Y. Pan, C. Sun et al.
The structure of product ion from the loss of HX was con-
firmed by multistage mass spectrometry. Taking ion a’’ (see
Table 3) as a representative example, a comparative ion
loss of HX. If the ortho-, meta- and para-monosubstituted
benzyl cations firstly rearrange to monosubstituted tropyli-
um (then losing their positional identity), they should show
no difference in the subsequent HX elimination reaction.
Therefore, the involvement of tropylium ion in these SNAr
reactions can be ruled out. As one of the most typical fea-
tures of an SNAr reaction, fluoro is the best leaving group
among the halogens in most SNAr reactions.[23] In fact, the
elimination of HF is much more efficient than that of HCl,
HBr, and HI (see the relative abundances in Tables 1–3),
which clearly indicates that these substitution reactions un-
dertake an SNAr mechanism. In the gas phase, the high sta-
bility of HF may be another reason for the efficient loss of
HF.
+
À
with a definite structure of CH3N
N
designed and synthesized from protonated 4-(4-methylpiper-
azino)benzylamine by the elimination of ammonia.[20] The
fragmentation pattern of this ion is exactly the same as that
of ion a’’ (Figure 2). Therefore, the structure of ion a’’ is un-
doubtedly confirmed to be 4-(4-methylpiperazino)benzylium
and the most likely reaction mechanism to form this ion is
nucleophilic aromatic substitution.
To gain more insights into the mechanism of the present
SNAr reaction, DFT calculations were carried out at the
B3LYP/6-311+ +GACTHNUTRGNEUNG(2d,p) level of theory. The reaction of 4-
chlorobenzyl cation with piperazine through the ion/neutral
complex is used as a representative example. A schematic
potential energy diagram for this reaction is given in
Figure 3 and full details of the structures and energies of in-
volved species are provided in the Supporting Information.
The ion/neutral complex (Int-1), formed by dissociation of
the precursor ion ([M+H]+), is located 26.3 kJmolÀ1 below
the separated piperazine and 4-chlorobenzyl cation (ion b).
Within Int-1, the piperazine nucleophile is able to attack the
para position of ion b to give rise to a much more stable in-
termediate Int-2. Int-2 lies in a local energy minimum that is
32.9 kJmolÀ1 lower in energy than Int-1. Int-2 is a covalently
bonded complex, that is, a cationic s complex, in which the
À
Figure 2. CID MS3 spectra of the ion at m/z=189 derived from a) proton-
ated 4-(4-methylpiperazino)benzylamine (m/z=206), b) protonated 1-(4-
fluorobenzyl)-4-methylpiperazine (m/z=209).
calculated CACTHNUGTRNEUNG(aryl) N bond length is 1.57ꢂ. The conversion
of Int-1 into Int-2 is almost barrierless (see the Supporting
Information; Figure S2 for more details). The subsequent
HCl elimination leading to the formation of 4-piperazino-
benzylium (ion a) is also favorable in terms of energy be-
cause the total energy of ion a and HCl is 94.5 kJmolÀ1
lower than that of Int-2, although an energy barrier (TS-1)
should be surmounted. The reaction for the nucleophilic
attack of chlorotropylium by piperazine was also calculated
(see dashed line in Figure 3), however the corresponding s
Moreover, the elimination of hydrogen halides from the
protonated N-(4-halobenzyl)piperazines was further con-
firmed by isotopic labeling experiments. When deuterated
substrates were employed, elimination of the corresponding
deuterium halides was observed (Table 2). The proton affini-
ty (PA)[21] of piperazine is 944 kJmolÀ1, whereas the PA of
halobenzene is approximate 755 kJmolÀ1, which suggests
that the ionizing proton is attached to the thermodynamical-
ly favored piperazine nitrogen under ESI conditions. There-
fore, it is impossible that the loss of HX is initiated by pro-
tonation at the halogen substituent giving rise to phenyl cat-
ions. Proton transfer from nitrogen to halogen through the
phenyl ring (ring-walk[22]) during collisional activation is also
obviated because the loss of DX was observed only in the
deuterium labeling experiments (Table 2). These results con-
solidate the reaction mechanism and the product structures
presented above.
As shown in Scheme 1, the positive charge of the benzyl
cation can be resided at the para- or ortho position, but not
at the meta position, so the nucleophilic substitution can
occur at the para- or ortho position rather than at the meta
position. The experimental results are in good accordance
with this criterion (see entries 2–4 in Table 1). The m-chloro-
benzyl cation does not show SNAr reactivity with piperazine,
which supports that the methylene arenium is involved in
Figure 3. Calculated potential energy diagram for the nucleophilic aro-
matic substitution reaction between piperazine and 4-chlorobenzyl cation
(b) at the B3LYP/6-311+ +GACTHNUTRGNE(NUG 2d,p) level. The solid line represents the
involvement of benzylium ion and the dashed line represents the involve-
ment of tropylium ion.
10822
ꢁ 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2011, 17, 10820 – 10824