Full Paper
Finally, monoradicals 1, 2, and 3 and triradical 7 all have
doublet ground states and would be expected, therefore, to
be approximately equally reactive were their EAs similar. How-
ever, this is not the case because 7 has a substantially greater
(calculated) EA (6.06 eV) than 1–3 (5.65, 4.87, and 5.06 eV, re-
spectively; Table 1). It is not surprising then that 7 reacts at
a higher efficiency than both 2 and 3 with cyclohexane, allyl
iodide, dimethyl disulfide, and tert-butylisocyanide (Table 1).
On the other hand, 7 reacts more slowly than monoradical
1 with all reagents studied, despite its greater EA. This finding
is rationalized by the fact that the radical site that reacts first
in the triradical (C5) is not the inherently most reactive radical
site (C4) due to spin–spin coupling, which prevents radical site
C4 from reacting first.
tive than triradical 7 due to its very high (calculated) EA
(7.20 eV),[20] which enhances the reactivity of both the most re-
active radical site and the ortho-benzyne moiety. All of these
findings serve to emphasize the importance of polar effects on
the chemical properties of bi- and triradicals. Furthermore, the
(calculated) D-Q splittings of the triradicals appear not to be
correlated with their reactivities because it is only À7.1 kcal
molÀ1 for
7
(calculated herein), but À30.9, À39.2, and
À49.9 kcalmolÀ1 for the 2,4,6-,[19] 3-hydroxy-2,4,6-,[19] and 3,4,5-
tridehydropyridinium ions, respectively;[20] the last three triradi-
cals were all more reactive than 7.
Experimental Section
Radical precursors and FT-ICR mass spectrometry
Conclusion
The precursor for monoradical 1, 4-iodoisoquinoline, was synthe-
sized according to a procedure reported in the literature.[30] The
precursor for monoradical 2, 5-nitroisoquinoline, was obtained
from Sigma-Aldrich and used as received. The precursor for mono-
radical 3, 8-iodoisoquinoline, was synthesized by dissolving 8-ami-
noisoquinoline (0.5 g, 3.36 mmol; Carbocore) in a cold solution
(ice–salt bath) of concentrated HCl (3 mL) and H2O (3 mL), and
then slowly adding a solution of NaNO2 (0.28 g, 4.03 mmol) in H2O
(2 mL). The resulting red solution was stirred for 15 min, and a solu-
tion of KI (1.12 g, 6.73 mmol) in H2O (3 mL) was added. The mixture
was then heated for 3 h at 1008C, cooled, and made basic with
aqueous NH3. The product was extracted with dichloromethane
and the organic layer washed with 5% sodium metabisulfite fol-
lowed by brine. Chromatography on silica gel (95:5 CH2Cl2/MeOH)
gave 8-iodoisoquinoline (0.52 g, 60%). 1H NMR (400 MHz, CDCl3):
d=7.30 (t, J=7.9 Hz, 1H), 7.44 (d, J=5.6 Hz, 1H), 7.71 (d, J=
7.1 Hz, 1H), 8.05 (d, J=7.7 Hz, 1H), 8.54 (d, J=5.7 Hz, 1H),
9.35 ppm (s, 1H); ESI-MS: m/z: 256 [M+H]+.
Ion 7 underwent typical radical reactions with all of the re-
agents studied and was more reactive than most of the mono-
and biradicals studied. The high reactivity of the triradical was
attributed mainly to the EA, which was greater than those for
any of the other radicals studied. This caused 7 to react faster
than, for example, two related monoradicals, 2 and 3, the reac-
tivities of which are primarily controlled by polar effects. How-
ever, triradical 7 reacted more slowly than the most reactive
monoradical, 1 (with the highest EA of the monoradicals), due
to a stabilizing interaction of the intrinsically most reactive rad-
ical site (C4) in the triradical with the radical site at C8, which
prevents this radical site from reacting first. Compared with
the triradical, the reactivity of a related singlet (ground state)
biradical, 4, is reduced by both a lower EA and greater cou-
pling of the biradical electrons. However, the triradical reacted
at comparable rates to that of 5 with a lower EA, but a triplet
ground state, since the most reactive radical site is not C4, but
C5, for the reasons discussed above.
The precursors for biradicals 4 and 5, 4-iodo-8-nitroisoquinoline
and 4-iodo-5-nitroisoquinoline, respectively, were synthesized by
dissolving 4-iodoisoquinoline (110 mg, 0.43 mmol) in H2SO4 (5 mL)
at 08C. An excess of HNO3 was added at À38C to the brown solu-
tion. The solution was stirred for 15 min at this temperature, al-
lowed to warm to room temperature, and then stirred for an addi-
tional 2 h. The solution was poured into ice and made basic with
aqueous NH3. After filtration and chromatography on silica gel
(20:80 ethyl acetate/hexanes), 4-iodo-5-nitroisoquinoline[21] (60 mg)
and 4-iodo-8-nitroisoquinoline (10 mg) were obtained. 4-Iodo-8-ni-
troisoquinoline: 1H NMR (400 MHz, CDCl3): d=7.89 (t, J=8.1 Hz,
1H), 8.39 (d, J=7.6 Hz, 1H), 8.43 (d, J=8.7 Hz, 1H), 9.14 (s, 1H),
9.85 ppm (s, 1H); ESI-MS: m/z: 301 [M+H]+.
Triradical 7 is best described as a relatively unreactive biradi-
cal (C4 and C8 radical sites) with a reactive radical site (C5). In-
terestingly, the monoradical with the analogous radical site, 2,
is the least reactive of the three isomeric monoradicals (it also
has the lowest EA). Hence, reactivity predictions for aromatic
carbon-centered s,s,s-type triradicals cannot be made based
solely on the chemical properties of analogous monoradicals,
even when the coupling between the radical sites is weak. This
finding is in contrast to that made for the 2,4,6-tridehydropyri-
dinium and 3-hydroxy-2,4,6-tridehydropyridinium ions, in
which the most reactive radical site (C2) is the same as that for
the analogous, most reactive monoradical.[19] Both of these tri-
radicals react with greater efficiencies than triradical 7 due to
their substantially greater EAs (by almost 1 eV; EA=7.01 and
6.86 eV, respectively).[19] They are best viewed as highly reactive
monoradicals with a relatively unreactive and an entirely un-
reactive meta-benzyne moiety, respectively. On the other hand,
the 3,4,5-tridehydropyridinium ion is best viewed as a highly
reactive ortho-benzyne with a reactive radical site, which reacts
through radical mechanisms at the C3 radical site, but through
nonradical mechanisms at either the C3 or C4 radical sites.[20]
The 3,4,5-tridehydropyridinium ion is substantially more reac-
The precursor for triradical 7, 4,5,8-triiodoisoquinoline, was synthe-
sized in the following manner. 4,5,8-Tribromoisoquinoline[31] (0.6 g,
1.64 mmol) in THF (3 mL) was added slowly to a solution of THF
(40 mL) and nBuLi (2.5m in hexanes; 4.3 mL, 10.8 mmol) at À788C
under nitrogen at a rate that kept the solution temperature at or
below À608C. The mixture was stirred for 90 min, and then a solu-
tion of I2 (2.78 g, 10.8 mmol) in THF (8 mL) was added over
a period of 20 min, while keeping the temperature below À608C.
The reddish suspension was stirred for 1 h at À788C, allowed to
warm slowly to room temperature, and then stirred for an addi-
tional 4 h. After dilution with diethyl ether (40 mL), the solution
was washed with 5% sodium metabisulfite. The organic layer was
dried and concentrated under reduced pressure to give an orange
solid that was purified by chromatography on silica gel (20:80
ethyl acetate/hexanes) to give 4,5,8-triiodoisoquinoline (330 mg,
Chem. Eur. J. 2016, 22, 809 – 815
813
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim