3094 J . Org. Chem., Vol. 62, No. 10, 1997
Arnone et al.
measured at three different temperatures together with
the relevant thermodynamic data are collected in Tables
1 and 2.
The reactivity ratios measured for benzenethiolate-
dehalogenation reactions in methanol of compounds 3
and 5 furnish results similar to those obtained by Miller
and Williams in the methoxy-dechlorination of com-
pounds 3 and 4. Moreover, the ratios are practically
independent of the halogen present [at 35 °C in 5: kCONH
/
2
k
CO Me 2.2, 3.0, and 2.5 for L ) Cl, Br, and I, respectively;
2
in 3: kCONH /kCO Me 0.14 for all the three halogens].
2
2
Bearing in mind the differences in the electronegativities,5e
in the abilities to give hydrogen bonding,5a in the
strengths5b and in the polarizabilities5c of the CAr-Halg
bond and in the volumes of the halogens used, the results
obtained can hardly agree with the hypothesis of Miller
and Williams.4
However, to gain information on the peculiar activation
of the o-carboxamido substituent we have extended our
study to the piperidino-dehalogenation reactions of sub-
strates 3 and 5 in methanol. The reactivity ratios
The results obtained by us by studying the kinetic
effects of carboxamido and carbomethoxy groups in
thiophenes 1 and 2 induced us to question this interpre-
tation. Indeed, the hydrogen bonding suggested by Miller
and Williams would be effective in activating SNAr with
both anionic (e.g., methoxide or benzenethiolate) and
neutral (amines: e.g., piperidine) nucleophiles. In con-
trast, the greater effectiveness of the o-carboxamido
group with respect to the o-carbomethoxy group was
observed only with anionic nucleophiles, i.e., with re-
agents that can give in some manner a local increase of
nucleophile concentration, eventually via a hydrogen
bond formation between the hydrogen atoms of carbox-
amido group and the anionic nucleophiles.
In order to gain further information on the relative
activating power of carbomethoxy and carboxamido
groups we have extended our studies in this field on the
basis of the following considerations. Firstly, the inter-
pretation of data offered by Miller and Williams makes
no differentiation between anionic and neutral nucleo-
philes, while our data with thiophene derivatives give a
different response; therefore, we have studied an anionic
and a neutral nucleophile to ascertain this point. Sec-
ondly, since these authors explain their results on the
grounds of Halg‚‚‚H‚‚‚N hydrogen bonding, which would
weaken the CAr-Halg bond, we have extended the
investigation to bromine and iodine in addition to chlo-
rine, because the three halogens have different hydrogen-
bonding forming abilities5a as well as different bond
strengths5b and different polarizabilities5c of the relevant
CAr-Halg bonds. Finally, Miller and Williams assumed
that the O‚‚‚O repulsive forces, particularly important
in the case of carbomethoxy group, disfavor the reactions
where this substituent is involved with respect to those
of the corresponding carboxamides: to lessen these
interactions we have tested an anionic sulfur nucleophile
instead of an anionic oxygen nucleophile.5d
obtained, always less than unity [at 35 °C, in 5: kCONH
/
2
kCO Me 0.27, 0.59, and 0.29 in 3: kCONH /kCO Me 0.20, 0.22,
2
2
2
and 0.24 for L ) Cl, Br, and I, respectively], are not
consistent with the hypothesis of Miller and Williams4
on the general activation of the o-carboxamido group.
Indeed, in the piperidino-dehalogenation reactions in
methanol, the carboxamido and carbomethoxy groups
show the same sequence of activating ability in both the
ortho- and the para-positions (the carbomethoxy group
is more activating than the carboxamido group), in
contrast with the proposed occurrence of a L‚‚‚H‚‚‚N
hydrogen bonding suggested by Miller and Williams,
which should be operating with both anionic and neutral
nucleophiles in 5. In every case, the activation by the
carboxamido group with respect to the activation by the
carbomethoxy group should be favored.
The ortho/para ratios6 (k5/k3) at 35 °C in methanol for
the two activating groups are consistent with our indi-
cations: indeed, they indicate a peculiar activation by
the o-carboxamido group only with sodium benzenethio-
late (R ) NH2: k5/k3 0.22-0.30; R ) OMe: k5/k3 0.013-
0.018) and not with piperidine (R ) NH2, OMe k5/k3
0.03-0.08).
P ip er id in o-Deh a logen a tion Rea ction s of 3 a n d 5
in Ben zen e. Kinetic data measured at three different
temperatures together with the relevant termodynamic
data are collected in Table 3.
Thus, we have studied the reactivity with sodium
benzenethiolate (in methanol) and with piperidine (in
methanol and benzene) of 1-halogeno-4-COR-2-nitroben-
zenes (in 3: L ) Cl, Br, I; R ) NH2, OMe) and 1-halogeno-
6-COR-2-nitrobenzenes (in 5: L ) Cl, Br, I; R ) NH2,
OMe). Substrates 5 have been preferred to 4 in order to
make a more homogeneous comparison. In fact, both
compounds 3 and 5 contain a nitro group ortho to the
reaction center which assures the occurrence of compa-
rable proximity effects, particularly in the case of neutral
nucleophiles (e.g., piperidine, see below).
The kinetic data obtained show that in piperidino-
dehalogenation reactions the reactivity of o- and p-COR
halogeno compounds 3 and 5 is affected by the nature of
the solvent used. Thus, the reactivity ratios calculated
for 3 in benzene (i.e., for the p-COR compounds) are
always less than unity (at 35 °C: kCONH /kCO Me 0.27, 0.29,
2
2
and 0.21 for L ) Cl, Br, and I, respectively), which is
analogous to the situation observed in methanol. In
contrast, reactivity ratios near unity are observed for 5
(i.e., for o-COR compounds, at 35 °C: kCONH /kCO Me 1.1,
2
2
1.3, and 1.5 for L ) Cl, Br, and I, respectively] that show
that for the o-COR compounds the ratios are significantly
solvent-dependent. Usually, on going from a protic and
polar solvent (e.g., dioxane/water, methanol) to an aprotic
Resu lts a n d Discu ssion
Ben zen eth iola te- a n d P ip er id in o-Deh a logen a tion
Rea ction s of 3 a n d 5 in Meth a n ol. Kinetic data
(6) It must be noted that all of the ortho/para ratios are lower than
unity, indicating the occurrence of a significant kinetic primary steric
effect (e.g., see: Miller, J . Aromatic Nucleophilic Substitution; Elsevi-
er: Amsterdam, 1968; pp 95 and 350 and references therein) in the
ortho-isomers because of the 2,6-disubstitution.
(5) Isaacs, N. S. In Physical organic chemistry; Longman Scientific
& Technical: Harlow, 1987; (a) pp 62-64; (b) pp 36-37; (c) pp 246-
247; (d) pp 456-457; (e) p 31; (f) pp 171-172.