Enhanced radical delivery from aldoxime esters for EPR and ring closure applications

Article abstract of DOI:10.1039/a910346p

Arylmethaniminyl and alkyl radicals were generated from di- and tri- methoxyphenyl aldoxime esters, by photolysis in the presence of 4- methoxyacetophenone, and were detected by EPR spectroscopy: good yields of cyclised products were isolated from suitably unsaturated alkyl substituents.

Full text of DOI:10.1039/a910346p

Enhanced radical delivery from aldoxime esters for EPR and ring closure
applications
Andrew J. McCarroll and John C. Walton*
University of St. Andrews, School of Chemistry, St. Andrews, Fife UK KY16 9ST. E-mail: jcw@st-and.ac.uk
Received (in Cambridge, UK) 24th December 1999, Accepted 26th January 2000
Arylmethaniminyl and alkyl radicals were generated from
di- and tri-methoxyphenyl aldoxime esters, by photolysis in
the presence of 4-methoxyacetophenone, and were detected
by EPR spectroscopy: good yields of cyclised products were
isolated from suitably unsaturated alkyl substituents.
parameters, are the signatures of rarely observed arylmethani-
minyl radicals,8 i.e. 7. Fig. 1(b) shows the well-resolved second-
order structure of the individual lines of the tert-butyl radical
and the partly resolved long range splitting from ring hydrogens
in iminyl radical 7 (Ar = Ph). The ratio of the concentrations of
the alkyl and iminyl radicals depended on temperature and on
the nature of R. For example, from 3e (R = allyl) very intense
spectra mainly of allyl were obtained, particularly for lower
power and low modulation amplitude [Fig. 2(a), whereas for 4a
(R = Bu) the spectra were dominated by iminyl signals [Fig.
2(b)] and in general primary alkyls were difficult to detect.
These differences in the concentrations of the two radicals were
probably due to minor differences in the rates of their
termination reactions. For highly reactive s-radicals, e.g.
cyclopropyl, generated from 4f and trifluoromethyl generated
from 4g, the EPR spectra showed only the 2,4,6-trimethoxy-
Molecules suitable for direct photochemical or thermal genera-
tion of free radicals, other than peroxides and azo compounds,
are quite limited, and therefore potential alternative types are of
special interest. Oxime esters 2 are easily made from aldehyde
(or ketone) oximes 1 and carboxylic acids (or acyl halides)
(Scheme 1) and are much easier to handle and characterise than
peroxides or azo compounds. These esters, which contain weak
N–O bonds, appear to be a category of precursors for carbon-
centred and iminyl radicals (7) with promise of considerable
generality, but so far exploitation has been minimal. Thus,
benzophenone oxime esters [Ph2CNNOC(O)R] were shown to
function as photolytic sources of alkyl and diphenylmethanimi-
nyl radicals and used in preparations of alkyl aromatics,1,2 alkyl
chlorides,2 and alkanes.2 Radical induced homolysis of oxime
benzoates by tributylstannane was found to be a useful method
for generating iminyl radicals,3,4 which were also produced,
without the need for toxic organotin compounds, by use of Ni
powder and acetic acid.5,6
Aldoxime esters 2 (Ar = Ph) displayed strong UV absorption
bands at 208 and 254 nm. Inclusion of methoxy substituents in
the aryl group (3, 4) induced additional strong absorption bands
at longer wavelengths (lmax = 273, 312 nm) and this implied
that photo-dissociation to radicals might be more efficient. We
tested this prospect by monitoring solution phase photolyses of
oxime esters containing a range of aromatic and alkyl
substituents by means of 9 GHz EPR spectroscopy. No
observable spectra were obtained from photolysis (500 W Hg
arc, unfiltered) of degassed solutions of oxime esters 2 to 6 in
tert-butylbenzene solution in the temperature range 200 to 320
K, except for a very weak spectrum of the allyl radical from 2e.
However, on addition of 4-methoxyacetophenone to the
solutions as a photosensitiser (!0.1 equiv.), good quality EPR
spectra were obtained, especially from methoxy-substituted
oxime esters of types 3 and 4, but not from types 5 and 6
containing electron-withdrawing substituents. In the sensitised
photolyses of 2d, 3d and 4d, for example, the EPR spectra
consisted of a superposition of two radicals [see Fig. 1(a), (b)].
The most prominent signals were due to the tert-butyl radical,
with EPR parameters identical to those reported in the
literature.7 Two widely spaced groups of nitrogen triplets could
also be discerned in the wings of the spectra [from 7 (Ar = Ph):
a(H) = 80.5, a(N) = 10.0 G, g = 2.0034 at 250 K]. The
exceptionally large a(H) values, as well as the other EPR
Fig. 1 (a) 9.4 GHz EPR spectrum obtained from photolysis of a solution of
2d and MAP in tert-butylbenzene at 225 K showing the tert-butyl and
iminyl radicals. (b) Scale expansion of a section of the above spectrum at
250 K showing the second order structure on the tert-butyl radical and partly
resolved long range splittings on the iminyl radical 7 (Ar = Ph).
Fig. 2 (a) EPR spectrum of the allyl radical obtained on photolysis of a
solution of 3e and MAP in tert-butylbenzene at 200 K. Note the absence of
iminyl 7 at low microwave power (0.2 mW) and low modulation amplitude
(0.3 Gpp). (b) EPR spectrum of the iminyl radical 7 (Ar = TMP) and the
butyl radical, from photolysis of a solution of oxime ester 4a and MAP in
tert-butylbenzene at 235 K.
Scheme 1
DOI: 10.1039/a910346p
Chem. Commun., 2000, 351–352
This journal is © The Royal Society of Chemistry 2000
351

sion of photosensitiser. Best yields of 9 were, however, obtained
from the dimethoxy oxime ester 8b when photosensitiser was
included. Variation of the photosensitiser concentration showed
that as little 0.1 equiv. could be used before serious loss of
efficiency was experienced. The bicyclic ether 14 was obtained
in good yield from photolysis of the trimethoxy oxime ester 13,
but the yield of 1-isopropyl-3-methylcyclopentane (12) from 11
was moderate, possibly because of poor hydrogen donation to
the tertiary cyclised radical by the solvent.10 No uncyclised
products from direct reduction of the initial unsaturated alkyl
radicals were detected. This is an advantage in that by-products
are minimised, but indicates that, as expected, hydrogen
donation is slow. An advantage of the method is that halogen-
donor solvents can also be used for the preparation of
functionalised rings. For example, photolysis of 8c in CCl4
afforded a good yield of chloromethylenecyclopentane (10).
The main products derived from the iminyl radicals were the
corresponding aldehydes, probably formed from intermediate
imines, ArCHNNH, which are known to be highly susceptible to
hydrolysis.
These results demonstrate that di- and tri-methoxy substitu-
tion of aryl oxime esters enhances their performance as
photolytic sources of carbon-centred and iminyl radicals and
that further improvement can be achieved by inclusion of a
photosensitiser. These compounds are promising radical pre-
cursors of use in spectroscopic studies and as ‘cleaner’ tin-free
reagents for preparative decarboxylative cyclisations of un-
saturated carboxylic acids.
Scheme 2
Table 1 Results of irradiation of oxime esters
Oxime ester Ara
Solvent
MAP/equiv. Product (%)b
We thank the EPSRC (grant GR/L49185) for financial
support.
8a
8a
Ph
Ph
PhCH3
PhCH3
PhCH3
PhCH3
CCl4
—
1
9 (6)
9 (28)
9 (39)
9 (77)
10 (60)
12 (34)
14 (72)
8b
8b
8c
11c
13c
DMP
DMP
TMP
TMP
TMP
—
1
1
Notes and references
PhCH3
CH2Cl2
1
1
1 M. Hasebe, K. Kogawa and T. Tsuchiya, Tetrahedron Lett., 1984, 3887;
M. Hasebe and T. Tsuchiya, Tetrahedron Lett., 1986, 3239.
2 M. Hasebe and T. Tsuchiya, Tetrahedron Lett., 1987, 6207; M. Hasebe
and T. Tsuchiya, Tetrahedron Lett., 1988, 6287.
3 J. Boivin, A.-M. Schiano and S. Z. Zard, Tetrahedron Lett., 1994, 35,
249.
a DMP = 2,4-dimethoxyphenyl, TMP = 2,4,6-trimethoxyphenyl. b Yields
determined by NMR.
4 J. Boivin, A.-C. Callier-Dublanchet, B. Quiclet-Sire, A.-M. Schiano and
S. Z. Zard, Tetrahedron, 1995, 51, 6517.
5 J. Boivin, A.-M. Schiano and S. Z. Zard, Tetrahedron Lett., 1992, 33,
7849.
phenylmethaniminyl radical [7, Ar = 2,4,5-(MeO)3C6H2]
except at low temperatures [ < ca. 220 K], where cyclohexa-
dienyl radicals from addition of cyclopropyl (or CF3) to the tert-
butylbenzene solvent were also visible.9
6 S. Z. Zard, Synlett, 1996, 1148.
The large doublet H-hfs and comparatively few lines of the
methaniminyl radicals provide a ‘window’ of ca. 60 G in the
centre of the EPR spectrum which minimises overlap with the
co-radical. Furthermore, the iminyl spectra act as useful
standards with known g-factors so that spectral analysis of the
co-radical is facilitated. Thus, these oxime esters constitute a
convenient new class of radical precursors suitable for spectro-
scopic studies. Unlike the alternative diacyl peroxides, reagents
2 are innocuous and convenient to handle.
7 R. W. Fessenden and R. H. Schuler, J. Chem. Phys., 1963, 39, 2147.
8 A. R. Forrester, M. Gill, C. J. Meyer, J. S. Sadd and R. H. Thomson,
J. Chem. Soc., Perkin Trans. 2, 1979, 606.
9 EPR experiments with 2g were also carried out in liquid cyclopropane
which enables lower temperatures to be accessed. However, due partly
to poor solubility at low temperatures, only weak spectra of 7 (Ar =
2,4,6-(MeO)3C6H2 were observed.
10 For example: oxime ester 13c (0.25 g, 0.69 mmol) and MAP (0.12 g, 0.8
mmol) in CH2Cl2 (5 cm3) were photolysed in a quartz tube for 3 h at
50 °C with light from a 400 W medium pressure Hg lamp. The solvent
was evaporated and the solid remaining was extracted several times with
Et2O. Concentration and microdistillation at ambient temperature/0.01
Torr gave bicyclic ether 14 (ref. 11) (0.01 g, 12%) (ref. 12), dH(300
MHz, CDCl3) 1.15–1.30 (2H, m), 1.39–1.68 (6H, m), 1.83–2.04 (3H,
m), 3.77–3.87 (2H, m), 3.96 (1H, q, J 8).
The EPR spectral results indicated that use of di- and tri-
methoxy oxime esters, in conjunction with a photosensitiser,
could enhance their efficiency in preparative reactions. Accord-
ingly, several ring closure reactions of oxime esters 8, 11 and 13
were investigated. (Scheme 2).
11 P. A. Baguley and J. C. Walton, J. Chem. Soc., Perkin Trans. 1, 1998,
2073.
12 Low yield due to small scale reaction and volatile product: see final row
Table 1 for a better measure of the true yield.
Solutions of individual oxime esters (ca. 0.13 mol dm23) in
a hydrogen donor solvent (PhCH3, CH2Cl2) were photolysed
with light from a 400 W medium pressure Hg lamp for ca. 3 h.
Table 1 shows that for 8a yields of the cyclised product,
methylenecyclopentane (9), were low, but improved on inclu-
Communication a910346p
352
Chem. Commun., 2000, 351–352