Origin of regioselectivity in the reactions of nitronate and enolate ambident anions

Article abstract of DOI:10.1021/jo302103c

The reactions of nitronates of ring-substituted phenylnitromethanes and enolates of ring-substituted 1-phenyl-2-propanones with MeOBs gave exclusively the O-methylated and C-methylated products, respectively. DFT calculations suggested that two factors, n

Full text of DOI:10.1021/jo302103c

Article
pubs.acs.org/joc
Origin of Regioselectivity in the Reactions of Nitronate and Enolate
Ambident Anions
Tomomi Sakata, Natsuko Seki, Kozue Yomogida, Hiroko Yamagishi, Akino Otsuki, Chie Inoh,
and Hiroshi Yamataka*
Department of Chemistry and the Research Center for Smart Molecules, Rikkyo University, Nishi-Ikebukuro, Toshima-ku 171-8501,
Tokyo, Japan
S
* Supporting Information
ABSTRACT: The reactions of nitronates of ring-substituted
phenylnitromethanes and enolates of ring-substituted 1-phenyl-2-
propanones with MeOBs gave exclusively the O-methylated and C-
methylated products, respectively. DFT calculations suggested that
two factors, namely, intrinsic barriers and metal-cation coordina-
tion, control the C/O selectivity. The kinetic preference for O-
methylation in the reactions of nitronates arises from the intrinsic
barriers, which are ca. 10 kcal/mol lower for O-methylation than
for C-methylation. The situation is the same for the gas-phase reaction of an enolate, in which the O-methylation is more
favorable than the C-methylation. The experimentally observed C-selectivity of enolate reactions in solution is due to the metal-
cation coordination, which hinders O-methylation for enolates. The effects of the enolate reactivity and the solvent on the C/O
selectivity are also rationalized to arise from the two factors.
Enolates give either or both O- and C-alkylated products
depending on the stability of the enolate, solvent, and the
counterion. (3) A more stable and less reactive enolate leads to
more O-alkylation. (4) Polar aprotic solvent favors O-
alkylation.
In the present study, we have carried out the reactions of a
series of ring-substituted phenylmethanenitronates and of
substituted 1-phenyl-2-propanone enolates with methyl p-
bromobenzenesulfonate (MeOBs). DFT calculations were
then performed to examine the origins of the preference for
C-alkylation (enolate) and O-alkylation (nitronate) and of the
different behavior of enolates and nitronates. It was found that
the size of intrinsic barriers of the alkylation reactions and the
metal-cation complexation play essential roles in determining
the C/O selectivity.
INTRODUCTION
■
Nitronates and enolates are useful reactive intermediates in
synthetic chemistry. In alkylation reactions, these ambident
anions may give two products: C-alkylated and O-alkylated
ones. Understanding of the origin of the selectivity is important
because C-alkylation is one of the most common ways to form
a carbon−carbon bond and O-alkylation provides useful
reactive intermediates, alkylnitronate, and vinyl ether. However,
nitronates and enolates behave differently in many respects and
the origin of this difference is not well recognized.
Alkylations of nitronates in aprotic solvents usually give the
O-alkylated products.1−4 Exceptional C-alkylation of nitronates
was reported for the reactions with stable carbocations,
although as in the case of protonation of nitronates,5 fast
reversible formation of low concentration of the O-alkylated
intermediate was not excluded.6 By contrast, the competition
between C- and O-alkylation in enolate reactions in solution
was more complex. The alkylation of enolates of simple
monoketones and aldehydes normally occurred on carbon, and
the alkylation on oxygen has not been observed. Enolates of
acidic ketones, such as 1,3-diketones and β-ketoesters, on the
other hand, gave both the C- and O-alkylated products.7,8
Protic solvents led to the thermodynamically more favored C-
alkylation products, whereas polar aprotic solvents favored
alkylation at the more electronegative O atom.9 It was also
shown that the C/O selectivity depends on the metal cation
and the alkylating reagent.10,11 In the gas phase, the reaction of
cyclohexanone enolate and MeBr gave only the O-methylated
product, with no evidence for the C-attack.12 The compiled
experimental results thus indicated several important points:
(1) Nitronates give in most cases O-alkylated products. (2)
RESULTS AND DISCUSSION
■
Experimental Results. A series of sodium nitronates (1-X,
X = p-MeO, p-Me, m-Me, H, p-Cl, m-F, m-Cl, p-CF3, m-NO2)
were prepared by the reactions of ring-substituted phenylnitro-
methanes and NaOEt in diethyl ether. The reactions of 1-X
with MeOBs were carried out in DMF, 80% (v/v) aqueous
DMF, and 90% (v/v) aqueous MeOH, and the products were
1
analyzed. Product yields were determined by H NMR in
deuterated solvents. Table 1 lists the yields and material
balances for the reactions of 1-H in three solvent systems, and
Tables S1−S3 in Supporting Information summarize the results
for 1-X. The results in Table 1 showed that in all solvent
Received: September 26, 2012
Published: November 8, 2012
© 2012 American Chemical Society
10738
dx.doi.org/10.1021/jo302103c | J. Org. Chem. 2012, 77, 10738−10744

The Journal of Organic Chemistry
Article
into account, which arises from stabilizing solvent effect on the
nitronate salts in the aqueous solvent. Correlation between the
two sets of rate constants in Figure 1 gave a linear plot with the
Table 1. Product Yields and Material Balances for the
Reactions of 1-H and MeOBs at 25 °C
a
O-
material
balance
(%)
c
reaction
time
conversion adduct
solvent
DMF
(%)
(%)
Z:E
5 min
59.3
79.6
94.7
16.7
32.8
48.0
79.9
4.6
59.0
78.2
95.9
13.5
29.9
44.2
65.3
2.5
6.7:1.0
6.4:1.0
6.6:1.0
3.6:1.0
3.6:1.0
3.3:1.0
3.3:1.0
1.8:1.0
2.4:1.0
2.4:1.0
2.3:1.0
2.3:1.0
4.7:1.0
99.5
98.2
101.3
96.8
97.1
96.2
85.4
97.9
95.0
88.3
83.5
87.7
57.5
15 min
50 min
80% DMF-d7/ 5 min
20% D2O
(v/v)
15 min
30 min
104 min
90% CD3OH/ 15 min
10% H O
b 2
72 min
2.5 h
4 h
14.2
30.1
33.4
46.4
87.0
9.2
(v/v)
18.4
27.0
34.1
Figure 1. Logarithmic plots of second-order rate constants for the
reaction of 1-X and MeOBs in DMF against those in 80% (v/v)
aqueous DMF at 25.0 °C.
5.5 h
1 day
44.5
a
b
Toluene was used as an internal standard. CD3OH/H2O mixture,
rather than CD3OD/D2O, was used as the solvent to avoid H-D
c
slope of 0.76, indicating that the substituent effect is smaller in
aqueous DMF. The Hammett plot in Supplementary Figure S1
showed a good linear correlation with ρ of −0.84. The
Brønsted-type plots for the reactions in DMF and aqueous
DMF are shown in Figures 2 and 3. Here, since the pKa values
−
exchange in PhCHNO2
.
% O-adduct is relative to initial
concentration of the starting nitronate.
systems only O-methylated products were obtained as a
mixture of E- and Z-isomers, with no C-methylated product
detected in 1H NMR. We ascribed the major isomer to Z, since
B3LYP/6-31+G* calculations showed that the Z-isomer is 18.9
kcal/mol more stable than the E-isomer. This is likely due to a
steric effect. The nature of the solvent did not influence the C/
O selectivity for the reactions of nitronates, in sharp contrast to
the enolate cases where aprotic solvent tends to lead to O-
alkylation.9 The material balance was nearly perfect in DMF but
decreased as the reaction proceeded in aqueous solvents. This
is likely due to the decomposition of the product especially of
the E-isomer in aqueous methanol, since it was reported that
O-alkylated products of nitronates, especially the less stable E-
isomer, were not stable and decomposed to give aldehydes.4
We indeed detected benzaldehyde in the reaction mixture as
expected for the decomposition and hydrolysis of the O-
methylated product.
Figure 2. Brønsted-type plot for O-methylation of 1-X and MeOBs in
DMF at 25.0 °C. The pKa values are those in DMSO.
The rates of the reactions were determined photometrically
in DMF and aqueous DMF and are listed in Table 2. It appears
that the reaction in DMF is 5 times faster than that in aqueous
DMF when the difference in the MeOBs concentration is taken
in DMF and in aqueous DMF were not available, plots were
made by using pKa values in DMSO13 and in aqueous MeOH,14
respectively. The use of the pKa values of related solvent
systems would cause no serious error since the pKa values in
Table 2. Pseudo-first-order Rate Constants, with Associated
Standard Deviations, for the Reactions of 1-X and MeOBs in
DMF and in 80% (v/v) Aqueous DMF at 25.0 °C
a
b
X
kobs (10−3 s−1
)
kobs (10−3 s−1
)
p-MeO
p-Me
m-Me
H
2.00 0.08
1.38 0.08
1.45 0.01
1.24 0.05
c
2.18 0.13
2.20 0.04
2.03 0.02
p-Cl
0.675 0.023
0.567 0.044
0.556 0.020
0.395 0.014
0.301 0.006
1.22 0.04
m-F
1.12 0.06
m-Cl
p-CF3
m-NO2
c
c
c
a
b
[1-X] = 0.046 mmol/L, [MeOBs] = 0.46 mmol/L, in DMF. [1-X] =
Figure 3. Brønsted-type plot for O-methylation of 1-X and MeOBs in
80% (v/v) aqueous DMF at 25.0 °C. The pKa values are those in 50%
(v/v) aqueous MeOH.
c
0.1 mmol/L, [MeOBs] = 5.0 mmol/L, in 80% aqueous DMF. Not
determined.
10739
dx.doi.org/10.1021/jo302103c | J. Org. Chem. 2012, 77, 10738−10744

The Journal of Organic Chemistry
Article
the C-protonated products were more stable than the O-protonated
ones, whereas the O-protonation was kinetically favored for the
DMF and DMSO, at least in some compounds, were reported
to vary in the same direction and correlate with the pKa values
of these species in water.15 Although the number of points is
small due to the lack of pKa values, there appears no anomaly in
the size of the slopes. Thus, no nitroalkane anomaly was
detected in the methyl-transfer reaction of nitronates at the O
atom. This is due to the absence of charge imbalance at the O-
alkylation TS, and the result, in turn, supports the TS
imbalance rationale16 for nitroalkane anomaly, which has
been observed for proton-transfer reactions of nitroalkanes.17
For comparison, a series of sodium enolates (2-X, X = p-
MeO, p-Me, m-Me, H, p-Cl, p-CF3) were prepared from the
reactions of ring-substituted 1-phenyl-2-propanones and
sodium hexamethyldisilazide (NaHMDS) in THF, and C/O
selectivity was determined for the reaction with MeOBs at 25
°C. As shown in Table 3, the reaction gave the C-methylated
product as the only product.
reaction of PhCHNO2 with various acids.5 The C/O selectivity is
−
similar for both methylating reagents, although the kinetic O-
preference is slightly larger for the reactions with MeOMs than with
MeCl.
The reactions of enolates in Table 5 showed the same trend as
those of nitronates in that the C-methylation is favored thermody-
namically, whereas the O-methylation is preferred kinetically. The
calculated kinetic O-methylation preference in the enolate reactions
did not reproduce the experimental results that only C-methylated
product was observed for the reactions of 2-X in THF. A possible
origin of this discrepancy will be discussed later.
In Figure 4 and Supplementary Figure S2 are illustrated the
Brønsted-type (rate-equilibrium) plots for the reactions with MeCl
and with MeOMs, respectively.
These Brønsted-type plots gave normal α values of about 0.4 for all
reactions. Several points are apparent from Figures 4 and S2: (1) The
points of the C-methylation and the O-methylation gave separated
Brønsted plots for both the nitronate and enolate series, indicating that
the C-methylation and O-methylation reactions belong to different
reaction families. (2) The lines for the O-methylation are located in
lower-right positions relative to the lines for the C-methylation in both
nitronates and enolates, reflecting the fact that although the C-
methylated product is more stable the barrier for C-methylation is
higher for a given anion. (3) In each C-attack and O-attack series, the
Brønsted lines of nitronates and enolates are located close to each
other, with the lines of nitronates being slightly above the enolate lines.
This means that nitronates are less reactive than enolates at the same
reaction energies. In relation to this, it is interesting to note that ΔH
for the Me-transfer reactions of nitronates and enolates with MeCl
gave excellent correlation against ΔH for the proton-transfer reaction
of these anions with CH2(NO2)2, with both nitronates and enolates
being in the same correlation lines (Figure 5). On the basis of the fact
that the carbon affinity and the proton affinity give an excellent single
correlation line with a unity slope for each reacting site, the above
Brønsted lines further mean that nitronates are less reactive than
enolates at the same pKa values. This, in turn, indicates that the kinetic
barriers are larger for nitronates than for enolates.
Table 3. Product Yields and Material Balances for the
a
Reactions of 2-X and MeOBs in THF at 25 °C
reaction
time
conversion
(%)
C-adduct
(%)
material balance
(%)
X
p-MeO
15 min
82.8
98.9
99.6
90.2
97.7
100.0
60.7
91.2
100.0
54.4
92.6
100.0
45.0
90.4
100.0
22.2
62.4
100.0
72.7
86.2
87.3
69.1
87.5
91.6
47.3
81.7
82.1
54.1
93.6
97.1
37.4
78.1
82.9
16.0
48.3
83.5
89.9
87.3
87.6
88.9
89.9
91.6
86.7
80.4
82.1
99.7
103.7
97.1
92.4
87.7
82.9
93.2
85.9
60 min
overnight
14 min
p-Me
m-Me
H
60 min
overnight
10 min
60 min
overnight
10 min
60 min
overnight
10 min
p-Cl
p-CF3
A more quantitative means of interpreting the kinetic barrier is
given by Marcus’ equation (eq 3), in which ΔE⧧ and ΔE have their
usual meanings and ΔE0⧧ is the intrinsic barrier.18 The intrinsic barrier
is the barrier for the hypothetical thermoneutral step of a given
reaction (kinetic barrier). Equation 3 indicates that a reaction barrier is
controlled by the intrinsic barrier and the reaction endothermicity; the
latter modifies the overall barrier in such a way that the barrier
increases when the reaction is endothermic and decreases if it is an
exothermic reaction (thermodynamic driving force).
60 min
overnight
10 min
60 min
overnight
83.5
a
b
Trimethoxybenzene was used as an internal standard. % C-adduct is
relative to initial concentration of the starting enolate.
⧧
⧧
ΔE⧧ = ΔE0 + 1/2ΔE + (ΔE)2/16ΔE0
Computational Results. In order to understand why nitronates
and enolates behave differently, DFT calculations (B3LYP/6-31+G*)
were carried out for the reactions of nitronates (1-X) and enolates (3-
X) with two methylating reagents (eqs 1 and 2). The calculated
activation and reaction enthalpies for the reactions with MeCl and
MeOSO2Me (MeOMs) are listed in Tables 4 and 5.
(3)
The intrinsic barrier of a given group-transfer reaction between R
and X fragments (ΔE0 R,X) can be calculated by the assumption of
arithmetic mean of the intrinsic barriers of two symmetry reactions, eq
⧧
4.18
⧧
⧧
⧧
ΔE0
= 1/2[(ΔE0
+ ΔE0 X,X)]]
(4)
R,X
R,R
In Tables 6 and 7 are listed the calculated intrinsic barriers for the
methylating reactions of three ring-substituted phenylmethanenitro-
⧧
nates and phenylacetaldehyde enolates, respectively. Here, ΔH0 (Y,Y)
⧧
is the barrier of a methyl-transfer reaction between Y's, ΔH0 (X,X) is
the barrier of a methyl-transfer reaction between nitronates (or
⧧
enolates) at the C or the O position, and ΔH0 (X,Y) is the intrinsic
barrier for the methylation of a nitronate (or an enolate).
It is seen from Table 4 that the C-methylated products are more
stable by 13 - 15 kcal/mol than the O-methylated ones, whereas the
activation barriers are lower for the O-methylation by 2 - 4 kcal/mol
for all substituted nitronates. The calculated kinetic O-preference is in
line with the experimental selectivity mentioned above. Similar trends
have previously been reported for the protonation of nitronates, where
The data in Tables 6 and 7 show that the intrinsic barriers were
much smaller (9−10 kcal/mol) for the O-methylation than for the C-
methylation both for the nitronate and enolate reactions. The smaller
intrinsic barriers for the O-methylation arise from smaller barriers for
the symmetrical methyl-transfer reactions (ΔH0⧧(X,X)) at O
compared to those at C. This explains the strong preference for O-
10740
dx.doi.org/10.1021/jo302103c | J. Org. Chem. 2012, 77, 10738−10744

The Journal of Organic Chemistry
Article
−
Table 4. Calculated Activation and Reaction Enthalpies for the Reactions of XC6H4CHNO2 and Methylation Reagents at
a
B3LYP/6-31+G*
MeCl
MeOMs
C-methylation
O-methylation
C-methylation
O-methylation
no.
X
ΔH⧧
ΔH
ΔH⧧
ΔH
ΔH⧧
ΔH
ΔH⧧
ΔH
1
2
p-NH2
p-MeO
p-OH
p-Me
m-Me
H
6.7
7.2
−19.7
−17.1
−16.5
−15.9
−15.2
−14.6
−9.7
−8.9
0.1
4.3
5.0
−4.6
−1.5
−2.0
−0.7
0.1
5.3
6.0
−32.1
−29.9
−28.9
−28.3
−27.6
−27.0
−22.1
−21.5
−12.4
−7.0
1.3
1.8
2.2
2.8
3.1
3.1
4.8
5.2
8.6
11.1
−17.0
−14.9
−13.9
−13.1
−12.4
−11.9
−7.4
−6.9
1.3
3
7.3
5.0
6.4
4
7.7
5.6
6.8
5
8.0
5.9
7.0
6
8.0
6.0
0.6
7.2
7
p-Cl
9.7
7.7
5.0
8.9
8
m-Cl
10.2
13.5
16.2
8.1
5.8
9.2
9
p-CN
p-NO2
11.5
14.1
13.7
18.7
12.9
15.9
10
5.4
6.2
a
Energies are relative to the separated reactants in kcal/mol.
Table 5. Calculated Activation and Reaction Enthalpies for the Reactions of XC6H4CHCHO− and Methylation Reagents at
a
B3LYP/6-31+G*
MeCl
MeOMs
C-methylation
O-methylation
C-methylation
O-methylation
no.
X
ΔH⧧
ΔH
ΔH⧧
ΔH
ΔH⧧
ΔH
ΔH⧧
ΔH
1
2
p-NH2
p-MeO
p-OH
p-Me
m-Me
H
3.5
4.0
−25.4
−22.8
−22.2
−21.7
−21.7
−20.5
−15.0
−14.4
−4.5
1.0
1.5
1.5
2.1
1.9
2.4
3.9
5.2
7.1
10.7
−11.0
−8.8
−8.4
−7.3
−6.5
−6.1
−1.5
0.9
2.3
3.2
−37.7
−33.4
−35.0
−34.1
−33.9
−34.2
−27.4
−28.1
−16.9
−10.6
−0.3
0.1
−21.3
−19.1
−18.9
−18.2
−18.9
−18.5
−12.1
−13.2
−3.0
3
4.0
2.9
−0.2
0.7
4
4.5
3.7
5
4.8
2.6
−0.4
−0.4
2.4
6
4.8
2.7
7
p-Cl
6.7
5.8
8
m-Cl
7.0
4.9
1.3
9
p-CN
p-NO2
10.8
14.1
8.4
10.1
13.1
5.7
10
1.7
14.6
8.2
2.5
a
Energies are relative to the separated reactants in kcal/mol.
Figure 4. Brønsted-type plots for the reactions of nitronates and enolates with MeCl at B3LYP/6-31+G*. Closed symbols are for nitronates and
open symbols are for enolates. Circles are for C-methylation and triangles are for O-methylation. Numbers correspond to the substituent in Table 4.
alkylation observed for the nitronate reactions and for the gas-phase
If the O-alkylation selectivity in the nitronate reactions arises from
the intrinsic barrier difference between the O- and C-alkylation steps,
the C/O selectivity for a reaction with a very small intrinsic barrier
should become controlled by the thermodynamic driving force. This
argument may apply to the reported kinetic C-alkylation selectivity for
the cation−anion combination reactions of nitronates with carboca-
tions.6
enolate reactions. It is interesting to note that the substituent effect on
⧧
ΔH0 (X,X) is small, because a substituent that makes the anion a
better nucleophile makes it a poorer leaving group at the same time.
⧧
On the other hand, the size of ΔH0 (Y,Y) depends on the identity of
Y, reflecting, for example, that Br− and Cl− are reactive both as a
nucleophile and a leaving group.
10741
dx.doi.org/10.1021/jo302103c | J. Org. Chem. 2012, 77, 10738−10744

The Journal of Organic Chemistry
Article
C/O selection in the enolate reactions cannot be explained by the
intrinsic barrier argument alone.
Since the C/O selectivity has been shown to depend on the
existence of a countercation, we have then examine the effect of a
metal cation on the C/O selectivity for the Me-transfer reactions of
aliphatic nitronates and enolates. Calculations were carried out for the
SN2 reactions of MeCl with CH2NO2− and CH2CHO− with and
without Li+ solvated by three Me2O. The results are summarized in
Table 8.
Since the reactant states are stabilized by the strong interaction
between the anions and Li+, the activation enthalpies greatly increased
for all reactions upon complexation. In the absence of the cation, the
C/O selectivity is very low for the simple aliphatic nitronate. This is
because the O-alkylated alkenes in these reactions do not have a
stabilizing group on the CH2 end of the molecules and are thus
relatively unstable compared to the aromatic counterparts. This can be
seen in the difference of the reaction enthalpies between C-
methylation and O-methylation, which are much larger for the simple
aliphatic nitronate and enolate (20−24 kcal/mol, Table 8) than for
aromatic counterparts (14−15 kcal/mol, Tables 4 and 5). Under these
circumstances, we only focus on the change of the C/O selectivity
upon metal coordination.
Figure 5. Correlation between carbon affinity (vs MeCl) and proton
affinity (vs CH2(NO2)2) of nitronates (closed symbol) and enolates
(open symbol) at the carbon (circle) and the oxygen (triangle)
reaction center.
The calculations showed that the relative activation enthalpies
(δΔH⧧) for the C- vs the O-methylation of the nitronate changed little
upon Li+ coordination to the nitronate. By contrast, the C/O
selectivity dramatically changed upon Li+ coordination for the enolate
reactions. Thus, the preference of enolates for O-alkylation switches to
C-alkylation in the presence of the countercation. Furthermore, the
coordination may introduce a larger effect if enolates react as
aggregates. Lithium enolates and other alkyl lithium reagents are
known to exist as aggregates, which introduce a larger steric hindrance
than a monomer,19 although whether these reagents react as an
aggregate or as a monomer that exists in equilibrium with the
aggregate is not well understood.20
The coordination effect explains the experimental observations that
the gas-phase reaction of the cyclohexanone enolate gave only the O-
alkylated product whereas the C-alkylation was preferred in solution
reactions. The fact that more C-alkylation occurs in protic solvents
than in polar aprotic solvents may also be explained partly as arises
from the coordination effect, since a polar aprotic solvent interacts
strongly with a metal cation, leaving an enolate anion less coordinated.
Another contribution from solvents is an anion solvation by protic
solvents, which tends to inhibit O-methylation.
⧧
Table 6. Intrinsic Barriers (ΔH0 (X,Y), in kcal/mol) for the
−
SN2 Reactions of XC6H4CHNO2 and Methylation
Reagents at B3LYP/6-31+G*
p-NH2
H
p-NO2
Me-Y
ΔH0⧧(Y,Y) at C
at O
at C
ΔH0⧧(X,X)
30.8 12.1
ΔH0⧧(X,Y)
at O
at C
at O
31.5
13.5
30.4
10.5
+
Me-OMe2
Me-OMs
Me-Br
4.2
5.4
17.9
18.5
12.9
14.9
17.5
29.5
8.8
9.5
17.5
18.1
12.5
14.6
17.2
29.1
8.2
17.3
17.9
12.3
14.4
16.9
28.9
7.4
8.0
8.8
3.2
−5.7
−1.7
3.5
3.9
2.4
Me-Cl
5.9
5.2
4.4
Me-ONO
Me-CN
8.5
7.8
7.0
27.4
20.4
19.8
19.0
⧧
Table 7. Intrinsic Barriers (ΔH0 (X,Y), in kcal/mol) for the
SN2 Reactions of XC6H4CHCHO− and Methylation
Reagents at B3LYP/6-31+G*
The origin of the dramatic difference in the Li+ coordination effect
for the enolate and the nitronate likely resides in the fact that enolates
have one negative oxygen in the molecule whereas nitronates have
two. Thus, the countercation coordination blocks the oxygen reaction
center leading to the enhanced C-selectivity for enolates, whereas one
oxygen atom is still reactive leaving the C/O selectivity nearly
unchanged upon cation coordination for nitronates. The calculated
transition structures for the O-methylation of the Li+-coordinated
nitronate and enolate clearly illustrate the situation (Figure 6).
In summary, the reactions of nitronates (1-X) of ring-substituted
phenylnitromethanes with MeOBs gave the O-methylated product in
nearly quantitative yields both in aqueous and nonaqueous solvents.
Analogous reactions of enolates (2-X) of ring-substituted 1-phenyl-2-
propanones in THF gave the C-methylated product exclusively. Thus,
the alkylation reactions of nitronates and enolates showed clear
difference in the C/O regioselectivity. The Brønsted plots for the
reactions of 1-X with MeOBs showed no nitroalkane anomaly.
DFT calculations for the reactions of 1-X and 3-X with MeCl and
MeOMs revealed that for both nucleophiles O-methylation is
kinetically favored, whereas C-methylation is thermodynamically
preferred in the gas phase. Two factors were considered to control
the C/O selectivity, namely the intrinsic barriers and the metal-cation
coordination. The kinetic preference for O-methylation for the
nitronate reactions arises from the difference of the intrinsic barriers,
which are ca. 10 kcal/mol lower for the O-methylation than for the C-
methylation. The same applies to the O-methylation in the gas-phase
reaction of an enolate. The experimentally observed C-selectivity of
p-NH2
H
p-NO2
Me-Y
ΔH0⧧(Y,Y) at C
at O
at C
ΔH0⧧(X,X)
29.1 8.3
ΔH0⧧(X,Y)
at O
at C
at O
31.3
10.7
28.9
8.5
+
Me-OMe2
Me-OMs
Me-Br
4.2
5.4
17.8
18.4
12.8
14.8
17.4
29.4
7.5
8.1
16.7
17.3
11.7
13.7
16.3
28.3
6.3
16.6
17.2
11.6
13.6
16.2
28.2
6.3
7.0
6.9
1.3
−5.7
−1.7
3.5
2.5
1.4
Me-Cl
4.5
3.3
3.4
Me-ONO
Me-CN
7.1
5.9
6.0
27.4
19.0
17.9
17.9
Although the intrinsic preference for the O-alkylation was observed
in the gas-phase enolate reaction, the enolates of simple ketones and
aldehydes give exclusively the C-alkylated products in polar aprotic
solvents, and the enolates of diketones and ketoesters, which can exist
in protic solvents, give the mixture of the C- and O-alkylated products
with more C-isomer in protic solvents than in aprotic solvents. The
preference for the C-alkylation for less stable and more reactive
enolates is in line with the above intrinsic barrier argument in that the
reactivities of more reactive enolates is controlled by the
thermodynamic driving force to a larger extent. However, the different
behavior in the gas-phase and in solution and the solvent effect on the
10742
dx.doi.org/10.1021/jo302103c | J. Org. Chem. 2012, 77, 10738−10744

The Journal of Organic Chemistry
Article
Table 8. Activation and Reaction Enthalpies for the SN2 Reactions of MeCl with CH2NO2 and CH2CHO− with and
−
a
without Li+(Me2O)3 at B3LYP/6-31+G*
CH2NO2
CH2CHO−
−
b
b
cation
ΔH⧧
ΔH
δΔH⧧
ΔH⧧
ΔH
δΔH⧧
C-methylation
O-methylation
C-methylation
O-methylation
no
no
yes
yes
−1.3
0.2
−32.3
−8.5
−25.8
−2.1
−1.5
−4.6
−4.7
13.5
21.0
−41.3
−20.7
−31.3
−14.7
0.1
20.7
22.9
−2.2
−7.5
a
Energies are relative to the separated reactants (anion + MeCl, or anion-Li(OMe)2 complex + MeCl) in kcal/mol. Product states are C- or O-
b
methylated product + LiCl. Relative activation enthalpy for C-methylation over O-methylation.
Product Analysis. The reaction of 1-X (ca. 50 mmol/L), a slightly
excess amount of MeOBs, and a drop of toluene dissolved in DMF-d7
was carried out at 25 °C in an NMR sample tube. At preset intervals,
the NMR spectra were recorded and the remaining 1-X and the O-
methylated products were quantified with toluene as an internal
standard. Assignment of E- and Z-methylated products was made
according to the literature.4 Reactions in other solvents were carried
out similarly. The reaction of 2-X (ca. 40 mmol/L) with 2 equiv of
MeOBs in THF was carried out at 25 °C under N2 in a flask. At preset
intervals, a part of the reaction solution was quenched with water and
the organic layer was extracted with ether and subject to NMR
analysis. Trimethoxybenzene was used as an internal standard.
Computational Methods. Electronic structure calculations were
carried out with the Gaussian 03 program suite.22 Geometries were
fully optimized at the B3LYP/6-31+G* level of theory. Vibrational
normal-mode analyses were performed to ensure that each optimized
structure was a true minimum or a saddle point on the PES. Unscaled
frequencies were used to obtain thermochemical quantities, the
thermal enthalpies at 298 K.
Figure 6. Transition structures for the O-methylation of the Li+-
coordinated (A) nitronate and (B) enolate calculated at B3LYP/6-
31+G*.
enolate reactions is due to the metal-cation coordination, which
hinders O-methylation for enolates. Variation in the C/O selectivity
with an enolate stability is rationalized by a variable relative importance
of the thermodynamic driving force over the kinetic barrier in enolate
reactions.
ASSOCIATED CONTENT
* Supporting Information
Tables (S1−S3), Figures (S1, S2), and calculated geometrical
coordinates and energies of the species discussed. This material
is available free of charge via the Internet at http://pubs.acs.org.
■
S
EXPERIMENTAL SECTION
■
Materials. DMF was distilled over MS 4A. Deuterated solvents
were used as received: DMF-d7 (ACROS, 99.5% atom D), D2O
(Merck, 99.9% atom D), and CD3OD (ACROS, 99.8% atom D).
CD3OH was prepared by mixing CD3OD and H2O and fractionally
distilled. Ring-substituted phenylnitromethanes and 1-phenyl-2-prop-
anones were prepared as reported previously.14,21 A series of sodium
nitronates (1-X, X = p-MeO, p-Me, m-Me, H, p-Cl, m-F, m-Cl, p-CF3,
m-NO2) were prepared from the reactions of ring-substituted
phenylnitromethanes and NaOEt in diethyl ether.
AUTHOR INFORMATION
Corresponding Author
*E-mail: yamataka@rikkyo.ac.jp.
Notes
■
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
Sodium Nitronate (1-X). In a 100 mL flask were placed ethanol
(3 mL), dry ether (3 mL), and sodium (0.25 g, 0.011 mol). After the
sodium dissolved completely, the solution was diluted with dry ether
(30 mL), and then phenylnitromethane (1.189 g, 8.68 mmol) in dry
ether (10 mL) was added dropwise with stirring. The resultant
precipitate was collected, washed with dry ether, and dried. Yield,
82.2%. Other substituted salts (1-X) were prepared in a similar
manner.
Sodium Enolate (2-X). 2-X (X = p-MeO, p-Me, m-Me, H, p-Cl, p-
CF3) was prepared by mixing a preset amount of substituted 1-phenyl-
2-propanone and NaHMDS in THF and stirred for 1 h under N2 at 25
°C, and used without isolation.
The study was in part supported by SFR aid by Rikkyo
University and a Grant-in-Aid for Scientific Research (No.
22350023) from the Ministry of Education, Science, Sports,
Culture and Technology, Japan.
REFERENCES
■
(1) Breuer, E.; Aurich, H. G.; Nielsen, A. In Nitrones, Nitronates and
Nitroxides: The Chemistry of Functional Groups; Patai, S.; Rappoport,
Z., Eds.;Jorn, Wiley & Sons: New York, 1989; Chapter 3.
(2) Linton, B. R.; Scott, M.; Hamilton, A. D. Chem.Eur. J. 2000, 6,
2449.
(3) Thurston, J. T.; Shriner, R. L. J. Org. Chem. 1937, 2, 183.
(4) Kornblum, N.; Brown, R. A. J. Am. Chem. Soc. 1964, 86, 2681.
(5) Sato, M.; Kitamura, Y.; Yoshimura, N.; Yamataka, H. J. Org.
Chem. 2009, 74, 1268.
(6) Bug, T.; Lemek, T.; Mayr, H. J. Org. Chem. 2004, 69, 7565.
(7) Heiszwolf, G. J.; Kloosterziel, H. Recl. Trav. Chim. Pays-Bas. 1970,
89, 1153.
Kinetics. The rates of the reactions 1-X in DMF were studied
photometrically at 25.0 0.1 °C with the substrate concentration of
0.046 mmol/L in the presence of 10 equiv of MeOBs. The decay of
absorbance of 1-X at λmax (332−370 nm, depending on the
substituent) was followed. Excellent pseudo-first-order rate plots (R2
> 0.9997) were obtained. Reactions in 80% (v/v) aqueous DMF were
carried out with the substrate concentration of 0.10 mmol/L in the
presence of 50 equiv of MeOBs. Rates were measured at least 3 times,
and the reproducibility is listed in Table 2.
(8) Suama, M.; Sugita, T.; Ichikawa, K. Bull. Chem. Soc. Jpn. 1971, 44,
1999.
10743
dx.doi.org/10.1021/jo302103c | J. Org. Chem. 2012, 77, 10738−10744

The Journal of Organic Chemistry
Article
(9) Kurts, A. L.; Macias, A.; Beletskaya, I. P.; Reutov, O. A.
Tetrrahedron 1971, 27, 4759.
(10) Kurts, A. L.; Genkina, N. K.; Macias, A.; Beletskaya, I. P.;
Reutov, O. A. Tetrrahedron 1971, 27, 4777.
(11) Sarthou, P.; Bram, G.; Guibe, F. Can. J. Chem. 1980, 58, 786.
(12) Jones, M. E.; Kass, S. R.; Filley, J.; Barkley, R. M.; Ellison, G. B.
J. Am. Chem. Soc. 1985, 107, 109.
(13) Keeffe, J. R.; Morey, J.; Palmer, C. A.; Lee, J. C. J. Am. Chem.
Soc. 1979, 101, 1295.
(14) Ando, K.; Shimazu, Y.; Seki, N.; Yamataka., H. J. Org. Chem.
2011, 76, 3937.
(15) Chmurzynski, L.; Warnke, Z. Aust. J. Chem. 1993, 46, 185.
(16) Bernasconi, C. F. Acc. Chem. Res. 1987, 20, 301. Bernasconi, C.
F. Acc. Chem. Res. 1992, 25, 9. Bernasconi, C. F. Adv. Phys. Org. Chem.
1992, 27, 116. Bernasconi, C. F.; Wenzel, P. J. J. Am. Chem. Soc. 1994,
116, 5405. Bernasconi, C. F.; Wenzel, P. J. J. Am. Chem. Soc. 1996, 118,
11446. Bernasconi, C. F. Adv. Phys. Org. Chem. 2010, 44, 223.
(17) Pearson, R. G.; Dillon, R. L. J. Am. Chem. Soc. 1953, 75, 2439.
Kresge, A. J. Can. J. Chem. 1974, 52, 1897. Kresge, A. J.; Drake, D. A.;
Chang, Y. Can. J. Chem. 1974, 52, 1889. Bordwell, F. G.; Boyle, W. J.,
Jr. J. Am. Chem. Soc. 1971, 93, 511. Bordwell, F. G.; Boyle, W. J., Jr. J.
Am. Chem. Soc. 1972, 94, 3907. Bordwell, F. G.; Boyle, W. J., Jr; Yee,
K. C. J. Am. Chem. Soc. 1970, 92, 5926. M. Fukuyama, M.; Flanagan, P.
W. K.; Williams, F. T., Jr.; Frainier, L.; Miller, S. A.; Shechter, H. J. Am.
Chem. Soc. 1970, 92, 4689. Grinblat, J.; Ben-Zion, M.; Hoz, S. J. Am.
Chem. Soc. 2001, 123, 10738. Eliad, L.; Hoz, S. J. Phys. Org. Chem.
2002, 15, 540. Yamataka, H.; Mustanir; Mishima, M. J. Am. Chem. Soc.
1999, 121, 10223.
(18) Marcus, R. A. J. Phys. Chem. 1968, 72, 891. Cohen, A. O.;
Marcus, R. A. J. Phys. Chem. 1968, 72, 4249. Marcus, R. A. J . Am.
Chem. Soc. 1969, 91, 7224.
(19) Liou, L. R.; McNeil, A. J.; Ramirez, A.; Toombes, G. E. S.;
Gruver, J. M.; Collum, D. B. J. Am. Chem. Soc. 2008, 130, 4859.
Seebach, D. Angew. Chem., Int. Ed. Engl. 1988, 27, 1624.
(20) Yamataka, H.; Yamada, K.; Tomioka, H. In The Chemistry of
Organolithium Compounds; Rappoport, Z., Marek, I., Eds.; Wiley: New
York, 2004; Vol. 2, p 908.
(21) Yamamoto, Y.; Hasegawa, H.; Yamataka., H. J. Org. Chem. 2011,
76, 4652.
(22) Frisch, M. J. et al. Gaussian 03, Revision E.01; Gaussian, Inc.:
Wallingford, CT, 2004.
10744
dx.doi.org/10.1021/jo302103c | J. Org. Chem. 2012, 77, 10738−10744