Structure-reactivity relationships in the pyridinolysis of N-methyl-N-arylcarbamoyl chlorides in dimethyl sulfoxide

Article abstract of DOI:10.1021/jo0108212

Nucleophilic substitution reactions of N-methyl-N-arylcarbamoyl chlorides (YC₆H₄N(CH₃)COCl) with pyridines (XC₅H₄N) have been investigated in dimethyl sulfoxide at 45.0 °C. A striking trend in the selectivity parameters is that they are constant within experimental errors, ρ_X = -2.25 ± 0.03, β_X = 0.42 ± 0.01, and ρ_X = 1.10 ± 0.06, with changing reactivities of the electrophiles (6δσ_Y) and nucleophiles (δσ_X), respectively, and this leads to a vanishingly small cross-interaction constant, ρ_XY β_XY 0. The rate data can be expressed in the Ritchie N₊ type equation. Based on this and other results, the mechanism of nucleophile (pyridine) addition to the resonance- stabilized carbocation is proposed. It has been shown from the definition β_XY (and ρ_XY) together with the Marcus equation that the high intrinsic barrier, ΔG^±, in the intrinsic-barrier controlled reaction series is a prerequisite for such reactions in which the cross-interaction vanishes and the N₊ relationship holds.

Full text of DOI:10.1021/jo0108212

J . Org. Chem. 2001, 66, 8549-8555
8549
Str u ctu r e-Rea ctivity Rela tion sh ip s in th e P yr id in olysis of
N-Meth yl-N-a r ylca r ba m oyl Ch lor id es in Dim eth yl Su lfoxid e
Ikchoon Lee,* Sung Wan Hong,† Han J oong Koh,‡ Yong Lee, Bon-Su Lee, and Hai Whang Lee
Department of Chemistry, Inha University, Inchon, 402-751 Korea, Department of Chemistry,
Woosuk University, Chonju, 565-701 Korea, and Department of Science Education,
Chonju National University of Education, Chonju, 560-757 Korea
ilee@inha.ac.kr
Received August 13, 2001
Nucleophilic substitution reactions of N-methyl-N-arylcarbamoyl chlorides (YC6H4N(CH3)COCl)
with pyridines (XC5H4N) have been investigated in dimethyl sulfoxide at 45.0 °C. A striking trend
in the selectivity parameters is that they are constant within experimental errors, FX ) -2.25 (
0.03, âX ) 0.42 ( 0.01, and FY ) 1.10 ( 0.06, with changing reactivities of the electrophiles (δσY)
and nucleophiles (δσX), respectively, and this leads to a vanishingly small cross-interaction constant,
FXY ≈ âXY ≈ 0. The rate data can be expressed in the Ritchie N+ type equation. Based on this and
other results, the mechanism of nucleophile (pyridine) addition to the resonance- stabilized
carbocation is proposed. It has been shown from the definition of âXY (and FXY) together with the
q
Marcus equation that the high intrinsic barrier, ∆G0 , in the intrinsic-barrier controlled reaction
series is a prerequisite for such reactions in which the cross-interaction vanishes and the N+
relationship holds.
In tr od u ction
nucleophile selectivities are observed with changing
electrophile (carbocation) reactivity. Although originally
such relationship is reported to apply for ring-substituted
diarylmethyl (D+), triarylmethyl (T+), and the aryl tro-
pylium ions (Tp+), it has since been extended to other
highly resonance-stabilized systems such as quinon me-
thide,5 4.
The mechanism of the nucleophilic substitution reac-
tions of N,N-disubstituted (1, N,N-dimethyl: 2, N,N-
diphenyl: 3, N-methyl-N-phenyl) carbamoyl chlorides has
long been controversial ranging from an SN11,2a to SN22
including an SN2 (intermediate)3 type. However there
does seem to exist a general consensus that the mecha-
nism for solvolysis is primarily SN1 in character, favored
by a strong resonance electron donation from nitrogen.
For some systems, a weak nucleophilic participation by
solvent and/or amine nucleophile has been observed in a
predominantly SN1 reaction.1h
For reactions of strongly resonance-stabilized systems,
e.g., carbocations, that follow the Ritchie N+ relation-
ship,4 eq 1 where N+ is a parameter characteristic of
In this work, we report structure-reactivity relation-
ships observed in the pyridinolysis of N-methyl-N-aryl-
carbamoyl chlorides 3 in dimethyl sulfoxide, eq 2. One
log kNu ) N+ + log k0
(1)
nucleophile reactivity and k0 is a constant dependent
solely on the identity of the cation, essentially constant
* To whom correspondence should be addressed. Fax: +82-32-865-
4855.
† Woosuk University.
‡ Chonju National University of Education.
(1) (a) Hall, H. K., J r.; Lueck, C. H. J . Org. Chem. 1963, 28, 1963.
(b) Hall, H. K., J r. J . Org. Chem. 1964, 29, 3539. (c) Queen, A. Can. J .
Chem. 1967, 45, 1619. (d) Queen, A.; Nour, T. A.; Paddon-Row, M. N.;
Preston, K. Can. J . Chem. 1970, 48, 522. (e) J ohnson, S. L. Adv. Phys.
Org. Chem. 1967, 5, 237. (f) J ohnson, S. L.; Giron, H. M. J . Org. Chem.
1972, 37, 1383. (g) Bacaloglu, R.; Daescu, C.; Ostrogovich, G. J . Chem.
Soc., Perkin Trans. 2 1972, 1011. (h) D’Souza, M. J .; Kevill, D. N.;
Bentley, T. W.; Devaney, A. C. J . Org. Chem. 1995, 60, 1632.
(2) (a) Kim, S. C.; Song, H. B.; Lee, I. J . Korean Chem. Soc. 1979,
23, 368. (b) Kevill, D. N.; D’Souza, M. J . J . Chem. Res. Synop. 1993,
174. (c) Liu, K.-T.; Chen, H.-I.; Lin, Y.-S.; J in, B.-Y. J . Phys. Org. Chem.
2000, 13, 322.
(3) Bentley, T. W.; Bowen, C. T.; Marten, D. H.; Schleyer, P. V. R.
J . Am. Chem. Soc. 1981, 103, 5466.
(4) (a) Ritchie, C. D. Acc. Chem. Res. 1972, 5, 348. (b) Ritchie, C. D.
J . Am. Chem. Soc. 1975, 97, 1170. (c) Ritchie, C. D. Can. J . Chem.
1986, 64, 2239.
of the principal objectives was to investigate how selec-
tivity changes as the substituents in the nucleophile (X)
and substrate (Y) are varied and to determine the cross-
interaction constant, FXY and âXY in eq 36 where ∆pKX )
e
e
pKX - pKH etc. and FXY ) âXY‚FX ‚FY with Fie ) ∆pKi/σi,
etc. Thus, âXY ) c‚FXY where c is a constant. This study
(5) Richard, J . P.; Toteva, M. M.; Crugeiras, J . J . Am. Chem. Soc.
2000, 122, 1664.
(6) (a) Lee,. I. Chem. Soc. Rev. 1990, 19, 317. (b) Lee. I. Adv. Phys.
Org. Chem. 1992, 27, 57. (c) Lee, I.; Lee, H. W. Collect. Czech. Chem.
Commun. 1999, 64, 1529. (d) Isaacs, N. S. Physical Organic Chemistry,
2nd ed.; Longman: Harlow, 1995; p 186.
10.1021/jo0108212 CCC: $20.00 © 2001 American Chemical Society
Published on Web 11/08/2001

8550 J . Org. Chem., Vol. 66, No. 25, 2001
Lee et al.
Ta ble 1. Secon d -Or d er Ra te Con sta n ts, k2 (×104 M-1 s-1), for th e Rea ction s of Y-Su bstitu ted N-Meth yl-N-a r ylca r ba m oyl
Ch lor id es w ith X-P yr id in es in Dim eth yl Su lfoxid e a t 45.0 °C
b
- d
X
pKa
Y ) p-CH3O
H
p-Cl
p-NO2
3090
FY
p-CH3O
6.58
6.03
5.59
5.21
4.87
3.17
2.81
2.30
1.86
1.35
56.2
132
74.1
191
115
1.12 ( 0.04
(0.999)c
1.11 ( 0.03
(0.999)
1.15 ( 0.01
(1.00)
1.08 ( 0.01
(1.00)
1.10 ( 0.02
(1.00)
1.04 ( 0.04
(0.999)
1.11 ( 0.03
(0.999)
1.12 ( 0.02
(1.00)
p-CH3
33.9
20.4
15.8
11.2
2.00
1.51
1.00
0.603
0.417
1690
1200
724
p-C6H5CH2
H
42.7
31.6
22.9
77.6
56.2
44.6
m-C6H5
m-CH3CO
m-Cl
561
3.98
6.03
97.7
2.63
2.09
1.26
0.871
4.90
3.39
2.24
1.51
72.4
52.5
28.8
17.8
p-CH3CO
p-CN
1.08 ( 0.02
(1.00)
1.05 ( 0.02
(1.00)
FXY ) -0.02
(0.997)c
m-CN
a
FX
-2.22 ( 0.08
(0.996)c
-2.27 ( 0.11
(0.993)
-2.24 ( 0.11
(0.992)
-2.27 ( 0.08
(0.996)
âX
0.41 ( 0.01
0.42 ( 0.01
(0.999)
0.41( 0.01
(0.998)
0.42 ( 0.01
(0.999)
(1.00)c
a
b
Excluding m-CN. The σ values were taken from: Hansch, C.; Leo, A.; Taft, R. W. Chem. Rev. 1991, 91, 165. In water at 25.0 °C.
Fischer, A.; Galloway, W. J .; Vaughan, J . J . Chem. Soc. 1964, 3591. Hong, S. W.; Koh, H. J .; Lee, I. J . Phys. Org. Chem. 1999, 12, 425.
c Correlation coefficients. The σ and σ- values from the same source as footnote a.
d
reveals that the FXY (and âXY) value is practically zero
and the Ritchie N+-type relationship holds for the present
series of reactions, eq 2. We discuss the mechanistic
implications of such relationships in the nucleophilic
substitution reactions of N,N-disubstituted 1-3 carbam-
oyl chlorides.
pyridines at the IPCM/B3LYP/6-31 G* level7 have shown
that there is a constant pKa difference of ∆pKa
)
pKa(MeCN) - pKa(H2O) ) 7.7 due mainly to the H+ ion
solvation free energy difference of 10.5 kcal mol-1 be-
tween acetonitrile and water. The plot of pKa(MeCN) vs
pKa(H2O) exhibited a straight line of near unity (1.02)
slope so that the Brønsted coefficients determined by the
plot of log k2(MeCN) against pKa(H2O) should be almost
the same as those against pKa(MeCN).
log(kXY/kHH) ) FXσX + FYσY + FXYσXσY
(3a)
A similar invariance of Brønsted coefficient for the
reactions of 4-nitrophenylsulfamate in chloroform with
eight pyridines has been reported using pKa(H2O) and
pKa(CH3CN) values with â ) 0.31 ( 0.02 (r ) 0.985) and
â ) 0.30 ( 0.02 (r ) 0.994), respectively.8 Moreover, the
plots of pKa(ꢀ) (in five solvents including water) vs σ gave
the slopes, Fs(ꢀ), which is linear with Onsager dielectric
function (ꢀ - 1)/(2ꢀ + 1), eq 5, with correlation coefficient
of 0.999 (n ) 5).7 This means that the specific
log(kXY/kHH) ) âX∆pKX + âY∆pKY + âXY∆pKX∆pKY
(3b)
Resu lts a n d Discu ssion
The pseudo-first-order rate constants observed (kobs
)
obeyed eq 4, for all the reactions with negligible k0 (≈0)
in DMSO. The second-order rate constants for pyridi-
nolysis, k2 (M-1 s-1) summarized in Table 1, were
obtained as the slopes of the plots of kobs against pyridine
concentrations, [Py], eq 4. No third-order or higher-order
terms were detected, and no complications were found
in the determination of kobs or in the linear plots of eq 4.
ꢀ - 1
Fs ) 14.6
- 16.1
(5)
(
)
2ꢀ + 1
hydrogen bonding solvation component is not important
in the solvation effect on the ionization equilibria of
pyridinium ions in water. The slope, Fs , is thus solely
dependent on the bulk solvent effect (ꢀ) and for ꢀ ) 78.39
(water) and ꢀ ) 37.99 (acetonitrile) the Fs values are quite
similar being -8.9 and -9.1, respectively. Since the ꢀ
value for DMSO is 46.5,9 the Fs, value of -9.0 is obtained
in DMSO from eq 5. This indicates that the Brønsted
coefficients determined by plotting log k2 (DMSO) against
pKa (H2O) should be practically the same as those by
plotting log k2 (DMSO) against pKa (DMSO). Spillane et
kobs ) k0 + k2 [Py]
(4)
The rate is faster with a stronger nucleophile and with
a stronger electron-withdrawing group in the substrate
3 as normally observed for a nucleophilic substitution
reaction in which negative charge develops at the reac-
tion center, carbonyl carbon, i.e., FY > 0, in the transition
state (TS). The Brønsted âX (ânuc) and Hammett FX (Fnuc
)
and FY values are also shown in Table 1. The pKa values
of pyridines used in the Brønsted plots were those
determined in water. Thus, the Brønsted coefficients in
Table 1 (âX ≈ 0.42) could be in error since the rate data
in Table 1 (in DMSO) should be plotted using pKa values
measured in DMSO. However our recent theoretical
studies of solvent effects on the basicity of substituted
(7) Lee, I.; Kim, C. K.; Han, I. S.; Lee, H. W.; Kim, W. K.; Kim, Y.
B. J . Phys. Chem. B 1999, 103, 7302.
(8) Spillane, W. J .; Hogan, G.; McGrath, P.; King, J .; Brack, C. J .
Chem. Soc., Perkin Trans. 2 1996, 2099.
(9) Reichardt, C. Solvents and Solvent Effects in Organic Chemistry,
2nd ed.; VCH: Weinheim, 1988; Table A-1, p 408.

Pyridinolysis of N-Methyl-N-arylcarbamoyl Chlorides
J . Org. Chem., Vol. 66, No. 25, 2001 8551
F igu r e 1. Brønsted plots (âX) for the (X)-pyridinolysis of
Y-substituted N-methyl-N-arylcarbamoyl chlorides (for Y )
p-CH3O, H, p-Cl, and p-NO2 in DMSO at 45 °C).
F igu r e 2. Hammett plots (FX) for the pyridinolysis of Y-
substituted N-methyl-N-arylcarbamoyl chlorides (for Y )
p-CH3O, H, p-Cl, and p-NO2 in DMSO at 45.0 °C).
al.10 reported that the Brønsted coefficients (âX) for the
reaction of N-phenylsulfamoyl chlorides (YC6H4NHSO2-
Cl) with anilines (XC6H4NH2) in acetonitrile are similar
when determined using pKa values of anilines measured
in water (âX ) 0.685) and in DMSO (âX ) 0.62). This
provides evidence in support of correlating the rate data
determined in DMSO (and in MeCN) with the pKa values
measured in water.
(σY)) FYσY in eq 6 is a constant so that eq 6a can now be
rearranged to eq 6b, which is the same form as eq 1. In
the present reaction series, a similar relation also holds
for log kY(X), since for a given X (σX) the rates are solely
log kX(Y) ) FXσX + (log kHH + FYσY) ) N+ + constant
(6b)
The excellent linearities found in the Brønsted plots
using 10 nucleophlies (r g 0.998, standard deviation e
0.01) in Figure 1 lend further credence to our procedure.
We therefore think that our âX values in Table 1
represent reasonable and meaningful values. The Ham-
mett plots for variation of substituent Y in the substrates
dependent on Y (σY) in eq 6a. This result is at variance
with the results of McClelland et al.,12 that there is a
clear trend for âX (ânuc) observed with primary amines to
decrease with increasing triarylmethyl cation reactivity.
For the ∆σY ≈ +1.5 change they obtained ∆ânuc (∆âX) e
-0.15; i.e., âXY is distinctly nonzero, which is in contrast
to our value of ∆âX ≈ 0, i.e., âXY ≈ 0 for the similar
increase in ∆σY; for change in σY from Y ) p-MeO to
p-NO2, (∆σY ) 1.27 - (-0.27) ) 1.54), ∆âX ≈ 0 in Table
1. This shows that the present results obey the Ritchie
N+ relationship better than their laser flash photolysis
results in aqueous acetonitrile solution.
Originally, the N+ equation was introduced for addition
of nucleophiles to the resonance-stabilized carbocations
such as substituted triarylmethyl carbocations and the
aryl tropylium ions.4 In view of the relatively strong
resonance stabilization of the N,N-disubstituted carbam-
oyl cations,1h,13 Scheme 1, a possibility of such a mech-
anism, i.e., nucleophile (pyridine) addition to the reso-
nance-stabilized acylium ion 5, cannot be excluded. In
fact, an SN1 mechanism has been proposed for the
solvolysis of N,N-dimethyl 1 and N,N-diphenyl 2 car-
bamoyl chlorides in various aqueous solutions by many
investigators in this field.1 However it has been well
established that there is a weak nucleophilic participa-
-
- 11
showed better correlation with σp (FY
)
rather than
with the normal σp (FY) constants. In the Hammett plots
for substituent (X) changes in pyridines, a negative
deviation was observed with X ) m-CN (Figure 2) so that
nine pyridines, excluding the m-CN, are used in the FX
determination.
A striking trend in the selectivity parameters, FX, âX,
and FY in Table 1, is that they are all constant, within
the experimental errors, with changing reactivities of the
reacting partners (electrophiles and nucleophiles) FX
)
-2.25 ( 0.03, âX ) 0.42 ( 0.01 and FY ) 1.10 ( 0.06. As
a result, the cross-interaction constant, FXY in eq 3a,
becomes vanishingly small (FXY ≈ -0.02 with r ) 0.997
and n ) 36) and the rate constant represents a simple
sum of the two substituent effects without the last term
of eq 3a,6a-c eq 6a.
log(kXY/kHH) = FXσX + FYσY
(6a)
This means naturally that the Ritchie N+ relationship,
eq 1, holds since for a given substrate (i.e., for a given Y
(12) McClelland, R. A.; Kanagasabapathy, V. M.; Banait, N. S.;
Steenken, S. J . Am. Chem. Soc. 1992, 114, 1816.
(13) Li, H. G.; Kim, C. K.; Lee, B.-S.; Kim, C. K.; Rhee, S. K.; Lee,
I. J . Am. Chem. Soc. 2001, 123, 2326. The acylium ion is known to be
strongly stabilized by the electron-donating substituent theoretically.
(10) Spillane, W. J .; McHugh, F. A.; Burke, P. O. J . Chem. Soc.,
Perkin Trans. 2 1998, 13.
(11) J ohnson, C. D. The Hammett Equation; Cambridge University
Press: Cambridge, 1973; p 28.

8552 J . Org. Chem., Vol. 66, No. 25, 2001
Lee et al.
Ta ble 2. Secon d -Or d er Ra te Con sta n ts, k2 (× 104 M-1 s-1),a for th e Rea ction s of Y-Su bstitu ted
N-Meth yl-N-a r ylca r ba m oyl Ch lor id es a n d X-P yr id in es in Dim eth yl Su lfoxid e a t 45.0 °C w ith KCl (M) Ad d ed
[KCl]/M
X ) Y
0
0.0015
0.002
0.003
0.006
0.01
0.02
0.03
0.04
k2/10-4 M-1 s-1
p-CH3O
H
56.2
31.6
57.4
32.9
58.0
33.8
58.6
34.8
59.2
35.7
61.3
36.6
62.0
37.3
63.4
38.0
64.2
38.5
a
Averages of three to five determinations.
Sch em e 1
tion by solvent toward a predominantly SN1 solvolysis.2
Kevill and co-workers1h have shown that the solvolysis
of 1 and 2 involves low values of l (0.2-0.3) as well as m
(0.5-0.6) in eq 7, which was interpreted to support the
SN1 mechanism in which positive charge is delocalized
from the developing acylium ion to the adjacent nitrogen
(which provides a substantial internal assistance) leading
to a relatively low sensitivity to changes in solvent
ionizing power. Their product analysis,1h however, showed
F igu r e 3. Plot of k2 (×104 M-1 s-1) vs [KCl] for X ) Y ) H:
(A) suppression of internal return of solvent-separated ion
pairs by the ion-triplet [Cl-||R+||Cl-] formation; (B) normal salt
effect with b ≈ 2.0.
log(k/k0) ) mY + lN +c
(7)
complexes are formed between anionic substrate (NHP-)
and cations of salts which are presumed to exist in
solvent-separated loose ion-pair forms. The initial rise
in kobs was found to become steeper with the increase in
the acetonitrile content of solvent mixture due to the
increase in ion-pair complex formation. The rate-deter-
mining step has been shown to be the hydroxide ion
attack at the carbonyl carbon of NHP-.
clearly that products are formed primarily by reaction
of the carbocation with solvent at the solvent-separated
ion-pair stage. These results also support the nucleophilic
attack at the resonance-stabilzed carbocations.
We have investigated salt effects on the rate by adding
KCl to the solution as shown in Table 2. The plot of k2 vs
[KCl] in Figure 3 shows an initial rapid nonlinear rise
of k2 which is quite similar to that observed in the special
salt effect on solvolyses proceeding through solvent-
separated ion-pairs.14 Such special salt effect is only
associated with long-lived carbocations, and the initial
rapid increase in k2 has been ascribed to the suppression
of internal return of the solvent-separated ion-pair by the
added salt.14 In the present reaction, however, the added
anion, Cl-, is the common ion and we think that the
solvent-separated ion-pairs become stabilized and the
internal return is suppressed by ion-triplet formation,
Cl-||R+||Cl-, where R+ is the carbamoyl cation. This is
reasonable since ion-pair formation becomes much more
stronger in aprotic solvents than in water and cations
are stabilized by the solvent, DMSO, and also by the two
counterions, Cl-. This type of ion-triplet formation in the
SN1 and SN2 processes has been discussed in detail by
Ritchie.15 Khan16 has observed a similar sharp initial rise
in kobs for the alkaline hydrolysis of N-hydroxyphthal-
imide (NHP) with addition of salts (NaCl, KCl etc.) in
aqueous acetonitrile solution. He proposed the ion-pair
formation mechanism in explaining the results: ion-pair
Addition of the relatively large amount of salt, [KCl]
> 0.01 M, causes only a normal salt effect increase in k2,
showing that there is no appreciable trapping of a free
carbocation by chloride ion. Thus the reaction seems to
proceed through a solvent-separated ion-pair, not through
a dissociated carbocation.14a Since the reaction is run in
a poor ionizing solvent (DMSO; YOTS ) -1.3117), the
dissociation of ion-pairs to free carbocations become
energetically unfavorable.15 A relatively large increase
in reaction rate due to normal salt effect is observed with
b ≈ 2 in eq 8, since ionization is involved in less polar
solvent, DMSO. For example in the ionization of 4-
methoxyneophyl tosylate in the presence of LiClO4, the
constant b has been shown to increase with the decrease
of solvent polarity;14b b values were 0.0 (DMSO), 1.4
(DMF), 47.0 (Me2CO), and 482 (THF).14b
log k ) log k0 + b[KCl]
(8)
Since the individual selectivities of the nucleophile and
substrate, FX (âX) and FY, are nonzero but there is no
interaction between them, FXY ≈ 0, such a mechanism,
i.e., the nucleophilic attack on the resonance-stabilized
(14) (a) Alder, R. W.; Baker, R.; Brown, J . M. Mechanism in Organic
Chemistry; Wiley: London, 1971; p 85. (b) Ruff, F.; Csizmadia, I. G.
Organic Reactions. Equilibria, Kinetics and Mechanism; Elsevier:
Amsterdam, 1994; p 300.
(15) Ritchie, C. D. Physical Organic Chemistry, 2nd ed.; Marcel
Dekker: New York, 1990; p 132.
(16) Khan, M. N. J . Phys. Org. Chem. 1994, 7, 412
(17) Peterson, P. E. In Nucleophilicity; Harris, J . M., McManus, S.
P. Eds.; American Chemical Society: Washington D. C., 1987; p 299.

Pyridinolysis of N-Methyl-N-arylcarbamoyl Chlorides
J . Org. Chem., Vol. 66, No. 25, 2001 8553
Ta ble 3. Activa tion P a r a m eter sa for th e Rea ction s of
Y-Su bstitu ted N-Meth yl-N-a r ylca r ba m oyl Ch lor id es w ith
X-P yr id in es in Dim eth yl Su lfoxid e
carbocation, seems quite appealing for the present series
of reactions, eq 9. The factors in favor of this proposed
mechanism are: (i) Constant selectivities of nucleophile
(âX ) 0.42 and FX ) -2.25) for variations of substituent
Y is observed, i.e., the N+ type relation can be applied,
which is known to hold primarily for the nucleophile
addition to the strongly resonance-stabilized carboca-
tions. (ii) The dipolar aprotic solvent used, DMSO, is
known to stabilize cationic species specifically.18 (iii) The
magnitude of the FX (Fnuc) values (≈-2.25) indicates that
there is ca. one-third positive charge development on the
pyridine nitrogen in the TS since the F value for the
protonation of pyridinium ions is -5.90.19 This means
that there is considerable charge transfer in the TS, with
substantial positive charge development on the pyridine
nitrogen, as has been reported for the amine addition to
the resonance-stabilized triarylmethyl carbocations.12 (iv)
A stronger electron donating substituent (Y) in the ring,
e.g., Y ) p-CH3O, reduces the addition rate (Table 1),
i.e., FY > 0. If a limiting SN1 mechanism were to apply,
bond cleavage should have progressed much ahead of
bond formation in the TS, so that the converse should
hold since an electron donor Y facilitates the bond
cleavage (FY < 0), as we normally find in the SN1
reactions. Similar magnitude of positive Hammett coef-
ficients FY ranging from FY ) +0.97 (with CH3CH2CH2-
NH2) to FY ) +1.37 (with CF3CH2NH2) has been observed
for the primary amine additions to the substituted (Y,
Y′, Y′′-T) triarylmethyl carbocations in aqueous aceto-
nitrile solution at 20.0 °C.12 In this series of reactions
the FY values showed a small but definite trend of
T
k2 (10-4
(°C) M-1 s-1
∆Hq
-∆Sq
X
Y
)
(kcal mol-1
)
(cal mol-1 K-1
)
p-CH3 p-CH3O
55
45
35
55
45
35
49.8
33.9
23.0
7.1 ( 0.3
9.9 ( 0.3
5.6 ( 0.2
9.4 ( 0.3
8.7 ( 0.4
10.9 ( 0.4
5.5 ( 0.2
48 ( 2
37 ( 1
54 ( 2
41 ( 1
37 ( 1
43 ( 2
54 ( 2
p-CH3
H
H
125
74.1
43.8
21.5
15.8
11.6
52.2
31.6
19.2
p-CH3CO 55
45
35
55
45
35
H
H
H
p-NO2
55 1150
45
35
55
45
35
55
45
35
724
455
1.55
0.871
0.489
24.2
m-CN
H
m-CN p-NO2
17.8
13.1
a
Calculated by the Eyring equation. Errors shown are standard
deviations.
(δσY < 0) stabilizes the cationic form 3+. The large
negative entropies of activation, ∆Sq, are consistent with
the rate-limiting addition of the two reacting species in
the TS.
increase as the basicity of the amine decreased, FY
)
+0.97 f +1.37 for ∆pKX ) -4.7,12 which is in contrast
to the trend found not only in this work (∆FY ≈ 0 for ∆pKX
) -5.2) but also in the original reactions from which
Ritchie introduced the N+ equation. (v) If the present
reaction series proceeded by an SN2 mechanism, the
positive FY values in Table 1 indicate that negative charge
development at the carbonyl carbon occurs as a result of
a greater degree of bond-making than bond-breaking in
the TS, which should lead to substantial interaction
between the nucleophile and substrate in the SN2 TS with
FXY ) -0.6 to -0.8,6a,b as we have normally observed.
Thus, the near-zero value of FXY precludes the SN2
mechanism for the present series of reactions. (vi) The
Hammett plots for variation of substituent Y in the
There is no general agreement on why nucleophile
selectivities are constant with changing electrophile
reactivity for reactions of strongly resonance-stabilized
carbocations that follow the Ritchie N+ equation.4,5,12,20
The following analysis is an attempt to provide some
insights into this problem.
Since we have obtained practically constant âX values,
δâX ) 0 (Table 1) with changing substrate (δpKY), the
cross-interaction constant âXY is zero since âXY ) ∂âX/
∂∆pKY from eq 3b. Thus our results in Table 1 indicate
that âXY ≈ 0.
-
substrate showed better correlation with σp (FY > 0)
rather than with σp suggesting that there is through
conjugation effect of the lone pair on the N toward the
electron withdrawing Y substituents in the second step
transition state which is absent in the cationic interme-
diate where the lone pair primarily delocalizes toward
the carbonyl group. This type of through conjugation has
been observed in the deprotonation equilibria of anilin-
ium ions, which has led to the introduction of σp- constant
with FY > 0. (vii) The results of salt effects (Table 2)
support the hypothesis that the reactions proceed via the
solvent-separated ion-pairs. (viii) Finally, the activation
parameters in Table 3 are in accord with the proposed
mechanism. The ∆Hq values are lower for both electron-
acceptor and donor Y with a given X. This is reasonable
since an electron acceptor (δσY > 0) facilitates the
addition step (ka in eq 9) whereas as an electron donor
It has been shown that for families of nucleophiles (X),
values of log KX (-∆GX°/2.3RT) are linearly related to
pKa’s of the conjugate acids of nucleophiles (∆GHX°/2.3RT)
with slopes not far from unity.4c Similar relations, δ∆GY°
) δ∆GR° where ∆GR° ) -2.3RT log KR for
R+ + H2O {KR} ROH + H+
are also observed for the electrophiles (Y) in the carboca-
tion addition reactions.4c Thus eqs 3 can be transformed
into free energy forms (e.g., eq 10), and the definitions
of âXY can be given by eq 1121 with 2.3RT ) 1.36 kcal
mol-1 at 298 K.
∆G q ) âX′∆GX° + âY′∆GY° + âXY′∆GX°∆GY° (10)
∂2∆Gq
âXY
(18) Reference 9, Chapter 3.
(19) Page, M.; Williams, A. Organic and Bio-organic Mechanisms;
Longman: Harlow, 1997; p 200.
â
XY′ ) -
)
(11)
1.36 ∂∆GX°∂∆GY°

8554 J . Org. Chem., Vol. 66, No. 25, 2001
Lee et al.
On the other hand, the Marcus equation22 can be
represented in the free energy form as eq 12, where ∆G°
comprises two component terms, δ∆G° ) δ∆GX° + δ∆GY°;
i.e., the changes in the thermodynamic driving force can
be disected into the contributions of the two reactants,
X and Y. Arnett et al., have shown that for the addition
of resonance-stabilized carbocations and organic anions
the free energy of reaction, ∆G°, can be expressed into
such an additive form.23
stabilized cation, or the reactions that follow strictly the
N+ relationship, proceed by an electron-transfer (ET)
mechanism.24 In the electron-transfer reaction, the trans-
fer of a single electron from D to A leads to D+ and A-,
(D+ + A-) but the two moieties are not linked by a
covalent bond in contrast to the nucleophilic addition
pathway in which there is a covalent bond (D+-A-). Since
in the ET process there is no covalent link in the TS, the
cross-interaction between D+ and A- should be negligible,
FXY ≈ 0. It has been shown that any factor (in this case
∆G°2
q
1
2
electronic due to high intrinsic barrier, ∆E0 ) large) that
∆Gq ) ∆G0q + ∆G° +
(12)
(13)
q
inhibits or hinders the bond coupling process tends to
favor the ET process.20a The proclivity of the substrate
to undergo ET reactions should increase with availability
of a high lying π* LUMO. This requirement is certainly
met with all the strongly resonance-stabilized carboca-
tions including the N-methyl-N-arylcarbamoyl cations.24
The higher the π* LUMO, the larger is the interfrontier
energy gap, ∆ꢀ ) ꢀLUMO - ꢀHOMO () ꢀπ* - ꢀn) and hence
the smaller is the charge transfer stabilization in the
incipient bond formation in the TS, i.e., the higher is the
kinetic (intrinsic) barrier. In such kinetically hindered
charge-transfer stage, the ET process should be favored
with an outer-sphere complex formation in which there
is no bonding interaction (FXY ≈ 0).24a
16∆G0
∂2∆Gq
1
)
q
∂∆GX°∂∆GY°
8∆G0
Partial second derivatives of ∆Gq in eq 12 with respect
to ∆GX° and ∆GY° give eq 13.5 Equations 11 and 13 link
the two relationships of eq 3 and 12 as eq 14, where c is
a constant. This equation tells us that the cross-interac-
tion constants, âXY (FXY), are dependent only on the
q
kinetic (intrinsic) barrier, ∆G0 , and independent of the
thermodynamic driving force, ∆G°.
- 1
6∆G0
â
XY( ) cFXY) =
(14)
q
Con clu sion s
The nucleophilic substitution reactions of Y-substituted
N-methyl-N-arylcarbamoyl chlorides (YC6H4N(CH3)COCl)
with pyridines (XC5H4N) in DMSO show constant FX (and
âX) with changing substrates (δσY), and constant FY with
changing nucleophiles (δσX). As a result, the cross-
interaction constant FXY (and âXY) is vanishingly small
and the Ritchie N+ type equation holds. These and other
results lead us to the nucleophile addition mechanism
on the resonance-stabilized carbocations, 5. From the
definition of cross-interaction constant, FXY (âXY), and the
Marcus equation, it can be shown that for such reactions
with vanishing cross-interaction and valid N+ type
Our experimental results of âXY ≈ FXY ≈ 0 lead us then
to eq 15, which shows that the cross-interaction constant,
âXY (and FXY), is zero when the intrinsic barrier is high
- 1
6∆G0
âXY
=
≈ 0
(15)
q
(∆G0 ≈ ∞)5 for the intrinsic-barrier controlled reaction
series. This is reasonable since for the reaction series with
extremely high kinetic barrier, discrimination of reac-
tivities between different reactants becomes almost
impossible, i.e., the selectivity becomes insignificantly
small, FXY ) 0. Thus a necessary condition for the valid
N+ relationship with negligibly small cross-interaction
q
q
relationship the high intrinsic barrier, ∆G0 , is a neces-
q
sary condition, âXY = 1/(6∆G0 ) ≈ 0. The electron-transfer
q
(ET) reactions provide a possible type of process in which
cross-interaction between the nucleophile and electro-
phile is insignificant despite the substantial charge
transfer from the nucleophile to the electrophile.
constant, FXY ≈ 0, is high intrinsic barriers, ∆G0 ) large,
in the intrinsic-barrier controlled reaction series. It is
natural that if the N+ relationship holds the cross-
interaction constant is zero and vice versa, since the
nucleophile selectivities are constant, δâX ) 0 or δFX ) 0
with changing electrophiles reactivity δpKY or δσY, when
the N+ relationship holds so that δâX/δpKY ) âXY ) 0 or
δFX/δσY ) FXY ) 0. It is, however, rather unexpected that
the negligible cross-interaction is obtained when the
Exp er im en ta l Section
Ma ter ia ls. International Specialty Chemical ACS grade
dimethyl sulfoxide (DMSO) was used after two distillations.
The pyridine nucleophiles, Aldrich GR, were used without
further purification. The substrate, N-methyl-N-arylcarbamoyl
chloride, was Aldrich GR purchased and was recrystallized
before use. Y-Substituted N-methyl-N-arylcarbamoyl chlorides
were prepared by reacting Y-substituted N-methyl-N-aryl-
anilines with triphosgene followed by small amount of triethyl-
amine in ethyl acetate. The product, p-methyl-N-methyl-N-
arylcarbamoyl chloride, was obtained by column chromatog-
raphy (silica gel, 30% ethyl acetate/n-hexane). The other
products, para-substituted (Y) N-methyl-N-arylcarbamoyl chlo-
rides, were purified by recrystallization from n-hexane. The
substrates synthesized were confirmed by spectral and el-
emental analysis as follows.
q
intrinsic barrier is high, ∆G0 ) large.
There are therefore two cases where the cross-interac-
tion becomes negligible, FXY ≈ 0. (i) No interaction due
to large distance (rXY ) ∞) involved between the two
reactants.6a,b (ii) Indiscrimination due to large intrinsic
q
barriers, ∆G0 ≈ ∞. However, these seemingly different
cases may be simplified into a unified condition of (i) if
we assume that the nucleophile additions to the resonance-
(20) Denton, P.; J ohnson, C. D. J . Chem. Soc., Perkin Trans. 2 1995,
477.
(21) Lee, I.; Lee, H. W. Bull. Korean Chem. Soc. 2001, 22, 732.
(22) (a) Marcus, R. A. J . Chem. Phys. 1956, 24, 966; 1963, 39, 1734.
(b) Marcus, R. A. Annu. Rev. Phys. Chem. 1964, 15, 155.
(23) (a) Troughton, E. B.; Molter, K. E.; Arnett, E. M. J . Am. Chem.
Soc. 1984, 106, 6726. (b) Arnett, E. M.; Chawla, B.; Motter, K.;
Amarnath, K.; Healy, M. J . Am. Chem. Soc. 1985, 107, 5288.
(24) (a) Pross, A. Theoretical and Physical Principles of Organic
Reactivity; Wiley: New York, 1995; Chapter 9. (b) Hoz, S. In Nucleo-
philicity; Harris, J . M., McManus, S. P., Eds.; American Chemical
Society: Washington D.C., 1987; p 181.

Pyridinolysis of N-Methyl-N-arylcarbamoyl Chlorides
J . Org. Chem., Vol. 66, No. 25, 2001 8555
p-CH3OC6H4N(CH3)COCl: liquid; δH (CDCl3) 6.7-7.0 (C6H4,
4H, m), 3.8 (CH3O, 3H, s), 3.2 (CH3, 3H, s); νmax (neat) 3040
(CH), 2850 (CH, aromatic), 1740 (CdO); m/z 199 (M+). Anal.
Calcd for C9H10ClNO: C, 54.3; H, 5.0. Found: C, 54.3; H, 5.1.
p-ClC6H4N(CH3)COCl: mp 68-70 °C; δH (CDCl3) 6.7-7.3
(C6H4, 4H, m), 3.2 (CH3, 3H, s); νmax (KBr) 3050 (CH), 2920
(CH, aromatic), 1740 (CdO); m/z 203 (M+). Anal. Calcd for
C8H7Cl2NO: C, 47.3; H, 3.4. Found: C, 47.2; H, 3.5.
p-NO2C6H4N(CH3)COCl: mp 102-104 °C; δH (CDCl3) 6.8-
7.6 (C6H4, 4H, m), 3.2 (CH3, 3H, s); νmax (KBr) 3100 (CH), 2950
(CH, aromatic), 1740 (CdO); m/z 214 (M+). Anal. Calcd for
C8H7ClN2O3: C, 44.9; H, 3.3. Found: C, 44.8; H, 3.2.
P r od u ct An a lysis. N-Methyl-N-arylcarbamoyl chloride
(C6H5N(CH3)COCl) was reacted with an excess of p-methylpy-
ridine (p-CH3C5H4N) with stirring for more than 15 half-lives
at 45.0 °C in dimethyl sulfoxide, and the products were isolated
by evaporating the solvent under reduced pressure. The
product mixture was separated by column chromatography
(silica gel, 15% ethyl acetate/n-hexane). The same method was
used for the other product. Analysis of the products gave the
following results.
C6H5N(CH3)CONC5H4CH3: δH, NMR (250 MHz, CDCl3),
8.0-8.8 (C5H4, 4H, m), 7.0-7.3 (C6H5, 5H, m), 3.2 (NCH3, 3H,
s), 2.2 (CH3, 3H, s,); m/z 213 (M+). Anal. Calcd for C14H15NO:
C, 91.8; H, 8.20. Found: C, 91.9; H, 8.19.
Ra te Con sta n ts. Rates were measured conductometrically
in dimethyl sulfoxide. The conductivity bridge used in this
work was a self-made computer automatic A/D converter
NO2C6H4N(CH3)CONC5H4CH3: δH, NMR (250 MHz, CDCl3),
8.0-8.8 (C5H4(pyridine), 4H, m), 7.0-7.3 (C6H4, 4H, m), 3.2
(NCH3, 3H, s), 2.2 (CH3, 3H, s); m/z 285 (M+); Anal. Calcd for
C14H14N2O3: C, 80.4; H, 6.70. Found: C, 80.3, H, 6.71.
conductivity bridge. Pseudo-first-order rate constants, kobs
,
were determined by the curve fitting analysis of the diskette
recorded data with a modified Origin program, which fits
conductance vs time data to the equation λ ) λ∞ + (λo - λ∞)
exp(-kobst), where λ∞, λo - λ∞, and kobs are iteratively optimized
to achieve the best possible least-squares fit with a large excess
of pyridine (Py); [N-methyl-N-arylcarbamoyl chloride] ≈ 1 ×
10-3 M and [Py] ) 0.03-0.24 M. Second-order rate constants,
k2, were obtained from the slope of a plot of kobs vs [Py] with
more than five concentrations of pyridine, eq 4. The k2 values
in Table 1 are the averages of more than three runs and were
reproducible to within (3%.
Ack n ow led gm en t. We thank Woosuk University
for support of this work. This work was also supported
by Grant No. 1999-2-123-0003-5 from the Basic Re-
search Program of the Korea Science and Engineering
foundation.
J O0108212