A R T I C L E S
Perrin et al.
Hz). GC- MS (EI) m/z calcd for C10H11Br (M+): 210.0; found 210.0,
plus peaks at M + 2 and M + 1 (partially deuterated).
Scheme 3. Proposed (Complete) Mechanism for Haloaromatic
Formation via Halide Addition to a p-Benzyne Derived from an
Enediyne
1-Iodo-5,6,7,8-tetrahydronaphthalene: 1H NMR (DMSO-d6) δ 1.61
(2H, m, CH2CH2CH2CH2), 1.70 (2H, m, CH2CH2CH2CH2), 2.51 (2H,
t, J ) 7.0, CH2CH2CH2CH2), 2.65 (2H, t, J ) 6.0, CH2CH2CH2CH2),
6.81 (1H, m), 7.05 (<1H, d, J ) 7.5 Hz), 7.60 (1H, d, J ) 7 Hz).
GC-MS (EI) m/z calcd for C10H11I (M+): 258.0; found 258.1, plus
peak at M + 1 (partially deuterated).
Kinetics. All studies were done in pairs that differed in only one
variable. Besides an initial 1H NMR spectrum each sample was analyzed
at intervals, until e5% of starting material remained. The time to
remove the sample and acquire each spectrum was 10-20 min, which
was subtracted from the time elapsed.
is where one molecule of the spin-trapping reagent phenyl tert-
butyl nitrone is added along with one hydrogen atom from
â-mercaptoethanol.8
To explain the unusual incorporation of hydrogen and chlorine
in 1 and 2 we here propose attack of halide on a p-benzyne to
generate a haloaryl anion that is trapped in situ by a proton
donor. For our studies we used cyclodeca-1,5-diyn-3-ene (5).9
Its 10-membered ring has sufficient strain that cyclization to a
p-benzyne (6) occurs at a convenient rate on slight heating.
Nucleophilic attack by halide on 6 can then lead to a haloaryl
anion (7, X ) Cl, Br, I, or the corresponding aryllithium), which
is readily protonated to give 1-halotetrahydronaphthalene (8),
as proposed in Scheme 3.
Disappearance of starting material was monitored by comparing the
integrated intensity of the olefinic CH singlet of the enediyne (δ 5.88)
to that of 1,3,5-trichlorobenzene (δ 7.61) as internal standard. The first-
order rate constant k was obtained from a logarithmic plot of [enediyne]
versus time. Nonlinear fitting to exponential decay gave nearly the same
values. Deviations from first-order kinetics were not explored for
reactions without added halide.
1
Yields. Percent yields of product were obtained from the H NMR
integrations of product signals at the last time point, relative to internal
standard. They are reported as percent conversions, relative to the
amount of reactant consumed, under conditions where >90% was
consumed. One of the product doublets was often weaker than the other
two signals, especially when the reaction was carried out with D2O
(which exchanges with the pivalic acid). This was due to deuterium
incorporation, which could be confirmed by GC-MS analysis, as noted
in the characterizations above. Those weaker signals were not used for
the analysis of product yields. With LiBr or LiCl yields were instead
based on the intensity of the most downfield doublet. With LiI this
overlapped with the internal standard, so the product CH2 intensity at
δ 2.65 was compared to the pivalic acid intensity at δ 2.35. Owing to
errors in integration, the accuracy of the percent yields is generally
(4.
We now show that heating 5 in the presence of halide anion
and an acid does indeed produce 8 (X ) Cl, Br, I) and that the
kinetics are consistent with 6 as the intermediate that is captured
by halide. Moreover, this mechanism can account for the 1:1
mixtures and for the single chlorine in enediyne-derived natural
products such as 1 and 2.
Experimental Section
Sample Preparation. Cyclodeca-1,5-diyn-3-ene (5) was synthesized
by a standard procedure.10 It was added to a solution of lithium halide
in wet DMSO-d6 containing an aliquot of 1,3,5-trichlorobenzene as
internal standard. Because formation of halotetralin (8) consumes HX,
the reaction produces an equivalent of base, which would complicate
the reaction by functioning as another nucleophile. To avoid this and
establish buffered conditions once the reaction starts, a carboxylic acid
(acetic or pivalic) was included. No effort was made to remove water
from the solvent, inasmuch as the carboxylic acid also hydrogen-bonds
to the halide. The sample in an NMR tube was degassed by freeze/
pump/thaw on a Schlenk line, sealed, and then immersed in a 37 °C
((0.1 °C) oil bath. Disappearance of starting material and formation
of product were monitored periodically by removing the sample and
analyzing it by NMR and after ∼80 h by mass spectrometry.
NMR spectra were obtained on a Varian 500 MHz UNITY
spectrometer. GC-mass spectra were obtained on a Thermo-Finnigan
Trace GC/MS Plus.
Characterization of Products. 1-Chloro-5,6,7,8-tetrahydronaph-
thalene: 1H NMR (DMSO-d6) δ 1.67 (2H, m, CH2CH2CH2CH2), 1.73
(2H, m, CH2CH2CH2CH2), 2.64 (2H, t, J ) 6.5, CH2CH2CH2CH2),
2.70 (2H, t, J ) 6.0, CH2CH2CH2CH2), 7.00 (<1H), 7.05 (1H, d, J )
7.5 Hz), 7.15 (1H, d, J ) 8 Hz). GC-MS (EI) m/z calcd for C10H10-
ClD (M+): 167.2; found 167.2, plus peaks at M + 2 and M - 1
(incompletely deuterated).
Computations. Ab initio density functional theory calculations on
the energy of interaction between p-benzyne (9) and F-, Cl-, OH-, or
H2O were performed at the B3LYP/6-31G(d,p) level using Gaussian
98, revision A.7.11 Basis-set superposition error was ignored.
Results
When a solution of 5 in DMSO-d6 at 37 °C is monitored by
1H NMR, the olefinic CH singlet at δ 5.88 slowly disappears.
In the presence of excess lithium halide plus excess weak acid
three new signals appear between δ 7.0 and 7.2 (Cl), 7.0 and
7.35 (Br), or 6.8 and 7.6 (I). In addition, the two CH2 signals
of 5 separate into two pairs of signals, indicating a symmetry
reduction, which excludes 8 (X ) H), the usual product from
6.5 Figure S1 in the Supporting Information shows representative
1H NMR spectra in the presence of LiBr. GC-MS analysis
showed volatiles with m/z corresponding to M+ ) C10H10ClD,
C10H11Br, or C10H11I. Therefore, the product is the 1-halotet-
rahydronaphthalene (8, X ) Cl, Br, or I).
According to both 1H NMR integration and GC-MS analysis,
8 is partially deuterated at C4. Even without added D2O there
is deuterium incorporation, from DMSO-d6. Percent deuterium
enrichments in 8 (X ) Cl, Br, I) were found to be 67%, 51%,
and 42%, respectively, by 1H NMR spectroscopy and 60%, 44%,
and 40% by GC-MS analysis. The increasing extent of
deuteration, from I to Br to Cl is such that 8 (X ) Cl) has
1-Bromo-5,6,7,8-tetrahydronaphthalene: 1H NMR (DMSO-d6) δ
1.65 (2H, m, CH2CH2CH2CH2), 1.72 (2H, m, CH2CH2CH2CH2), 2.60
(2H, t, J ) 6.5, CH2CH2CH2CH2), 2.69 (2H, t, J ) 6.0, CH2CH2-
CH2CH2), 6.98 (1H, m), 7.04 (<1H, d, J ) 8 Hz), 7.32 (1H, d, J ) 8
(7) Bross, P. F.; Beitz, J.; Chen, G.; Chen, X. H.; Duffy, E.; Kieffer, L.; Roy,
S.; Sridhara, R.; Rahman, A.; Williams, G.; Pazdur, R. Clin. Cancer Res.
2001, 7, 1490.
(8) Usuki, T.; Nakanishi, K.; Ellestad, G. A. Org. Lett. 2006, 8, 5461.
(9) Nicolaou, K. C.; Zuccarello, C.; Riemer, C.; Estevez, V. A.; Dai, W.-M. J.
Am. Chem. Soc. 1992, 114, 7360.
(10) Jones, G. B.; Wright, J. M.; Plourde, G. W., II; Hynd, G.; Huber, R. S.;
Mathews, J. E. J. Am. Chem. Soc. 2000, 122, 1937.
(11) Frisch, M. J.; et al. Gaussian 98, revision A.7; Gaussian, Inc.: Pittsburgh,
PA, 1998.
9
4796 J. AM. CHEM. SOC. VOL. 129, NO. 15, 2007