Carbon-carbon bond formation by radical addition-fragmentation reactions of O-tert-alkyl enols and O-cyclopropylcarbinyl enols

Article abstract of DOI:10.1016/S0040-4039(03)01092-X

Terminal alkenes of the type H₂C=C(OR¹)X, in which R¹ is a tertiary alkyl or a 1-cyclopropylethyl group and X=Ph, OSiMe₂Bu^t, OEt or H, undergo radical-chain reactions with organic halides R²Hal to give carbonyl compounds R²CH₂C(=O)X.

Full text of DOI:10.1016/S0040-4039(03)01092-X

TETRAHEDRON
LETTERS
Pergamon
Tetrahedron Letters 44 (2003) 4645–4648
Carbonꢀcarbon bond formation by radical
addition–fragmentation reactions of O-tert-alkyl enols and
O-cyclopropylcarbinyl enols
Yudong Cai and Brian P. Roberts*
Department of Chemistry, University College London, 20 Gordon Street, London WC1H 0AJ, UK
Received 28 March 2003; accepted 25 April 2003
Abstract—Terminal alkenes of the type H2CꢁC(OR1)X, in which R1 is a tertiary alkyl or a 1-cyclopropylethyl group and X=Ph,
OSiMe2But, OEt or H, undergo radical-chain reactions with organic halides R2Hal to give carbonyl compounds R2CH2C(ꢁO)X.
© 2003 Elsevier Science Ltd. All rights reserved.
The addition–fragmentation sequence generalised in
Scheme 1 is of central importance in several radical-
based methods for the construction of carbonꢀcarbon
bonds.1,2 For example, allylation can be achieved using
allylstannanes or allylic sulfides as the radical acceptor
1. Compounds of the type 1 have also found extensive
use as addition–fragmentation chain-transfer agents
(AFCTAs) in the radical polymerisation of vinylic
monomers, where they provide an effective tool for
controlling the molar mass of the polymer produced, as
well as for introducing functional end groups.3 Meijs,
Rizzardo and co-workers4 have described the use of
O-benzyl enols, including a-benzyloxystyrene 2, as
AFCTAs for the preparation of low molecular weight
polystyrene and poly(methyl methacrylate), when b-
scission of an intermediate benzyloxyalkyl radical 3
terminates one polymer chain and provides a benzyl
radical to seed the growth of another (Scheme 2).
The thermal rearrangement of an a-alkoxystyrene 4 to
give an a-alkylated acetophenone 5, which involves a
formal 1,3-shift of the O-alkyl group, was first
described by Claisen in 1896.5 The mechanism was
investigated in the 1950s by Wiberg and co-workers6
who concluded that a radical-chain pathway is fol-
lowed, by way of the addition–fragmentation cycle
shown in Scheme 3. Although the potential of this type
of reaction as a synthetic tool for the construction of
carbonꢀcarbon bonds was explicitly recognised some
time ago,7 it has received little attention outside the
area of polymer chemistry. Our own interest in the
b-scission of a-alkoxyalkyl radicals, both with regard to
mechanistic studies8 and applications in synthesis,9
encouraged us to explore the use of O-alkylated enols
of the type 6 for carbonꢀcarbon bond formation by the
radical addition–fragmentation mechanism set out in
Scheme 4. However, while this work was in progress, a
Scheme 1.
Scheme 2.
Keywords: radicals and radical reactions; alkylation; addition reactions; fragmentation reactions.
* Corresponding author.
0040-4039/03/$ - see front matter © 2003 Elsevier Science Ltd. All rights reserved.
doi:10.1016/S0040-4039(03)01092-X

4646
Y. Cai, B. P. Roberts / Tetrahedron Letters 44 (2003) 4645–4648
Scheme 3.
report by Roepel10 appeared describing the radical reac-
tions of a-phenylselenyl-malononitriles or -malonic
esters 9 with the O-benzyl enols 10 according to
Scheme 5. Prompted by this disclosure, we now present
a preliminary account of our studies of the reactions of
halogen-atom donors with the O-tert-alkyl enols 6a–d
to give carbonyl compounds of the type 8 (Scheme 4,
Y=Br or I).
group, as generalised by the equilibrium process shown
in Eq. (1) of Scheme 4.
We directed our attention first to radical-chain reac-
tions of a-tert-butoxystyrene13 6a, which were expected
to yield phenyl ketones of the type 11. When a-tert-
butoxystyrene was heated alone in refluxing benzene
under argon, no change was detected by NMR spec-
troscopy after 3 h. However, in the presence of 5 mol%
azobis(isobutyronitrile) (AIBN) or dilauroyl peroxide
(DLP), similar treatment resulted in the complete con-
version of 6a to neopentyl phenyl ketone 11a, which
was isolated in ca. 95% yield.14 When ethyl bromo-
acetate (3 equiv.) was present along with AIBN, NMR
analysis after removal of the solvent showed complete
conversion of 6a to a mixture of ethyl 4-oxo-4-
phenylbutanoate 11b and 11a in the ratio 70:30.15 The
tert-butyl radical resulting from the b-scission of 7 now
abstracts bromine from BrCH2CO2Et in competition
with its addition to the styrene 6a. When the latter was
added slowly by syringe pump to the bromoester and
AIBN (10 mol%) in refluxing benzene, the product
ratio 11b:11a increased to 95:5. For all reactions in the
presence of halogen-atom donors, a small amount of
2,4,6-collidine or 2,2,6,6-tetramethylpiperidine was also
added as a sterically hindered base to suppress acid-
catalysed elimination of alkene from (or adventitious
hydrolysis of) the O-alkylated enols. Slow addition of
6a was not necessary when the bromoacetate was
replaced as halogen-atom donor by the much more
reactive ethyl iodoacetate (2 equiv.) and 11b was then
produced in high yield without competitive formation
of 11a. Several similar reactions of a-tert-butoxystyrene
6a and of the O-tert-alkyl enols 6b–d with sources of
relatively electrophilic carbon-radical addenda were
carried out and the results are summarised in Table 1.16
Although the silyl esters 12 could be isolated by careful
chromatography on Florisil®, because of their sensitiv-
ity to hydrolysis it was usually more convenient to
convert them first to alkyl esters.18 For example, the
mixed succinate ester 12a was readily desilylated by
DMF-catalysed reaction with oxalyl chloride and, with-
out isolation, the resulting acid chloride was converted
to diethyl succinate 13a (83% from 6b) by treatment
with ethanol and pyridine.18a A similar radical-chain
Our initial choice of an O-tert-alkyl substituent for 6
was guided by two considerations. First, previous
kinetic studies of the b-scission of cyclic and acyclic
tert-alkoxyalkyl radicals have shown that such species
undergo relatively rapid C–O cleavage to give tert-alkyl
radicals,8 in accord with the ease with which tertiary
alcohols can be deoxygenated by making use of this
process.9 Second, although the C–H bond strength
decreases appreciably along the series Me–H<1°-R–H<
2°-R–H<3°-R–H, the corresponding bonds to more
electronegative elements, including bromine and iodine,
vary much less in strength.8c,11 For example, the CꢀI
bond dissociation enthalpies for MeꢀI, EtꢀI, PriꢀI and
ButꢀI are reported to be 239, 236, 234 and 231 kJ
mol−1, respectively; for comparison DH(PhCH2ꢀI) is
significantly smaller at 215 kJ mol−1.12 Thus, not only
should an adduct radical of the type 7 undergo rela-
tively rapid b-scission to give the tertiary alkyl radical
t
’
R , but the latter should (in contrast to a benzyl
radical10) be fairly easily transformed into a new adden-
dum radical, by transfer of an electronegative atom or
Scheme 4.
Scheme 5.

Y. Cai, B. P. Roberts / Tetrahedron Letters 44 (2003) 4645–4648
4647
Table 1. Addition–fragmentation reactions of O-tert-alkyl enols in refluxing benzene or cyclohexanea
Entry
Radical acceptor
Halogen-atom donor
R–Y (equiv.)
Conditionsb,c
Product(s)d
Isolated yield (%)
H2CꢁC(ORt)X
1
2
3
4
5
6
7
8
6a
6a
6a
6a
6a
6b
6b
6c
6c
6d
6d
BrCH2CO2Et (3)
ICH2CO2Et (2)
AIBN,e Ben, SP
AIBN, Ben
AIBN, Ben
AIBN, Ben
AIBN, Ben
DLP, Cyc
DLP, Cyc
DLP, Cyc
DLP, Cyc
DLP, Cyc
11b+11a (95:5)
85 (11b)
82
85
11b
11c
11d
11e
12a
12b
13a
13b
14a
14b
BrCMe2CO2Et (3)
BrCH(CO2Et)2 (2)
BrCMe(CO2Et)2 (2)
BrCH2CO2Et (2)
BrCMe(CO2Et)2 (1.2)
BrCH2CO2Et (1.2)
BrCMe(CO2Et)2 (1.2)
BrCH2CO2Et (0.5)
BrCMe(CO2Et)2 (0.5)
84
91
83f
73f
64
9
10
11
52
30g
50g
DLP, Cyc
a Internal temperature ca. 85°C; total reaction time generally 3 h.
b Unless stated otherwise, 5 mol% of initiator (based on acceptor) was used; 2,4,6-collidine (10 mol%) was present in each case.
c Ben=benzene solvent, Cyc=cyclohexane solvent; SP=slow addition of the acceptor during the first 1.5 h using a syringe pump.
d Product ratio estimated by 1H NMR spectroscopy.
e 10 Mol% initiator present.
f Isolated as the all-ethyl esters (see text).
g Conversion of bromide estimated by 1H NMR spectroscopy.
the primary n-butyl radical. Slow addition of 6a using a
syringe pump raised the proportion of 15 to 34%, but it
appears that the equilibrium (1) in Scheme 4 lies
impractically far to the left in this case. Replacing the
n-butyl iodide with s-butyl iodide (3 equiv.), with all
the reagents present initially, improved the situation
such that the product ketone ratio 16:11a was 35:65. In
refluxing octane solvent, in the presence of 2,2,6,6-tetra-
methylpiperidine (TMP; 20 mol%) and using di-tert-
pentyl peroxide (DTPP; 30 mol%) as initiator, the ratio
16:11a could be raised to 52:48.
reaction of O-tert-butyl O-ethyl ketene acetal 6c with
ethyl bromoacetate yielded diethyl succinate directly
(64%), but the greater acid sensitivity of 6c compared
with 6b often proved to be a disadvantage.
According to Roepel,10 the chemistry shown in Scheme
5 was not successful for the production of aldehydes
from benzyl vinyl ether (X=H). Therefore, it is note-
worthy that the aldehydes 14a and b could be obtained
from reactions of O-3-methyl-3-pentyl vinyl ether 6d
(entries 10 and 11),19 albeit so far in relatively low yield.
Phenyl ketones derived from reactions of 6a with
sources of nucleophilic radical addenda proved more
difficult to obtain in preference to the styrene-rear-
rangement product 11a. Thus, treatment of 6a with
n-butyl iodide (5 equiv.) in refluxing benzene, contain-
ing AIBN (5 mol%) and collidine (10 mol%), resulted in
its complete conversion to a 19:81 mixture of ketones
15 and 11a after 3 h (Scheme 6). Evidently here the
tert-butyl radical is being trapped by 6a in preference to
Scheme 6.

4648
Y. Cai, B. P. Roberts / Tetrahedron Letters 44 (2003) 4645–4648
Attachment of a cyclopropyl group at an alkyl-radical
centre leads to some stabilising delocalisation of the
unpaired electron12,20 and would also be expected to
increase the nucleophilicity (reduce the ionisation
energy) of the radical, compared with an acyclic alkyl
substituent, suggesting that a secondary cyclopropyl-
carbinyl radical will behave in many respects like a
simple tertiary alkyl radical. Therefore, we reasoned
that radical adducts of the O-cyclopropylcarbinyl enol
17 should undergo b-scission with similar facility to the
acyclic tert-alkoxyalkyl radical analogues 7, to give the
1-cyclopropylethyl radical 18 that will undergo rapid
ring opening to give the primary homoallylic radical
19.21 The latter should then abstract iodine more
efficiently than the tert-butyl radical in an exchange
process of the type shown Eq. (1) of Scheme 4.
Chem. Soc., Perkin Trans. 2 2002, 155–163.
9. (a) Dang, H.-S.; Roberts, B. P. J. Chem. Soc., Perkin
Trans. 1 2002, 1161–1170; (b) Cai, Y.; Dang, H.-S.;
Roberts, B. P. J. Chem. Soc., Perkin Trans. 1 2002,
2449–2458.
10. Roepel, M. G. Tetrahedron Lett. 2002, 43, 1973–1976.
11. (a) Laurencelle, N.; Pacey, P. D. J. Am. Chem. Soc. 1993,
115, 625–631; (b) Harvey, J. N.; Viehe, H. G. J. Prakt.
Chem. 1995, 337, 253–265; (c) Luo, Y.-R.; Benson, S. W.
J. Phys. Chem. A 1997, 101, 3042–3044.
12. CRC Handbook of Chemistry and Physics, 82nd ed.; CRC
Press: Boca Raton, 2001.
13. Kostikov, R. R.; Drygailova, E. A.; Golovkina, E. A.;
Komendantov, A. M.; Molchanov, A. P. J. Org. Chem.
USSR (Engl. Trans.) 1987, 23, 1917–1920. 2-tert-Butoxy-
2-phenylethyl iodide, prepared from styrene, tert-butyl
alcohol, iodine and yellow mercuric oxide according to
the general method of Wiberg,6 was dehydroiodinated by
treatment with KOBut in THF (see Middleton, D. S.;
Simpkins, N. S. Synth. Commun. 1989, 19, 21–29).
14. Under the same conditions, the O-benzyl analogue, a-
benzyloxystyrene 2, rearranged only partially (40%) to
PhCH2CH2C(O)Ph.
We were pleased to find that when 17 was treated with
n-butyl iodide (5 equiv.), DTPP (30 mol%) and TMP
(20 mol%) in refluxing octane, with all reagents present
initially, the enol ether was completely converted to
mainly the n-butyl adduct 15 (85%) along with 15% of
the ketone 20. A similar reaction using s-butyl iodide (3
equiv.) afforded the ketone 16 containing only 4% of
20.
15. Under the same conditions, a-benzyloxystyrene afforded
only PhCH2CH2C(O)Ph (38%) and no 11b.
16. Representative procedure: Diethyl bromomalonate (0.956
g, 4.0 mmol), a-tert-butoxystyrene 6a (0.352 g, 2.0
mmol), AIBN (16.4 mg, 0.10 mmol), 2,4,6-collidine (26.4
ml, 0.20 mmol) and dry benzene (4 mL) were added to a
dry, argon-filled flask, containing a magnetic stirrer bar
and equipped with a condenser. The flask was then
immersed in an oil bath, pre-heated to 90°C, and the
reaction mixture was stirred at reflux under argon for 3 h.
The solvent was removed by rotary evaporation and the
residue was purified by flash chromatography on silica
gel, using petroleum (bp 40–60°C)-diethyl ether (4:1 v/v)
as eluent, to give diethyl phenacylmalonate17 11d (0.466
g, 84%) as a colourless oil. NMR (500 MHz for 1H, 125.7
MHz for 13C; CDCl3 solvent, J in Hz): lH 1.29 (6H, t, J
7.1, Me), 3.62 (2H, d, J 7.1, CH2COPh), 4.06 (1H, t, J
7.1, CH), 4.23 (4H, m, CH2Me), 7.46–8.00 (5H, m, Ph);
lC 14.0, 37.8, 47.2, 61.7, 128.1, 128.6, 133.5, 136.0, 169.0,
196.5. Found: C, 64.9; H, 6.4. C15H18O5 requires C, 64.7;
H, 6.5%. Satisfactory spectroscopic and analytical data
were obtained for all new compounds described herein.
17. Fuentes, J.; Molina, J. L.; Pradera, M. A. Tetrahedron:
Asymmetry 1998, 9, 2517–2532.
We conclude that O-tert-alkyl and O-cyclopropyl-
carbinyl enols offer considerable promise for the forma-
tion of carbonꢀcarbon bonds under tin-free
conditions.22 Synthetic and kinetic (radical-clock) stud-
ies using these compounds are on-going and will be
reported in a full paper.
References
1. Rosenstein, I. J. In Radicals in Organic Synthesis;
Renaud, P.; Sibi, M. P. Eds.; Wiley-VCH: Weinheim,
2001; Vol. 1, Chapter 1.4, pp. 50–71.
2. Motherwell, W. B.; Crich, D. Free Radical Chain Reac-
tions in Organic Synthesis; Academic Press: London,
1992.
3. (a) Colombani, D.; Chaumont, P. Prog. Polym. Sci. 1996,
21, 439–503; (b) Colombani, D.; Chaumont, P. Acta
Polym. 1998, 49, 225–231; (c) Colombani, D. Prog.
Polym. Sci. 1999, 24, 425–480.
4. (a) Meijs, G. F.; Rizzardo, E. Makromol. Chem. Rapid
Commun. 1988, 9, 547–551; (b) Rizzardo, E.; Meijs, G.
F.; Thang, S. H. US Patent 6197905 (2001).
5. Claisen, L. Chem. Ber. 1896, 29, 2931–2932.
6. (a) Wiberg, K. B.; Rowland, B. I. J. Am. Chem. Soc.
1955, 77, 1159–1163; (b) Wiberg, K. B.; Shryne, T. M.;
Kinter, R. R. J. Am. Chem. Soc. 1957, 79, 3160–3164; (c)
Wiberg, K. B.; Kinter, R. R.; Motell, E. L. J. Am. Chem.
Soc. 1963, 85, 450–454.
18. (a) Wissner, A.; Grudzinskas, C. V. J. Org. Chem. 1978,
43, 3972–3974; (b) Sakaitani, M.; Ohfune, Y. J. Org.
Chem. 1990, 55, 870–876.
19. Because of its higher boiling point, the vinyl ether 6d was
used in preference to commercially available tert-butyl
vinyl ether.
20. Roberts, C.; Walton, J. C. Chem. Commun. 1984, 1109–
1111.
21. (a) Bowry, V. W.; Lusztyk, J.; Ingold, K. U. J. Am.
Chem. Soc. 1991, 113, 5687–5698; (b) Beckwith, A. L. J.;
Bowry, V. W. J. Am. Chem. Soc. 1994, 116, 2710–2716.
22. For a related type of reaction that utilises O-stannyl
enolates, see: Miura, K.; Fujisawa, N.; Saito, H.; Wang,
D.; Hosomi, A. Org. Lett. 2001, 3, 2591–2594.
7. Wang, S.; Morrow, G. W.; Swenton, J. S. J. Org. Chem.
1989, 54, 5364–5371.
8. (a) Perkins, M. J.; Roberts, B. P. J. Chem. Soc., Perkin
Trans. 2 1975, 77–84; (b) Dang, H.-S.; Franchi, P.;
Roberts, B. P. Chem. Commun. 2000, 499–500; (c) Field-
ing, A. J.; Franchi, P.; Roberts, B. P.; Smits, T. M. J.