Mechanisms of Acid-Base Catalysis of β-Elimination Reactions in Systems Activated by a Pyridine Ring

Article abstract of DOI:10.1021/jo030352z

β-Elimination reactions from 1 (in quinuclidine/quinuclidinium chloride, imidazole/imidazolium, and acetate/acetic acid buffers) and from 2 (in imidazole/imidazolium and acetate/acetic acid buffers) with formation of 4-vinylpyridine and 2-vinylpyridine, respectively, were studied. The results of a kinetic study of acid-base catalysis and H/D exchange are consistent with NH⁺, the protonated substrate, as the species that undergoes carbon deprotonation with an E1cb mechanism. The comparison with previously studied reactions in acetohydroxamate/acetohydroxamic acid buffer confirms this assignment. The high proton activating factor, PAF, value observed (PAF = 1.2 × 10 ⁶ with isomer 1 in quinuclidine/quinuclidinium buffer) can be explained with the high stability by the resonance of the intermediate carbanion.

Full text of DOI:10.1021/jo030352z

Mech a n ism s of Acid -Ba se Ca ta lysis of â-Elim in a tion Rea ction s in
System s Activa ted by a P yr id in e Rin g
Sergio Alunni* and Laura Ottavi
Dipartimento di Chimica. Universita` degli Studi Perugia, Via Elce di Sotto 8, 06123 Perugia Italy
alunnis@unipg.it
Received November 17, 2003
â-Elimination reactions from 1 (in quinuclidine/quinuclidinium chloride, imidazole/imidazolium,
and acetate/acetic acid buffers) and from 2 (in imidazole/imidazolium and acetate/acetic acid buffers)
with formation of 4-vinylpyridine and 2-vinylpyridine, respectively, were studied. The results of a
kinetic study of acid-base catalysis and H/D exchange are consistent with NH+, the protonated
substrate, as the species that undergoes carbon deprotonation with an E1cb mechanism. The
comparison with previously studied reactions in acetohydroxamate/acetohydroxamic acid buffer
confirms this assignment. The high proton activating factor, PAF, value observed (PAF ) 1.2 × 10
6
with isomer 1 in quinuclidine/quinuclidinium buffer) can be explained with the high stability by
the resonance of the intermediate carbanion.
In tr od u ction
nism7,10 is shown in Scheme 1 with 1 in acetohydrox-
amate/acetohydroxamic acid buffer at 50 °C, µ ) 1 M KCl.
The pKNa of conjugated acid of 1, NH+, in the reaction
condition, is 4.85 and the range of pH studied was 8.45-
9.15. Under these conditions it has been demonstrated7
that the reacting species, which undergoes carbon depro-
tonation, is NH+, even if this species is present in solution
at a very low concentration with respect to that of the
unprotonated substrate of 1, N. For this mechanism eq
1 is valid7 (in the reaction conditions of the study the
contribution of the elimination reaction induced by H2O
was negligible).
The mechanism of base-induced 1,2-elimination reac-
tions1 has been studied owing to the various interesting
mechanistic possibilities: the concerted process E2, the
mechanisms involving the intermediate carbanion, or the
mechanism involving the intermediate carbocation. In
previous studies2-5 of systems activated by a nitrophenyl
group, the mechanism of base-induced â-elimination
reactions has been shown to be E1cb (ΑxhDH + DN)6 with
the intermediate carbanion formed with different degrees
of reversibility depending on the structure of the sub-
strate. The substrates studied were activated by p-nitro-
phenyl2-4 or o-nitrophenyl5 groups and had a tertiary
amine as leaving group. Using a combination of three
techniques, acid-base catalysis studies, H/D exchange,
and solvent isotope effect, it was possible to demonstrate
a carbanion mechanism and to evaluate the kinetic
parameters of the process. An extension of these studies
to systems activated by a pyridine ring has shown the
importance of catalysis by protonation of the nitrogen
atom of the aromatic ring.7-10 The proposed mecha-
+
NH+
OH
(k -[OH-] + kNH [B])
[H+]
KaN
+
B
kobs ) kN2 H
‚
(1)
NH+
H2O
NH+
2
NH+
BH
k
+ k
+ k
[BH]
The reactivity of NH+ was much larger than that of
+
+
N; the kNBH / kBN ratio (kBNH and kNB are the second-order
rate constants for the elimination reaction by a base B
from NH+ and N, respectively) is the proton activating
factor, PAF (2.7 × 105 for isomer 1 and 5.2 × 106 for
isomer 2), and the large activation by protonation of the
nitrogen atom can be explained7 by the strong stabiliza-
tion by resonance of the intermediate carbanion formed
from NH+. The PAF parameter is important in chemis-
try11 and biochemistry.12 Kinetic analysis allows the
(1) (a) Saunders: W. H., J r.; Cockerill, A. F. Mechanism of Elimina-
tion Reaction; Wiley-Interscience Publication, New York, 1973; Vol.
II, pp 60-68. (b) Baciocchi, E. The Chemistry of Halides, Pseudo
Halides and Azides, Supplement D; Patai, S., Rappoport, Z., Eds.;
Wiley: Chichester, UK, 1983. (c) Gandler, J . R. The Chemistry of
Double Bonded Functional Group; Patai, S., Ed., Wiley: Chichester,
UK, 1989. (d) Saunders, W. H., J r. Acc. Chem Res. 1976, 9, 19. (e)
Stirling, C. J . M. Acc. Chem. Res. 1979, 12, 198. (f) Marshall, D. R.;
Thomas, P. J .; Stirling, C. J . M. J . Chem. Soc., Perkin Trans. 2 1977,
1914. (g) Cavestri, R. C.; Fedor, L. R. J . Am. Chem. Soc. 1970, 92,
4610. (h) Fedor, L. R.; Glave, W. R. J . Am. Chem. Soc. 1971, 93, 985.
(i) Gandler, J . R.; J encks, W. P. J . Am. Chem. Soc. 1982, 104, 1937.
(2) Alunni, S.; J encks, W. P. J . Am. Chem. Soc. 1980, 102, 2052.
(3) Keeffe, J . R.; J encks, W. P. J . Am. Chem. Soc. 1983, 105, 265.
(4) Alunni, S.; Tijskens, P. J . Org. Chem. 1995, 60, 8371.
(5) Alunni, S.; Ruzziconi, R.; Teofrasti, O. Res. Chem. Intermed.
1999, 25, 483.
(8) Alunni, S.; Busti, A. J . Chem. Soc., Perkin Trans. 2 2001, 778.
(9) Alunni, S.; Laureti, V.; Ottavi, L.; Ruzziconi, R. J . Org. Chem.
2003, 68, 718.
(10) Alunni, S.; Conti, A.; Palmizio Errico, R. Res. Chem. Intermed.
2001, 27, 653.
(11) (a) Stewart, R.; Srinivasan, R. Acc. Chem. Res. 1978, 11, 271.
(b) McCann, G. M.; More O’Ferrall, R. A.; Walsh, S. M. J . Chem. Soc.,
Perkin Trans. 2 1997, 2761. (c) Eustace, S. J .; McCann, G. M.; More
O’Ferrall, R. A.; Murphy, M. G.; Murray, B. A.; Walsh, S. M. J . Phys.
Org. Chem. 1998, 11, 519.
(12) Abeles, R. H.; Frey, P. A.; J encks, W. P. Biochemistry; J ones
and Bartlett: Sudbury, MA, 1992.
(6) Guthrie, R. D.; J encks, W. P. Acc. Chem. Res. 1989, 22, 343.
(7) Alunni, S.; Conti, A.; Palmizio Errico, R. J . Chem. Soc., Perkin
Trans. 2 2000, 453.
10.1021/jo030352z CCC: $27.50 © 2004 American Chemical Society
Published on Web 03/03/2004
2272
J . Org. Chem. 2004, 69, 2272-2283

Acid-Base Catalysis of â-Elimination Reactions
SCHEME 1
F IGURE 1. Plot of the pseudo-first-order rate constants kE
against [B] for the elimination reaction with 1 in quinuclidine/
quinuclidinium buffers at 50 °C, µ ) 1 M KCl: (9) pH ) 10.1,
(b) pH ) 10.24, (4) pH ) 10.41, (0) pH ) 10.68, (1) pH ) 11.
F IGURE 2. Plot of kA/[B] against [BH] for the addition
reaction with 4-vinylpyridine in quinuclidine/quinuclidinium
buffers at 50 °C, µ ) 1 M KCl: (9) pH ) 10.1, (b) pH ) 10.24,
(4) pH ) 10.41, (0) pH ) 10.68, (1) pH ) 11.
+
+
NH+
BH
kNBH , k
/kN2 H , and k∞ values to be determined; k∞ is
kobs at [buffer] f ∞ (eq 10). The reacting species, NH+,
was identified using acid-base catalysis studies. The plot
of kobs against [B] showed curvatures indicating a step-
wise mechanism with a change, induced by [BH], in the
rate-determining step from carbon deprotonation to leav-
ing group expulsion. However, the leveling-off at high
[buffer] (k∞) was constant and independent of pH, ac-
cording to eq 10; this is evidence that NH+ is the species
that undergoes carbon deprotonation. H/D exchange
experiments showed the incorporation of deuterium at
the â-carbon during the elimination reaction in D2O, in
acetohydroxamate/acetohydroxamic acid buffer, at 50 °C
and µ ) 1 M KCl, in agreement with a stepwise reversible
mechanism, (E1cb)R(AxhDH + DN*).6 Further support for
the proposed mechanism has come from studies of the
solvent kinetic isotope effect.10
Resu lt a n d Discu ssion
Stu d y of Acid -Ba se Ca ta lysis in Qu in u clid in e/
Qu in u clid in iu m (Q/QH+) Bu ffer s. Equilibration of
4-vinylpiridine in Q/QH+ buffer, at 50 °C and µ ) 1 M
KCl, was followed by monitoring the decrease of absor-
bance of 4-vinylpiridine at 280 nm. The process of
addition of quinuclidine to 4-vinylpiridine in Q/QH+
buffer to give 1 is an equilibrium reaction involving the
reverse elimination reaction from 1 to give quinuclidine
and 4-vinylpyridine (P and PH+). The observed pseudo-
first-order rate constant, kobs, is the sum kE + kA, where
kE and kA are the observed pseudo-first-order rate
constants for elimination and addition reactions, respec-
tively. The two observed rate constants, kE and kA, were
calculated as shown in the Experimental Section from
the values of A0 (absorbance at λ ) 280 nm, t ) 0), A∞
(absorbance at λ ) 280 nm, t ) ∞, equilibrium), and At
(absorbance at λ ) 280 nm, t ) t). In Figure 1 the plot of
kE against [B], the concentration of the basic buffer
component, at various pH values is shown.
Linearity is observed, but the slope at each pH in-
creases by increasing the [H+], while the intercept
increases by increasing the [OH-]. We have observed an
exponential trend in the plots of kA against [B], while the
plot of kA/[B] against [BH] is linear (Figure 2).
Good linearity is observed in the plot of Figure 2 and
both the intercept and the slope depend on [H+]. As will
be discussed later, these results are in agreement with
an equilibrium involving the elimination-addition reac-
In this paper, we extend these studies on N-[2-(4-
pyridyl)ethyl]quinuclidinium (1) and N-[2-(2-pyridyl)-
ethyl]quinuclidinium (2), at 50 °C, µ ) 1 M KCl, in
different buffer systems and different pH ranges. The
pKNa of 2 is 3.81 at 50 °C, µ ) 1 M KCl.7 Because the
competition between N and NH+ depends on the ratio
between the rate constants of the two species and on the
[NH+]/[N] ratio, we wanted to check the consistency of
the previously made mechanistic assignment under dif-
ferent conditions of pH and buffers. The results obtained
support the mechanistic assignment; in fact, by analyzing
the kinetic parameters obtained and comparing them
with those of the other substrates and/or different reac-
tion conditions a significant consistency is found.
J . Org. Chem, Vol. 69, No. 7, 2004 2273

Alunni and Ottavi
SCHEME 2
be kinetically distinguished, and we have no specific
evidence on this point. However, the value of kNOH with
-
isomer 1, kNOH- ) 3.45 × 10-3 M-1 s-1 at 50 °C and µ ) 1
M KCl, is close5 to the second-order rate constant for the
OH--induced elimination reaction at 50 °C with N-[2-(o-
-3
-
nitrophenyl)ethyl]quinuclidinium ion, kOH ) 1.58 × 10
M-1 s-1. With this second substrate, an E1cb mechanism
was demonstrated for the base-induced â-elimination
reaction. This is an important point to be considered even
if it is not clear whether the same mechanism can operate
in systems with similar levels of activation. In the
following kinetic treatment and related discussion, an
(E1cb)I mechanism with N (Scheme 2) will be assumed.
However, if the mechanism with N is E2 concerted, the
same expression of kE and kA will be obtained, but with
the rate constants kNOH-, kN, kHN O(k-N2/kN2 ), and kNBH(kN-2
/
B
2
kN2 ), which refer to the concerted process. Equation 3
explains the nonlinearity of the plot kA against [B], owing
to the significant contribution of the term containing the
[B][BH] product, while it is in agreement with the
linearity observed in the plot of Figure 2. From eq 2 and
a plot of kE vs [B], it is possible to evaluate the intercept,
iE, and the slope, sE:
tion, with a competition between the unprotonated N and
protonated NH+ substrate in the elimination reaction and
with the related competition between the unprotonated
P and the protonated PH+ in the addition direction. The
mechanism shown in Scheme 2 is E1cb with both the
unprotonated N and protonated NH+ substrate.
Our results are in agreement with an (E1cb)I mecha-
+ [OH-][H+]
nism. In fact, under these conditions, kN2 . kHN O + kBNH
iE ) kONH-[OH-] + kNH
(4)
(5)
2
-
OH
+
NH+
H2O
NH+
BH
KaN
and kN2 H . k
+ k
, the expression of kE is given by
+ [OH-][H+]
+ [H+]
kE ) kONH-[OH-] + kNH
+
sE ) kNB + kNBH
-
OH
KaN
KaN
+ [H+]
kNB + kNBH
[B] (2)
From the plots iE vs [OH-] and sE vs [H+]/KaN the
KaN
(
)
+
second-order rate constants kNOH-, kN, and kNH can be
B
B
evaluated (Table 1). kNOH and kNB are the rate constants
where KNa ) 1.41 × 10-5 M is the acid dissociation
constant of NH+ at 50 °C and µ ) 1 M KCl.7 It follows
that the rate-determining step in the reverse direction,
the addition reaction, is the protonation of the intermedi-
ate carbanions, I1 and I2, formed respectively from P and
PH+ by H2O and BH. The expression of kA for this
mechanism is then given by
-
for the formation of the intermediate carbanion by OH-
or quinuclidine base, if the mechanism is (E1cb)I (Scheme
2). They will represent kNOH (concerted) and kBN (con-
-
certed), the rate constants for a E2 concerted process from
+
N, if this is the operating mechanism. The kNBH /kBN ratio
is the proton activating factor (PAF) with the base
quinuclidine, PAF ) 1.2 × 106. Again, this parameter
represents the quantification of the catalysis of the
elimination reaction by protonation of the nitrogen atom
of the pyridine ring within an (E1cb)I mechanism as in
Scheme 2; in the case of an E2 mechanism with N the
PAF value is associated with a change in mechanism
from E2 with N to (E1cb)I with NH+. The kinetic analysis
is, then, consistent with an E1cb irreversible mechanism
from NH+, (E1cb)I (Scheme 2), where carbon deprotona-
tion is the rate-determining step. The alternative pos-
sibility of an (E1cb)R mechanism, with rate-determining
the expulsion of the leaving group, is kinetically incon-
sistent with our results. The implication for the reverse
reaction is that the mechanism for the addition of
quinuclidine to 4-vinylpyridine (protonated and unpro-
tonated, PH+ and P) involves the formation of the
intermediate carbanion in a reversible process, followed
by rate-determining carbon protonation of the carbanion
by BH and H2O and the formation of 1. For this
mechanism (Scheme 2), eq 3 can be derived by steady-
state approximation. The plots (kA/[B] vs [BH]) are shown
in Figure 2, and good linearity is observed. The intercepts
kN-2
k2N
kA ) (kNH O + kNBH[BH]) [B] +
2
NH+
-2
[H+]
KaP
k
NH+
H2O
NH+
BH
(k
+ k
[BH])
[B] (3)
NH+
2
k
where KPa ) 2.29 × 10-6 M is the acid dissociation
constant of PH+ at 50 °C and µ ) 1 M KCl. It must be
considered that in Scheme 2 an (E1cb)I mechanism is
assumed for the elimination reaction with the unproto-
nated N and protonated NH+ substrate. However, we
assign the (E1cb)I mechanism with NH+, because of the
large value of PAF observed, a consistent value of kB and
analogy to the reaction in acetohydroxamate/acetohy-
droxamic acid7 or imidazole/imidazolium or acetate/acetic
acid buffers, where the observed change in the rate-
determining step can conclusively be related to the E1cb
mechanism (see the related discussion in the text). The
mechanism of the elimination reaction with N can
instead be E1cb or E2 concerted (AxhDHDN).6 They cannot
2274 J . Org. Chem., Vol. 69, No. 7, 2004

Acid-Base Catalysis of â-Elimination Reactions
TABLE 1. Ra te Con sta n ts for th e Elim in a tion Rea ction w ith Isom er 1 Accor d in g to Sch em e 1 in Va r iou s Bu ffer
System s a t 50 °C, µ ) 1 M KCl
acetate/
imidazole/
acetohydroxamate/
quinuclidine/
acetic acid
imidazolium
acetohydroxamic acid
quinuclidinium
pKBa
a
4.65
6.78
9.15
11.0
103 × kNOH- (M-1 s-1
)
3.5b (3.45)c,d
105 × kNB (M-1 s-1
)
7.8d,e
20.8
64
5.21
63.3f
(0.53)
f
+
8.8 × 10-4
74.6
1.87
1.4 × 10-2
7.64
2.2
kNBH (M-1 s-1
)
+
+
kNBH /kN2 H (M-1
)
105 × k∞ (s-1
)
1.62
pKBa of the conjugated acid of the buffer at 50 °C, µ ) 1 M KCl. Obtained from the plot iE vs [OH-] (eq 4): y ) 0.3 × 10-6 (SD ) 0.6
a
b
× 10-6) + 0.0035 (SD ) 0.002)x (r ) 0.9946). c Value directly measured in OH-/H2O at 50 °C, µ ) 1 M KCl. Reference 7. e Value estimated
d
NH+
OH
with a previously proposed LFER; the change in the rate constant k
from this work with respect to that previously reported (ref 7)
-
does not change the parameters of the LFER. f Obtained from the plot sE vs [H+]/KNa (eq 5): y ) 5.21 × 10-5 (SD ) 1.2 × 10-5) + 63.3 (SD
) 3.5)x (r ) 0.9957).
in Figure 2, iA, and the slopes, sA, are both dependent on
for isomer 2) were measured by following the formation
of 4-vinylpyridine (P) in equilibrium with its conjugated
acid (PH+) from 1 at λ ) 280 nm or 2-vinylpyridine (P in
equilibrium with its conjugated acid PH+) from 2 at λ )
290 nm, H2O, µ ) 1 M KCl, 50 °C. The total concentration
of the product is C ) [P] + [PH+]. As will be shown by
kinetic analysis of the data, the results are consistent
with the mechanism of Scheme 1, with both isomers 1
and 2. For this mechanism the kinetic expression is given
by
[H+]:
NH+
kN-2
k2N
k
-2
[H+]
KaP
NH+
H2O
iA ) kNH O
+ k
(6)
(7)
NH+
2
k
2
NH+
kN-2
k2N
k
-2
[H+]
KaP
NH+
BH
sA ) kNBH
+ k
NH+
k
2
Using linear regression analysis, the plot of intercepts
iA vs [H+]/KPa (eq 6) gives y ) 126.4 × 10-5 (SD ) 10 ×
10-5) + 35.14 (SD ) 5)x (r ) 0.9728). The new intercept
C )
+
NH+
H2O
NH+
OH-
k′
+ k
+
[OH-] + kNH [B]
B
+
kN2 H
‚
‚[NH+]‚t
NH+
H2O
NH+
+
of this secondary plot is i′ ) 126.4 × 10-5 M-1 s-1 and
A
k
+ kN2 H + kNHH [H+] + k
[BH]
BH
+
the new slope is s′ ) 35.14 M-1 s-1. Again, using linear
A
regression analysis, the secondary plot of sA vs [H+]/KaP
(eq 7) gives y ) 100 × 10-5 (SD ) 110 × 10-5) + 371 (SD
(8)
We have chosen to use C ) kobs‚[N]‚t, where [N] is
calculated by [N] ) {KNa /([H+] + KaN)}‚[S]0 and [NH+] )
[N]‚[H+]/KaN. With this treatment, kobs was calculated by
a plot of [C] vs t (kobs ) slope/[N]); the expression for kobs
(Scheme 1) is then given by
) 52)x (r ) 0.9711). The value for the new intercept, i′′,
A
is 100 × 10-5 M-2 s-1, the related slope value, s′′, is 371
A
M-2 s-1. The ratio i′ /i′′ is kHN O/kBNH ) 1.26 M, while the
A
A
2
+
NH+
BH
ratio s′ /s′′ is kNHHO /k
) 0.095 M. It can be seen that
A
A
the protonation 2of I2 by BH is faster than by H2O, in
agreement with the higher acidity of BH. The related
comparison for the protonation of I1 must take into
kobs
)
NH+
H2O
NH+
+
+ k -[OH-] + kNH [B]
[H+]
KaN
k′
account the large standard deviation of i′′. However, the
OH B
A
+
kN2 H
‚
‚
parameter obtained would also be consistent owing to the
expected lower selectivity of I1 with respect to I2 and the
NH+
H2O
NH+
BH
+
+
k
+ kN2 H + kNHH [H+] + k
[BH]
+
fact that kNH O is a first-order rate constant with stan-
(9)
2
dard state for [H2O] ) 55 M. In the case of an E2
mechanism with N, i′ would be kNH O (concerted) and i′′
Equation 9 is the equation of a curve for a plot of kobs
vs [B], and at high [buffer] it becomes
A
A
would be kNBH (concerted); the ratio 2i′ /i′′ is then related
A
A
+
NH+
BH
to the concerted addition reaction. The term k
/kN2 H
+
+
kNBH [B]
kNBH KBa
is an interesting parameter that can be estimated. In fact,
[H+]
KaN
+
+
k∞ ) kN2 H
‚
‚
) kN2 H
‚
‚
)
+
+
a value of kNHHO /kN2 H ) 0.05 was previously reported,7 so
NH+
BH
NH+
KaN
2
k
[BH]
k
BH
+
NH+
BH
NH+
BH
from the ratio kNH /k
) 0.095 M, a value for k
/
NH+
2
H2O
+
kN2 H ) 0.53 M-1can be calculated. As discussed in a
previous paper,7 this ratio controls the degree of curva-
ture that can be observed in a plot of kE vs [B] (Figure
1). The value of 0.53 M-1, estimated in a Q/QH+ buffer
system, is consistent with the lack of curvature found in
Figure 1.
k
‚KTAUT (10)
where KBa is the acid dissociation constant of the buffer
BH. For the imidazolium cation KBa ) 1.66 × 10-7
M
(pKB ) 6.78) at 50 °C and µ ) 1 M KCl and KTAUT is the
a
equilibrium constant for the tautomerization from N to
I2 (Scheme 1). The k∞ value is independent of pH and is
the same as that derived in a previous work7 from eq 1
at pH ≈ 9 in acetohydroxamate/acetohydroxamic acid
buffer. Our choice to calculate kobs (eq 9) is related to the
important point in our discussion that the term k∞ is
Acid -Ba se Ca t a lysis in Im id a zole/Im id a zoliu m
Bu ffer s. The pseudo-first-order rate constants (kobs
,
initial rate) for the elimination reaction in imidazole/
imidazolium buffers at different pH values (pH ) 6.29,
6.71, 6.96, 7.27 for isomer 1 and pH ) 6.48, 6.81, 7.07
J . Org. Chem, Vol. 69, No. 7, 2004 2275

Alunni and Ottavi
TABLE 2. Ra te Con sta n ts for th e Elim in a tion Rea ction
w ith Isom er 2 Accor d in g to Sch em e 1 in Va r iou s Bu ffer
System s a t 50 °C, µ ) 1 M KCl
acetate/
acetic
acid
acetohydroxamate/
acetohydroxamic
acid
imidazole/
imidazolium
pKBa
a
4.65
6.78
9.15
4.8b,c
24.8
38
106 × kNB (M-1 s-1
)
+
5.97 × 10-4 2.14 × 10-2
kNBH (M-1 s-1
)
+
+
kNBH /kN2 H (M-1
)
27.63
3.12
9.48
2.42
106 × k∞, (s-1
)
2.98
pKBa of the conjugated acid of the buffer at 50 °C, µ ) 1 M
a
b
KCl. Value estimated with a previously proposed LFER; the
F IGURE 3. Dependence of kobs on [buffer] for the elimination
reaction with 1 in imidazole/imidazolium buffer at different
pH values, 50 °C, µ ) 1 M KCl: (9) pH ) 6.29, (O) pH ) 6.71,
(2) pH ) 6.96, (0) pH ) 7.27. Solid lines are calculated from
eq 14 and rate constants from Table 1.
NH+
-
change in the rate constant k
from this work with respect to
OH
that previously reported (ref 7) does not change the parameters
of the LFER. c Reference 7.
NH+
OH
for isomer 1 and k
) 290 M-1 s-1 for isomer 2 can be
-
expected to be consistently independent of pH in several
buffer systems. The plots of kobs against [B] at various
pH values in imidazolium/imidazole buffer for substrate
1 is shown in Figure 3. Similar behavior was observed
for isomer 2.
calculated from eq 12. These values can be compared with
those previously7 determined by a mathematical extra-
polation at [buffer] ) 0 in acetohydroxamate/aceto-
NH+
OH
hydroxamic acid buffer pH ) 8.45-9.45: k
) 191.1
-
NH+
M-1 s-1 (isomer 1) and kOH- ) 442.05 M-1 s-1 (isomer 2).
We consider these values in agreement, owing to the
The parameters of the mechanism reported in Scheme
1 were calculated from eq 9 with the approximation that
+
different procedures and conditions used. The calculated
the term kNHH [H+] as the denominator is not signifi-
+
NH+
H2O
values of k′
are 2.54 × 10-6 s-1 (isomer 1) and 8.75 ×
cant. This term should be present conceptually for the
principle of microscopic reversibility, but it is numerically
not significant with respect to the others terms, so it is
neglected in order to simplify the calculations. Evidence
of the validity of this approximation is the consistency
of the model obtained (Figure 3).
The experimental determination of the observed pseudo-
first-order rate constants for the reaction of NH+ in OH-/
H2O, in absence of buffer (the solution was buffered by
the reactants), by initial rates was done using the kinetic
expression
10-7 s-1 (isomer 2).
The expression for kobs, as described in the Experimen-
tal Section, is given by eq 9. This eq can be transformed
into eqs 13 and 14:
NH+
H2O
NH+
2
k
+ k
NH+
2
(kobs - k0)
[B]
[H+]
KaN
+
‚
)
‚ kNBH
‚
-
[BH]
[BH]
k
NH+
BH
NH+
2
k
kobs
k
(13)
(14)
NH+
H2O
NH+
OH
+ k -[OH-]
[H+]
KaN
k′
+
C ) kN2 H
‚
‚
‚[N]‚t (11)
[H+]
+
NH+
H2O
+
+
k′ + kNBH ‚ [B] ‚
k
+ kN2 H + kNHH [H+]
+
0
KaN
kobs
)
NH+
H2O
NH+
2
NH+
BH
NH+
2
From the plot of C vs time the term slope/[N] is k0. A
study of k0 with isomer 1 and 2 at different pH values,
in the 5.2-6.35 range, showed some scatter but was
consistent with a linear dependence of k0 vs [H+]/KaN, in
agreement with an (E1cb)I mechanism, where the base
H2O competes with OH-, and with eq 12:
k
+ k
NH+
2
k
k
+
+
‚ [BH]
k
+
+
where k′ ) k0(kNH + k2NH )/kN2 H . Kinetic parameters
0
H2O
were calculated from the intercept and from the slope of
+
plot {(kobs - k0)/[BH]}‚1.05 vs kobs (eq 13). The kNBH and
+
kN2 H
KW
KaN
+
NH+
BH
k
/kN2 H values found for isomers 1 and 2 are reported
NH+
OH
k0 )
‚ k
‚
+
-
NH+
H2O
NH+
2
in Tables 1 and 2, respectively.
Figure 3 shows the fit of kobs with these rate constants
and eq 14 at the various pH values of work for isomer 1
k
+ k
NH+
2
[H+]
KaN
k
NH+
H2O
NH+
H2O
+
NH+
BH
and 2, respectively. Alternatively, the kNBH and k
/
‚ k′
‚
(12)
NH+
2
+
kN2 H values can be calculated by eq 14 with nonlinear
fitting of the data. Good agreement ((5%) between the
two methods was found. The curvature at each pH is
consistent7 with a carbanion mechanism and with a
change in the rate-determining step from carbon depro-
k
+ k
+
+
NH+
H2O
The term (k
+ kN2 H )/kN2 H ) 1.05 was known from a
previous work.7 The linear regression with eq 12 of k0 vs
[H+]/KNa is k0 ) 0.958 × 10-6 (SD ) 0.11 × 10-6) + 2.42
× 10-6 (SD ) 0.51 × 10-6)x with isomer 1 and k0 ) 1.05
× 10-7 (SD ) 0.08 × 10-7) + 8.33 × 10-7 (SD ) 2.6 ×
+
NH+
tonation at low [BH] (kN2 H > k
‚[BH], mechanism
(E1cb)I) to leaving group expulsion at high [BH], where
BH
10-7)x with isomer 2. The standard deviations are
reprotonation of the intermediate carbanion by BH
NH+
OH
+
reported in parentheses. A value of k - ) 241.3 M-1 s-1
is faster than the leaving group expulsion (k2NH
<
2276 J . Org. Chem., Vol. 69, No. 7, 2004

Acid-Base Catalysis of â-Elimination Reactions
The curvatures observed are in agreement with Scheme
1 (with the two isomers) and eq 9. The calculated values
of the rate constants (eq 13) are reported in Tables 1 and
2.
H/D Exch a n ge. H/D exchange experiments confirmed
the reversible E1cb mechanism at appropriate [buffer]
for the â-elimination reaction from 1 or 2 in imidazole/
imidazolium buffers. It was shown that 1 incorporated
deuterium in the â-position with respect to the leaving
group during the elimination reaction in D2O, imidazole/
imidazolium buffers. The technique used was previously
described.4,5 In an experiment with 1 in D2O at 50 °C, µ
) 1 M KCl, [B] ) [BD] ) 0.09 M, when the elimination
reached 49% of the 4-vinylpyridine formed, H/D exchange
was 38%. Similar experiments gave consistent results.
With 2 in D2O at 50 °C, µ ) 1 M KCl, [B] ) 0.05 M, [BD]
) 0.1 M, when the elimination reached 41% of the
2-vinylpyridine formed, H/D exchange was 14%. Consis-
tent results were also obtained with the same technique
in acetate/acetic acid buffer, D2O at 50 °C, µ ) 1 M KCl.
For example, with 1 (D2O, at 50 °C, µ ) 1 M KCl, [B] )
[BD] ) 0.05 M) the formation of ∼15% of 4-vinylpyridine
corresponded to 28% H/D exchange in the substrate. With
2 (D2O, at 50 °C, µ ) 1 M KCl, [B] ) 0.2 M, [BD] ) 0.1
M) the formation of 22% of 2-vinylpyridine corresponded
to 15% of H/D exchange in the substrate. It should be
noted that the consistency of H/D exchange experiments
could be qualitatively evaluated by the known term
F IGURE 4. Dependence of kobs on [buffer] for the elimination
reaction with 1 in acetate/acetic acid buffer at different pH
values, 50 °C, µ ) 1 M KCl: (9) pH ) 4.14, (O) pH ) 4.69, (2)
pH ) 4.94, (0) pH ) 5.12. Solid lines are calculated from eq
14 and rate constants from Table 1.
SCHEME 3
+
NH+
k
/kN2 H ; in fact the ratio (% H/D exchange)/(% vi-
BH
nylpyridine formed) should be consistent with the term
+
NH+
k
/kN2 H ‚[BD]. However, the lack of knowledge about
BD
NH+ NH+
the primary kinetic isotope effect, k
/kBD , and, in
BH
general, the difficulty in getting quantitative results for
H/D exchange experiments in our systems makes this a
qualitative techniques, but it strongly supports the
assigned mechanism. The presence of H/D exchange, at
[BD] where some reversibility in the formation of the
carbanion is expected, is evidence of the presence of the
intermediate carbanion.
Com p a r ison of Kin etic P a r a m eter s of th e Elim i-
n a tion Rea ction fr om 1 a n d 2 w ith Differ en t Bu ffer
System s. The rate constants calculated for 1 and 2 in
NH+
BH
k
‚[BH], mechanism (E1cb)R). This mechanistic pos-
sibility is a partially reversible E1cb mechanism, where
the two barriers for leaving group expulsion and for
reprotonation of the carbanion are close enough to make
possible the change in the rate-determining step within
the range of [buffer] used; it will be indicated also as
(E1cb)R. The result that the leveling-off at high [buffer]
(Figure 3) is independent of pH is in agreement with
carbon deprotonation occurring from NH+ (eq 10). The
large reactivity of NH+ with respect to N can be related
to the operating E1cb mechanism and to strong stabiliza-
tion by resonance of the intermediate I2 formed from NH+
(Scheme 3).
If the base-induced carbon deprotonation was from the
unprotonated substrate, N, the value of k∞ would be
dependent on OH- (this point is discussed later). The
process (Scheme 1) from N to I2 is a tautomerization
process, and at high [BH] it will be a real equilibrium.
The overall process (and k∞) then becomes pH-indepen-
dent, since both the tautomerization equilibrium and the
leaving group expulsion are pH-independent.
Acid -Ba se Ca ta lysis in Aceta te/Acetic Acid Bu ff-
er s. Pseudo-first-order rate constants (kobs, initial rate)
in acetate/acetic acid buffers, 50 °C and µ ) 1 M KCl,
were measured as previously described with imidazole/
imidazolium buffers in the Experimental Section. Kinetic
treatment and rate constants evaluation were also the
same. The plot of kobs against [B] for isomer 1 are reported
in Figure 4. Similar behavior was observed with isomer
2.
different buffer systems are reported in Table 1 and Table
+
2, respectively. The values of kNBH , the second-order-
rate constant for carbon deprotonation from NH+, are
consistently related to the strength of the attacking base.
+
+
The ratio kNBH (1)/kNBH (2) is 1.47 in acetate, 0.65 in
imidazole, and 0.84 in acetohydroxamate. Application of
+
the Brønsted eq13,14 with kNBH for isomer 1 is shown in
Figure 5.
The â value obtained is 0.8. For substrate 2 there are
+
only three kNBH values, anyway the linear regression
+
analysis (log kNBH against pKBa + log(p/q)) gives: y )
-8.2 (SD ) 1.7) + 1.0 (SD ) 0.2)x (r ) 0.9731). However,
these values are only to be considered indicative, because
the set of bases used is not homogeneous, with N or O as
the attacking atom and with acetohydroxamate base that
could give the R effect.14 It can be noted that the values
(13) Bell, R. P. The Proton in Chemistry; Cornell University Press:
Ithaca, NY, 1973.
(14) Chapman, N. B.; Shorter, J . Advances in Linear Free Energy
Relationships; Plenum Publishing Company Ltd.: London, 1972; pp
301-303.
J . Org. Chem, Vol. 69, No. 7, 2004 2277

Alunni and Ottavi
SCHEME 4
with the two isomer are high enough so that some
reversibility in the formation of the carbanion is experi-
mentally observed in the range of the [BH] used. The
+
NH+
BH
variation of k
/kN2 H with the structure of BH, the
+
NH+
F IGURE 5. Brønsted plot of log kNBH against pKaB + log(p/q)
protonating acid, is related to the variation of k
,
BH
for isomer 1 (50 °C, µ ) 1 M KCl). pKBa is related to the acid of
the buffer; p is the number of equivalent protons in the
conjugate acid, q is the number of equivalent basic nitrogens
in the base. The bases are acetate, imidazole, acetohydrox-
amate, and quinuclidine. Linear regression analysis gives the
equation y ) -7.1(SD ) 1.3) + 0.84(SD ) 0.16)x, r ) 0.9526.
+
because kN2 H is independent of the buffer species. The
NH+
BH
+
values of k
/kN2 H , estimated in Q/QH+, is 0.53 M-1
(Table 1); this value is related to the low acidity of QH+
NH+
BH
(then the low value expected for k
) and determines
the lack of curvature in the plot of kE vs [B]. The order
+
NH+
BH
NH+
BH
of k
(or k
/kN2 H ) is acetate > acetohydroxamate >
+
of kNBH for isomer 1 are very close to those of isomer 2
imidazole > quinuclidine for isomer 1 and acetohydrox-
amate > acetate > imidazole for isomer 2. There is no
direct relationship to the pKBa of the acid (BH); this
could be explained by a specific interaction due to the
geometry of the different structure of BH.
with the same base, whereas when the reacting species
is the unprotonated substrate N and the base is OH-,
kNOH ) 3.45 × 10-3 M-1 s-1 (isomer 1)7 and kONH ) 0.27
-
-
NH+
× 10-3 M-1 s-1 (isomer 2).7 The kOH- value with isomer 1
is 241.3 M-1 s-1 and 290 M-1 s-1 with isomer 2 (this
For the elimination reaction from NH+ in different
buffer systems and in the pH range used, an important
check of the consistency of the mechanistic assignment,
E1cb with some degree of reversibility, is related to the
fact that k∞, the observed first-order rate constants at
high [buffer], when the leveling-off occurs, is expected to
be independent of the pH and buffer (eq 10). In a
+
NH+
OH
work), the ratio k -(1)/kNH-(2) is 0.83. When the react-
OH
ing species is NH+, the E1cb mechanism, involving an
intermediate carbanion (Scheme 1), can be assigned,
whereas with the less activated unprotonated substrate,
N, it is not clear if the mechanism is (E1cb)I or E2. The
different behaviors of the two isomers with respect to
+
NH+
BH
NH+
OH
and kNOH could be explained by the different reso-
condition of full reversibility, when k
[BH] . kN2 H
,
k
-
-
+
nance energies and different electrostatic interactions15
if the mechanism is E1cb, or it could also be related to
the different operating mechanism, if there is a change
in mechanism from N to NH+.
kobs becomes k∞ ) kN2 H KTAUT, where KTAUT is the equi-
librium constant for tautomerization from N to I2; in this
condition the overall process can be written as in Scheme
4.
+
+
NH+
Because kN2 H and KTAUT are independent of pH, k∞
The ratio k
[BH]/kN2 H is related to the partition of
BH
will be independent of pH even if other equilibrium
constants and other rate constants vary with pH. In fact,
the variation of the buffer system B/BH and pH implies
variations in the positions of the first and second equi-
librium (Scheme 4) with the related rate constants, but
the ratio [I2]/[N] ) KTAUT remains constant. The calcu-
lated values of k∞ for isomer 1 and 2 are reported in
Tables 1 and 2. The values found with different buffer
systems are in agreement, and the average value is k∞ )
(1.9 ( 0.3) × 10-5 s-1 with isomer 1 and (2.8 ( 0.4) ×
10-6 s-1 with isomer 2.
the intermediate carbanion. The E1cb mechanism with
NH+ and the presence of an intermediate carbanion were
assigned in our work by focusing experiments on the
minimum of the energy in the energy/reaction coordinate
diagram, where the intermediate is located.16 So the
curvature observed in the studies of the acid-base
catalysis is related to the competition between the two
+
NH+
terms k
[BH] and kN2 H , then to the presence of the
BH
carbanion intermediate in the reaction path. Even the
H/D experiments (discussed in the previous section) can
identify the presence of an intermediate carbanion by
focusing on its partition ratio in D2O. It was previously
Another rate constant is consistent within the proposed
mechanistic model, kNOH-, the second-order rate constant
for the reaction with 1 in Q/QH+ buffer with OH- base.
The value extrapolated at [buffer] ) 0 in the plot kE vs
showed that the technique of solvent kinetic isotope
3,10
effect,
focusing again on the partition ratio of the
carbanion in the reaction with 1 or 2, can be used as a
quantitative support of the mechanistic model assigned.
With this technique the pseudo-first-order rate constant
kobs for the acid-base catalysis study are compared in
H2O and D2O. The differences observed depend on the
contribution of the term kBH(D)[BH(D)] with respect to k2.
[B] is kNOH ) 3.5 × 10-3 M-1 s-1, this value is in good
-
agreement with that directly measured experimentally
in OH-/H2O in the same reactions conditions as previ-
ously reported,7 kONH- ) 3.45 × 10-3 M-1 s-1. The validity
of the E1cb mechanistic assignment with 1 and 2 is
strongly supported by the consistency of the kinetic
parameters obtained: the rate constants that are ex-
pected to remain constant in different buffer systems and
pH range, k∞ and kNOH-, are effectively in good ap-
+
NH+
BH
Apart from the Q/QH+ buffer, all the k
/kN2 H values
(15) Tobin, J . B.; Frey, P. A. J . Am. Chem. Soc. 1996, 118, 12253.
(16) J encks, W. P. Acc. Chem. Res. 1976, 9, 425.
2278 J . Org. Chem., Vol. 69, No. 7, 2004

Acid-Base Catalysis of â-Elimination Reactions
TABLE 3. Exp ected P lots of kobs a ga in st [B] for
Elim in a tion Rea ction fr om Su bstr a tes Activa ted by a
P yr id in e Rin ga
TABLE 4. Exp r ession of kobs for Elim in a tion Rea ction
Rela ted to Ta ble 3a
a
kobs calculated with initial rate by a plot of C against t, kobs
)
slope/[N]. In the case of I, L, M, N plots (Q/QH+ buffer) kobs ) kE,
calculated following the total reaction at pH ≈ 11, with isomer 1.
The E1cb mechanism from N refers to process N f I1 f P.
a
kobs calculated by initial rate by a plot of C against t, kobs
)
slope/[N]. In the case of I, L, M, N plots (Q/QH+ buffer) kobs ) kE,
calculated following the total reaction at pH ≈ 11, with isomer 1.
proximation constant, whereas those expected to vary,
+
+
NH+
BH
kNBH and k
/kN2 H , show changes that are consistent
Str u ctu r e a n d u n d er Differ en t Rea ction Con d i-
tion s. For the systems giving â-elimination reaction with
structure Y-CH2-CH2-X, where Y is the â-activating group
and X is the leaving group, the expected dependence of
the plots kobs vs [B] (the basic component of the buffer),
for an acid-base catalysis study where Y is a pyridine
ring, is shown in Table 3. In Table 4 the derived kinetic
expressions of kobs, related to Table 3 are reported.
Various possible situations are considered, depending
on whether the reacting species is N or NH+ or if there
is competition between N and NH+ with the possible
mechanism E2, E1cb irreversible, E1cb reversible, or
E1cb partially reversible. A point of mechanistic interest,
in these systems, is to clarify if the reacting species
undergoing carbon deprotonation is the unprotonated
substrate N or the protonated substrate NH+. It is also
important to know whether the mechanism of the reac-
tion is E2 concerted, E1cb irreversible, or E1cb reversible.
It can be seen from Table 3 that there are some diagnostic
differences related to the operating reaction mechanism
and to the reacting species that undergoes carbon depro-
tonation (whether N or NH+). Substrates 1 and 2 in
acetate/acetic acid, imidazole/imidazolium, and acetohy-
with the structure of the reactants (Tables 1 and 2). It is
interesting to note that the reaction path that we have
proposed for the acid-base catalysis of our reactions,
represented by the sequence of steps N f NH+ f I2 f
P, is the only path kinetically consistent with the
experimentally observed acid-base catalysis. In fact, if
the tautomerization process occurs by first carbon depro-
tonation followed by nitrogen protonation, the reaction
path will be represented by the sequence of steps N f I1
f I2 f P. In this case, a general base catalysis would be
observed, independently if the rate-determining step is
the first or the second one. This discussion is valid for
all studied systems except isomer 1 in quinuclidine/
quinuclidinium buffer (Scheme 2).
In Table 3 is shown the expected acid-base catalysis
for the various possible situations in solution, depending
on if the reacting species is N, NH+, or the competition
N + NH+ and depending on the mechanism of the
reaction.
Our systems show acid-base catalysis described by
plots F or H, Table 3, (kE values in quinuclidine/
quinuclidinium buffer with isomer 1 for the elimination
from NH+ are described by plot F) where the slopes
increase by increasing the [H+]; this is in agreement with
the mechanism described in Scheme 1.
7
droxamate/acetohydroxamic acid buffers show acid-
base catalysis as H (Table 3). This is in agreement with
NH+ as reacting species and with an E1cb partially
reversible mechanism. These conclusions are confirmed
by the expected presence of H/D exchange7 and consistent
Mech a n ism Assign m en t for Elim in a tion Rea c-
tion s w ith
a Ser ies of Su bstr a tes of Differ en t
J . Org. Chem, Vol. 69, No. 7, 2004 2279

Alunni and Ottavi
values of the solvent isotope effect.10 The proton activat-
ing factors determined with acetohydroxamate/aceto-
hydroxamic acid7 buffer at 50 °C, µ ) 1 M KCl are 2.7 ×
105 for isomer 1 and 5.2 × 106 for isomer 2; with the base
hydroxamic acid buffer, 50 °C and µ ) 1 M KCl). It is to
be considered that the kBN value in acetohydroxamate/
acetohydroxamic acid with these substrates was esti-
mated by the previously proposed LFER;7 this value is
then an approximation, but it is useful to calculate the
order of magnitude of PAF. These high PAF values are
similar to those determined with 1 or 2, and they can be
put in relation to the presence of a resonance stabilized
intermediate carbanion, formed from NH+, and an (E1cb)I
mechanism. The change in leaving group from the
tertiary amine to Cl or F shows the expected mechanistic
OH- they are 0.7 × 105 for isomer 1 and 1.1 × 106 for
NH+
OH
isomer 2 (calculated with the k
values from this
-
work). Substrate 1, in quinuclidine/quinuclidinium buffer,
shows acid-base catalysis as I or L (Table 3). So at pH
10.1-11 there is a competition between N and NH+ as
reacting species, but when NH+ is the reacting species
the consistent mechanism can be E2 or (E1cb)I. A
distinction between these two mechanisms can be made
considering the PAF value determined, PAF ) 1.2 × 106.
This value is similar to the value determined in aceto-
hydroxamate/acetohydroxamic acid7 buffer, where an
E1cb mechanism was demonstrated with NH+, so an
E1cb irreversible mechanism can be assigned in quinu-
clidine/quinuclidium buffer with 1. In agreement with
this interpretation is the fact that in acetate/acetic acid
and imidazole/imidazolium buffers an E1cb partially
reversible mechanism can be demonstrated; the change
to quinuclidine/quinuclidium buffer can be interpreted
as the expected mechanistic change to (E1cb)I, owing to
the lower acidity of QH+ (BH) and then to the increase
in the energetic barrier for the reprotonation of the
intermediate carbanion (decrease of the rate constant,
change associated to decrease in the barrier for leaving
+
group expulsion (increased value of kN2 H ) from (E1cb)R
to (E1cb)I mechanism. The catalysis observed with these
substrates is described by plot F (Table 3), and this is in
agreement with the sequences of steps N f NH+ f I2 f
P, as previously described. It is relevant to report that a
study of the non-steady-state kinetics of the elimination
reaction of HBr from 2-(p-nitrophenyl)ethyl bromide in
alchol/alkoxide media has been interpreted with two-
steps mechanism involving an intermediate carbanion;19
this substrate has a lower â-activation with respect to a
protonated pyridine ring and Br is a better leaving group
with respect to Cl or F. It is to be considered that an E2
concerted mechanism can have a transition state with
partial negative charge development at the â-carbon1
(E1cb-like transition state). However, we favor the simple
interpretation that to the same values of PAF is associ-
ated the same mechanism E1cb. The substrates studied
present different, but related structures, with variations
of Y and X. Mechanism assignment for a single substrate
is supported by applying, in most cases, the combined
use of two or three techniques. As previously discussed,
the interrelation between the various systems offers a
further possibility of mechanism assignment.
An interesting variation in the system is represented
by the variation of the buffer (consequently of the pH).
This point has been previously discussed. What we would
like to point out now is that, in our model, the variation
of the buffer system represents a mechanistic test of self-
consistency of the model. In fact, with substrates 1 and
2 we have an E1cb mechanism as mechanistic model. To
check conclusively the consistency of the model, an
important variation of the reaction system is made
(change of the buffer and pH). The variation is selected
to focalize the consistency of key parameters of the model
(in our case k∞). The analysis of the results due to the
variations made in the model can be a conclusive test of
the validity of the model. This situation in our systems
is exemplified in Figure 6.
NH+
BH
NH+
BH
k
). We consider, also, that the fit of k
with isomer
1 in quinuclidine in the Brønsted plot (Figure 5) can be
in agreement with an (E1cb)I mechanism in quinuclidine/
quinuclidinium buffer; in fact, the Brønsted plot is related
to the E1cb mechanism. A conclusion from the results of
an E1cb mechanism in quinuclidine/quinuclidinium buffer
is that the mechanism remains E1cb when the strength
of the base increases from acetate (pKa ) 4.65 at 50 °C
and µ ) 1 M KCl) to quinuclidine (pKa ) 11 at 50 °C and
µ ) 1 M KCl). It has been reported in the literature a
change in mechanism from E1cb to concerted E2 by
increasing the strength of the base.17
Substrates 2-(2-chloroethyl)pyridine,8 2-(2-fluoroethyl)-
pyridine,9 and 2-(4-fluoroethyl)pyridine9 from previous
studies show acid-base catalysis in acetate/acetic acid
and acetohydroxamate/acetohydroxamic acid buffers as
E or F (Table 3). From these results, it can be deduced
that the reacting species is NH+, and both an E2 or
(E1cb)I mechanisms are in agreement. To distinguish
between these two possibilities is generally very dif-
ficult.9,18 In our interrelated systems, however, these
substrates can be put in relation with 1 or 2. The PAF
value calculated with 2-(2-chloroethyl)pyridine8 is PAF
) 1.38 × 105 (acetohydroxamate/acetohydroxamic acid
buffer, 50 °C and µ ) 1 M KCl), PAF ) 3.6 × 105 with
2-(2-fluoroethyl)pyridine9 and PAF ) 0.65 × 105 with
The consistency of the results obtained can be seen in
Tables 1 and 2; the parameters that have to remain
constant, with the variation of pH and buffer systems,
remain constant, while those that have to change with
the buffer structure change consistently.
2-(4-fluoroethyl)pyridine9
(acetohydroxamate/aceto-
(17) J ia, Z. S.; Rudzinski, J .; Paneth, P. and Thibblin, A. J . Org.
Chem. 2002, 67, 177.
All of the interrelated systems, where Y is a pyridyl
or a quinolyl group,7-10,20 react by an E1cb mechanism,
with NH+ being the reacting species owing to the high
PAF values (a part 1 in quinuclidine/quinuclidium buffer
at pH 10.1-11, where there is a competition between N
(18) (a) More O’Ferrall, R. A.; Warren, P. J . J . Chem. Soc., Chem.
Commun. 1975, 483. (b) More O’Ferrall, R. A.; Slae, S. J . Chem. Soc.,
B 1970, 260. (b) More O’Ferrall, R. A.; Larkin, F.; Walsh, P. J . Chem.
Soc., Perkin Trans. 2 1982, 1573. (c) More O’Ferrall, R. A.; Larkin, F.
Aust. J . Chem. 1983, 36, 1831. (d) Larkin, F. G.; More O’Ferrall, R.
A.; Murphy, D. G. Collect. Czech. Chem. Commun. 1999, 64, 1833. (e)
More O’Ferrall, R. A.; Larkin, F.; Walsh, P. J . Chem. Soc., Perkin
Trans. 2 1982, 1573. (f) Carey, E.; More O’Ferrall, R. A.; Vernon, N.
M. J . Chem. Soc., Perkin Trans. 2 1982, 1581. (g) Kelly, R. P.; More
O’Ferrall, R. A.; O’Brien Myles J . Chem. Soc., Perkin Trans. 2 1982,
211.
(19) Handoo, K. L.; Lu, Y.; Zhao, Y.; Parker, V. D. Org. Biomol.
Chem. 2003, 1, 24.
(20) Alunni, S.; Orazi, C. J . Phys. Org. Chem. 2001, 14, 879.
2280 J . Org. Chem., Vol. 69, No. 7, 2004

Acid-Base Catalysis of â-Elimination Reactions
TABLE 5. P AF a n d Meth ylAF Va lu es for HX
â-Elim in a tion Rea ction s for th e System Y-CH2-CH2-X in
H2O a t 50 °C, µ ) 1 M KCl
+
The PAF values are calculated as the ratio kNBH /kBN in
a
b
acetohydroxamate/acetohydroxamic acid buffers. Values calcu-
lated in OH-/H2O. c Value determined at 25 °C and µ ) 1 M KCl.
It can be seen that there is good agreement between
the two parameters as expected within an E1cb mecha-
nism involving the protonated or the methylated pyridine
+
ring (activating system). A comparison of kNBH with the
same activating system (Y ) 2-pyridyl) and leaving group
Cl, F shows that the second-order rate constant for carbon
deprotonation from NH+ is larger (acetohydroxamate/
acetohydroxamic buffer, 50 °C and µ ) 1 M KCl) with
+
the leaving group X ) chlorine (kNBH ) 1.05 M-1 s-1
)
F IGURE 6. Schematic representation of the consistency of
the model.
NH+
B
than with X ) fluorine (k
) 0.35 M-1 s-1). The same
and NH+). Systems activated by p-nitrophenyl or o-
nitrophenyl groups, previously studied3-5 react by the
E1cb mechanism, and the effect of the variation of the
structure of the leaving group can be understood in terms
of a relief of steric strain in the expulsion of the leaving
group from the intermediate carbanion. When Y )
p-nitrophenyl in acetohydroxamate/acetohydroxamic acid
buffer, 50 °C and µ ) 1 M KCl at pH ) 9.05, a
mechanistic change from an E1cb partially reversible
mechanism (X ) N-methylpyrrolidine, pKa ) 10.77 in
H2O, at 25 °C and µ ) 1 M KCl) to an E1cb irreversible
mechanism (X ) N-isopropylpyrrolidine, pKa ) 11.25 in
H2O, at 25 °C and µ ) 1 M KCl) has been assigned.4 We
can conclude that the activation by p-nitrophenyl or
o-nitrophenyl or protonated pyridine ring provides enough
stabilization of the intermediate carbanion to make the
carbanion mechanism energetically favored with respect
to the concerted E2 process.16
situation is found with the methylated substrates: when
Y ) 2-pyridyl and X ) Cl, kNBCH ) 17.9 M-1 s-1, whereas
3
when X ) F, kNBCH ) 10.25 M-1 s-1 (OH-/H2O 25 °C and
3
µ ) 1 M KCl). Because of the fact that in our interrelated
model the mechanism with NH+ or the methylated
substrate can be assigned as (E1cb)I, it follows that there
is some lengthening of the carbon-leaving group bond in
the intermediate carbanion18d,21 (and related transition
state). In fact, the higher electronegativity of F, with
respect to Cl, should imply a reverse order of reactivity
(even if the possible significant contribution of the
enamine type structure to the delocalization of the
negative charge could limit this effect). Thus the larger
reactivity when the leaving group is Cl with respect to F
indicates that there is a favored lengthening of the
carbon-chlorine bond with respect to that of the stronger
carbon-fluorine bond. This is a point of interest for
interpreting a heavy atom isotope effect.17,22
Another motif of consistency of the proposed model
comes from the comparison of PAF values determined
Regarding the possibility of an E1cb ion pairs mech-
anism,1a,3 where the carbanion intermediate is bonded
with the related methyl activating factor. In fact, while
+
the kNBH rate constant is extracted from a mathematical
(21) (a) Thibblin, A. J . Am. Chem. Soc. 1988, 110, 4582. (b)
O¨ lwegård, M.; McEwen, I.; Thibblin, A.; Ahlberg, P. J . Am. Chem. Soc.
1985, 107, 7494. (c) Thibblin, A.; Ahlberg, P. J . Am. Chem. Soc. 1977,
99, 7926. (d) Ahlberg, P. Chem. Scr. 1973, 3, 183. (f) Thibblin, A.;
Ahlberg, P. J . Am. Chem. Soc. 1979, 101, 7311. (g) Thibblin, A. Scr.-
Chem. 1980, 15, 121. (h) Thibblin, A. J . Am. Chem. Soc. 1983, 105,
853. (i) Hoffmann, R.; Radom, L.; Pople, J . A.; Schleyer, P. v. R.; Hehre,
W. J .; Salem, L. J . Am. Chem. Soc. 1972, 94, 6221. (j) Apeloig Y.;
Rappoport, Z. J . Am. Chem. Soc. 1979, 101, 5095. (k) Saunders, W.
H., J r. J . Org. Chem. 1997, 62, 244. (l) Saunders, W. H., J r. J . Org.
Chem. 1999, 64, 861.
model describing the E1cb mechanism from NH+ (Scheme
1), it is possible to directly measure kNBCH (in OH-/H2O,
3
50 °C) following the elimination reaction with 1-methyl-
2-(2-chloroethyl)pyridinium iodide, 1-methyl-2-(2-fluoro-
ethyl)pyridinium iodide, or 1-methyl-4-(2-fluoroethyl)-
pyridinium iodide. The MethylAFs previously deter-
mined8,9 with several substrates are reported in Table 5
along with the related PAF values for comparison.
(22) Ryberg, P.; Matsson, O. J . Am. Chem. Soc. 2001, 123, 2712.
J . Org. Chem, Vol. 69, No. 7, 2004 2281

Alunni and Ottavi
P r od u ct An a lysis. Rea ction in Qu in u clid in e/Qu in u -
clid in iu m (Q/QH+) Bu ffer s. Several experiments with a
solution of 4-vinylpyridine in buffers, at different [buffer] and
[Q]/[QH+] ratios, left at 50 °C for different reaction times (6
and 20 h) showed that, after extraction with n-hexane and
GLC analysis, the only products were 4-vinylpyridine and
quinuclidine. Also, a mixture of 4-vinylpyridine and Q/QH+
buffers in D2O in a NMR tube left at 50 °C and analyzed at
various intervals of time showed the formation of 1. The same
products were obtained in similar experiments following the
elimination reactions from 1.
Rea ction s in Im id a zole/Im id a zoliu m Bu ffer s. A solu-
tion of 1 or 2 in several buffer solutions at different pH values
were left to react at 50 °C for ∼6 or 30 h, respectively. The
solution was taken to pH ≈ 10 before extraction with n-hexane;
the GLC analysis showed the presence of only 4- or 2-vinyl-
pyridine and quinuclidine.
with BH, in systems where the curvature in the plots
kobs against [B] is observed, it can be excluded. When
acid-base catalysis is observed as E or F in Table 3 this
mechanistic possibility would be consistent; however, the
possibility of rate-determining the expulsion of the leav-
ing group should be associated with a change from
quinuclidine E1cb partially reversible mechanism to the
much better leaving group fluorine or chlorine and this
fact seems inconsistent. Also the PAF values, in the case
of E1cb (ion pair) mechanism with the rate-determining
step the expulsion of the leaving group, are related to a
combination of rate constants, whereas we have found
PAF values, with substrates with Cl or F as leaving
group, consistent with those of substrates with quinu-
clidine as leaving group. The possibility of E1cb (ion pair)
mechanism with rate-determining carbon deprotonation
is also consistent with catalysis as E or F in Table 3. In
this case the transition state for the E1cb (ion pair) or
(E1cb)I mechanism would be the same, and our kinetics
techniques do not allow the distinction.
Rea ction s in Aceta te/Acetic Acid Bu ffer s. A solution of
1 or 2 in several buffer solutions at different pH values was
left to react at 50 °C for ∼3 or 25 h, respectively. The solution
was taken to pH ≈ 10 before extraction with n-hexane; the
GLC analysis showed the presence of only 4- or 2-vinylpyridine
and quinuclidine.
Regarding the possibility of competition E1cb/E2 it can
be noted that a significant contribution of E2 can be
excluded in the cases where the curvature is observed in
the plots kobs against [B] and the leveling-off is reached
and shown to be independent of pH. In the case that the
curvature is observed but the leveling-off is not reached,
a large contribution by the E2 process would imply a
significant distortion from the E1cb model. In the case
of substrates with Cl or F as leaving group (catalysis E
or F in Table 3) the assignment of the (E1cb)I mechanism
with not significant contribution of E2 is related to the
observed PAF values and their similarity with substrates
with quinuclidine as leaving group.
Kin et ic Mea su r em en t s in Qu in u clid in e/Qu in u cli-
d in iu m Bu ffer . The kinetic studies in Q/QH+ buffers were
carried out following the decrease in absorbance at λ ) 280
nm of a solution = 8.6 × 10-4 M of 4-vinylpyridine in 2.8 mL
of buffer in a thermostated cuvette at 50 °C. The ionic strength
was maintained at 1 M with potassium chloride; slight
variations in pH within the buffer series required some
adjustments of the pH. The process of the addition reaction of
quinuclidine to 4-vinylpyridine in Q/QH+ buffers to give 1
(substrate 1 will be indicated as S; in Q/QH+ it will correspond
to the unprotonated substrate N) is an equilibrium involving
the reverse elimination reaction. The observed pseudo-first-
order rate constants were evaluated by the equation ln[(A0 -
A∞)/(At - A∞)] ) kobst, where kobs ) kE + kA, where kA is the
observed pseudo-first-order rate constant for the addition
reaction and kE is the observed pseudo-first-order rate constant
of the reverse elimination reaction from 1 to form 4-vinylpy-
ridine (P). The value of kA is calculated by kA ) kobs/[1 + ([P]∞/
[S]∞)] and [P]∞/[S]∞ ) 1/(A0/ꢀP[P]∞ - 1), where ꢀP is the
extinction coefficient7 of 4-vinylpyridine at 50 °C, µ ) 1 M KCl
and λ ) 280 nm, ꢀP ) 1634 M-1 cm-1; A0 is the value of
absorbance at t ) 0; [S]∞ is the [S] at equilibrium; [P]∞ is the
[4-vinylpyridine] at equilibrium, and [P]∞ ) [A∞ - (ꢀS/ꢀP)A0]/
(ꢀP - ꢀS), where ꢀS ) 7.9 M-1 cm-1 is the extinction coefficient
of 1 in the reaction condition at λ ) 280 nm. The [OH-] at 50
°C, µ ) 1 M KCl was calculated from the measured pH and
Exp er im en ta l Section
Ma ter ia ls. Glass-distilled and freshly boiled water was used
throughout. Reagent-grade potassium chloride, quinuclidinium
hydrochloride, potassium acetate, imidazole, and 2-vinylpyri-
dine were used without further purification. 4-Vinylpyridine
was purified by column chromatography with silica gel (Et2O).
N-[2-(4-P yr id yl)et h yl]q u in u clid in iu m b r om id e a n d
N-[2-(2-P yr id yl)eth yl]qu in u clid in iu m tosila te. These com-
pounds were prepared and characterized as previously re-
ported.7
from the empirical7 equation [OH-] × 10-pH ) 5.89 × 10-14
,
pKw ) 13.23. Some equilibration experiments, starting from
1 and following the formation of 4-vinylpyridine at 280 nm,
p Ka Deter m in a tion . The pKa of quinuclidinium chloride
or imidazolium chloride, 50 °C and µ ) 1 M KCl, was
determined23 by potentiometric titration with KOH 1 M or HCl
1 M, respectively. The pKa of 4-vinylpyridine and 2-vinylpy-
ridine were previously reported9 as 5.64 and 5.06, respectively,
measured in acetate/acetic acid buffer 0.1 M, 50 °C and µ ) 1
M KCl. These values are used for the study of acid-base
catalysis in the various buffer systems. However, we have
found that at very low [buffer] or in the absence of buffer, the
pKa is 5.23 and 4.77, respectively. Therefore, we have found
that it is more correct to use these last values to determine
gave consistent results (in this case, kobs ) kE + kA, and kA
obs/[1 + (A∞ - ꢀS[S]0)/(ꢀP[S]0 - A∞)], where A∞ is the absorbance
at t f ∞ and [S]0 is the concentration of 1 at t ) 0).
)
k
Kin etic Mea su r em en ts in Im id a zole/ Im id a zoliu m or
Aceta te/Acetic Acid Bu ffer s. Elimination reactions from 1
or from 2 in imidazole/imidazolium or acetate/acetic acid
buffers at 50 °C, µ ) 1 M KCl were followed by monitoring
the formation of 4-vinylpyridine at λ ) 280 nm or 2-vinylpy-
ridine at λ ) 290 nm, by initial rates.7 The concentration of 1
was =1.5 × 10-3 M, the concentration of 2 was =1.7 × 10-3
M, and the reactions were followed up to =3%. At the pH used,
the 4- or 2-vinylpyridine product, P, is at equilibrium with its
conjugated acid, PH+. In this condition the total concentration,
C, of the product (C ) [P] + [PH+]) was calculated from
+
NH+
OH
the rate constants in the absence of buffer (k′NHHO , k -).
2
H/D Exch a n ge. H/D exchange experiments in D2O imida-
zole/imidazolium buffer, or D2O, CH3COO-/CH3COOD buffer,
were carried out using the previously described techniques4,5,7
in order to determine if deuterium was incorporated in the
â-position of 1 or 2, with respect to the leaving group during
the elimination reaction.
[H+]
(At - A0)
C )
1 +
(15)
+
(
)
[H ]
KaP
KaP
ꢀP + ꢀPH
(23) Albert, A.; Serjeant, E. P. The Determination of Ionization
Constant; Chapman and Hall Ltd.: London, 1971.
2282 J . Org. Chem., Vol. 69, No. 7, 2004

Acid-Base Catalysis of â-Elimination Reactions
where At is the absorbance at time t; A0 is the absorbance at
t ) 0; KPa is the dissociation constant of PH+ (KPa ) 2.29 ×10-6
Ack n ow led gm en t. The authors thank the Minis-
tero dell’Universita` e della Ricerca Scientifica e Tecno-
logia (MURST) for financial support.
M for 4-vinylpyridinium,9 pKaP) 5.64, or KaP ) 8.71 × 10-6
M
for 2-vinylpyridinium,9 pKaP ) 5.06, at 50 °C, µ ) 1 M KCl); ꢀP
is the extinction coefficient of 4-vinylpyridine7 (1634 M-1 cm-1
at λ ) 280 nm, 50 °C and µ ) 1 M KCl) or of 2-vinylpyridine7
(3584 M-1 cm-1 at λ ) 290 nm, 50 °C and µ ) 1 M KCl), and
ꢀPH is the extinction coefficient of 4-vinylpyridinium9 (8987 M-1
cm-1 at λ ) 280 nm, 50 °C and µ ) 1 M KCl) or of
2-vinylpyridinium9 (9773 M-1 cm-1 at λ ) 290 nm, 50 °C and
µ ) 1 M KCl). A plot of C against t was linear, and the values
of kobs were calculated from kobs ) slope/[N]; the [N] was
calculated from [N] ) {KNa /(KaN + [H+])}[S]0, where [S]0 is the
initial concentration of 1 or 2 ([S]0 ) [N] + [NH+]). 4-Vinyl-
pyridine or 2-vinylpyridine were shown to be stable under the
experimental reaction conditions.
Su p p or tin g In for m a tion Ava ila ble: Pseudo-first-order
rate constants kobs, kE, and kA, for the elimination reactions
from 1 in quinuclidine/quinuclidium buffer at different pH
values (H2O, 50 °C and µ ) 1 M KCl); kobs, for the elimination
reactions from 1 and 2 in imidazole/imidazolium or acetate/
acetic acid buffers at different pH values (H2O, 50 °C and µ )
1 M KCl); k0, for the elimination reactions from 1 and 2 in
absence of buffer (H2O, 50 °C and µ ) 1 M KCl); and the plots
for the elimination reactions from 1 and 2 in acetate/acetic
acid buffers. This material is available free of charge via the
Internet at http://pubs.acs.org.
J O030352Z
J . Org. Chem, Vol. 69, No. 7, 2004 2283