Photolysis of 1-alkylcycloalkanols in the presence of (diacetoxyiodo) benzene and I2. Intramolecular selectivity in the β-scission reactions of the intermediate 1-alkylcycloalkoxyl radicals

Article abstract of DOI:10.1021/jo049524y

The C-C β-scission reactions of 1-alkylcycloalkoxyl radicals, generated photochemically by visible light irradiation of CH₂Cl ₂ solutions containing the parent 1-alkylcycloalkanols, (diacetoxy)-iodobenzene (DIB), and I₂, have been investigated through the analysis of the reaction products. The 1-alkylcycloalkoxyl radicals undergo competition between ring opening and C-alkyl bond cleavage as a function of ring size and of the nature of the alkyl substituent. With the 1-propylcycloheptoxyl, 1-propylcyclooctoxyl, and 1-phenylcyclooctoxyl radicals, formation of products deriving from an intramolecular 1,5-hydrogen atom abstraction reaction from the cycloalkane ring has also been observed. The results are discussed in terms of release of ring strain associated to ring opening, stability of the alkyl radical formed by C-alkyl cleavage, and with cycloheptoxyl and cyclooctoxyl radicals, also in terms of the possibility of achieving a favorable geometry for intramolecular hydrogen atom abstraction.

Full text of DOI:10.1021/jo049524y

P h otolysis of 1-Alk ylcycloa lk a n ols in th e P r esen ce of
(Dia cetoxyiod o)ben zen e a n d I2. In tr a m olecu la r Selectivity in th e
â-Scission Rea ction s of th e In ter m ed ia te 1-Alk ylcycloa lk oxyl
Ra d ica ls
Carla S. Aureliano Antunes,1a Massimo Bietti,*,1a Osvaldo Lanzalunga,1b and
Michela Salamone1a
Dipartimento di Scienze e Tecnologie Chimiche, Universita` “Tor Vergata”, Via della Ricerca Scientifica,
1 I-00133 Rome, Italy, and Dipartimento di Chimica, Universita` “La Sapienza”, P.le A. Moro,
5 I-00185 Rome, Italy
bietti@uniroma2.it
Received March 23, 2004
The C-C â-scission reactions of 1-alkylcycloalkoxyl radicals, generated photochemically by visible
light irradiation of CH2Cl2 solutions containing the parent 1-alkylcycloalkanols, (diacetoxy)-
iodobenzene (DIB), and I2, have been investigated through the analysis of the reaction products.
The 1-alkylcycloalkoxyl radicals undergo competition between ring opening and C-alkyl bond
cleavage as a function of ring size and of the nature of the alkyl substituent. With the
1-propylcycloheptoxyl, 1-propylcyclooctoxyl ,and 1-phenylcyclooctoxyl radicals, formation of products
deriving from an intramolecular 1,5-hydrogen atom abstraction reaction from the cycloalkane ring
has also been observed. The results are discussed in terms of release of ring strain associated to
ring opening, stability of the alkyl radical formed by C-alkyl cleavage, and with cycloheptoxyl and
cyclooctoxyl radicals, also in terms of the possibility of achieving a favorable geometry for
intramolecular hydrogen atom abstraction.
Alkoxyl radicals are important intermediates which
play a major role in the photooxidation of hydrocarbons
in the atmosphere2 and in several processes occurring in
biological systems.3 Moreover, following the increasing
importance experienced in recent years by radicals in
organic synthesis, alkoxyl radical chemistry has been
successfully exploited in synthetically useful processes,
mainly through cyclization,4 â-fragmentation,5 and 1,5
hydrogen atom transfer reactions.6 Among these pro-
cesses, C-C â-scission leading to a carbonyl compound
and an alkyl radical is one of the most important rea-
ctions of alkoxyl radicals (Scheme 1).
With tertiary alkoxyl radicals bearing different alkyl
groups it is possible, at least in principle, to obtain three
different C-C bond fragmentations and accordingly the
intramolecular selectivity in the â-scission reactions of
tertiary alkoxyl radicals has been the subject of detailed
investigation.7-13 The main conclusion of these studies
was that cleavage generally leads to the most stable
possible alkyl radical. However, when considering the
ring-opening reactions of cycloalkoxyl radicals it was
suggested that other factors may play a role. For ex-
ample, it was found that ring opening of the cyclopentoxyl
radical is more than 100 times faster than ethyl radical
ejection from the 2-butoxyl radical.14 A similar behavior
was observed when the ring opening of the 1-methylcy-
clopentoxyl radical was compared with ethyl radical
(1) (a) Universita` “Tor Vergata”. (b) Universita` “La Sapienza”.
(2) See, for example: Orlando, J . J .; Tyndall, G. S.; Wallington, T.
J . Chem. Rev. 2003, 103, 4657-4689. Atkinson, R. Int. J . Chem. Kinet.
1997, 29, 99-111.
(3) See, for example: Davies, M. J .; Headlam, H. A. Free Rad. Biol.
Med. 2002, 32, 1171-1184. Halliwell, B.; Gutteridge, J . M. C. Free
Radicals in Biology and Medicine, 3rd ed.; Oxford University Press:
New York, 1999. Cumming, J . N.; Ploypradith, P.; Posner, G. H. Adv.
Pharmacol. 1997, 37, 2253-2297. Ullrich, V.; Brugger, R. Angew.
Chem., Int. Ed. Engl. 1994, 33, 1911-1919.
(4) Hartung, J .; Kneuer, R.; Laug, S.; Schmidt, P.; Sˇpehar, K.;
Svoboda, I.; Fuess, H. Eur. J . Org. Chem. 2003, 4033-4052. Hartung,
J . Eur. J . Org. Chem. 2001, 619-632. Hartung, J . In Radicals in
Organic Synthesis; Renaud, P., Sibi, M. P., Eds.; Wiley-VCH: Wein-
heim, 2001; Vol. 2, pp 427-439.
(5) (a) Sua´rez, E.; Rodriguez, M. S. In Radicals in Organic Synthesis;
Renaud, P., Sibi, M. P., Eds.; Wiley-VCH: Weinheim, 2001; Vol. 2, pp
440-454. (b) Zhang, W. In Radicals in Organic Synthesis; Renaud,
P., Sibi, M. P., Eds.; Wiley-VCH: Weinheim, 2001; Vol. 2, pp 234-
245. (c) Yet, L. Tetrahedron 1999, 55, 9349-9403. (d) Dowd, P., Zhang,
W. Chem. Rev. 1993, 93, 2091-2115.
(6) (a) Cˇ ekovic´, Z. Tetrahedron 2003, 59, 8073-8090. (b) Feray,
L.; Kuznetsov, N.; Renaud, P. In Radicals in Organic Synthesis;
Renaud, P., Sibi, M. P., Eds.; Wiley-VCH: Weinheim, 2001; Vol. 2, pp
246-278. (c) Majetich, G.; Wheless, K. Tetrahedron 1995, 51, 7095-
7129.
(7) (a) Nakamura, T.; Watanabe, Y.; Suyama, S.; Tezuka, H. J .
Chem. Soc., Perkin Trans. 2 2002, 1364-1369. (b) Nakamura, T.;
Busfield, W. K.; J enkins, I. D.; Rizzardo, E.; Thang, S. H.; Suyama, S.
J . Org. Chem. 2000, 65, 16-23.
(8) Wilsey, S.; Dowd, P.; Houk, K. N. J . Org. Chem. 1999, 64, 8801-
8811.
(9) Choo, K. Y.; Benson, S. W. Int. J . Chem. Kinet. 1981, 13, 833-
(10) (a) Walling, C. Pure Appl. Chem. 1967, 15, 69-80. (b) Walling,
C.; Padwa, A. J . Am. Chem. Soc. 1963, 85, 1593-1597.
(11) (a) Bacha, J . D.; Kochi, J . K. J . Org. Chem. 1965, 30, 3272-
3278. (b) Kochi, J . K. J . Am. Chem. Soc. 1962, 84, 1193-1197.
(12) Greene, F. D.; Savitz, M. L.; Osterholtz, F. D.; Lau, H. H.;
Smith, W. N.; Zanet, P. M. J . Org. Chem. 1963, 28, 55-64.
(13) Gray, P.; Williams, A. Chem. Rev. 1959, 59, 239-328.
(14) Walling, C.; Clark, R. T. J . Am. Chem. Soc. 1974, 96, 4530-
4533.
10.1021/jo049524y CCC: $27.50 © 2004 American Chemical Society
Published on Web 07/03/2004
J . Org. Chem. 2004, 69, 5281-5289
5281

Aureliano Antunes et al.
SCHEME 1
SCHEME 2
CHART 1
SCHEME 3
SCHEME 4
ejection from the 2-methyl-2-butoxyl radical.10b Results
which were rationalized in terms of the relief of ring
strain associated to cyclopentane ring opening.
For what concerns instead the comparison between
ring-opening reactions of cycloalkoxyl radicals having
different ring size, Beckwith studied the reversibility of
the exo cyclization reaction of ω-formyl alkyl radicals by
means of competitive kinetics (Scheme 2).15
Rate constants for â-scission (ring-opening) of the
cyclopentoxyl and cyclohexoxyl radicals were determined
as ko ) 4.7 × 108 and 1.1 × 107 s-1, respectively, in line
with the greater ring strain associated with a five-
membered ring as compared to a six-membered one.16 In
agreement with this conclusion are also the results of
relative reactivity studies showing that the 1-methylcy-
clopentoxyl radical undergoes ring-opening 10 times
faster than the 1-methylcyclohexoxyl one.10b Along this
line, in a study of the intramolecular selectivity in the
C-C â-scission reactions of 1-isopropylcyclopentoxyl and
1-R-cyclohexoxyl radicals (R ) Et, iPr) it was shown that
in the former radical ring opening is the major fragmen-
tation pathway, whereas in the 1-R-cyclohexoxyl radicals
ring opening is the major fragmentation pathway when
R ) Et while C-R cleavage predominantes when R )
iPr (Scheme 3).12
Thus, variations in both ring size and nature of the
alkyl ring substituent R may influence the competition
between ring opening and C-R cleavage in the 1-alkyl-
cycloalkoxyl radical.
In view of the importance of cycloalkoxyl radical ring-
opening reactions in organic synthesis,5,17,18 of the limited
amount of data available on the intramolecular selectivity
in C-C â-scission reactions of 1-alkylcycloalkoxyl radi-
cals, and in order to provide additional information on
the factors governing this selectivity, we have carried out
a systematic product study for the â-scission reactions
of an extended series of 1-alkylcycloalkoxyl radicals,
where both ring size and nature of the alkyl group have
been varied. The alkoxyl radicals have been generated
photochemically by visible light irradiation of CH2Cl2
solutions containing the parent 1-alkylcycloalkanols (Chart
1), (diacetoxy)iodobenzene (DIB), and I2 (Suarez’s re-
agent).
The DIB/I2 reagent is a well-known reagent based on
a hypervalent iodine compound (DIB) which, in the
presence of visible light, is able to convert hydroxyl-
containing substrates into products deriving from inter-
mediate oxygen-centered radicals.5a,17 The main advan-
tage of this reagent is represented by the fact that the
alkoxyl radicals can be easily generated from the parent
alcohols. Moreover, irradiation with visible light prevents
the occurrence of undesired photochemical reactions of
the carbonyl products formed by alkoxyl radical â-scis-
sion, which can instead occur when UV light is used for
alkoxyl radical generation.19
A mechanism for alkoxyl radical formation, proceeding
through the photochemical decomposition of an interme-
diate hypoiodite, which is generated by thermal reaction
of DIB, I2, and the alcohol, has been recently proposed.20
With 1-alkylcycloalkanols, the overall process can be
described to occur according to Scheme 4, where an
intermediate 1-alkylcycloalkoxyl radical is initially formed
by visible light irradiation of 1-alkylcycloalkanol/DIB/I2
mixtures and then undergoes competition between ring
opening and C-R cleavage as a function of ring size and
nature of the R group. The carbon-centered radicals
formed by alkoxyl radical â-scission are efficiently trapped
(15) (a) Beckwith, A. L. J .; Hay, B. P. J . Am. Chem. Soc. 1989, 111,
2674-2681. (b) Beckwith, A. L. J .; Hay, B. P. J . Am. Chem. Soc. 1989,
111, 230-234.
(16) Eliel, E. L.; Wilen, S. H. Stereochemistry of Organic Compounds;
J ohn Wiley & Sons: New York, 1994.
(17) Togo, H.; Katohgi, M. Synlett 2001, 5, 565-581.
(18) Ryu, I. Chem. Soc. Rev. 2001, 30, 16-25.
(19) Avila, D. V.; Brown, C. E.; Ingold, K. U.; Lusztyk, J . J . Am.
Chem. Soc. 1993, 115, 466-470.
(20) Courtneidge, J . L.; Lusztyk, J .; Page`, D.Tetrahedron Lett. 1994,
35, 1003-1006.
5282 J . Org. Chem., Vol. 69, No. 16, 2004

Photolysis of 1-Alkylcycloalkanols in the Presence of DIB and I2
TABLE 1. P r od u ct Distr ibu tion s Obser ved a fter
In the 1-alkylcyclohexanol series (Table 3), exclusive
formation of the corresponding ring-opened iodoketone
was observed in the reactions of substrates 3a ,b,h ,i. With
1-tert-butylcyclohexanol (3e) and 1-benzylcyclohexanol
(3g), exclusive formation of the corresponding alkyl iodide
(2-iodo-2-methylpropane and benzyl iodide, respectively)
and cycloalkanone was instead observed. The reaction
of 1-propylcyclohexanol (3c) led to the formation of
9-iodononan-4-one as major product accompanied by
smaller amounts of 1-iodopropane and cyclohexanone.
The reaction of 1-isopropylcyclohexanol (3d ) led to the
formation of 2-iodopropane and cyclohexanone as major
products accompanied by a smaller amount of 8-iodo-2-
methyloctan-3-one.
In the 1-alkylcycloheptanol series (Table 4), the reac-
tion of 1-phenylcycloheptanol (4h ) led to the exclusive
formation of 7-iodo-1-phenylheptan-1-one. With 1-isopro-
pylcycloheptanol (4d ), 1-tert-butylcycloheptanol (4e), and
1-benzylcycloheptanol (4g), exclusive formation of the
corresponding alkyl iodides (2-iodopropane, 2-iodo-2-
methylpropane, and benzyl iodide, respectively) and
cycloalkanones was instead observed. The reaction of
1-propylcycloheptanol (4c) led to the formation of com-
parable amounts of 10-iododecan-4-one, cycloheptanone,
and 1-iodopropane as major products accompanied by
smaller amounts of products (together accounting for ≈
23% of the detected products) identified as 1-propyl-3-
cyclohepten-1-ol, 1-propyl-4-cyclohepten-1-ol, and 4-iodo-
1-propylcycloheptanol.
In the 1-alkylcyclooctanol series (Table 5), the reactions
of 1-isopropylcyclooctanol (5d ), 1-tert-butylcyclooctanol
(5e), and 1-benzylcyclooctanol (5g) led to the exclusive
formation of the corresponding alkyl iodides (2-iodopro-
pane, 2-iodo-2-methylpropane, and benzyl iodide, respec-
tively) and cycloalkanones. The reaction of 1-propylcyclo-
octanol (5c) led to the formation of 11-iodoundecan-4-one,
cyclooctanone, and 1-iodopropane (together accounting
only for ∼19% of the detected products), accompanied by
greater amounts of products identified as 1-propyl-3-
cycloocten-1-ol, 1-propyl-4-cycloocten-1-ol, and 4-iodo-1-
propylcyclooctanol and by a smaller amount (∼2%) of a
product which, on the basis of the GC-MS data, can be
tentatively assigned to 6-iodo-1-propyl-4-cycloocten-1ol.
With 1-phenylcyclooctanol (5h ), formation of 8-iodo-1-
phenyloctan-1-one as major product was observed, ac-
companied by comparable amounts of products identified
as 1-phenyl-3-cycloocten-1-ol, 1-phenyl-4-cycloocten-1-ol,
and 4-iodo-1-phenylcyclooctanol (together accounting for
∼43% of the detected products).
Ir r a d ia tion of Ar -Sa tu r a ted CH2Cl2 Solu tion s (T ) 20 °C)
a
Con ta in in g 1-Alk ylcyclobu ta n ols (1f,g), DIB, a n d I2
a
[substrate] ) 10 mM; [DIB] ) 11 mM; [I2] ) 10 mM; λirr
)
480 nm. b Ratio between C-R cleavage and ring-opening products.
c Irradiation time ) 1 min. No C-R cleavage products detected.
d
e Irradiation time ) 2 min.
by iodine,21-23 minimizing the competition with side
reactions which may influence product distribution.
Resu lts
Argon-saturated CH2Cl2 solutions containing the 1-alkyl-
cycloalkanol (1-5) (10 mM), DIB (11 mM), and I2 (10
mM) were irradiated with visible light (λmax ≈ 480 nm)
at T ) 20 °C. The irradiation time was chosen in such a
way as to avoid complete substrate conversion. After
workup of the reaction mixture, the reaction products
1
were generally identified by GC-MS and H NMR and
quantitatively determined, together with the unreacted
1
substrate by GC and H NMR, using bibenzyl as internal
standard.
The reaction products are reported, together with sub-
strate conversion and ratio between C-R cleavage and
ring-opening products (Q), in Tables 1-5 for 1-alkyl-
cyclobutanols (1f,g), 1-alkylcyclopentanols (2a -i),
1-alkylcyclohexanols (3a -e,g-i), 1-alkylcycloheptanols
(4c-e,g,h ) and 1-alkylcyclooctanols (5c-e,g,h ), respec-
tively.
The results collected in Table 1 for 1-allylcyclobutanol
(1f) and 1-benzylcyclobutanol (1g) show the formation of
the corresponding ring-opened iodoketone (7-iodo-1-hep-
ten-4-one and 5-iodo-1-phenylpentan-2-one, respectively)
as the exclusive reaction product.
The results obtained for the 1-alkylcyclopentanol series
(Table 2) show the exclusive formation of the correspond-
ing ring-opened iodoketone in the reactions of substrates
2a -d ,h ,i. With 1-tert-butylcyclopentanol (2e), 1-allylcy-
clopentanol (2f), and 1-benzylcyclopentanol (2g) forma-
tion of comparable amounts of the corresponding ring-
opened iodoketones (7-iodo-2,2-dimethylheptan-3-one,
8-iodo-1-octen-4-one, and 6-iodo-1-phenylhexan-2-one, re-
spectively), alkyl iodides (2-iodo-2-methylpropane, 3-iodo-
propene, and benzyl iodide, respectively), and cycloal-
kanones were instead observed.
It is interesting to observe that the reactions of the
1-tert-butylcycloalkanols (2e-5e) led in all cases to
significantly lower conversions than those observed in the
reactions of the other substrates under the same experi-
mental conditions. Conversion was found to increase by
increasing irradiation time or, with 5e, by increasing DIB
and I2 concentrations. Observations which suggest that
in the presence of a bulky tert-butyl substituent formation
of the intermediate hypoiodite by thermal reaction
between DIB, I2, and the alcohol is disfavored.
(21) Minisci, F.; Recupero, F.; Gambarotti, C.; Punta, C.; Paganelli,
R. Tetrahedron Lett. 2003, 44, 6919-6922.
(22) Hook, S. C. W.; Saville, B. J . Chem. Soc., Perkin Trans. 2 1975,
589-593.
(23) The possibility that iodoalkane formation occurs through an
iodine transfer step from the hypoiodite, precursor of the alkoxyl
radical, should also be considered. We thank a reviewer for drawing
our attention on this point.
Discu ssion
As mentioned above, ring-opening of cycloalkoxyl radi-
cals to give ω-formyl alkyl radicals is a reversible process
J . Org. Chem, Vol. 69, No. 16, 2004 5283

Aureliano Antunes et al.
TABLE 2. P r od u ct Distr ibu tion s Obser ved a fter Ir r a d ia tion of Ar -Sa tu r a ted CH 2Cl2 Solu tion s (T ) 20 °C) Con ta in in g
a
1-Alk ylcyclop en ta n ols (2a -i), DIB, a n d I2
a
b
[substrate] ) 10 mM; [DIB] ) 11 mM; [I2] ) 10 mM; λirr ) 480 nm; irradiation time ) 5 min. Ratio between C-R cleavage
(cyclopentanone) and ring-opening products. c No C-R cleavage products detected. Irradiation time ) 15 min. e Irradiation time )
10 min.
d
(Scheme 2);15 thus, for relative reactivity studies, condi-
tions must be chosen in such a way that the follow-up
reaction of the carbon-centered radical produced by ring
opening is much faster than its cyclization. In the
cyclobutoxyl and cyclopentoxyl radicals the equilibrium
is strongly displaced to the fragmented carbon-centered
radical (i.e., K ) ko/kc . 2000 for the cyclopentoxyl
radical) while a significantly smaller equilibrium con-
stant (K . 10) has been measured for the cyclohexoxyl
radical.15a No data are instead available for the cyclo-
heptoxyl and cyclooctoxyl radicals for which, however, it
should be taken into account that formation of seven- and
eight-membered rings is usually disfavored as compared
to five- and six-membered ones.24 In this context, it is
also important to point out that both the efficient
trapping by iodine of the carbon-centered radical pro-
duced by ring opening (Scheme 4),21,22 and the steric and
electronic effects resulting from the presence of an alkyl
5284 J . Org. Chem., Vol. 69, No. 16, 2004

Photolysis of 1-Alkylcycloalkanols in the Presence of DIB and I2
TABLE 3. P r od u ct Distr ibu tion s Obser ved a fter Ir r a d ia tion of Ar -Sa tu r a ted CH 2Cl2 Solu tion s (T ) 20 °C) Con ta in in g
a
1-Alk ylcycloh exa n ols (3a -e,g-i), DIB, a n d I2
a
b
[substrate] ) 10 mM; [DIB] ) 11 mM; [I2] ) 10 mM; λirr ) 480 nm; irradiation time ) 10 min. Ratio between C-R cleavage
(cyclohexanone) and ring-opening products. c No C-R cleavage products detected. No ring-opening products detected. e Traces (<1%) of
d
a product assigned by GC-MS analysis to 1-propyl-3-cyclohexenol were also detected. f Irradiation time ) 30 min. Irradiation time )
g
5 min.
substituent bound to the carbonyl group,25,26 should
strongly depress the relative importance of the cyclization
reaction. Accordingly, in the reactions of substrates 3d
and 4c, no increase in the relative amount of ring-opening
products was observed when the concentration of iodine
was increased by a factor 5. Results which confirm that,
under the experimental conditions employed, cyclization
of the carbon-centered radical produced by ring opening
does not compete with its reaction with iodine. On the
basis of these considerations, it is reasonable to conclude
that reversibility should not influence to any significant
extent the intramolecular selectivity (competition between
ring opening and C-R cleavage) in the C-C â-scission
reactions of the 1-alkylcycloalkoxyl radicals reported
above.
Starting from the 1-alkylcyclobutanols, the results of
Table 1 show that the exclusive fragmentation pathway
of the 1-allylcyclobutoxyl and 1-benzylcyclobutoxyl radi-
cals, generated, respectively, from substrates 1f and 1g,
is cyclobutane ring opening (Scheme 5), and even though
C-R cleavage would lead to relatively stable radicals
(CH2dCHCH2 from 1f and PhCH2 from 1g),27 no evi-
dence for the formation of products deriving from this
pathway was obtained, a behavior which suggests that
with 1-alkylcyclobutoxyl radicals the fragmentation re-
gioselectivity is governed by the very high ring strain
associated with a four-membered ring.16
•
•
(24) Mandolini, L. Adv. Phys. Org. Chem. 1986, 22, 1-111.
Illuminati, G.; Mandolini, L. Acc. Chem. Res. 1981, 14, 95-102.
(25) Henry, D. J .; Coote, M. L.; Go´mez-Balderas, R.; Radom, L. J .
Am. Chem. Soc. 2004, 126, 1732-1740.
(26) Spellmeyer, D. C.; Houk, K. N. J . Org. Chem. 1987, 52, 959-
974.
(27) Wayner, D. D. M.; Griller, D. In Advances in Free Radical
Chemistry; Tanner, D. D., Ed.; J AI Press: Greenwich, 1990; Vol. 1, p
159.
J . Org. Chem, Vol. 69, No. 16, 2004 5285

Aureliano Antunes et al.
TABLE 4. P r od u ct Distr ibu tion s Obser ved a fter Ir r a d ia tion of Ar -Sa tu r a ted CH 2Cl2 Solu tion s (T ) 20 °C) Con ta in in g
a
1-Alk ylcycloh ep ta n ols (4c-e,g,h ), DIB, a n d I2
a
b
[substrate] ) 10 mM; [DIB] ) 11 mM; [I2] ) 10 mM; λirr ) 480 nm; irradiation time ) 5 min. Ratio between C-R cleavage
(cycloheptanone) and ring-opening products. c Ratio between cycloheptanone and 10-iododecan-4-one. Irradiation time ) 10 min. e No
d
ring-opening products detected. f Irradiation time ) 8 min. No C-R cleavage products detected.
g
The situation changes as we move to the 1-alkylcyclo-
pentanols. The results of Table 2 show that while the
exclusive fragmentation pathway of the alkoxyl radicals
generated from substrates 2a -d ,h ,i is cyclopentane ring
opening, competition between ring opening and C-R
cleavage is instead observed for the alkoxyl radicals
generated from substrates 2e-g. This result indicates
that the importance of the C-R cleavage pathway
increases with the stability of the alkyl radical R•.
However, since it is well-known that both the allyl and
benzyl radicals are significantly more stable than the tert-
butyl radical,28 the observation that the importance of
the C-R cleavage pathway decreases on going from the
1-tert-butylcyclopentoxyl radical (for which Q ) 2.2) to
the 1-benzylcyclopentoxyl and 1-allylcyclopentoxyl ones
(for which Q ) 1.03 and 0.58, respectively) indicates that
release of ring strain and alkyl radical stability are not
the only factors which govern the fragmentation regio-
selectivity, but also steric effects are likely to play a
significant role.
CH3CH2, CH3CH2CH2, Ph and (CH3)3CCH2, respectively).
With alkoxyl radicals derived from substrates 3d , 3e, and
3g (R ) (CH3)2CH, (CH3)3C and PhCH2, respectively),
predominant or exclusive C-R cleavage is instead ob-
served. Clearly, in a strain-free system the fragmentation
regioselectivity is essentially governed by the stability
of the radical formed. Unfortunately, the results obtained
for the 1-alkylcyclohexoxyl radicals do not provide infor-
mation on the role of steric effects because no evidence
for the formation of ring opening products was obtained
in the reactions of the 1-tert-butylcyclohexoxyl and 1-ben-
zylcyclohexoxyl radicals.
Interestingly, comparison between the results obtained
for the 1-benzylcyclobutoxyl, 1-benzylcyclopentoxyl and
1-benzylcyclohexoxyl radicals (for which Q < 0.01, Q )
1.03, and Q > 200, respectively) indicates that the ring
opening reaction of cycloalkoxyl radicals follows the
order: cyclobutoxyl > cyclopentoxyl > cyclohexoxyl, in
line with the ring strain associated with four-, five-, and
six-membered rings16 and with the computed activation
energies for the ring opening reactions of cycloalkoxyl
radicals (7.7, 15.4, and 20.7 kcal/mol, for the cyclobutoxyl,
cyclopentoxyl, and cyclohexoxyl radical, respectively).8
As we move to the 1-alkylcyclohexanol series, the
results of Table 3 show that now exclusive or predomi-
nant ring opening is observed for the alkoxyl radicals
generated from substrates 3a -c,h ,i (for which R ) CH3,
In the 1-alkylcycloheptanol series, the results of Table
4 show that exclusive ring opening is observed only for
the 1-phenylcycloheptoxyl radical in line with the rela-
tively low stability of the phenyl radical which disfavors
(28) The following homolytic bond dissociation energies have been
determined for (CH3)3C-H, PhCH2-H and CH2dCHCH2-H: 93.6,
87.9, and 86.8 kcal mol-1, respectively (see ref 27).
5286 J . Org. Chem., Vol. 69, No. 16, 2004

Photolysis of 1-Alkylcycloalkanols in the Presence of DIB and I2
TABLE 5. P r od u ct Distr ibu tion s Obser ved a fter Ir r a d ia tion of Ar -Sa tu r a ted CH 2Cl2 Solu tion s (T ) 20 °C) Con ta in in g
a
1-Alk ylcycloocta n ols (5c-e,g,h ), DIB, a n d I2
a
b
[substrate] ) 10 mM; [DIB] ) 11 mM; [I2] ) 10 mM; λirr ) 480 nm; irradiation time ) 5 min. Ratio between C-R cleavage
(cyclooctanone) and ring-opening products. c A small amount (∼2%) of a product assigned by GC-MS analysis to 6-iodo-1-propyl-4-
cyclooctenol was also detected. Ratio between cyclooctanone and 11-iodoundecan-4-one. e Irradiation time ) 10 min. f No ring-opening
d
products detected. g [substrate] ) 10 mM; [DIB] ) 40 mM; [I2] ) 20 mM. Irradiation time ) 8 min. No C-R cleavage products detected.
h
i
SCHEME 5
c). This radical can be trapped by iodine forming 4-iodo-
1-propylcycloheptanol or disproportionate to give the two
isomeric 1-propylcycloheptenols.
The observation that in the 1-propylcycloheptoxyl
radical 1,5-HAT competes with C-C â-scission whereas
with the corresponding cyclohexoxyl radical (derived from
3c) exclusive â-scission is observed can be explained in
terms of the greater flexibility of the former radical16,30
which allows the achievement of the quasilinear arrange-
ment between the oxygen, the migrating hydrogen, and
the carbon center, proposed for the transition state for
1,5-HAT.29,31 The cyclohexoxyl radical would have instead
to achieve an unfavorable boat conformation in order to
undergo 1,5-HAT.29 Another factor in this respect may
be represented by differences in the rate of ring-opening
of the 1-propylcyclohexoxyl and 1-propylcycloheptoxyl
radicals, but unfortunately the experimental data do not
the C-Ph bond cleavage pathway.27 With alkoxyl radicals
derived from substrates 4d ,e,g (R ) (CH3)2CH, (CH3)3C,
and PhCH2, respectively) exclusive C-R cleavage is
observed. Competition between ring opening and C-R
cleavage is instead observed for the 1-propylcycloheptoxyl
radical as shown by the formation of comparable amounts
of 10-iododecan-4-one (Scheme 6, path a), cycloheptanone
and 1-iodopropane (path b). In this reaction, formation
of smaller amounts of products identified as 1-propyl-3-
cyclohepten-1-ol, 1-propyl-4-cyclohepten-1-ol, and 4-iodo-
1-propylcycloheptanol was also observed.
Formation of the latter products can be rationalized
in terms of an intramolecular 1,5-hydrogen atom transfer
(1,5-HAT) from the cycloheptane ring in the intermediate
alkoxyl radical to give a carbon centered radical (path
(29) Dorigo, A. E.; Houk, K. N. J . Org. Chem. 1988, 53, 1650-1664.
(30) Wiberg, K. B. J . Org. Chem. 2003, 68, 9322-9329.
(31) Dorigo, A. E.; Houk, K. N. J . Am. Chem. Soc. 1987, 109, 2195-
2197.
J . Org. Chem, Vol. 69, No. 16, 2004 5287

Aureliano Antunes et al.
SCHEME 6
allow a direct comparison between the reactivities of the
two radicals.32
in terms of an intramolecular 1,5-HAT from the cyclooc-
tane ring in the intermediate alkoxyl radical. Competi-
tion between ring opening and 1,5-HAT is observed also
with the 1-phenylcyclooctoxyl radical, as shown by the
formation of 8-iodo-1-phenyloctan-1-one as major product
accompanied by comparable amounts of products identi-
fied as 1-phenyl-3-cycloocten-1-ol, 1-phenyl-4-cycloocten-
1-ol, and 4-iodo-1-phenylcyclooctanol. Again, as discussed
above for the cycloheptoxyl radicals, the presence of a
phenyl group should determine an increase in the relative
importance of the ring opening pathway as compared to
the intramolecular hydrogen atom transfer one.35 The
observation that in both the 1-propylcyclooctoxyl and
1-phenylcyclooctoxyl radicals the relative importance of
the 1,5-HAT pathway is significantly higher than in the
corresponding cycloheptoxyl radicals points toward a
more favorable geometry for the intramolecular 1,5-HAT
reaction in the former radicals.16,30,38
On the basis of the comparable strain energies associ-
ated with seven- and eight-membered rings,16 and of the
similar values of the Q ratio measured for the fragmen-
tation reactions of the 1-propylcycloheptoxyl and 1-pro-
pylcyclooctoxyl radicals (0.75 and 0.44, respectively), it
seems reasonable to suggest that the two radicals un-
dergo ring-opening with similar rates.
Finally, in full agreement with the results presented
above are also the results obtained previously in the
photolysis of a series of 1-vinylcycloalkanols in the
presence of DIB and I2.39 The product distributions shown
in Scheme 7 indicate that the intermediate alkoxyl
radicals undergo competition between ring opening and
intramolecular addition to the double bond depending on
ring size, while products deriving from C-R bond cleav-
age are not observed, as expected on the basis of the
relatively low stability of the vinyl radical.27
The observation that with the 1-propylcycloheptoxyl
radical no products deriving from a 1,5-HAT reaction
from the propyl group are formed indicates that intramo-
lecular HAT occurs significantly faster from the cyclo-
heptane ring than from the propyl side chain, in line with
the well-known preference for abstraction at more highly
substituted carbon atoms.18,33,34
Finally, the observation that no products deriving from
a 1,5-HAT reaction are formed with the 1-phenylcyclo-
heptoxyl radical can be explained in terms of an increase
in the rate of ring opening determined by the presence
of a phenyl substituent,35 which accordingly decreases the
relative importance of the 1,5-HAT pathway.
In the 1-alkylcyclooctanol series, the results of Table
5 show that with the alkoxyl radicals derived from
substrates 5d , 5e, and 5g (R ) (CH3)2CH, (CH3)3C and
PhCH2, respectively) exclusive C-R cleavage is observed.
Competition between ring opening and C-R cleavage is
instead observed for the 1-propylcyclooctoxyl radical as
shown by the formation of comparable amounts of 11-
iodoundecan-4-one, cyclooctanone and 1-iodopropane.
With this radical however, â-scission reactions appear
to be minor pathways as clearly shown by the formation
of significant amounts of products identified as 1-propyl-
3-cycloocten-1-ol, 1-propyl-4-cycloocten-1-ol, and 4-iodo-
1-propylcyclooctanol. In analogy, with the results dis-
cussed above for the 1-propylcycloheptoxyl radical (Scheme
6), formation of these products can be again rationalized
(32) In a first approximation, by assuming similar rate constants
for C-R bond cleavage in the 1-propylcyclohexoxyl and 1-propylcyclo-
heptoxyl radicals, comparison between the Q ratios obtained from the
reactions of the two radicals (Q ) 0.06 for the 1-propylcyclohexoxyl
radical and Q ) 0.75 for the 1-propylcycloheptoxyl one) suggests that
ring-opening of the former radical is faster than that of the latter one,
a finding which, however, is in contrast with the significantly greater
ring strain associated to a seven-membered ring as compared to a six-
membered one. Moreover, such an assumption does not take into
account that the differences in strain between the radical and the
product cycloalkanone should be significantly different in the two
systems (see ref 16), and accordingly different rates of C-R cleavage
can be reasonably expected for the two radicals.
(33) Heusler, K.; Kalvoda, J . Angew. Chem., Int. Ed. Engl. 1964, 3,
525-538.
(34) Weavers, R. T. J . Org. Chem. 2001, 66, 6453-6461.
(35) For example an increase between 10 and 20 times (depending
on solvent) in the rate constant for C-Me â-scission is observed on going
from the tert-butoxyl radical (see ref 36) to the cumyloxyl one (see refs
19 and 37).
Thus, ring-opening is observed only for the 1-vinylcy-
clopentoxyl radical, while for the other radicals the rate
of intramolecular addition is significantly faster than that
(36) (a) Tsentalovich, Y. P.; Kulik, L. V.; Gritsan, N. P.; Yurk-
ovskaya, A. V. J . Phys. Chem. A 1998, 102, 7975-7980. (b) Weber,
M.; Fischer, H. J . Am. Chem. Soc. 1999, 121, 7381-7388.
(37) Baciocchi, E.; Bietti, M.; Salamone, M.; Steenken, S. J . Org.
Chem. 2002, 67, 2266-2270.
(38) Petrovic´, G.; Cˇ ekovic´, Z. Tetrahedron 1999, 55, 1377-1390.
(39) (a) Galatsis, P.; Millan, S. D. Tetrahedron Lett. 1991, 32, 7493-
7496. (b) Galatsis, P.; Millan, S. D.; Faber, T. J . Org. Chem. 1993, 58,
1215-1220.
5288 J . Org. Chem., Vol. 69, No. 16, 2004

Photolysis of 1-Alkylcycloalkanols in the Presence of DIB and I2
were used without further purification. Details of the synthesis
of the other 1-alkylcycloalkanols are given in the Supporting
Information. The purity of the 1-alkylcycloalkanols employed
in the product studies was always g99%.
SCHEME 7
P r od u ct An a lyses. All the reactions were carried out under
an argon atmosphere. Dichloromethane was purified prior to
use by column chromatography over basic alumina. Irradia-
tions were performed with visible light (10 × 15 W lamps with
emission between 400 and 550 nm, λmax ≈ 480 nm). The reactor
was a cylindrical flask equipped with a water-cooling jacket
thermostated at 20 °C. Irradiation times were chosen in such
a way as to avoid complete substrate consumption. In a typical
experiment, a solution of the alcohol (10 mM) in CH2Cl2 (5
mL) containing (diacetoxy)iodobenzene (11 mM) and iodine (10
mM) was irradiated for times ranging between 1 and 30 min
under Ar bubbling. The reaction mixture was then poured into
water and extracted with CH2Cl2 (2 × 10 mL). The combined
organic layers were washed with a 10% aqueous thiosulfate
solution (2 × 30 mL) and water (2 × 30 mL) and dried over
anhydrous sodium sulfate. Reaction products were generally
identified by GC-MS and 1H NMR (for details see the
Supporting Information) and quantitatively determined by GC
and 1H NMR (comparison with authentic samples for the
cycloalkanones and alkyl iodides). In the reactions of sub-
strates 4c, 5c, and 5h , the reaction products were not isolated
and their structures were assigned on the basis of the GC-
MS and, when possible, 1H NMR data. The quantitative
analysis of the reaction mixture was performed by GC and by
1H NMR after removal of the solvent at atmospheric pressure.
In both cases, bibenzyl was used as internal standard. In the
latter method, the product RC(O)(CH2)nCH2I (where n ) 2-6,
and R ) Me, Et, Pr, i-Pr, t-Bu, benzyl, allyl, Ph, and neopentyl)
was quantitatively determined through the characteristic CH2I
signal at 3.17 ppm. Good to excellent mass balances (g 85%)
were obtained in all experiments. With the exception of the
1-benzylcycloalkanols for which both benzyl iodide and cy-
cloalkanone were quantitatively determined, no quantitative
determination of the alkyl iodide was carried out for the
reactions of the other substrates. Thus, the ratio Q between
R cleavage and ring-opening products was generally deter-
mined from the amounts of cycloalkanone and ring-opened
iodinated ketone formed. Two to four experiments were carried
out for each substrate showing excellent reproducibility.
for ring opening. As a matter of comparison, an analogous
trend has been observed with the alkoxyl radicals formed
from the 1-benzylcycloalkanols (substrates 2g-5g), where
competition between ring-opening and C-R cleavage is
observed only for the 1-benzylcyclopentoxyl radical,
whereas with the other 1-benzylcycloalkoxyl radicals
exclusive C-R cleavage occurs. Interestingly, comparison
of the results obtained for the 1-vinylcyclopentoxyl
radical39a with those obtained for the 1-allylcyclopentoxyl
and 1-benzylcyclopentoxyl ones suggests that the rate of
intramolecular addition to the double bond in the former
radical is comparable with those for C-R cleavage in the
latter ones.
In conclusion, the results presented above clearly show
that in the fragmentation reactions of 1-alkylcycloalkoxyl
radicals the regioselectivity is governed by the interplay
between release of ring strain associated to ring opening
and stability of the alkyl radical formed by C-R cleavage,
even though in the presence of bulky substituents such
as a tert-butyl group steric effects appear to play a role.
The importance of the ring-opening pathway decreases
by increasing ring size, and, with the exclusion of the
cyclobutoxyl radicals, by increasing the stability of R•.
Accordingly, the rates of ring-opening follow the order
cyclobutoxyl > cyclopentoxyl > cyclohexoxyl, in line with
the ring strain associated with four-, five-, and six-
membered rings. Unfortunately, the results obtained do
not allow a direct comparison between the rates of ring-
opening of the cyclohexoxyl, cycloheptoxyl, and cyclooc-
toxyl radicals, pointing however toward similar rates of
ring-opening for the latter radicals (cycloheptoxyl ≈
cyclooctoxyl).
With the 1-propylcycloheptoxyl, 1-propylcyclooctoxyl,
and 1-phenylcyclooctoxyl radicals â-scission competes
with an intramolecular 1,5-HAT reaction, a behavior that
has been interpreted in terms of the possibility of
achieving a more favorable geometry in the transition
state for intramolecular 1,5-HAT in cycloheptoxyl and
cyclooctoxyl radicals as compared to cyclohexoxyl ones.
Ack n ow led gm en t . Financial support from the
European Union (Contract No. QLK5-CT-1999-01277)
and the Ministero dell’Istruzione dell’Universita` e
della Ricerca (MIUR) is gratefully acknowledged.
Su p p or tin g In for m a tion Ava ila ble: Details on the syn-
thesis and characterization of the 1-alkylcycloalkanols and of
their reaction products (GC-MS and 1H NMR data). This
material is available free of charge via the Internet at
http://pubs.acs.org.
Exp er im en ta l Section
Ma ter ia ls. (Diacetoxy)iodobenzene (DIB) and iodine were
of the highest commercial quality available. Commercial
samples of 1-methylcyclopentanol and 1-methylcyclohexanol
J O049524Y
J . Org. Chem, Vol. 69, No. 16, 2004 5289