Structural effects on the β-scission reaction of tertiary arylcarbinyloxyl radicals. The role of α-cyclopropyl and α-cyclobutyl groups

Article abstract of DOI:10.1021/jo050883i

A product and time-resolved kinetic study on the reactivity of tertiary arylcarbinyloxyl radicals bearing α-cyclopropyl and α-cyclobutyl groups has been carried out. Both the 1-cyclopropyl-1-phenylethoxyl (1 ^.) and α,α-dicyclopropylphenylmethoxyl (2^.) radicals undergo β-scission to give cyclopropyl phenyl ketone as the major or exclusive product with rate constants higher than that measured for the cumyloxyl radical. It is proposed that in the transition state for β-scission of 1^. and 2^., formation of the C=O double bond is assisted by overlap with the C-C bonding orbitals of the cyclopropane ring. With tertiary arylcarbinyloxyl radicals bearing α-cyclobutyl groups such as the 1-cyclobutyl-1-phenylethoxyl (4^.) and 1-cyclobutyl-1-phenylpropoxyl (5^.) radicals, the fragmentation regioselectivity is essentially governed by the stability of the radical formed by β-scission. Accordingly, 4^. undergoes exclusive C-cyclobutyl bond cleavage to give acetophenone, whereas with 5^., competition between C-cyclobutyl and C-ethyl bond cleavage, leading to propiophenone and cyclobutylphenyl ketone in a 2:1 ratio, is observed.

Full text of DOI:10.1021/jo050883i

Structural Effects on the â-Scission Reaction of Tertiary
Arylcarbinyloxyl Radicals. The Role of r-Cyclopropyl and
r-Cyclobutyl Groups
Massimo Bietti,* Giacomo Gente, and Michela Salamone
Dipartimento di Scienze e Tecnologie Chimiche, Universita` “Tor Vergata”, Via della Ricerca Scientifica,
1 I-00133 Rome, Italy
bietti@uniroma2.it
Received May 2, 2005
A product and time-resolved kinetic study on the reactivity of tertiary arylcarbinyloxyl radicals
bearing R-cyclopropyl and R-cyclobutyl groups has been carried out. Both the 1-cyclopropyl-1-
phenylethoxyl (1•) and R,R-dicyclopropylphenylmethoxyl (2•) radicals undergo â-scission to give
cyclopropyl phenyl ketone as the major or exclusive product with rate constants higher than that
measured for the cumyloxyl radical. It is proposed that in the transition state for â-scission of 1•
and 2•, formation of the CdO double bond is assisted by overlap with the C-C bonding orbitals of
the cyclopropane ring. With tertiary arylcarbinyloxyl radicals bearing R-cyclobutyl groups such as
the 1-cyclobutyl-1-phenylethoxyl (4•) and 1-cyclobutyl-1-phenylpropoxyl (5•) radicals, the fragmenta-
tion regioselectivity is essentially governed by the stability of the radical formed by â-scission.
Accordingly, 4• undergoes exclusive C-cyclobutyl bond cleavage to give acetophenone, whereas with
5•, competition between C-cyclobutyl and C-ethyl bond cleavage, leading to propiophenone and
cyclobutylphenyl ketone in a 2:1 ratio, is observed.
SCHEME 1
Recently, while investigating the O-neophyl rearrange-
ment of ring-substituted 1,1-diarylalkoxyl radicals, we
observed that among the radicals studied the cyclopropyl-
[bis(4-methoxyphenyl)]methoxyl radical displayed the
lowest reactivity, undergoing 1,2-aryl shift with k ) 2.6
× 105 s-1 1
. No evidence for the formation of products
deriving from C-cyclopropyl â-scission but only of the
rearrangement product cyclopropyl 4-methoxyphenyl
ketone was obtained. Quite interestingly, the rearrange-
ment rate constant measured for this radical is lower
than the rate constant measured previously for C-methyl
â-scission of the 4-methoxycumyloxyl radical, k ) 1.0 ×
106 s-1 (Scheme 1: An ) 4-MeOC6H4),2,3 leading to the
suggestion that in alkoxyl radicals C-cyclopropyl â-scis-
sion occurs at a significantly lower rate than C-Me
â-scission, in line with the slightly higher stability of the
methyl radical as compared to the cyclopropyl one (on
the basis of the CH3-H and c-C3H5-H BDEs: 105.0 and
106.3 kcal mol-1, respectively).4,5
(1) Aureliano Antunes, C. S.; Bietti, M.; Ercolani, G.; Lanzalunga,
O.; Salamone, M. J. Org. Chem. 2005, 70, 3884-3891.
(2) Baciocchi, E.; Bietti, M.; Salamone, M.; Steenken, S. J. Org.
Chem. 2002, 67, 2266-2270.
(3) Unfortunately, it was not possible to study the reactivity of the
1,1-di(4-methoxyphenyl)ethoxyl radical as the instability of the precur-
sor 1,1-di(4-methoxyphenyl)ethanol did not allow the synthesis of the
corresponding tert-butyl peroxide (see ref 1).
C-C â-scission is one of the most important reactions
of alkoxyl radicals, and accordingly, this reaction has
(4) Luo, Y.-R. Handbook of Bond Dissociation Energies in Organic
Compounds; CRC Press LLC: Boca Raton, FL, 2003.
(5) For a critical discussion on the relative stabilities of alkyl
radicals, see refs 6 and 7.
10.1021/jo050883i CCC: $30.25 © 2005 American Chemical Society
Published on Web 07/16/2005
6820
J. Org. Chem. 2005, 70, 6820-6826

Reaction of Tertiary Arylcarbinyloxyl Radicals
SCHEME 2
CHART 1
been thoroughly investigated8-16 and has found increas-
ing application in synthetically useful procedures.17-21
With tertiary alkoxyl radicals bearing different alkyl
groups it is possible, at least in principle, to obtain three
different C-C bond fragmentations (Scheme 2), and
consequently, the intramolecular selectivity in the â-scis-
sion reactions of alkoxyl radicals has been the subject of
detailed investigation.
SCHEME 3
The main conclusion of these studies was that cleavage
generally leads to the most stable possible alkyl radical,
and it was also shown that other factors such as the
stability of the product ketone (in some â-scission reac-
tions leading to the same radical R•)8a,13b or, when dealing
with cycloalkoxyl radicals, the release of ring strain
associated to ring opening13a,c,15,22-24 may play an impor-
tant role. Along this line, following our continuous
interest in the role of structural effects on the â-scission
reactions of alkoxyl radicals and taking advantage of the
fact that arylcarbinyloxyl radicals display an absorption
band in the visible region of the spectrum,2,25 it seemed
particularly interesting to study the reactivity of tertiary
arylcarbinyloxyl radicals bearing one or two R-cyclopropyl
groups to test whether the presence of these groups
influences the reactivity toward â-scission of these radi-
cals. For this purpose, we have carried out a product and
time-resolved kinetic study on the reactivity of the
1-cyclopropyl-1-phenylethoxyl and dicyclopropylphenyl-
methoxyl radicals (1• and 2•, respectively) whose struc-
tures are shown in Chart 1. The results obtained for these
radicals have been compared with those obtained analo-
gously for the cumyloxyl radical (3•).
The study has been also extended to tertiary arylcarbi-
nyloxyl radicals bearing one or two R-cyclobutyl groups
(radicals 4•-6• in Chart 1). However, because time-
resolved experiments have shown that 4• displays a
reactivity which exceeds the time resolution of the ns
laser flash photolysis apparatus employed (see below),
with radicals 5• and 6•, which can be reasonably expected
to display a higher reactivity than 4•, only product studies
have been carried out.
(6) Coote, M. L.; Pross, A.; Radom, L. Org. Lett. 2003, 5, 4689-4692.
(7) Matsunaga, N.; Rogers, D. W.; Zavitsas, A. A. J. Org. Chem.
2003, 68, 3158-3172.
(8) (a) Newcomb, M.; Daublain, P.; Horner, J. H. J. Org. Chem. 2002,
67, 8669-8671. (b) Horner, J. H.; Choi, S.-Y.; Newcomb, M. Org. Lett.
2000, 2, 3369-3372.
(9) (a) Nakamura, T.; Watanabe, Y.; Suyama, S.; Tezuka, H. J.
Chem. Soc., Perkin Trans. 2 2002, 1364-1369. (b) Nakamura, T.;
Busfield, W. K.; Jenkins, I. D.; Rizzardo, E.; Thang, S. H.; Suyama, S.
J. Org. Chem. 2000, 65, 16-23.
(10) Wilsey, S.; Dowd, P.; Houk, K. N. J. Org. Chem. 1999, 64, 8801-
8811.
(11) (a) Avila, D. V.; Brown, C. E.; Ingold, K. U.; Lusztyk, J. J. Am.
Chem. Soc. 1993, 115, 466-470. (b) Neville, A. G.; Brown, C. E.;
Rayner, D. M.; Ingold, K. U.; Lusztyk, J. J. Am. Chem. Soc. 1989, 111,
9269-9270.
(12) (a) Banks, J. T.; Scaiano, J. C. J. Am. Chem. Soc. 1993, 115,
6409-6413. (b) Baigne´e, A.; Howard, J. A.; Scaiano, J. C.; Stewart, L.
C. J. Am. Chem. Soc. 1983, 105, 6120-6123. (c) Mendenhall, G. D.;
Stewart, L. C.; Scaiano, J. C. J. Am. Chem. Soc. 1982, 104, 5109-
5114.
Results
Product Studies. Arylcarbinyloxyl radicals 1•-6•
have been generated photochemically by visible light
irradiation of CH2Cl2 solutions containing the parent
arylalkanols (1a-6a), (diacetoxy)iodobenzene (DIB), and
I2. It is well established that under these conditions the
DIB/I2 reagent converts alcohols into intermediate hy-
poiodites which are then photolyzed to give alkoxyl
radicals,18,19,22,26 precursors of the observed reaction
products. With arylalkanols 1a-6a, the alkoxyl radicals
so formed are expected to undergo â-scission to give the
corresponding aryl ketones (Scheme 3: cyp ) cyclopropyl;
cyb ) cyclobutyl).
(13) (a) Walling, C.; Clark, R. T. J. Am. Chem. Soc. 1974, 96, 4530-
4533. (b) Walling, C. Pure Appl. Chem. 1967, 15, 69-80. (c) Walling,
C.; Padwa, A. J. Am. Chem. Soc. 1963, 85, 1593-1597.
(14) (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.
(15) 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.
(16) Gray, P.; Williams, A. Chem. Rev. 1959, 59, 239-328.
(17) Yoshimitsu, T.; Sasaki, S.; Arano, Y.; Nagaoka, H. J. Org. Chem.
2004, 69, 9262-9268.
(18) Boto, A.; Herna´ndez, R.; Montoya, A.; Sua´rez, E. Tetrahedron
Lett. 2004, 45, 1559-1563. Francisco, C. G.; Gonzalez, C. C.; Paz, N.
R.; Sua´rez, E. Org. Lett. 2003, 5, 4171-4173. Alonso-Cruz, C. R.;
Kennedy, A. R.; Rodriguez, M. S.; Sua´rez, E. Org. Lett. 2003, 5, 3729-
3732. Sua´rez, E.; Rodriguez, M. S. In Radicals in Organic Synthesis;
Renaud, P., Sibi, M. P., Eds.; Wiley-VCH: Weinheim, Germany, 2001;
Vol. 2, pp 440-454.
Argon-saturated CH2Cl2 solutions containing arylal-
kanols 1a-6a (10 mM), DIB (22 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
(19) Togo, H.; Katohgi, M. Synlett. 2001, 5, 565-581.
(20) (a) Zhang, W. In Radicals in Organic Synthesis; Renaud, P.;
Sibi, M. P., Eds.; Wiley-VCH: Weinheim, Germany, 2001; Vol. 2, pp
234-245. (b) Dowd, P.; Zhang, W. Chem. Rev. 1993, 93, 2091-2115.
(21) Yet, L. Tetrahedron 1999, 55, 9349-9403.
(22) Aureliano Antunes, C. S.; Bietti, M.; Lanzalunga, O.; Salamone,
M. J. Org. Chem. 2004, 69, 5281-5289.
(25) (a) Avila, D. V.; Ingold, K. U.; Di Nardo, A. A.; Zerbetto, F.;
Zgierski, M. Z.; Lusztyk, J. J. Am. Chem. Soc. 1995, 117, 2711-2718.
(b) Avila, D. V.; Lusztyk, J.; Ingold, K. U. J. Am. Chem. Soc. 1992,
114, 6576-6577.
(26) Courtneidge, J. L.; Lusztyk, J.; Page`, D. Tetrahedron Lett. 1994,
35, 1003-1006.
(23) (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.
(24) Bietti, M.; Lanzalunga, O.; Salamone, M. J. Org. Chem. 2005,
70, 1417-1422.
J. Org. Chem, Vol. 70, No. 17, 2005 6821

Bietti et al.
TABLE 1. Product Distributions and Yields Observed
after Irradiation of Argon-Saturated CH2Cl2 Solutions
a
Containing Arylalkanols 1a-6a, DIB, and I2
FIGURE 1. Time-resolved absorption spectra observed after
266-nm LFP of 1p (10.4 mM) in an argon-saturated MeCN
solution at 192 ns (empty circles), 352 ns (filled circles), and
1.3 µs (empty squares) after the 8 ns, 10 mJ laser pulse.
Insets: (a) Zoom-in of the 260-600 nm region. (b) First-order
decay of radical 1• monitored at 485 nm. (c) Corresponding
first-order buildup of absorption at 240 nm, assigned to the
product cyclopropyl phenyl ketone.
a [DIB] ) 22 mM, [I2] ) 10 mM, [substrate] ) 10 mM,
irradiation wavelength ) 480 nm, irradiation time ) 15 min, T )
20 °C. b Product yields: cyclopropyl phenyl ketone 45%, acetophe-
none e 0.5%. c Irradiation time ) 2 min. Under these conditions,
complete conversion of the substrate (4a) was observed. d Irra-
diation time ) 1 min, using only 4 of the 10 15-W lamps available.
e Product yields: propiophenone 34%, cyclobutyl phenyl ketone
17%.
scribed procedure (see Supporting Information).2,27 To
obtain information on the effect of solvent polarity on the
â-scission reaction, radicals 1• and 2• have been also
generated in MeCN/H2O 1:1 and CF3CH2OH.
to avoid complete substrate conversion. After workup of
the reaction mixture, the reaction products were identi-
fied by GC (comparison with authentic samples) and GC-
MS and were quantitatively determined, together with
the unreacted substrate by GC, using bibenzyl as internal
standard. The product distributions and yields observed
after irradiation of argon-saturated CH2Cl2 solutions
containing substrates 1a-6a, DIB, and I2 are collected
in Table 1.
The reactions of R,R-dicyclopropylphenylmethanol (2a)
and 2-phenyl-2-propanol (3a) led to the formation of
cyclopropyl phenyl ketone and acetophenone, respec-
tively, as the exclusive reaction product. With 1-cyclo-
propyl-1-phenylethanol (1a), formation of cyclopropyl
phenyl ketone as the major product, accompanied by
traces of acetophenone, was instead observed, and a lower
limit for the product ratio cyclopropyl phenyl ketone/
acetophenone g90 was determined.
The reactions of 1-cyclobutyl-1-phenylethanol (4a) and
R,R-dicyclobutylphenylmethanol (6a) led to the formation
of acetophenone and cyclobutyl phenyl ketone, respec-
tively, as the exclusive reaction product. With 1-cyclobu-
tyl-1-phenylpropanol (5a), formation of propiophenone
and cyclobutyl phenyl ketone in a 2:1 ratio was observed.
Time-Resolved Kinetic Studies. Alkoxyl radicals 1•,
2•, and 4• have been generated in MeCN by 266-nm laser
flash photolysis (LFP) of the parent tert-butyl peroxides
1p, 2p, and 4p (equation 1: cyp ) cyclopropyl; cyb )
cyclobutyl), which in turn have been synthesized by
reaction of arylalkanols 1a, 2a, and 4a with tert-butyl
hydroperoxide in the presence of p-toluenesulfonic acid
according to a slight modification of a previously de-
The cumyloxyl radical 3• was instead generated in
MeCN and CF3CH2OH by 266-nm LFP of dicumyl
peroxide (5.0 × 10-3 M) as described previously.25
In MeCN, alkoxyl radicals 1• and 2• display a broad
absorption band in the visible region of the spectrum
centered around 485 nm, which is unaffected by oxygen,
in line with previous observations on the spectral proper-
ties and reactivity toward oxygen of the cumyloxyl radical
3•.2,11a,25,28 The visible absorption band undergoes a first-
order decay, leading in both cases to a corresponding
strong increase in absorption in the UV region of the
spectrum, assigned, on the basis of the product studies
described above, to the product cyclopropyl phenyl ketone
formed by â-scission in radicals 1• and 2•. Quite interest-
ingly, as compared to MeCN, in MeCN/H2O 1:1 and CF3-
CH2OH, a red-shift in the position of the visible absorp-
tion band maximum of radicals 1•-3• of ≈30 nm is
observed.2
As an example, Figure 1 displays the time-resolved
absorption spectra observed after LFP of an argon-
(27) Hendrickson, W. H.; Nguyen, C. C.; Nguyen, J. T.; Simons, K.
T. Tetrahedron Lett. 1995, 36, 7217-7220.
(28) Actually, the time-resolved absorption spectrum of the R,R-
dicyclopropylphenylmethoxyl radical (2•) recorded in MeCN (see Figure
S1 in the Supporting Information) shows a very broad absorption band
extending from approximately 550 nm down to the UV region,
characterized by maxima at ≈490, 420, and 360 nm.
6822 J. Org. Chem., Vol. 70, No. 17, 2005

Reaction of Tertiary Arylcarbinyloxyl Radicals
TABLE 2. Visible Absorption Band Maximum
Wavelengths, Decay, and Buildup Rate Constants for
Alkoxyl Radicals (1•-4•) Measured at T ) 25 °C
radical
conditions
λmax/nma
kV/s-1 b
kv/s-1 c
1•
MeCN
MeCN
Ar
O2
485
485
515
515
3.0 × 106
3.1 × 106
9.3 × 106
2.1 × 107
3.2 × 106
3.0 × 106
8.9 × 106
2.2 × 107
MeCN/H2Od Ar
CF3CH2OH
Ar
2•
MeCN
MeCN
Ar
O2
490e
1.3 × 106
1.1 × 106
4.6 × 106
1.6 × 107
1.3 × 106
490e
1.4 × 106
MeCN/H2Od Ar
4.5 × 106
f
f
g
CF3CH2OH
Ar
3•
4•
MeCN
CF3CH2OH
Ar
Ar
485
515
7.1 × 105
6.1 × 106
MeCN
Ar
g8 × 107 h
a Visible absorption band maximum of the alkoxyl radical.
b Measured following the decay of the alkoxyl radical at the visible
absorption band maximum. c Measured following the buildup of
the product aryl ketone at 240 or 250 nm. d MeCN/H2O 1:1 (v/v).
FIGURE 2. Time-resolved absorption spectra observed after
266-nm LFP of dicumyl peroxide (5.0 mM) in an argon-
saturated CF3CH2OH solution at 48 (filled circles), 120 (empty
circles), 258 (filled squares), and 810 ns (empty squares) after
the 8 ns, 10 mJ laser pulse. Insets: (a) Zoom-in of the 260-
600 nm region. (b) First-order decay of radical 3• monitored
at 515 nm. (c) Corresponding first-order buildup of absorption
at 250 nm, assigned to the product acetophenone.
e See ref 28 and Figure S1 in the Supporting Information. λmax
f
between 480 and 530 nm. The low intensity of the visible
absorption band did not allow a precise determination of the band
maximum under these conditions. g A fast buildup was observed,
but because of an initial bleaching in the kinetic traces, it was
not possible to measure the buildup rate constant. h Measured at
T ) -10 °C. Under these conditions, the radical was too reactive
to be studied by ns LFP and, on the basis of the buildup at 250
nm, only a lower limit could be determined.
saturated MeCN solution containing peroxide 1p, show-
ing after 192 ns (empty circles) a broad absorption band
centered at 485 nm, which is assigned to the radical 1•.
This species undergoes a first-order decay (inset b)
accompanied by a corresponding buildup of optical den-
sity at 240 nm (inset c) assigned to cyclopropyl phenyl
ketone (an isosbestic point is visible at 290 nm) which,
accordingly, is characterized by an absorption maximum
at this wavelength. As expected, both the 485-nm and
the 240-nm bands were not affected by the presence of
oxygen. Also visible is a relatively stable absorption band
centered at 280 nm which, on the basis of the comparison
with literature data,2,29 is assigned to the tert-butoxyl
radical, formed according to eq 1.30
region, but only the formation, immediately after the
laser pulse, of a stable species absorbing around 250 nm
assigned on the basis of the results of product studies
described above to acetophenone formed after C-cyclobu-
tyl â-scission in 4•.
The reactivity of alkoxyl radicals 1• and 2• was mea-
sured spectrophotometrically by monitoring the decrease
in optical density at their visible absorption maxima and
the corresponding buildup in the UV region because of
the formation of cyclopropyl phenyl ketone, which were
found to follow in all cases first-order kinetics. The
corresponding rate constants thus obtained are reported
in Table 2. Radical 4• was instead too reactive to be
studied by ns LFP and, on the basis of the buildup of the
product acetophenone at 250 nm described above, even
at T ) -10 °C only a lower limit for its reactivity could
be determined (k g 8 × 107 s-1).
The time-resolved absorption spectra observed after
LFP of an argon-saturated MeCN solution containing
peroxide 2p are instead displayed in Figure S1 (see
Supporting Information).
Figure 2 displays the time-resolved absorption spectra
observed after LFP of an argon-saturated CF3CH2OH
solution containing dicumyl peroxide. The spectrum
recorded after 48 ns is characterized by a broad absorp-
tion band in the visible region centered at 515 nm and
by a band in the UV region between 300 and 350 nm,
which are assigned to the cumyloxyl radical 3•.2,25 As
In the same table, the decay rate constants for the
cumyloxyl radical 3• measured at 485 and 515 nm in
MeCN and CF3CH2OH, respectively, are also included.
compared to the spectrum of 3• recorded in MeCN,2,25
a
Discussion
30-nm red-shift in the position of the visible absorption
band maximum is observed (see later). This species
undergoes a first-order decay (inset b) accompanied by a
corresponding buildup of optical density at 250 nm (inset
c) and 280 nm assigned to acetophenone (an isosbestic
point is visible at 300 nm).
With peroxide 4p, the time-resolved absorption spec-
trum observed after LFP of an argon-saturated MeCN
solution did not show an absorption band in the visible
The results of product studies collected in Table 1 show
that the reaction of 1-cyclopropyl-1-phenylethanol (1a)
leads to the formation of cyclopropyl phenyl ketone and
acetophenone in a g90 ratio. This result clearly indicates
that in 1• cleavage of the C-methyl bond is strongly
favored over that of the C-cyclopropyl one, as expected
on the basis of the higher stability of the methyl radical
as compared to the cyclopropyl one discussed above.4,5
However, since the difference in BDE between CH3-H
and c-C3H5-H has been reported to be around 1 kcal
mol-1,4 the large product ratio observed in the reaction
of 1a suggests that the stability of the radical formed
after â-scission is not the only factor which governs the
fragmentation regioselectivity in 1•.
(29) Tsentalovich, Y. P.; Kulik, L. V.; Gritsan, N. P.; Yurkovskaya,
A. V. J. Phys. Chem. A 1998, 102, 7975-7980.
(30) The slight increase in absorption observed at 280 nm is
attributed to the formation of cyclopropyl phenyl ketone, which, in
addition to the 240-nm band described above, is also characterized by
a weak absorption band centered at 276 nm.
J. Org. Chem, Vol. 70, No. 17, 2005 6823

Bietti et al.
SCHEME 4
SCHEME 5
Quite surprisingly, the results collected in Table 2 show
that as compared to 3•, an increase in the rate of
â-scission is observed with both 1• and 2•. In particular,
taking into account the results of product studies, it
appears that both C-methyl bond cleavage in 1• and
C-cyclopropyl bond cleavage in 2• occur faster than
C-methyl bond cleavage in 3•, whereas C-cyclopropyl bond
cleavage is significantly slower in 1• than in 2•. These
observations clearly indicate that the presence of an
R-cyclopropyl group plays an important role in the alkoxyl
radical fragmentation reaction.
In this context, it is well-known that cyclopropyl
substituents strongly stabilize carbocations by interaction
of the cyclopropyl C-C bonding orbitals with the vacant
p orbital of the carbocation.31 These orbitals, because of
angle strain, have an abnormal amount of p character,
and thus when the cyclopropyl group adopts a bisected
conformation (with respect to the nodal plane of the
vacant p orbital),32 effective overlap may occur. Accord-
ingly, the efficacy of cyclopropane as a neighboring group
in solvolytic reactions is well documented.34
measured for 1• (Table 2), the partial rate constants for
C-methyl bond cleavage assisted by the presence of an
adjacent cyclopropyl group (kMe(ass) ) 3.1 × 106 s-1) and
for unassisted C-cyclopropyl bond cleavage (kcyp(unass)
e 3.4 × 104 s-1) can be derived (Scheme 4).39,40
)
The rate constants measured for 2• and 3• (Table 2),
corrected for the statistical factor, provide instead the
assisted rate constant for C-cyclopropyl bond cleavage
(kcyp(ass) ) 6.5 × 105 s-1) and the unassisted rate constant
for C-methyl bond cleavage (kMe(unass) 3.6 × 105 s-1),
respectively (Scheme 5).
Along this line, it is reasonable to propose that in the
transition state for C-methyl â-scission of 1• and C-
cyclopropyl â-scission of 2•, formation of the CdO double
bond is assisted by overlap with the C-C bonding orbitals
of the cyclopropane ring. An analogous assistance is
clearly not possible for C-cyclopropyl â-scission in 1• and
for C-methyl â-scission in the cumyloxyl radical 3•.
Quite importantly, cyclopropyl participation involving
overlap between the C-C bonding orbitals of the cyclo-
propane ring and developing p orbitals has been also
proposed in other reactions, such as for example, the
thermolysis of azo compounds35 and the decarbonylation
of tricyclic alkenones.36,37
Since C-methyl cleavage in 1• and C-cyclopropyl cleav-
age in 2•, as well as C-cyclopropyl cleavage in 1• and
C-methyl cleavage in 3•, lead to the formation of the same
ketone (cyclopropyl phenyl ketone and acetophenone,
respectively), comparison between the pertinent rate
constants provides the kinetic effect of the stability of
the alkyl radical on the â-scission reaction.
In both the assisted and unassisted reaction, C-methyl
cleavage is favored over C-cyclopropyl cleavage, in line
with the slightly higher stability of the methyl radical
as compared to the cyclopropyl one.4,5 However, the
reactivity ratio decreases on going from the unassisted
(kMe(unass)/kcyp(unass) ) g10.6) to the assisted reaction
(kMe(ass)/kcyp(ass) ) 4.8), clearly indicating that, as compared
to the methyl radical, formation of the cyclopropyl radical
benefits to a greater extent from cyclopropyl assistance.
The same conclusion arises from the comparison between
the assisted and unassisted â-scission reactions leading
By combining the product ratio determined for the
reaction of 1a (Table 1) with the decay rate constant
(31) See, for example: Olah, G. A.; Prakash Reddy, V.; Surya
Prakash, G. K. Chem. Rev. 1992, 92, 69-95 and references therein.
(32) The preference for the bisected conformation over the perpen-
dicular one in cyclopropyl substituted cations has been confirmed in
several studies (see, for example, ref 33).
(33) Childs, R. F.; Faggiani, R.; Lock, C. J. L.; Mahendran, M.;
Zweep, S. D. J. Am. Chem. Soc. 1986, 108, 1692-1693. Buss, V.;
Gleiter, R.; Schleyer, P. v. R. J. Am. Chem. Soc. 1971, 93, 3927-3933.
Baldwin, J. E.; Fogelsong, W. D. J. Am. Chem. Soc. 1968, 90, 4303-
4310.
(34) See, for example: Roberts, D. D. J. Org. Chem. 1999, 64, 1341-
1343. Lowry, T. H.; Richardson, K. S. Mechanism and Theory in
Organic Chemistry, 3rd ed.; Harper & Row: New York, 1987, and
references therein.
(35) Schmidt, H.; Schweig, A.; Trost, B. M.; Neubold, H. B.; Scudder,
P. H. J. Am. Chem. Soc. 1974, 96, 622-624. Allred, E. L.; Voorhees,
K. J. J. Am. Chem. Soc. 1973, 95, 620-621. Allred, E. L.; Johnson, A.
L. J. Am. Chem. Soc. 1971, 93, 1300-1301.
(36) Battiste, M. A.; Visnick, M. Tetrahedron Lett. 1978, 19, 4771-
4774. Halton, B.; Battiste, M. A.; Rehberg, R.; Deyrup, C. L.; Brennan,
M. E. J. Am. Chem. Soc. 1967, 89, 5964-5965.
(37) Cyclopropyl participation has been also proposed in the ring-
opening of tricyclic cyclopropyl ketones, rationalized in terms of the
overlap between the scissile cyclopropane C-C bond and the adjacent
p orbital (see ref 38).
(38) Monti, S. A. J. Org. Chem. 1970, 35, 380-383. Monti, S. A.;
Bucheck, D. J.; Shepard, J. C. J. Org. Chem. 1969, 34, 3080-3084.
(39) The product ratio for 1a has been determined in CH2Cl2, while
the decay rate constant for 1• has been measured in MeCN. In this
respect, no significant difference in product ratio should be observed
on going from CH2Cl2 to MeCN as the solvent is expected to behave in
the same way on the two competing fragmentation reactions of the
alkoxyl radical.
(40) The decay rate constants for radicals 1•-3• have been also
measured in CH2Cl2 as 3.3 × 106, 1.4 × 106, and 1.3 × 106 s-1
,
respectively; values which are slightly higher than the corresponding
ones measured in the more polar solvent MeCN (see Table 2). However,
as the rate constants for â-scission of alkoxyl radicals are known to
decrease by decreasing solvent polarity (see, for example, refs 2, 11a,
and 29), the observation of an increase in rate on going from MeCN to
CH2Cl2 can be rationalized in terms of the contribution to the decay
rate constant of a hydrogen abstraction reaction from the solvent CH2-
Cl2.
6824 J. Org. Chem., Vol. 70, No. 17, 2005

Reaction of Tertiary Arylcarbinyloxyl Radicals
to the same radical, where, neglecting differences in
stability between the product ketones, cyclopropyl as-
sistance is relatively more important in C-cyclopropyl
than in C-methyl cleavage (kcyp(ass)/kcyp(unass) ) g19.1 and
(kMe(ass)/kMe(unass) ) 8.6).
basis of the c-C4H7-H and CH3-H BDEs: 96.8 and 105.0
kcal mol-1, respectively).4 Differences in radical stability
can also be invoked to explain the fragmentation regi-
oselectivity observed with 5•, as the c-C4H7-H BDE is
reported to be lower than the CH3CH2-H one (96.8 and
100.5 kcal mol-1, respectively).4 An observation that is
also supported by kinetic data where, however, even
though the high reactivity of 4• did not allow the
determination of its fragmentation rate constant but only
of a lower limit (kâ g 8 × 107 s-1 at T ) -10 °C, see Table
2), the comparison of this limit with the rate constant
measured for C-ethyl â-scission in the 2-(4-methylphe-
nyl)-2-butoxyl radical (kâ ) 1.1 × 107 s-1 at T ) - 15
°C)24 indicates that in tertiary arylcarbinyloxyl radicals
cleavage of the C-cyclobutyl bond is favored over that of
the C-ethyl one.
In conclusion, by means of product and time-resolved
kinetic studies, convincing evidence has been provided
for cyclopropyl participation in the â-scission reactions
of tertiary arylcarbinyloxyl radicals bearing R-cyclopropyl
groups. It is proposed that â-scission proceeds through a
transition state where formation of the CdO double bond
is assisted by overlap with the C-C bonding orbitals of
the cyclopropane ring. The kinetic data clearly show that
both C-methyl and C-cyclopropyl bond cleavage are
significantly accelerated by cyclopropyl assistance which,
moreover, is relatively more important in the latter than
in the former â-scission reaction. In particular, C-
cyclopropyl bond cleavage, which is expected to be a
relatively slow reaction, when assisted by the presence
of an additonal cyclopropyl group becomes faster than
C-methyl cleavage in the cumyloxyl radical. With tertiary
arylcarbinyloxyl radicals bearing R-cyclobutyl groups, the
fragmentation regioselectivity is essentially governed by
the stability of the radical formed by â-scission, with
cleavage of the C-cyclobutyl bond being favored over that
of the C-ethyl one and being the exclusive fragmentation
pathway when in competition with C-methyl bond cleav-
age.
As observed previously for the cumyloxyl radical 3•,2,11a
also the rate constants for â-scission of radicals 1• and 2•
increase by increasing solvent polarity. Solvent effects
on the rate constant for â-scission of the cumyloxyl
radical have been studied in detail,11a and the increase
in rate observed with increasing solvent polarity has been
interpreted in terms of the stabilization of the transition
state for â-scission through increased solvation of the
incipient carbonyl product. Quite interestingly, a fairly
good correlation between log kâ and the Dimroth-Rei-
chardt ETN parameter has been reported for the â-scission
reactions of the cumyloxyl radical 3• in several solvents.11a
In this context, also the kâ value measured for 3• in CF3-
CH2OH (characterized by ET ) 0.898)41 fits very well
N
into the log kâ versus ET correlation.42
N
Going from MeCN to MeCN/H2O 1:1 and CF3CH2OH
results in a red-shift in λmax (vis) of approximately 30 nm
for radicals 1•-3•.2 The observation of this effect in H2O
and CF3CH2OH but not in other protic solvents such as
AcOH, tBuOH, and dipolar aprotic ones such as MeCN
H
requires some comment. On the basis of Abraham’s R2
parameters,43 CF3CH2OH and AcOH are characterized
by similar hydrogen bond acidities (R2 ) 0.567 and
H
0.550, respectively), values that are significantly higher
than that determined for H2O (R2H ) 0.353). Clearly, this
solvent property alone cannot be invoked to explain the
observed solvent effect on λmax (vis). On the other hand,
both H2O and CF3CH2OH are characterized by ETN values
that are significantly higher than those measured for
MeCN, tBuOH, and AcOH,44 suggesting that the combi-
nation of different solute/solvent interaction forces (well
represented by this empirical parameter) contributes to
this effect.45
The reactions of 1-cyclobutyl-1-phenylethanol (4a) and
R,R-dicyclobutylphenylmethanol (6a) led to the formation
of acetophenone and cyclobutyl phenyl ketone, respec-
tively, as the exclusive reaction product. With 1-cyclobu-
tyl-1-phenylpropanol (5a), formation of propiophenone
and cyclobutyl phenyl ketone in a 2:1 ratio was instead
observed. Clearly in 4•, cleavage of the C-cyclobutyl bond
is strongly favored over that of the C-methyl one, as
expected on the basis of the large difference in stability
between the cyclobutyl and the methyl radical (on the
Experimental Section
Materials. Spectroscopic grade MeCN and CF3CH2OH were
used as received. CH2Cl2 was purified prior to use by column
chromatography over basic alumina. (Diacetoxy)iodobenzene
(DIB), iodine, R,R-dicyclopropylphenylmethanol (2a), and 2-phe-
nylpropan-2-ol (3a) were of the highest commercial quality
available and were used as received. Dicumyl peroxide was
recrystallized twice from methanol.
Details of the synthesis of arylalkanols 1a, 4a-6a, and of
the tert-butyl arylalkyl peroxides (1p, 2p, and 4p) are given
in the Supporting Information. The purity of the arylalkanols
(1a-6a) employed in the product studies was always g99%.
Product Analysis. All the reactions were carried out under
an argon atmosphere. Irradiations were performed with visible
light (4-10 × 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 T ) 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 (22 mM) and iodine (10 mM) was
irradiated for times ranging between 1 and 15 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 were dried over
(41) Reichardt, C. Solvents and Solvent Effects in Organic Chemistry,
3rd ed.; Wiley-VCH: Weinheim, Germany, 2003.
N
(42) As it has been pointed out that the ET parameter is mainly
related to the solvent anion solvating ability (see ref 41, pp 462-463),
N
the observation of a log kâ versus ET correlation reasonably reflects
the development of negative charge on the oxygen atom of the forming
carbonyl compound. We thank a referee for this suggestion.
(43) Abraham, M. H.; Grellier, P. L.; Prior, D. V.; Duce, P. P.; Morris,
J. J.; Taylor, P. J. J. Chem. Soc., Perkin Trans. 2 1989, 699-711.
(44) (CH3)3COH, MeCN, AcOH, CF3CH2OH, and H2O are character-
ized by the following ETN values: 0.389, 0.460, 0.648, 0.898, and 1.000,
respectively (see ref 41).
(45) In this connection, it is important to point out that in water
the red-shift in the cumyloxyl radical λmax (vis) is expected to be
significantly greater than 30 nm. See, for example, ref 46 for a
discussion on the effect of water on the 4-methoxycumyloxyl radical
λmax (vis).
(46) Baciocchi, E.; Bietti, M.; Lanzalunga, O.; Steenken, S. J. Am.
Chem. Soc. 1998, 120, 11516-11517.
J. Org. Chem, Vol. 70, No. 17, 2005 6825

Bietti et al.
anhydrous sodium sulfate. Reaction products were identified
by GC (comparison with authentic samples) and GC-MS and
were quantitatively determined by GC using bibenzyl as
internal standard. Excellent mass balances (g95%) were
obtained in all experiments. Blank experiments performed in
the absence of irradiation, keeping the solution in the dark,
showed the formation of negligible amounts of reaction prod-
ucts.
Laser Flash Photolysis Studies. Laser flash photolysis
experiments were carried out with a laser kinetic spectrometer
using the fourth harmonic (266 nm) of a Q-switched Nd:YAG
laser. The laser energy was adjusted to e10 mJ/pulse by the
use of the appropriate filter. A 3-mL Suprasil quartz cell (10
mm × 10 mm) was used for all experiments. Argon- or oxygen-
saturated MeCN, MeCN/H2O 1:1 (v/v), or CF3CH2OH solutions
of peroxides 1p, 2p, and dicumyl peroxide (between 5.0 × 10-3
and 1.0 × 10-2 M, A266 ≈ 1.0) were used. All the experiments
were carried out at T ) 25 ( 0.5 °C under magnetic stirring.
LFP of peroxide 4p (1.0 × 10-2 M) was instead carried out in
argon-saturated MeCN solution at T ) -10 °C. Rate constants
were obtained by averaging four to eight values each consisting
of the average of at least three laser shots and were reproduc-
ible to within 10%.
Acknowledgment. Financial support from the Min-
istero dell’Istruzione dell’Universita` e della Ricerca
(MIUR) is gratefully acknowledged. We thank Prof.
Osvaldo Lanzalunga for assistance in the synthesis and
characterization of the peroxides, Prof. Basilio Pispisa
and Dr. Lorenzo Stella for the use and assistance in the
use of a LFP equipment, and Mr. Luigi Gastaldo (Helios
Italquartz srl) for providing us with a photochemical
reactor.
Supporting Information Available: Details on the syn-
thesis and characterization of the precursor arylalkanols and
tert-butyl arylalkyl peroxides. Time-resolved absorption spec-
tra observed after LFP of peroxide 2p. This material is
available free of charge via the Internet at http://pubs.acs.org.
JO050883I
6826 J. Org. Chem., Vol. 70, No. 17, 2005