Scheme 1
Figure 2. EPR signal of the spin-adduct 5a and (t-Bu)2N(Oi).
The six other smaller EPR peaks (cf. Figure 2) possess the
typical nitroxides g values (2.0063-2.0065),5a,b which are
consistent with spin adducts 5a-e, generated from the spin-
trapping of 4a-e with t-BuNO (Scheme 1). The six-line
pattern is caused by the hyperfine splitting of the EPR signal
with one nitrogen (triplet) and one hydrogen at the â-position
(doublet).6 The aN (triplet) and aHâ (doublet) values vary little
among spin adducts 5a-e; thus, the aN value ranges from
15.4 G (for 5c) to 15.5 G (for 5b) and the aHâ value ranges
from 4.4 G (for 5a) to 4.5 G (for 5e).
Different molar ratios of 1a-e to SmI2 (e.g., 1.5:1, 1:1,
1:1.3) were used but were found to have little influence on
the g values or the aN and aHâ values. However, adding more
SmI2/THF solution did adversely affect the EPR signals,
because the solvent THF strongly absorbs microwaves due
to its high polarity. In addition, different temperatures (rt or
-70 °C) in the EPR cavity were also utilized. The final EPR
signals were unaffected, although the formation of the spin
adducts 5a-e was slower at -70 °C.
involve radicals. We believe that the radicals detected are
4a-d, but the signal-noise ratios are remarkably low and
the signals disappear within 8-15 min. This is probably due
to the instability of carbon-centered radicals bearing R-hy-
drogens.
No EPR signal was observed for the SmI2-promoted
coupling of 1e. Since the radical intermediate 4e has no
phenyl group attached to the carbon-centered radical, no EPR
signal was to be expected.
As the instability of the radicals 4a-e precludes obtaining
well-resolved spectra, the spin-trapping technique was ap-
plied. In spin-trapping, unstable, short-lived carbon-centered
radicals react with spin-trapping agents, such as nitroso
compounds, to form more stable, longer-lived nitroxides
which are easier to detect by EPR. N-tert-Butyl-R-phenylni-
trone (PBN) when used as a spin-trapping agent gave no
EPR signal, presumably because sterically hindered PBN
does not efficiently trap the radicals 4a-e. However, nitroso-
tert-butane (t-BuNO) as a spin-trapping agent gave similar
EPR signals for each substrate (cf. Figure 2). The strong
triplet peaks are attributed to di-tert-butyl nitroxide [t-Bu-
N(O‚)-Bu-t], which is always present in the t-BuNO solution.
(5) (a) He, R. H. Y.; Zhao, C. X.; Zhou, C. M.; Jiang, X. K. Tetrahedron
1999, 55, 2263. (b) Kojima, T.; Tsuchiya, J.; Nakashima, S.; Ohya-
Nishiguchi, H.; Yano, S.; Hidai, M. Inorg. Chem. 1992, 31, 2333.
(6) Janzen, E. G.; Davis, E. R.; Dubose, C. M. Magn. Reson. Chem.
1995, 33, S166.
(7) Katritzky, A. R.; Lan, X.; Yang, J. Z.; Denisko, O. V. Chem. ReV.
1998, 98, 409.
(8) General Experimental Details. SmI2 (0.1 M in THF) was purchased
from Aldrich and used directly without further treatment. Nitroso-tert-butane
(t-BuNO) and N-tert-butyl-R-phenylnitrone (PBN) were also purchased from
Aldrich. CH2Cl2 was distilled from sodium-benzophenone prior to use. In
a typical experiment, the substrate 1 (0.1 M in CH2Cl2, 0.2 mL) was added
into a deoxygenated EPR tube (2 mm in diameter) and cooled in a dry
ice-acetone bath (-70 °C). Then previously cooled SmI2 solution (0.2 mL)
was injected into the tube using a syringe. After being shaken rigorously
several times, the EPR tube was inserted into the EPR cavity (-70 °C) and
the EPR spectra for radical intermediates were recorded immediately. An
Oxford Instruments (CF900) helium flow cryostat was used to keep the
sample below -70 °C. For the spin-trapping experiment, the substrate 1
(0.1 M in CH2Cl2, 0.15 mL) and the spin trapper (t-BuNO, 0.1 M in CH2-
Cl2, 0.15 mL; or PBN, 0.1 M in CH2Cl2, 0.15 mL) were added to a
deoxygenated EPR tube and cooled in a dry ice-acetone bath (-70 °C).
An SmI2 solution (0.10, 0.15, or 0.20 mL) was then injected into the tube
using a syringe. After being shaken rigorously several times, the EPR tube
was inserted into the EPR cavity and the EPR spectra for radical
intermediates were recorded immediately. The spectra were recorded at -70
°C as well as at room temperature. EPR spectra were recorded by a Bruker
EPR Elexsys 580 spectrometer in CW mode using a rectangular cavity
(TE102). The conditions employed were as follows: modulation, 100 kHz;
frequency, 9.764 GHz; microwave power, 2 mW; modulation amplitude,
1- 5 G; time constant, 0.04 s; sweep width, 100 G. The magnetic field
was determined by a Hall probe and the microwave frequency by the built-
in frequency counter of the Bruker Bridge model E580-1010.
(4) (a) Katritzky, A. R.; Yannakopoulou, K.; Lue, P.; Rasala D.; Urogdi,
L. J. Chem. Soc., Perkin Trans. 1 1989, 225. (b) Katritzky A. R.; Fan, W.
Q. J. Fluorine Chem. 1991, 51, 33. (c) Katritzky, A. R.; Latif, M.; Urogdi,
L. J. Chem. Soc., Perkin Trans. 1 1990, 667. (d) Katritzky, A. R.; Chang,
H. X.; Wu, J. Synthesis 1994, 907.
1756
Org. Lett., Vol. 1, No. 11, 1999
Katritzky, Alan R.
