1 kcalmol , corre-
spondingly) indicate the significant kinetic activation of
ꢁ
NꢀCMe upon its binding to [B H ] . Such activation can be
10
9
accounted for in terms of electrostatic arguments as a result of
an increase of the positive atomic charge on the C nitrile atom
ongoing from NꢀCMe to I, while the FMO theory fails to ex-
plain this effect.
Figure 4. Relative energies of the interacting HOMOs and LUMOs of reac-
tants (the HOMO–LUMO gaps are indicated in eV).
The nucleophilic addition of II to I is exoergonic by
3.6 kcalmol , with slightly lower activation barrier compared
ꢁ
1
ꢁ
to the CA pathway (19.6 kcal
ꢁ
1
mol ). Thus, the experimentally
observed formation of the nu-
cleophilic addition product in-
stead of CA adduct can be ra-
tionalized by the thermodynam-
ic factors rather than by the ki-
netic arguments. The lower
thermodynamic stability of CA
adduct compared to NA prod-
Figure 5. Plots of selected frontier molecular orbitals.
uct is explained by the steric re-
ꢁ
charge on the nitrile C atom dramatically increases upon bind-
pulsion between the bulky [B H ] cluster and the ylide mole-
10
9
ꢁ
ing of NꢀCMe with [B H ] (from 0.28 e to 0.45 e). Such an
cule. The mechanism of NA of II to I is associative and includes
the generation of acyclic intermediate INTIV1a1, the protona-
tion of the nitrile N atom leading to INTIV1a1H , and deproto-
10
9
enhancement should facilitate the attack of a nucleophile (or
of a nucleophilic center of 1,3-dipole) at the nitrile C atom.
Thus, the great activation of the nitriles upon borylation may
be explained in terms of electrostatic arguments. It is also in-
teresting that the charge on the nitrile N atom becomes less
negative (from ꢁ0.33 e in NꢀCMe to ꢁ0.24 e in I). Therefore,
the borylation results in a release of the electron density from
the entire CꢀN group of nitriles.
+
nation of the C atom to finally afford V.
The ability to isolate enamino ketones from the reactions
between the closo-decaborate clusters and the azomethine
ylides gives us the opportunity to prepare new classes of
boron-containing bidentate ligands for use in the areas of co-
ordination and organometallic chemistry. The complexation of
these chelates with metal centers (for recent studies see
Refs. [64–68]) generating complexes functionalized with the
closo-decaborate moiety should open up yet another method
for facile derivatization of the boron clusters and research in
this direction is currently underway by our group.
Conclusion
We observed the novel reaction between the closo-decaborate
1
clusters [nBu N][B H (NCR )] and the azomethine ylides, which
4
10
9
selectively affords products of the nucleophilic addition, that
1
+
3
is, the borylated enamino ketones [B H {NCR =C(N C H R -
10
9
5
4
2
ꢁ
Experimental Section
p)COC H R -p}] , and no species derived from DCA was detect-
6
4
ed. This coupling represents the first example of new carbon–
carbon bond making in the reaction of nitrilium derivatives of
boron clusters with any nucleophile so far tested. The forma-
tion of new carbon–carbon bonds is of central importance in
organic chemistry as skeleton-forming processes. In main
group element chemistry, the studied reaction provides an
easy access to modification of boron clusters under mild condi-
tions and this is useful for, for example medicinal chemistry
and for syntheses of libraries of potential agents for the boron
Instrumentation and Materials
The reagents RC H COMe (R=H, F: Acros Organics; Me, OMe: Alfa
6
4
Aesar), pyridine, 4-picoline, and the employed solvents were ob-
tained from commercial sources and used as received. closo-Deca-
[69]
[11]
[6]
borate clusters 1a, 1b, 1c, and 1d were prepared in accord
with the published methods. Pyridinium salts 2b, 2b, and 2c were
synthesized by reaction of the substituted acetophenone with pyri-
[70]
dine in the presence of iodine by using known protocols, 2a
and 2d were prepared by reaction of the corresponding a-bro-
ChemPlusChem 2012, 77, 1075 – 1086
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chempluschem.org
1083

V. Yu. Kukushkin, N. A. Bokach et al.
[
42]
moacetophenone with pyridine or 4-picoline. RC H COCH Br (R=
6
4
2
[71]
Table 4. Crystal Data.
H, Me) were obtained accordingly to known protocols.
4
a
4b
Elemental analysis for boron was performed by the FSUE IREA 291
Center (Moscow) on a iCAP 6300 Duo ICP spectrometer using In as
formula
Mr
T (K)
C H B N O
683.87
100(2)
0.71073
Triclinic
P1
3
5
49 10
3
4
C H B N O
4
37 53 10
3
2
92 an internal standard. Infrared spectra were recorded on a Shi-
711.92
120(2)
0.71073
Triclinic
P1˜
8.3414(15)
14.338(3)
16.561(3)
93.244(6)
96.105(5)
100.330(6)
1931.8(6)
2
1
13
1
11
1
madzu FTIR 8400S instrument in KBr pellets. H, C{ H}, and B{ H}
NMR spectra were measured on a Bruker-DPX 300 spectrometer at
ambient temperature; BF ·Et O was used as the external standard
l(ꢂ)
crystal system
space group
a [ꢃ]
b [ꢃ]
c [ꢃ]
3
2
˜
11
1
for B{ H} NMR analysis. Electrospray ionization mass spectra were
obtained on a Bruker micrOTOF spectrometer equipped with elec-
trospray ionization (ESI) source and MeOH or MeCN were used as
the solvents. The instrument was operated both at positive and
negative ion modes using a m/z range of 50–3000. In the isotopic
pattern, the most abundant peak is reported. Molar conductivities
8.3048(10)
14.3736(17)
17.191(2)
109.546(6)
95.222(7)
100.532(7)
1875.4(4)
2
1.211
0.073
47852
a [deg]
b [deg]
g [deg]
ꢁ3-
ꢁ4
3
of 10 –10 m solutions in acetonitrile were measured on a Mettler
Toledo conductometer FE30 using Inlab 710 sensor. Melting points
were determined in a capillary on a Bꢃchi Melting Point 530 Appa-
ratus.
V [ꢂ ]
Z
ꢁ
3
1calc [Mgcm
]
1.224
0.074
24572
ꢁ
1
m(MoKa) [mm
No. reflns.
]
unique reflns.
GOOF (F )
9344
1.024
6779
0.901
2
R
R
wR
int
a]
0.0534
0.0588
0.0994
0.0824
0.0667
0.1457
X-ray diffraction studies
[
1
(Iꢂ2s)
(Iꢂ2s)
[
b]
The crystals of complexes 4a and 4b were immersed in cryo-oil,
mounted in a Nylon loop, and measured at a temperature of 100 K
or 120 K. The X-ray diffraction data were collected on a Bruker
Smart Apex II or Bruker Kappa Apex II Duo diffractometer using
MoKa radiation (l=0.71073 ꢂ). The APEX2 software package was
used for cell refinements and data reductions. The structures were
2
2
2
2
2 2 1/2
[
a] R
1
=Sj jF
o
j-jF
c
j j/SjF
o
j. [b] wR
2
=[S[w(F
o
ꢁF
c o
) ]/S[w(F ) ]] .
[
72]
The Hessian matrix was calculated analytically for the optimized
structures to prove the location of correct minima (no imaginary
frequencies) or saddle points (only one imaginary frequency), and
to estimate the thermodynamic parameters, the latter being calcu-
lated at 258C. The nature of all transition states was investigated
by the analysis of vectors associated with the imaginary frequency
[
73]
solved by direct methods using the SHELXS-97 program with
[74]
the Olex2 graphical user interface. A semiempirical absorption
[75]
correction (SADABS) was applied to all data. Structural refine-
[73]
ments were carried out using SHELXL-97. The crystal of 4b was
diffracting only weakly. The butyl propionate of crystallization was
heavily disordered and therefore omitted from the final structure
model. The contribution of the disordered solvent to the calculated
structure factors was taken into account by using the solvent mask
and by the calculations of the intrinsic reaction coordinates (IRC)
using the Gonzalez-Schlegel method.
[79–81]
Sometimes, the IRC cal-
culations failed due to low imaginary frequency of a transition
state. In these cases, the analysis of the TS nature was performed
in accord with the following procedure. First, the atoms of TS were
shifted from the equilibrium positions along the vectors corre-
sponding to the imaginary frequency, in both directions. Then, ge-
ometry optimization with small size of an optimization step was
carried out.
[74]
routine of Olex2. The missing solvent was also taken into ac-
count in the unit cell content. In structure 4a the NH hydrogen
atom was located from the difference Fourier and refined isotropi-
cally. All other hydrogen atoms were positioned geometrically and
constrained to ride on their parent atoms, with CꢁH=0.95–0.99 ꢂ,
NꢁH=0.88 ꢂ, BꢁH=1.12 ꢂ, and U =1.2–1.5 U (parent atom).
iso
eq
Total energies corrected for solvent effects (E ) were estimated at
The crystallographic details are summarized in Table 4.
CCDC 902348 (4a) and 902349 (4b) contain the supplementary
crystallographic data for this paper. These data can be obtained
free of charge from The Cambridge Crystallographic Data Centre
via www.ccdc.cam.ac.uk/data_request/cif.
s
the single-point calculations on the basis of gas-phase geometries
at the CPCM-B3LYP/6–311+G(d,p)//gas-B3LYP/6–31G(d) level of
theory using the polarizable continuum model in the CPCM ver-
[
82,83]
sion
with CH NO as solvent. The UAKS model was applied for
3 2
the molecular cavity. The entropic term in solution (S ) was calcu-
lated according to the procedure described by Wertz
Cooper and Ziegler using [Eqs. (1)–(4)]
s
[84]
and
85
Computational details
The full geometry optimization of all structures and transition
states has been carried out at the DFT/HF hybrid level of theory
using Becke’s three-parameter hybrid exchange functional in com-
bination with the gradient-corrected correlation functional of Lee,
s
DS ¼ RlnV =Vm;gas
ð1Þ
ð2Þ
1
m;liq
[
76,77]
[78]
Yang and Parr (B3LYP)
with the help of the Gaussian-03 pro-
gram package. No symmetry operations have been applied. The
geometry optimization was carried out using the 6–31G(d) basis
set followed by the single-point calculations with the 6–311+
o
m
s
DS ¼ RlnV =V
2
m;liq
[55–58]
G(d,p) basis set. As was shown previously,
this approach is suf-
ꢀ
ꢁ
o;s
liq
o;s
gas
s
ficiently accurate for the description of CAs to the CꢀN bond pro-
viding results close to those obtained by such methods as MP2,
MP4, CCSD(T), CBS-Q, and G3B3.
S
ꢁ S þ RlnV
=Vm;gas
m;liq
ð3Þ
a ¼
ꢂ
ꢃ
o;s
s
S
gas þ RlnV =Vm;gas
m;liq
1
084
www.chempluschem.org
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemPlusChem 2012, 77, 1075 – 1086

Reaction Mechanisms
S ¼ S þ DS
and 12-03-33071). We acknowledge Saint Petersburg State Uni-
versity for a research grant (2011–2013) and grant for applied re-
searches (2012–2013), and RAS Presidium Subprogram. A.L.M. ex-
presses his gratitude to the Government of Saint Petersburg for
grants for graduate and undergraduate students (2011, 2012).
M.L.K. is grateful to the FCT and IST for a research contract
within the Ciencia 2007 scientific program. We thank the Com-
puter Center (at Petrodvorets) of Saint Petersburg State University
for providing their computer facilities for the theoretical calcula-
tions. We are obliged to the FSUE IREA Shared Research Analytic
Center (Moscow, Russia) sponsored by the Federal Targeted Pro-
gram “Research and Development in the Priority Areas for the
Development of Scientific and Technological Complex of Russia
for 2007–2013” (contract No. 16.552.11.7054 from 12/06/2012)
for providing ICP-MS analysis.
s
g
sol
¼
¼
S þ ½DS þ aðS þ DS Þ þ DS ꢃ
g
1
g
1
2
ꢁ
1
ꢁ1
S þ ½ðꢁ12:16 cal mol KÞ-0:233ðS ꢁ12:16 cal mol KÞ
g
g
ꢁ
1
þ5:81 cal mol Kꢃ
ð4Þ
where S =gas-phase entropy of solute, DS =solvation entropy,
g
,s
sol
,
s
s
S8 liq, S8 gas, and V
=standard entropies and molar volume of
m,liq
ꢁ
1
the solvent in liquid or gas phases (171.8 and 275.0 Jmol K and
ꢁ
1
5
3.65 mLmol , respectively, for CH NO ), V
=molar volume of
3
2
m,gas
ꢁ1
the ideal gas at 258C (24450 mLmol ), V8 =molar volume of the
solution corresponding to the standard conditions (1000 mLmol ).
The enthalpies and Gibbs free energies in solution (H and G ) were
estimated using the [Eqs. (5) and (6)]
m
ꢁ
1
s
s
H ¼ E ð6-311 þ Gðd,pÞÞ ꢁ E ð6-311 þ Gðd,pÞÞ
s
s
g
ð5Þ
ð6Þ
þ H ð6-31GðdÞÞ
Keywords: boranes
calculations · nucleophilic addition · reaction mechanisms
·
cycloaddition
· density functional
g
G ¼ H ꢁTS
s
s
s
[
1] I. B. Sivaev, V. V. Bregadze, Eur. J. Inorg. Chem. 2009, 1433–1450.
where E , E and H are the total energies in solution and in gas
phase and gas-phase enthalpy calculated at the corresponding
level.
[2] I. B. Sivaev, V. I. Bregadze, N. T. Kuznetsov, Russ. Chem. Bull. 2002, 51,
1362–1374.
[3] I. B. Sivaev, V. I. Bregadze, S. Sjçberg, Collect. Czech. Chem. Commun.
s
g
g
2
002, 67, 679–727.
The atomic charges were computed by using the natural bond or-
[
4] I. B. Sivaev, N. A. Votinova, V. I. Bragin, Z. A. Starikova, L. V. Goeva, V. I.
Bregadze, S. Sjoberg, J. Organomet. Chem. 2002, 657, 163–170.
[86]
bital (NBO) partitioning scheme. The synchronicity of CAs (Sy)
[87–90]
was calculated using the formula:
[5] B. Ringstrand, P. Kaszynski, V. G. Young, Z. Janou, Inorg. Chem. 2010, 49,
166–1179.
1
[
6] I. N. Polyakova, K. Y. Zhizhin, N. T. Kuznetsov, Crystallogr. Rep. 2007, 52,
271–274.
n
X
dB ꢁ dB
Sy ¼ 1 ꢁ ð2n ꢁ 2Þꢁ
1
ꢄ
i
av
ð7Þ
dB
[7] G. Froehner, K. Challis, K. Gagnon, T. Getman, R. Luck, Synthesis and Re-
activity in Inorganic, Metal-Organic and Nano-Metal Chemistry 2006, 36,
av
i¼1
7
77–785.
8] V. Sꢄcha, J. Plesek, M. Kvꢄcalovꢅ, I. Cꢄsarovꢅ, B. Grꢃner, Dalton Trans.
009, 851–860.
where n is the number of bonds directly involved in the reaction
n=5 for 1,3-dipolar cycloadditions), dB is the relative variation of
a given Wiberg bond index B at the transition state relative to re-
[
(
i
2
i
[9] A. P. Zhdanov, M. V. Lisovsky, L. V. Goeva, G. A. Razgonyaeva, I. N. Polya-
actants (R), and products (P) and it is calculated as:
kova, K. Y. Zhizhin, N. T. Kuznetsov, Russ. Chem. Bull. 2009, 58, 1694–
1700.
[
10] A. P. Zhdanov, I. N. Polyakova, G. A. Razgonyaeva, K. Y. Zhizhin, N. T. Kuz-
TS
R
i
R
Bi ꢁ B
netsov, Russ. J. Inorg. Chem. 2011, 56, 847–855.
dB ¼
ð8Þ
i
P
B ꢁ B
[11] A. L. Mindich, N. A. Bokach, F. M. Dolgushin, M. Haukka, L. A. Lisitsyn,
A. P. Zhdanov, K. Y. Zhizhin, S. A. Miltsov, N. T. Kuznetsov, V. Y. Kukushkin,
Organometallics 2012, 31, 1716–1724.
i
i
If the dB value is negative it is assumed to be zero. The average
i
[
[
12] N. A. Bokach, Russ. Chem. Rev. 2010, 79, 89–100.
13] N. A. Bokach, M. L. Kuznetsov, V. Y. Kukushkin, Coord. Chem. Rev. 2011,
value of dB (dB ) is defined as:
i
av
2
55, 2946–2967.
n
X
[
[
[
14] V. Y. Kukushkin, A. J. L. Pombeiro, Chem. Rev. 2002, 102, 1771–1802.
15] M. L. Kuznetsov, Russ. Chem. Rev. 2006, 75, 1045–1073.
16] A. S. Kritchenkov, N. A. Bokach, M. L. Kuznetsov, F. M. Dolgushin, T. Q.
Tung, A. P. Molchanov, V. Y. Kukushkin, Organometallics 2012, 31, 687–
699.
dBav ¼ nꢁ
1
dBi
ð9Þ
i¼1
+
ꢁ
For ylide C H N CH C(O)Me, two conformers were calculated (II
5
5
and IIa, Table S2 and Table S3), and the first one was found to be
the most stable. The activation and reaction energies were estimat-
ed relative to II.
[
17] A. S. Kritchenkov, N. A. Bokach, M. Haukka, V. Y. Kukushkin, Dalton Trans.
2011, 40, 4175–4182.
[18] N. A. Bokach, I. A. Balova, M. Haukka, V. Y. Kukushkin, Organometallics
011, 30, 595–602.
19] N. A. Bokach, A. V. Khripoun, V. Y. Kukushkin, Inorg. Chem. 2003, 42,
96–903.
2
[
[
[
[
8
Acknowledgements
20] G. Wagner, M. Haukka, J. J. R. Frausto da Silva, A. J. L. Pombeiro, V. Y. Ku-
kushkin, Inorg. Chem. 2001, 40, 264–271.
21] K. V. Luzyanin, A. G. Tskhovrebov, M. F. C. Guedes da Silva, M. Haukka,
A. J. L. Pombeiro, V. Y. Kukushkin, Chem. Eur. J. 2009, 15, 5969–5978.
22] F. Risitano, G. Grassia, G. Brunob, F. Nicolbb, Liebigs Ann. 1997, 441–
We thank the Federal Targeted Program “Scientific and Scientific-
Pedagogical Personnel of the Innovative Russia in 2009–2013”
(contract No. P1294 from 09/06/2010), Federal Grant-in-Aid Pro-
445.
gram “Human Capital for Science and Education in Innovative
Russia” for 2009–2013, (agreement No. 14.B37.21.0794 from 03/
[
[
23] W. Zecher, F. Krohnke, Chem. Ber. 1961, 94, 690–697.
24] M. F. Aldersley, S. H. Chishti, F. M. Dean, M. E. Douglas, D. S. Ennis, J.
Chem. Soc. Perkin Trans. 1 1990, 2163–2174.
09/2012), Russian Fund for Basic Research (grants 11-03-00262
ChemPlusChem 2012, 77, 1075 – 1086
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chempluschem.org
1085

V. Yu. Kukushkin, N. A. Bokach et al.
[
[
[
25] O. Tsuge, S. Kanemasa, S. Takenaka, Bull. Chem. Soc. Jpn. 1985, 58,
137–3157.
26] O. Tsuge, S. Kanemasa, S. Takenaka, Bull. Chem. Soc. Jpn. 1985, 58,
320–3336.
27] I. I. Druta, M. A. Andrei, C. I. Ganj, P. S. Aburel, Tetrahedron 1999, 55,
3063–13070.
[58] M. L. Kuznetsov, V. Y. Kukushkin, A. J. L. Pombeiro, J. Org. Chem. 2010,
75, 1474–1490.
[59] G. Wagner, A. J. L. Pombeiro, V. Y. Kukushkin, J. Am. Chem. Soc. 2000,
122, 3106–3111.
[60] P. H. H. Hermkens, J. H. von Maarseveen, C. G. Kruse, H. W. Scheeren, Tet-
rahedron 1988, 44, 6491–6504.
3
3
1
[
[
28] G. Matusiak, Aust. J. Chem. 1999, 52, 149–151.
29] S. Kanemasa, S. Takenaka, H. Watanabe, O. Tsuge, J. Org. Chem. 1989,
[61] R. Sustmann, Tetrahedron Lett. 1971, 12, 2717–2720.
[62] R. Sustmann, Tetrahedron Lett. 1971, 12, 2721–2724.
[63] A. H. Soloway, W. Tjarks, B. A. Barnum, F. Rong, R. F. Barth, I. M. Codogni,
J. G. Wilson, Chem. Rev. 1998, 98, 1515–1562.
[64] L. Zhang, M. Brookhart, P. S. White, Organometallics 2006, 25, 1868–
1874.
[65] U. Beckmann, G. Hꢇgele, W. Frank, Eur. J. Inorg. Chem. 2010, 1670–
1678.
[66] A. F. Lugo (nꢆe Gushwa), A. F. Richards, Eur. J. Inorg. Chem. 2010, 2025–
2035.
5
4, 420–424.
30] Y. Shang, M. Zhang, S. Yu, K. Ju, C. Wang, X. He, Tetrahedron Lett. 2009,
0, 6981–6984.
31] F. Dumitrascu, C. I. Mitan, C. Draghici, M. T. Caproiu, D. Raileanu, Tetrahe-
dron Lett. 2001, 42, 8379–8382.
[
[
5
[
[
32] R. S. Tewari, A. Bajpai, J. Chem. Eng. Data 1985, 30, 505–507.
33] Y. Kobayashi, I. Kumadaki, E. Kobayashi, Heterocycles 1981, 15, 1223–
1225.
[
[
[
[
34] R. E. Banks, S. N. Mohialdin, J. Fluorine Chem. 1986, 34, 275–279.
35] R. E. Banks, S. N. Khaffaff, J. Fluorine Chem. 1991, 51, 407–418.
36] R. E. Banks, J. Thomson, J. Chem. Soc. Perkin Trans. 1 1986, 1769–1776.
37] J. Alvarez-Builla, E. Galvez, A. M. Cuadro, F. Florencio, S. Garcia-Blanco, J.
Heterocycl. Chem. 1987, 24, 917–926.
[67] M. Basato, E. Caneva, C. Tubaro, A. C. Veronese, Inorg. Chim. Acta 2009,
362, 2551–2555.
[68] L. A. Lesikar, A. F. Gushwa, A. F. Richards, J. Organomet. Chem. 2008, 693,
3245–3255.
[69] K. Y. Zhizhin, V. N. Mustyatsa, E. Y. Matveev, V. V. Drozdova, N. A. Votino-
va, I. N. Polyakova, N. T. Kuznetsov, Russ. J. Inorg. Chem. 2003, 48, 671–
675. Chem. Abstr. 2003, 140, 77181.
[70] I. Sasaki, L. Vendier, A. Sournia-Saquet, P. G. Lacroix, Eur. J. Inorg. Chem.
2006, 3294–3302.
[71] a) V. G. Chekhuta, L. A. Kucherova, Metody Poluch. Khim. Reakt. Prep.
1970, 22, 40–41.
[72] BrukerAXS, APEX2-Software Suite for Crystallographic Programs, Bruker
AXS, Inc., Madison, WI, USA, 2009.
[38] N. S. Prostakov, V. I. Kuznetsov, G. Dattaray, N. D. Sergeyeva, Khim. Geter-
otsikl. Soedin. 1980, 6, 806–807. Chem. Abs. 1980, 93, 168088.
[39] L. Depature, G. Surpateanu, Spectrochim. Acta Part A 2003, 59, 3029–
[
3
039.
40] A. M. Shestopalov, Y. A. Sharanin, V. P. Litvinov, V. N. Nesterov, A. S. De-
merkov, V. E. Shklover, Y. T. Struchkov, Izvestiya Akad. Nauk SSSR Seriya
Khim. 1991, 697–701. Chem. Abs. 1991, 116, 128606.
[
41] K. Dickorꢆ, F. Krçhnke, Chem. Ber. 1960, 93, 2479–2485.
[
42] J. Alvarez-Builla, J. L. Novella, E. Gꢅlvez, P. Smith, F. Florencio, S. Garciꢅ-
Blanco, J. Bellanato, M. Santos, Tetrahedron 1986, 42, 699–708.
43] M. Yamazaki, K. Noda, M. Hamana, Chem. Pharm. Bull. 1970, 18, 901–
[73] G. M. Sheldrick, Acta Crystallogr. 2008, A64, 112–122.
[74] O. V. Dolomanov, L. J. Bourhis, R. J. Gildea, J. A. K. Howard, H. Pusch-
mann, J. Appl. Crystallogr. 2009, 42, 339–341.
[
9
07.
[75] G. M. Sheldrick, SADABS-Bruker AXS Scaling and Absorption Correction,
Bruker AXS, Inc., Madison, Wisconsin, USA, 2008.
[76] A. D. Becke, J. Chem. Phys. 1993, 98, 5648–5652.
[77] C. Lee, W. Yang, R. G. Parr, Phys. Rev. B 1988, 37, 785–789.
[78] M. J. Frisch, et al. Gaussian 03, Gaussian Inc., Pittsburgh PA, 2003. See
the Supporting Information of full citation.
[79] C. Gonzalez, H. B. Schlegel, J. Chem. Phys. 1991, 95, 5853–5860.
[80] C. Gonzalez, H. B. Schlegel, J. Chem. Phys. 1989, 90, 2154–2161.
[81] C. Gonzalez, H. B. Schlegel, J. Phys. Chem. 1990, 94, 5523–5527.
[82] J. Tomasi, M. Persico, Chem. Rev. 1994, 94, 2027–2094.
[83] V. Barone, M. Cossi, J. Phys. Chem. A 1998, 102, 1995–2001.
[84] D. H. Wertz, J. Am. Chem. Soc. 1980, 102, 5316–5322.
[85] J. Cooper, T. Ziegler, Inorg. Chem. 2002, 41, 6614–6622.
[86] A. E. Reed, L. A. Curtiss, F. Weinhold, Chem. Rev. 1988, 88, 899–926.
[87] A. Moyano, M. A. Pericꢈs, E. Valentꢄ, J. Org. Chem. 1989, 54, 573–582.
[88] B. Lecea, A. Adeta, G. Rea, J. M. Ugalde, F. P. Cossꢄo, D. Toma, J. Am.
Chem. Soc. 1994, 116, 9613–9619.
[
[
44] W. Augstein, F. Kroehnke, Justus Liebigs Ann. Chem. 1966, 697, 158–170.
45] A. M. Shestopalov, Y. A. Sharanin, I. A. Aitov, V. N. Nesterov, V. E. Shklover,
Y. T. Struchkov, V. P. Litvinov, Khim. Geteritsikl. Soedin. 1991, 1087–1094.
Chem. Abstr. 1991, 116, 214304.
46] Y. Tominaga, Y. Miyake, H. Fujito, K. Kurata, H. Awaya, Y. Masuda, G. Ko-
bayashi, Chem. Pharm. Bull. 1977, 25, 1528–1533.
47] C. K. Ghosh, S. Sahana, Indian J. Chem. Sect. B: Org. Chem. Incl. Med.
Chem. 1996, 35, 203–206. Chem. Abs. 1996, 124, 260793.
48] W. J. Geary, Coord. Chem. Rev. 1971, 7, 81–122.
49] A. Szwajca, A. Katrusiak, M. Szafran, J. Mol. Struct. 2004, 705, 159–165.
50] D. Dou, I. J. Mavunkal, J. A. K. Bauer, C. B. Knobler, M. F. Hawthorne, S. G.
Shore, Inorg. Chem. 1994, 33, 6432–6434.
[
[
[
[
[
[
51] F. H. Allen, O. Kennard, D. G. Watson, L. Brammer, A. G. Orpen, R. Taylor,
J. Chem. Soc. Perkin Trans. 2 1987, S1–S19.
[
[
52] J.-C. Zhuo, K. Schenk, Helvetica Chimica Acta 1997, 80, 2137–2147.
53] M. Basato, B. Corain, M. Cofler, A. C. Veronese, G. Zanotti, Chem.
Commun. 1984, 1593–1594.
54] C. O. Kappe, E. Terpetschnig, G. Penn, G. Kollenz, K. Peters, E.-M. Peters,
H. G. von Schnering, Liebigs Ann. 1995, 537–543.
[89] I. Morao, B. Lecea, F. P. Cossꢄo, J. Org. Chem. 1997, 62, 7033–7036.
[90] F. P. Cossꢄo, I. Morao, H. Jiao, P. v. R. Schleyer, J. Am. Chem. Soc. 1999,
121, 6737–6746.
[
[55] M. L. Kuznetsov, V. Y. Kukushkin, A. I. Dement’ev, A. J. L. Pombeiro, J.
Phys. Chem. A 2003, 107, 6108–6120.
[
[
56] M. L. Kuznetsov, V. Y. Kukushkin, J. Org. Chem. 2006, 71, 582–592.
57] M. L. Kuznetsov, A. A. Nazarov, L. V. Kozlova, V. Y. Kukushkin, J. Org.
Chem. 2007, 72, 4475–4485.
Received: September 27, 2012
Published online on November 23, 2012
1
086
www.chempluschem.org
ꢀ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemPlusChem 2012, 77, 1075 – 1086