Ingold-Fischer "Persistent Radical Effect", Solvent Effect, and Metal Salt Oxidation of Carbon-Centered Radicals in the Synthesis of Mixed Peroxides from tert-Butyl Hydroperoxide

Article abstract of DOI:10.1021/jo9700944

Mixed peroxides are formed from tert-butyl hydroperoxide (TBH), tert-butyl peroxalate (TBP), and a variety of substrates (p-cresol, cyclohexene, styrene, α-methylstyrene, acrylonitrile, 2-methylcyclohexanone). Also, the oxidation of THF in the presence of acrylonitrile under the same conditions gives the mixed peroxide, generated by addition of the tetrahydrofuranyl radical to the double bond and the cross-coupling of the radical adduct with the tert-butylperoxyl radical. Similarly, benzoyl peroxide, TBH, and acrylonitrile give the mixed peroxide by oxidative arylation of the double bond. Paradoxically, TBH acts as effective inhibitor of the polymerization of vinyl monomers (acrylonitrile, styrene). An overall kinetic evaluation suggests that the conditions for the Ingold-Fischer "persistent radical effect", characterized by the simultaneous formation of a persistent and a transient radical, are fulfilled in all cases. The reactions are strongly affected by solvents, which form hydrogen bonds with TBH. Catalytic amounts of Cu(II) and Fe(III) salts influence the selectivity; the possibility that the mixed peroxides can also be generated by metal salt oxidation of carbon-centered radicals is discussed.

Full text of DOI:10.1021/jo9700944

J . Org. Chem. 1997, 62, 3849-3857
3849
In gold -F isch er “P er sisten t Ra d ica l Effect”, Solven t Effect, a n d
Meta l Sa lt Oxid a tion of Ca r bon -Cen ter ed Ra d ica ls in th e
Syn th esis of Mixed P er oxid es fr om ter t-Bu tyl Hyd r op er oxid e
Anna Bravo, Hans-Rene´ Bjørsvik, Francesca Fontana, Lucia Liguori, and Francesco Minisci*
Dipartimento di Chimica del Politecnico, via Mancinelli 7, I-20131 Milano, Italy
Received J anuary 17, 1997X
Mixed peroxides are formed from tert-butyl hydroperoxide (TBH), tert-butyl peroxalate (TBP), and
a variety of substrates (p-cresol, cyclohexene, styrene, R-methylstyrene, acrylonitrile, 2-methylcy-
clohexanone). Also, the oxidation of THF in the presence of acrylonitrile under the same conditions
gives the mixed peroxide, generated by addition of the tetrahydrofuranyl radical to the double
bond and the cross-coupling of the radical adduct with the tert-butylperoxyl radical. Similarly,
benzoyl peroxide, TBH, and acrylonitrile give the mixed peroxide by oxidative arylation of the double
bond. Paradoxically, TBH acts as effective inhibitor of the polymerization of vinyl monomers
(acrylonitrile, styrene). An overall kinetic evaluation suggests that the conditions for the Ingold-
Fischer “persistent radical effect”, characterized by the simultaneous formation of a persistent and
a transient radical, are fulfilled in all cases. The reactions are strongly affected by solvents, which
form hydrogen bonds with TBH. Catalytic amounts of Cu(II) and Fe(III) salts influence the
selectivity; the possibility that the mixed peroxides can also be generated by metal salt oxidation
of carbon-centered radicals is discussed.
In tr od u ction
of these is transient10 while the other is persistent:10 the
main reaction product is formed by cross-coupling be-
tween transient (Rt) and persistent (Rp) radicals (eq 1).
The formation of mixed peroxides has been observed
in a variety of reactions of hydroperoxides catalyzed by
metal salts complexes (copper,1,2 manganese,1 iron,3,4
including Gif5 and metalloporphyrin6 catalysis, etc.). The
interest of these reactions is also related to the fact that
hydroperoxides are intermediates in one of the most
important reactions of organic chemistry, the autoxida-
tion, in which transition-metal salts often play a key
catalytic role. Occasionally, the prevailing formation of
mixed peroxides has been observed in reactions of hy-
droperoxides also in the absence of transition-metal salts,
but in the presence of different sources of radicals, such
as acetyl peroxide7 (a clean source of methyl radicals) or
di-tert-butyl hyponitrite8 (a convenient source of tert-
butoxyl radical at moderate temperature). In these latter
cases, the only reasonable mechanism for the mixed
peroxide formation is based on the Ingold-Fischer
“persistent radical effect”,4,8,9 characterized by the fact
that two radicals are generated at similar rates and one
•
Rt + Rp• a RtRp
(1)
In preliminary communications,2,5,6,11 we have reported
new synthetic approaches to mixed peroxides from hy-
droperoxides concerning phenols, simple and conjugated
alkenes, and alkylbenzenes, always catalyzed by metal-
salt complexes, and we have explained their formation
by a ligand-transfer oxidation of carbon-centered radicals
according to the original Kochi interpretation12 (eq 2).
R• + MOOBu-t f ROOBu-t + M
M ) Cu(I), Fe(II), Mn(II)
(2)
In this paper, we describe new methods of synthesis
of mixed peroxides, catalyzed by metal salt complexes,
further new synthetic developments in the absence of
metal salt catalysis, a macroscopic solvent effect, and a
comparison between the catalytic and noncatalytic routes
in an attempt to define whether the metal salt oxidation
of the carbon-centered radical (eq 2) is a possible mech-
anism, considering that more recently13 Kochi has seem-
ingly abandoned the idea that mixed peroxides are
formed via a ligand-transfer process (eq 2).
X Abstract published in Advance ACS Abstracts, May 1, 1997.
(1) Kharasch, M. S.; Fono, A. J . Org.Chem. 1958, 23, 324; 1959, 24,
72.
(2) Araneo, S.; Fontana, F.; Minisci, F.; Recupero, F.; Serri, A. J .
Chem. Soc., Chem. Commun. 1995, 1399.
(3) Leising, R. A.; Norman, R. E.; Que, L., J r. Inorg. Chem. 1990,
29, 2553. Leising, R. A.; Zang, Y.; Que, L., J r. J . Am. Chem. Soc. 1991,
113, 8555. Kojima, T.; Leising, R. A.; Yan, S.; Que, L., J r. J . Am. Chem.
Soc. 1993, 115, 11328. Leising, R. A.; Kim, J .; Pe´rez, M. A.; Que, L.,
J r. J . Am. Chem. Soc. 1993, 115, 9524.
(4) Arends, I. W. C. E.; Ingold, K. U.; Wayner, D. D. M. J . Am. Chem.
Soc. 1995, 117, 4710.
Resu lts a n d Discu ssion
(5) Minisci, F.; Fontana, F.; Araneo, S.; Recupero, F. J . Chem. Soc.,
Chem. Commun. 1994, 1823.
(6) Minisci, F.; Fontana, F.; Araneo, S.; Recupero, F.; Banfi, S.; Quici,
S. J . Am. Chem. Soc. 1995, 117, 226.
(7) Kharasch, M. S.; Fono, A.; Nudenberg, W. J . Org. Chem. 1950,
15, 753.
Mixed P er oxid e fr om p-Cr esol. We have recently
reported, in a preliminary communication,11 the oxidation
of phenols bearing unsubstituted p-positions to p-quino-
(8) MacFaul, P. A.; Arends, I. W. C. E.; Ingold, K. U.; Wayner, D.
D. M. J . Chem. Soc., Perkin Trans. 2, in press. We thank K. U. Ingold
for the personal communication and the useful suggestions about the
Ingold-Fischer effect.
(9) Fischer, H. J . Am. Chem. Soc. 1986, 108, 3925. Daikh, B. E.;
Finke, R. G. J . Am. Chem. Soc. 1992, 114, 2938.
(10) Griller, D.; Ingold, K. U. Acc. Chem. Res. 1976, 9, 13.
(11) Bravo, A.; Fontana, F.; Minisci, F. Chem. Lett. 1996, 401.
(12) Kochi, J . K. Tetrahedron 1962, 18, 483. Kochi, J . K.; Mains, H.
E. J . Org. Chem. 1965, 30, 1862.
(13) Srinavasan, K.; Perrier, S.; Kochi, J . K. J . Mol. Catal. 1986,
38, 297.
S0022-3263(97)00094-7 CCC: $14.00 © 1997 American Chemical Society

3850 J . Org. Chem., Vol. 62, No. 12, 1997
Ta ble 1. Oxid a tion of p- a n d m -Cr esols by TBH a n d TBP
Bravo et al.
cresol
(mmol)
TBH
(mmol)
TBP
(mmol)
metal salt
(mmol)
solvent
(mL)
conversna
(%)a
T (°C)
1 (%)
2 + 3 (%)
para (1)
para (5)
para (1)
para (5)
para (5)
para (5)b
para (5)b
para (25)b
para (5)b
para (5)b
para (5)
para (5)
meta (5)b
1
1
1
1
1
1
benzene (5)
benzene (5)
CH2Cl2 (5)
30
37
13
45
45
45
45
45
20
70
20
20
20
20
50
20
78.3
21.8
89.3
21.7
78.2
10.7
10
benzene (10)
benzene (5)
benzene (10)
benzene (10)
benzene (10)
benzene (10)
benzene (10)
benzene (10)
benzene (10)
benzene (10)
1
16
100
98
88
66
38
18
64
82
100
12.6
17.8
89.4
13
10
10
10
10
10
10
10
10
FeTPPCl (10-3
FeTPPCl (10-3
FeTPPCl (10-3
FeTPPCl (10-3
)
87.4
82.2
10.6
87
89
100
100
)
)
)
MnTDClPPCl (10-3
Cu(OAc)2 (0.5)
)
11
Cu(OAc)2 (0.5)
FeTPPCl (10-3
)
2-methyl-p-benzo-quinone 4 (80%)
a
b
Conversion based on peroxides. 0.3 mmol of pyridine are added.
nes by TBH, catalyzed by the metalloporphyrins Fe-
tetraphenylporphyrin chloride (FeTPPCl) and Mn-tetrakis-
(dichlorophenyl)porphyrin chloride (MnTDClPPCl) (eq 3).
TBH has been evaluated15 at 1.3 × 108 M-1 s-1 in benzene
at 298 K) giving the tert-butylperoxyl radical (eq 6).
RO• + t-BuOOH f t-BuOO• + ROH
(6)
O
4
OH
+ 2 t-BuOOH
+ 2 t-BuOH + H2O
(3)
(4)
On the other hand, the rate constant for hydrogen
abstraction from phenol by cumyloxyl radical has been
evaluated15 at 2.8 × 108 M-1 s-1 in benzene at 298 K (eq
7).
O
OH
O
RO• + ArOH f ArO•+ ROH
(7)
+ 2 t-BuOOH
+ t-BuOH + H2O
5
OOBu-t
1
By using a 2:1 ratio (stoichiometric ratio of eqs 3 and
4) between TBH and phenol, one of the two conditions
for the Ingold-Fischer “persistent radical effect” is
plainly fulfilled (the peroxyl radical 4 and the phenoxyl
radical 5 are generated at the same rate). Moreover, the
bimolecular self-reactions of the phenoxyl radical leading
to the Pummerer ketone 2 and the diphenol 3 (eqs 8 and
9) are very fast reactions (transient radicals),16 whereas
the bimolecular self-reaction of tert-butylperoxyl radical
is a relatively slow radical-radical reaction (persistent
radical) (a value of 4 × 102 M-1 s-1 has been evaluated
for eq 10 in CCl4 at 303 K).17
In order to understand the mechanism of this oxida-
tion, we have particularly investigated the oxidation of
p-cresol under the same conditions in which phenols with
free p-positions are converted into benzoquinones.
The main reaction product is the mixed peroxide 1 with
a stoichiometric ratio of p-cresol and TBH (1:2, eqs 3 and
4) (Table 1). The Pummerer ketone 2 and the diphenol
3 are byproducts (Table 1).
O
O
OH
OH
O
O
O
O
+
2
3
Clearly, compounds 1-3 are formed by a free-radical
mechanism involving the phenoxyl radical intermediate.
Recently, we6,14 have reported evidence concerning the
involvement of t-BuO• and t-BuOO• radicals in the
oxidation and halogenations of alkanes, alkenes, and
alkylbenzenes by TBH, catalyzed by Fe(III) and Mn(III)
porphyrins. The tert-butoxyl radical would be generated
by the oxidation of the metalloporphyrin (PFe(III)) with
TBH (eq 5).
O
O
OH
O
(8)
2
O
O
O
OH
OH
O
(9)
+
PFe(III) + t-BuOOH f
PFe(IV) + t-BuO• + OH- (5)
3
2 t-BuOO•
t-BuOOBu-t + O2
(10)
Alkoxyl radicals react very rapidly with TBH (i.e., the
rate constant of H-abstraction by cumyloxyl radical from
(15) Avila, D. V.; Ingold, K. U.; Lusztyk, J .; Green, W. A.; Procopio,
D. R. J . Am. Chem. Soc. 1995, 117, 2929.
(16) Howard, A.; Scaiano, J . C. Landolt-Bo¨rnstein; 1984; Vol. 13, p
143.
(14) Minisci, F.; Fontana, F.; Zhao, L.; Banfi, S.; Quici, S. Tetrahe-
dron Lett. 1994, 35, 8033.

Synthesis of Mixed Peroxides from TBH
J . Org. Chem., Vol. 62, No. 12, 1997 3851
Thus, the other condition for a radical cross-coupling
(eq 11) is also fulfilled.
(ii) The Fe(III) complex could oxidize TBH to the
peroxyl radical, thus generating a Fe(III)/Fe(II) redox
chain (eqs 15 and 16).
O
PFe(III) + HOOBu-t f
+ •OOBu-t
1
(11)
PFe(II) + •OOBu-t + H+ (15)
PFe(II) + HOOBu-t f
Actually, the dimers 2 and 3 of the phenoxyl radical
become the main reaction products by increasing the
phenol:TBH ratio, whereas the mixed peroxide 1 is a
minor product of the reaction (Table 1).
In order to obtain direct evidence concerning the
mechanism of reaction 11, we have carried out the same
reaction with TBH in the absence of metalloporphyrins,
but using TBP (di-tert-butyl peroxyoxalate) as a clean
source of tert-butoxyl radical at moderate temperature
(40 °C) (eq 12).
PFe(III) + •OBu-t + OH- (16)
A further possible mechanism, which could superim-
pose to the cross-coupling, is related to the original Kochi
interpretation of the phenoxyl radical oxidation (eq 17),
with the generation of a different Fe(IV)/Fe(III) redox
chain.
O
t-BuOOCOCOOOBu-t f 2 t-BuO• + 2 CO2 (12)
1
+ PFe(IV)OOBu-t
(17)
The results were similar to those obtained for the
reaction catalyzed by metalloporphyrins: 1-3 were the
reaction products, but the ratio between the mixed
peroxide 1 and the dimers of the phenoxyl radical (2 and
3) is somewhat lower compared to the one observed in
the catalyzed reaction. Again, the amount of phenoxyl
radical, and consequently of its dimers (2 and 3), is
increased by increasing the p-cresol/TBH ratio (Table 1).
The formation of 2 and 3 by oxidation of p-cresol with
a variety of monoelectron oxidants (K3Fe(CN)6, Ce(IV),
Mn(III), Ag(I)) is well-known,18 and the ratio between the
ketone 2 and the diphenol 3 depends on temperature,
solvent, concentration, and nature of the oxidant. This
variation has been attributed to the reversibility of the
coupling step (eqs 8 and 9).
The results obtained with TBH and TBP represent
unequivocal evidence of the Ingold-Fischer “persistent
radical effect”; the only reasonable explanation for the
mixed peroxide formation is eq 11. By mere chance, the
stoichiometric ratio between phenol and TBH (eqs 3 and
4) corresponds to the same rates of eqs 6 and 7, thus
fulfilling the conditions for the Ingold-Fischer effect; the
increase of this ratio is reflected by an increase of the
rate of eq 7 and of the amount of phenoxyl radical dimers
2 and 3.
In order to have further indications about the catalytic
mechanism by metal salts, we have investigated the
catalysis by Cu(OAc)2 in the oxidation of p-cresol under
the same conditions utilized with metalloporphyrins. The
reaction is slower at room temperature, and it requires
a much higher amount of catalyst (5-10%), but it is
completely selective: the mixed peroxide 1 is the only
reaction product. By increasing the temperature to 50
°C, the oxidation is obviously faster, but always quite
selective. This represents a new type of catalysis in the
oxidation of phenols by TBH.
A simple explanation of this result could be related to
a superimposition of the phenoxyl radical oxidation (eq
18), according to the original Kochi mechanism, to the
cross-coupling of eq 11. This way, all of the phenoxyl
radicals, which exceed the cross-coupling, are scavenged
by Cu(II) salt, thus preventing the formation of 2
and 3; the Cu(I) salt is oxidized by TBH, generating tert-
butoxyl radical and giving a Cu(II)/Cu(I) redox chain
(eq 19).
O
+ Cu(II) + t-BuOOH
1 + Cu(I) + H+
(18)
If the cross-coupling (eq 11) is the only mechanism for
the mixed peroxide formation, then a fundamental ques-
tion arises: considering that the ratio between the
catalyst and TBH is 1:104, what is the operating mech-
anism for the metalloporphyrin catalysis? We can envis-
age two possible explanations:
Cu(I) + t-BuOOH
Cu(II) + t-BuO• + OH–
(19)
However, the cross-coupling (eq 11) could still be the
only mechanism for the formation of 1 also in this case,
if we admit that the Cu(II) catalysis is different from the
one provided by metalloporphyrins and that the copper
salt is more effective in generating a higher steady-state
concentration of peroxyl radicals. It is anyway quite
difficult to envisage a redox chain leading to the peroxyl
radical, in which Cu(I) and the tert-butoxyl radical are
not involved.
In any case, these results clarify the mechanism of
p-benzoquinone formation when the para-position of
phenol is free (i.e., oxidation of m-cresol): it occurs by
t-BuOH elimination from the mixed peroxide. The reac-
tion is more selective in the presence of small amounts
of pyridine, which catalyzes this elimination (eq 20).
(i) The Fe(IV) complex generated in eq 5 could abstract
hydrogen from both the phenol and TBH (eqs 13 and 14)
PFe(IV)dO + HOAr f PFe(III)OH + •OAr (13)
PFe(IV)dO + HOOBu-t f
PFe(III)OH + •OOBu-t (14)
with similar rates, thus regenerating the Fe(III) complex
and giving rise, consequently, to a Fe(IV)/Fe(III) redox
chain.
(17) Howard, J . A.; Ingold, K. U. Can. J . Chem. 1969, 47, 3797.
Adamic, K.; Howard, J . A.; Ingold, K. U. Can. J . Chem. 1969, 47, 3803.
(18) Anderson, R. A.; Dalgleish, D. T.; Nonhebel, D. C.; Pauson, P.
L. J . Chem. Res., Synop. 1977, 12.

3852 J . Org. Chem., Vol. 62, No. 12, 1997
Bravo et al.
Ta ble 2. Solven t a n d Con cen tr a tion Effects in th e
O
Oxid a tion of Cycloh exen e by TBH (1 m m ol) a n d TBP
(0.5 m m ol)a
Py +
PyH+ +
cyclohexene (mmol)
solvent (mL)
5 (%)
6 (%)
H
OOBu-t
5
10
20
5b
5c
5
benzene (6)
benzene (6)
benzene (6)
benzene (6)
benzene (6)
CH2Cl2 (6)
t-BuOH (6)
benzene (4)
pyridine (2)
benzene (3)
pyridine (3)
benzene (3)
pyridine (3)
benzene (2)
pyridine (4)
pyridine (6)
pyridine (10)
100
100
97
O–
O
O
O
3
+ t-BuO–
100
(20)
100
68
86
–
5
5
32
14
OOBu-t
OOBu-t
Mixed P er oxid e fr om Cycloh exen e. The synthesis
of mixed peroxides from alkenes and hydroperoxides by
the Kharasch reaction is well-known.1 The original Kochi
interpretation involves a Cu(I)/Cu(II) redox chain with
formation of alkoxyl radicals (eq 21), allylic hydrogen
abstraction (eq 22), and a ligand transfer oxidation of the
allyl radical (eq 23).
5
10
5
66
57
44
34
43
56
5
20
24
15
76
85
a
The conversions of cyclohexene, based on the peroxides, are
ROOH + Cu(I)
RO• + Cu(II)OH
in the range 40-50%. In the absence of TBP. c In the absence of
TBH.
b
(21)
•
RO• +
ROH +
(22)
(23)
“persistent radical effect” is fulfilled. In benzene solution
at room temperature the rate constant for eq 6 has been
evaluated15 at 1.3 × 108 M-1 s-1, whereas those for eqs
22 and 24 are, respectively,20 5.7 × 106 M-1 s-1 and 3.7
× 10-1 M-1 s-1. Thus, a persistent radical (t-BuOO•) and
a transient radical (cyclohexenyl) are simultaneously
formed, even with a 20:1 ratio of cyclohexene/TBH.
Reaction 6 is strongly affected by the nature of the
solvent, the rate constant being 6.7 × 106 M-1 s-1 at
room15 temperature in t-BuOH solution. Reactions 22
and 24 are substantially unaffected by the solvent, so that
reaction 22 should become competitive with reaction 6
in t-BuOH solution. In agreement with these rate
constants, the reaction in t-BuOH solution gives the
mixed peroxide 5 as the main reaction product (68%) but
also forms a significant amount of dicyclohexenyl (6)
(32%) by using the same ratio cyclohexene/TBH, 5:1,
utilized in benzene solution. Under these conditions,
reaction 22 is somewhat faster than reaction 6, which
explains the prevailing formation of cyclohexenyl radical
and its homocoupling (eq 26), in competition with the
cross-coupling (eq 25).
OOR
•
+ Cu(II)OOR
+ Cu(I)
5
Recently, we have reported6,19 that this mechanism is
not consistent with the chemoselectivity of the reaction
and that, at least, eq 22 should be substituted by eqs 6
and 24 in solvents that do not form hydrogen bonds with
hydroperoxides.
•
ROO• +
ROOH +
(24)
However, in a much more recent work, Kochi13 appar-
ently abandoned the idea that mixed peroxides were
formed via ligand transfer. Thus, the same two ques-
tions, discussed above for phenols, also arise with alk-
enes: are the conditions for the Ingold-Fischer effect
fulfilled for the cross-coupling formation of the mixed
peroxide 5 (eq 25)? What is the mechanism for the metal
salt catalysis if eq 23 is not involved?
•
(26)
2
•
t-BuOO• +
(25)
5
6
First, we have investigated the possibility that 5 could
be formed in the absence of metal salt catalysis by using
the same methodology reported above for phenols. Thus,
the decomposition of TBP in benzene solution at 45 °C
in the presence of cyclohexene and TBH cleanly leads to
the mixed peroxide 5; the process is even more selective
than the Kharasch reaction (Table 2). The only reason-
able explanation for this result is the decomposition of
TBP according to eq 12, followed by hydrogen abstraction
from TBH by the tert-butoxyl radical (eq 6) and from
cyclohexene by the tert-butyl peroxyl radical (eq 24), and
finally the cross-coupling between cyclohexenyl and per-
oxyl radicals (eq 25). The known absolute rate constants
for eqs 6, 22, and 24 indicate that the Ingold-Fischer
The effect of the solvent on the rate of eq 6 is due6,15,19
to hydrogen bond formation between the solvent and
TBH. We have recently observed, while discussing the
mechanism of the Gif reaction,6,19 that this solvent effect
is particularly marked in pyridine, which forms strong
hydrogen bonds with TBH (evaluated21 at 8 kcal/mol).
Thus, we have investigated the behavior of cyclohexene
with TBH and TBP in benzene solution in the presence
of pyridine. The ratio 5:6 decreases as the concentration
(20) Howard, A.; Scaiano, J . C. Landolt-Bo¨rnstein; 1984; Vol. 13, p
19. Encina, M. V.; Rivera, M.; Lissi, E. A. J . Polym. Sci., Polym. Chem.
Ed. 1978, 16, 1709. Paul, H.; Small. R. D.; Scaiano, J . C. J . Am. Chem.
Soc. 1978, 100, 4520. Encina, M. V., Scaiano, J . C. J . Am. Chem. Soc.
1981, 103, 6393. Chenier, J . H. B.; Tong, S. B.; Howard, J . A. Can. J .
Chem. 1978, 56, 3047.
(19) Minisci, F.; Fontana, F.; Araneo, S.; Recupero, F.; Zhao, L.
Synlett 1996, 119.
(21) Yablonskii, O. P.; Kobiakov, A. K.; Beolyaev, V. A. Neftekhimiya
1974, 14, 449.

Synthesis of Mixed Peroxides from TBH
J . Org. Chem., Vol. 62, No. 12, 1997 3853
Ta ble 3. Oxid a tion of Cycloh exen e by TBH (1 m m ol)
a n d TBP (0.5 m m ol) in th e P r esen ce of Cu (OAc)2 a n d
F e(OAc)2OHa
Ta ble 4. Oxid a tion of Cycloh exen e by TBH (1 m m ol)
a n d Cop p er Sa ltsa
cyclohexene
(mmol)
copper salt
(mmol)
T
5
6
7
cyclohexene
(mmol)
metal salt
(mmol)
solvent (°C) (%) (%) (%)
solvent 5 (%) 6 (%) 7 (%)
50
5
5
Cu(OAc)2 (0.1) benzene 70 87 13
Cu(OAc)2 (0.05) benzene 45 100
Cu(OAc)2 (0.05) benzene 70 100
5
10
5
10
5
Cu(OAc)2 (0.05)
Cu(OAc)2 (0.05)
Cu(OAc)2 (0.05)
Cu(OAc)2 (0.05)
Cu(OAc)2 (0.2)
Cu(OAc)2 (0.05)
AcOH (0.1)
benzene 100
benzene 100
t-BuOH
t-BuOH
t-BuOH
t-BuOH
77
75
57
68
2
3
11
12
43
32
5
5
20
5
5
5
5
20
5
5
Cu(OAc)2 (0.05) CH3CN
70
93 <1
6
CuSO4 (0.05)
CuSO4 (0.05)
pyridine 70
pyridine 70
96
91
4
9
5
Cu(OAc)2 (0.05) pyridine 45
Cu(OAc)2 (0.2) pyridine 45
80 <1 19
51
93
79
76
66
61
47
5
49
5
5
CuSO4 (0.05)
pyridine
pyridine
pyridine
pyridine
97
75
74
54
70
3
2
3
CuCl (0.01)
t-BuOH
70
70
70
70
70
70
70
70
70
7
5
Cu(OAc)2 (0.05)
Cu(OAc)2 (0.05)
Cu(OAc)2 (0.2)
13
13
46
2
Cu(OAc)2 (0.05) t-BuOH
Cu(OAc)2 (0.05) t-BuOH
Cu(OAc)2 (0.1) t-BuOH
Cu(OAc)2 (0.15) t-BuOH
Cu(OAc)2 (0.2) t-BuOH
Cu(OAc)2 (0.5) t-BuOH
Cu(OAc)2 (0.5) t-BuOH
Cu(OAc)2 (0.1) t-BuOH
AcOH (0.1)
21
19
34
39
53
95
94
51
20
5
5b
Fe(OAc)2OH (0.05) pyridine
18
5
5
20
5
a
The conversions of cyclohexene, based on the peroxides, are
b
in the range 60-70%. 10% of cyclohexenyl alcohol is also formed.
6
49
of pyridine increases; in pyridine solution 6 is largely
prevailing on 5 (Table 2).
5
5
5
Cu(OAc)2 (0.1) t-BuOH
AcOH (0.2)
Cu(OAc)2 (0.1) t-BuOH
AcOH (0.5)
Cu(OAc)2 (0.1) t-BuOH
AcOH (4)
70
70
70
32
5
68
95
99
If we assume that reaction 22 is not substantially
influenced by the solvent,22 we can evaluate from the
results of Table 2 that the rate constant for eq 6 is
roughly four times smaller in pyridine than in t-BuOH,
in agreement with the stronger hydrogen bond. Thus,
the rate constant for reaction 6 in pyridine can be
approximately evaluated at 1.6 × 106 M-1 s-1, compared
with 1.3 × 108 M-1 s-1 and 6.7 × 106 M-1 s-1, respectively,
in benzene15 and t-BuOH.15
<1
a
The conversions of cyclohexene, based on TBH, are in the
range 60-65%.
for the formation of 7 is therefore the oxidation of
cyclohexenyl radical by Cu(II) salt (eq 27).
However, the presence of a catalytic amount of CuSO4
in pyridine solution completely changes the reaction
course: the mixed peroxide 5 is the main reaction product
(97%) and only 3% of the dicyclohexenyl 6 is formed, vs
24% of 5 and 76% of 6 in the absence of CuSO4 under
the same conditions. If CuSO4 is substituted by Cu(OAc)2
under the same conditions, the formation of 6 is again
minimized (2%) and 5 is the main reaction product, but
a significant amount of cyclohexenyl acetate 7 is also
formed. 6 is minimized (3%) in pyridine solution and Cu-
(OAc)2 catalysis even with a large excess of cyclohexene
(20:1 ratio cyclohexene/TBH), whereas in the absence of
Cu(OAc)2 6 is the main reaction product (85%) (Tables 2
and 3). The 5:7 ratio decreases by increasing the amount
of Cu(OAc)2 or of AcOH. The presence of a catalytic
amount of Fe(OAc)2OH has a similar effect in pyridine
solution: 6 becomes a minor reaction product, whereas
5 is the main product, with a smaller amount of cyclo-
hexenol and traces of 7 (Table 3).
The reaction takes place with TBH also in the absence
of TBP, but in the presence of CuSO4, Cu(OAc)2, or CuCl,
and the results are quite similar in t-BuOH, pyridine, or
benzene solutions. In all cases, 6 is formed in traces or
it is absent; 7 is not formed in benzene solution, but it is
an increasing reaction product as the amount of Cu(OAc)2
increases or by increasing the amount of AcOH at low
Cu(II) salt concentration in t-BuOH or pyridine solution
(Table 4).
OAc
•
+ Cu(II) + AcOH
+ Cu(I) + H+ (27)
7
The Cu(I) is rapidly oxidized to Cu(II) by the peroxides,
generating tert-butoxyl radical and a Cu(II)/Cu(I) redox
chain.
The Fe(OAc)2OH catalyst appears to be less effective
and selective compared with Cu(OAc)2 in pyridine solu-
tion. The cyclohexenyl radical is less effectively scav-
enged, and minor amounts of 7 and of cyclohexenol are
formed (eq 28) (in parentheses are the results in the
absence of iron catalyst under the same conditions).
•
+ t-BuOOH + (t-BuOOCO)2 + Fe(OAc)2OH
OAc
OH
(28)
5
+
6
+
+
73% (24)
16% (76)
1% (–)
10% (–)
All of these results indicate that the Ingold-Fischer
“persistent radical effect” is certainly operating in the
formation of the mixed peroxide 5 (eq 25) in the absence
of metal salt catalysis, depending on the ratios between
cyclohexene and TBH and on the nature of the solvent,
but they strongly suggest that 5 can be also formed by
oxidation of the cyclohexenyl radical by Cu(II) or Fe(III)
salt (eqs 28 and 29). On the other hand, if eq 27
represents the mechanism for cyclohexenyl acetate for-
mation, we do not see any reason why a similar reaction
(eq 29) could not occur with TBH, even if clearly reaction
29 appears to be slower than reaction 27.
Cyclohexenyl acetate 7 cannot be formed by a cross-
coupling process: it is difficult to envisage a mechanism
for the acetoxyl radical formation, but above all the fast
23
decarboxylation of acetoxyl radical (∼109 s-1
) prevents
its intermolecular reactions. The reasonable mechanism
(22) Avila, D. V.; Brown, C. E.; Ingold, K. U.; Lusztyk, J . J . Am.
Chem. Soc. 1993, 115, 466.
(23) Hilborn, J . W.; Pincock, J . A. J . Am. Chem. Soc. 1991, 113, 2683.
Pincock, J . A. Acc. Chem. Res. 1997, 30, 43.
•
5 + Cu(I) + H+
(29)
+ Cu(II) + t-BuOOH

3854 J . Org. Chem., Vol. 62, No. 12, 1997
Bravo et al.
Moreover, the redox chains (eqs 30-32 and 33-34),
which can be envisaged with copper salt catalysis, always
involve both alkoxyl and peroxyl radical. This is not
consistent with the results of Tables 3 and 4, if we
exclude the involvement of eq 29 or the analogous eq
32.
9 easily loses t-BuOH, giving the ketonitrile (eq 40).
9 f PhCH2COCN + t-BuOH
(40)
In order to verify if the Cu(II) catalysis is really
necessary for the formation of 9, we have now carried
out the same reaction in benzene solution, but in the
absence of Cu(OAc)2. No polymerization occurs, but 9 is
the main reaction product from acrylonitrile. Biphenyl
and phenyl benzoate are byproducts, clearly arising from
the homolytic phenylation of benzene and the cage
coupling of the phenyl and benzoyloxyl radicals. The
diperoxide 10 is a minor reaction product; its formation
will be discussed later. The paradox of these results is
that the presence of TBH (an usual free-radical initiator,
particularly in association with metal salts) completely
inhibits the free-radical polymerization of acrylonitrile by
benzoyl peroxide either in the presence or in the absence
of copper salt catalysis.
Cu(II) + t-BuOOH
Cu(III) + t-BuOOH
Cu(III) + t-BuO• + OH–
Cu(II) + t-BuOO• + H+
(30)
(31)
•
+ Cu(III) + t-BuOOH
5 + Cu(II) + H+
(32)
Cu(II) + t-BuOOH
Cu(I) + t-BuOOH
Cu(I) + t-BuOO• + H+
Cu(II) + t-BuO• + OH–
(33)
(34)
An open problem concerns the ligand-transfer or
electron-transfer character of the cyclohexenyl oxidation
(eqs 27-29 and 32). The original Kochi mechanism12 was
related to a ligand-transfer process (eq 23). The following
results and discussions would suggest a more likely
“inner-sphere” electron-transfer mechanism.
Again, the only reasonable explanation in the absence
of Cu(II) salt is the Ingold-Fischer effect, leading to the
cross-coupling of the R-cyanoalkyl and peroxyl radicals
(eq 41).
Mixed P er oxid es fr om Acr ylon itr ile. In prelimi-
nary communications, we have reported2,24 how the
copper salt catalysis can substantially modify the free-
radical polymerization of acrylonitrile. When the polym-
erization is carried out by using a large amount of benzoyl
peroxide as radical source in the presence of Cu(OAc)2,
significant amounts of the dinitrile 8 (two stereoisomers)
have been isolated. We have explained24 the formation
of 8, which is not formed in the absence of Cu(OAc)2, by
a reversible intramolecular homolytic aromatic addition
(eqs 35-37); Cu(II) salt determines a fast irreversible
oxidation of the cyclohexadienyl radical (eq 38), shifting
the equilibrium of eq 37 to the right.
PhCH2C˙ H(CN) + t-BuOO• f
PhCH2CH(CN)OOBu-t (41)
9
The peroxyl radical would be generated according to
eq 42.
PhCOO• (Ph•) + HOOBu-t f
PhCOOH (PhH) + t-BuOO• (42)
We expect reaction 42 to be very fast and to compete
with the decarboxylation of benzoylperoxyl radical on the
basis of fast reaction 6 (∼108 M-1 s-1 in benzene),15
mainly due to enthalpic effects (bond energies 88, 103,
and 110 kcal/mol, respectively, for ROOH, ROH, and
ArCOOH): Thus, the kinetic conditions for the Ingold-
Fischer effect would be fulfilled also in this case. The
persistent radical effect will operate when the addition
of phenyl to acrylonitrile is about as fast as the reaction
of phenyl with TBH.
(PhCOO)2
CHCN
2Ph• + 2 CO2
PhCH2CHCN
(35)
Ph• + 2 CH2
(36)
CN
Ph CH2 CH
CH2
CH–CN
NC CH CH2
CN
CN
(37)
We have not obtained any evidence that reaction 39
could be an alternative mechanism to eq 41: no acetate
or benzoate 11 has been observed by the use of increasing
amounts of Cu(OAc)2, AcOH, or PhCOOH in various
solvents (benzene, t-BuOH, pyridine), contrary to the
behavior of cyclohexene (eq 27).
CN
CN
CN
+ Cu(II)
CN
+ Cu(I) + H+
(38)
CN
CN
8
PhCH2CH(CN)OCOR
In the absence of Cu(II) salt, the equilibrium of eq 37
is shifted to the left by irreversible acrylonitrile polym-
erization. In the presence of TBH, under the same
conditions, neither the polymerization nor the formation
of 8 take place, and the mixed peroxide 9 is the main
reaction product. We have explained2 its formation by
the oxidation of the radical adduct by Cu(II) salt (eq 39),
which would favorably compete with the polymeriza-
tion.
11
R = Me, Ph
This is not surprising, if an electron-transfer process
is involved in eqs 27, 28, 29, and 32, considering the
electrophilic character of the R-cyanoalkyl radical. Equa-
tion 41 would be the mechanism for the formation of 9
also with Cu(II) catalysis.
We have obtained further evidence of this behavior by
reaction of acrylonitrile with TBP. A fast polymerization
of the monomer takes place in benzene solution; however,
PhCH2C˙ HCN + t-BuOOH + Cu(II) f
PhCH2CH(OOBu-t)CN + Cu(I) + H+ (39)
(24) Araneo, S.; Fontana, F.; Minisci, F.; Recupero, F.; Serri, A.
9
Tetrahedron Lett. 1995, 36, 4307.

Synthesis of Mixed Peroxides from TBH
J . Org. Chem., Vol. 62, No. 12, 1997 3855
in the presence of TBH the polymerization is again
inhibited and the mixed peroxide 10 is cleanly formed
(eq 43).
Mixed P er oxid es fr om Styr en es. When TBP is
decomposed in benzene solution in the presence of
styrene or R-methylstyrene, the polymerization of the
alkene occurs. If TBH is added to the solution, the
polymerization is again inhibited, and diperoxides 13 and
14, epoxides 15 and 16, and the corresponding rear-
ranged aldehydes 17 and 18 are significant reaction
products.
CH2dCHCN + (t-BuOOCO)2 + 2 t-BuOOH f
t-BuOOCH2CH(CN)OOBu-t + 2 CO2 + 2 t-BuOH
(43)
10
The Ingold-Fischer effect is again responsible for the
formation of 10. The peroxyl radical, generated according
to eq 6, adds to acrylonitrile (eq 44), and the radical
adduct is trapped by another peroxyl radical (eq 45)
PhCH(OOBu-t)CH2OOBu-t
PhC(Me)(OOBu-t)CH2OOBu-t
13
14
PhCHCH2
PhC(Me)CH2
PhCH2CHO
PhCH(Me)CHO
O
O
18
17
t-BuOO• + CH2dCHCN f
t-BuOOCH2C˙ H(CN) (44)
15
16
Moreover, benzaldehyde, acetophenone, and minor
amounts of the alkoxylperoxyl derivatives 19 and 20 are
respectively formed from styrene and R-methylstyrene.
t-BuOOCH2C˙ H(CN) + •OOBu-t f 10
(45)
The presence of variable amounts of Cu(OAc)2 and
AcOH does not modify the reaction course in various
solvents (benzene, t-BuOH, pyridine): no acetic ester is
formed. This supports the hypothesis that no oxidation
takes place according to eqs 27-29, always due to the
electrophilic character of the R-cyanoalkyl radical.
The formation of 10 smoothly takes place also in the
absence of TBP, but only if Cu(II) catalysis2 is present
(eq 46).
PhCH(OOBu-t)CH2OBu-t
19
PhC(Me)(OOBu-t)CH2OBu-t
20
13 and 14 are clearly formed by the same mechanism
reported above with acrylonitrile (eqs 6, 12, 44, and 45);
15 and 16 arise from the fragmentation of the radical
adduct in a chain process according to eq 51, which
competes with the formation of the diperoxides.
3 t-BuOOH + CH2dCHCN Cu(II)8
10 + t-BuOH + H2O (46)
PhCHCH2
O
PhCHCH2 + •OBu-t
O
(51)
On the grounds of the above results, we believe that
also for reaction 46 only eqs 44 and 45 are responsible
for the formation of 10, contrary to our preliminary
communication,2 in which eq 46 was explained by a
ligand-transfer oxidation of the radical adduct. A further
synthetic development has been achieved by carrying out
the reaction in THF solution (eq 47).
O–Bu–t
The aldehydes 17 and 18 are well-known products of
the rearrangement of epoxides 15 and 16; benzaldehyde
and acetophenone arise from oxidative degradation,
probably through â-scission of benzyloxyl radicals, and
19 and 20 are generated by addition of tert-butoxyl
radical to the alkene, in competition with eq 6, followed
by cross-coupling of the radical adduct with the peroxyl
radical.
+ (t-BuOOCO)2 + CH2
CHCN + t-BuOOH
+ 2 t-BuOH + 2 CO2
(47)
O
CH2CH(CN)OOBu-t
O
As with cyclohexene, the reaction is strongly affected
by the solvent and the metal salt catalysis. In benzene
solution in the presence of Fe(III)- or Mn(III)-porphyrins,
13 is the main reaction product (∼90%) with styrene,
whereas 19 is a minor product (∼10%).6 In solvents,
which form hydrogen bonds with TBH, the rate of
reaction 6 is lower and the main reaction products in the
presence of iron, manganese, or copper salt catalysis are
19 (pyridine),6 21 (acetic acid),12,25 and 22 (methanol).25
12
Also, the formation of 12 (two stereoisomers in about
1:1 ratio) is clean, and it is well explained by hydrogen
abstraction from THF (eq 48), fast addition of the
nucleophilic radical to acrylonitrile (eq 49), and cross-
coupling of the radical adduct with the peroxyl radical
(eq 50).
+ t-BuO• (t-BuOO•)
+ t-BuOH (t-BuOOH) (48)
PhCH(OAc)CH2OBu-t
PhCH(OMe)CH2OBu-t
O
O
21
22
(49)
+ CH2
O
CHCN
O
CH2CHCN
Under these conditions, the addition of tert-butoxyl
radical to styrene (eq 52) prevails on reaction 6; the
benzylic radical adduct, due to its nucleophilic character,
is oxidized by the metal salt (eq 53), as is the allyl radical
in eqs 27-29.
+ t-BuOO•
12
(50)
O
CH2CHCN
We believe that eq 50 is responsible for the formation
of 12 also from TBH with Cu(II) catalysis, as we have
reported in a recent preliminary communication.2
(25) Minisci, F.; Cecere, M.; Galli, R; Gazz. Chim. Ital. 1963, 93,
1288.

3856 J . Org. Chem., Vol. 62, No. 12, 1997
Bravo et al.
PhCHdCH2 + •OBu-t f PhC˙ HCH2 - OBu-t (52)
mechanism with electrophilic carbon-centered radicals,
whereas with nucleophilic carbon-centered radicals there
is evidence that the oxidation by metal salts is a viable
alternative mechanism (eqs 29, 32, and 53). The nature
of the solvent plays a fundamental role in fulfilling the
Ingold-Fischer conditions and, therefore, in determining
the cross-coupling or the oxidation mechanism.
PhC˙ HCH2OBu-t + M(II) + ROH f
PhCH(OR)CH2OBu-t + M(I) + H+ (53)
M(II) ) Fe(III), Mn(III), Cu(II)
ROH ) t-BuOOH, AcOH, MeOH
Exp er im en ta l Section
Thus, also with styrene, as with cyclohexene, there is
evidence that two mechanisms of formation of mixed
peroxides might be operating: a cross-coupling between
the benzyl and peroxyl radicals when the Ingold-Fischer
conditions are fulfilled and an oxidation of the intermedi-
ate nucleophilic benzyl radical by metal salts.
Gen er a l Meth od s. Mass spectra were performed on a
GLC-MS Finnigan TSQ70 instrument, using a Varian 3700
gas chromatograph, equipped with a SBP-1 fused silica column
(30 m × 0.2 mm i.d., 0.2 µm film thickness) and helium as
carrier gas.
GLC analyses were performed on a Dani 6500 capillary gas
chromatograph, equipped with SBP-5 fused silica column (25
m × 0.25 mm i.d., 1 µm film thickness) at a hydrogen flow
rate of 8 cm3 min-1, PTV injector, flame ionization detector.
All of the solvents and the reagents were obtained from
commercial sources and were used without further purifica-
tion.
tert-Butyl peroxalate (TBP) was prepared according to the
literature procedure.26 All of the reaction products were known
compounds; some of them were commercially available, while
others were prepared following the literature procedures; they
were then used for the qualitative identification and the
quantitative analyses (GLC and GLC-MS): 1,11 2,18 3,18 5,1 8,24
9,2 11,2 12,2 13,5 14,5 19,5 20,5 21,25 22,26 and 23.1 Com-
mercially available: 4, 6, 7, 15, 16, 17, and 18.
Mixed P er oxid es fr om 2-Meth ylcycloh exa n on e.
In benzene solution at 45 °C, 2-methylcyclohexanone
reacts with TBH and TBP only in traces, giving the mixed
peroxide 23. However, at 67 °C the reaction gives 23
with good conversion and selectivity (eq 54).
O
+ t-BuOOH + (t-BuOOCO)2
O
(54)
+ 2 t-BuOH + 2 CO2 + H2O
OOBu-t
Gen er a l P r oced u r es for th e Oxid a tion of p- a n d m -
Cr esols. (A) By TBH a n d TBP . p-Cresol, TBP, and TBH
in the ratios and with the solvents reported in Table 1 were
stirred for 2 h at 45 °C under N2. The solution was directly
analyzed by GLC (biphenyl as internal standard) and GLC-
MS. Authentic samples of 1, 2, and 3 were utilized for the
identification and the quantitative analysis of the reaction
products. In the absence of TBP no reaction takes place,
whereas in the absence of TBH 2 and 3 were the only reaction
products. (B) By TBH a n d Meta llop or p h yr in s. p- or
m-cresols, TBH, FeTPPCl (commercial product), or MnTDClP-
PCl (prepared according to the literature procedure)27 in the
ratios and with the solvents reported in Table 1 were stirred
for 5 h at 20 °C or for 2 h at 70 °C under N2 in the presence of
0.03 mmol of pyridine for mmol of TBH. The solution was
directly analyzed as in A. m-Cresol does not lead to mixed
peroxides or dimers of the phenoxyl radical, but 2-methylben-
zoquinone 4 was the substantial reaction product; it was
identified by comparison with an authentic sample. (C) By
TBH a n d Cu (OAc)2. p-Cresol, TBH, and Cu(OAc)2 in the
ratios reported in Table 1 in benzene solution were stirred at
20 °C under N2; after 5 h the conversion was 3% and after 72
h it was 18%. 1 is the only reaction product. At 50 °C under
the same conditions the conversions were 23% after 3 h and
64% after 20 h; again, 1 was the only reaction product.
The results are reported inTable 1.
Gen er a l P r oced u r e for th e Oxid a tion of Cycloh exen e.
(A) By TBH a n d TBP . One mmol of TBH, 0.5 mmol of TBP,
and the amounts of cyclohexene and solvents reported in Table
2 were stirred for 3 h at 45 °C under N2. The solution was
directly analyzed by GLC (biphenyl as internal standard) and
GLC-MS. Authentic samples of 5 and 6 were utilized for the
identification and the quantitative analysis of the reaction
products. The conversions of cyclohexene are in the range 40-
50%. In the absence of TBP no reaction occurs under the same
conditions, whereas in the absence of TBH 6 is the only
reaction product. The results are reported in Table 2. (B)
By TBH a n d TBP in th e P r esen ce of Meta l Sa lts. The
reaction was carried out as in A in the presence of Cu(OAc)2
23
At 45 °C, the homolysis of TBP is fast, and a lack of
reaction, contrary to the behavior of phenols, cyclohexene,
acrylonitrile, styrene, and THF, must be ascribed to slow
hydrogen abstraction from 2-methylcyclohexanone by
peroxyl radical. At 67 °C, this hydrogen abstraction
effectively takes place, and once again, the conditions for
the Ingold-Fischer effect are fulfilled, so that cross-
coupling according to eq 55 leads to 23.
O
+ •OOBu-t
23
(55)
The same peroxide 23 was obtained by the Kharasch
reaction1 for 7 h at 70 °C, but not at 45 °C. No acetic
ester was obtained in the presence of Cu(OAc)2 or AcOH.
We believe that, also in the copper-catalyzed reaction,
23 is formed by eq 55, the only role of the copper salt
being the generation of alkoxyl and peroxyl radicals from
TBH. This is due to the electrophilic character of the
R-ketoalkyl radical, which prevents its oxidation by metal
salts, as previously discussed for R-cyanoalkyl radicals
from acrylonitrile.
Con clu sion s
The bimolecular self-reactions, the hydrogen abstrac-
tions, and the additions to unsaturated substrates are
relatively slow reactions for tertiary peroxyl radicals. This
fulfills the conditions for the Ingold-Fischer “persistent
radical effect” with several compounds (phenols, cyclo-
hexene, acrylonitrile, THF and acrylonitrile, styrenes,
2-methylcyclohexanone, and certainly a large variety of
other substrates) in thermal and catalytic reactions for
the synthesis of mixed peroxides from hydroperoxides by
cross-coupling of peroxyl with carbon-centered radicals
(eqs 11, 25, 41, 45, 50, and 55). This would be the only
(26) Bartlett, P. D.; Benzing, E. P.; Pincock, R. E. J . Am. Chem. Soc.
1960, 82, 1762.
(27) Banfi, S.; Maiocchi, A.; Montanari, F.; Quici, S. Gazz. Chim.
Ital. 1990, 120, 123. Quici, S.; Banfi, S.; Pozzi, G. Gazz. Chim. Ital.
1993, 123, 597.

Synthesis of Mixed Peroxides from TBH
J . Org. Chem., Vol. 62, No. 12, 1997 3857
and Fe(OAc)2OH, as reported in Table 3. Cyclohexenyl acetate
7 is formed in addition to 5 and 6. The conversions of
cyclohexene, based on the peroxides, are in the range 60-70%.
The results are reported in Table 3. By TBH a n d Cop p er
Sa lts. One mmol of TBH, cyclohexene, and copper salts in
the solvents reported in Table 4 were stirred for 2 h at 70 °C
or for 4 h at 45 °C under N2. The solution was directly
analyzed as in A and B, and the results are reported in Table
4.
TBP in 5 mL of THF and 5 mL of benzene were stirred for 2
h at 45 °C under N2. GLC analysis (biphenyl as internal
standard) revealed a yield of 63% of 12 (two stereoisomers in
45:55 ratio) as the only reaction product from acrylonitrile:
12 was identified by comparison with an authentic sample.
(B) Five mmol of TBH, 25 mmol of acrylonitrile, and 0.25 mmol
of Cu(OAc)2 in 25 mL of benzene and 5 mL of THF were stirred
for 18 h at 50 °C under N2. GLC revealed a yield of 80% of 12
based on THF.
Oxid a tion of Acr ylon itr ile by Ben zoyl P er oxid e a n d
TBH. Ten mmol of acrylonitrile, 10 mmol of TBH, and 10
mmol of benzoyl peroxide were refluxed for 4 h in 6 mL of
benzene under N2. The solution was washed with 10%
aqueous NaOH in order to separate benzoic acid, and then the
benzene solution was analyzed by GLC (PhCH2CH2CN as
internal standard) and GLC-MS. The four reaction products
(9, 34.9%; 10, 20.6%, biphenyl, 27.5%, and phenyl benzoate,
16.9%), obtained in overall 58% yield, were identified by
comparison with authentic samples.
Oxid a tion of Acr ylon itr ile by TBP a n d TBH. (A) Five
mmol of acrylonitrile and 0.5 mmol of TBP in 5 mL of benzene
were stirred for 2 h at 45 °C under N2. Only polymerization
of acrylonitrile took place. (B) As in A, in the presence of 1
mmol of TBH. No polymerization occurred, but 10 (identified
by comparison with an authentic sample) was the only reaction
product from acrylonitrile; the conversion (based on the
peroxides, GLC) was 27%. (C) Five mmol of TBH, 25 mmol of
acrylonitrile, and 0.25 mmol of Cu(OAc)2 in 25 mL of benzene
were stirred for 18 h at 50 °C under N2. GLC revealed a yield
of 55% of 10 based on TBH. No polymerization of acrylonitrile
took place.
Oxid a tion of Styr en es by TBH a n d TBP . (A) Five mmol
of styrene or R-methylstyrene and 0.5 mmol of TBP in 5 mL
of benzene were stirred for 2 h at 45 °C under N2. Only the
polymerization of the monomers took place. (B) As in A, in
the presence of 1 mmol of TBH. Styr en e. No polymerization
occurred. The reaction solution was directly analyzed by GLC
(biphenyl as internal standard) and GLC-MS by using
authentic samples for the identification of the main reaction
products: 13 (32.2%), 15 (23.1%), 17 (9.8%), 19 (6.5%), and
benzaldehyde (22.5%); yields are based on the peroxides. (C)
As in B. r-Meth ylstyr en e. The main reaction products were
analyzed by GLC and GLC-MS (comparison with authentic
samples): 14 (10.1%), 16 (19.7%), 18 (10.3%), 20 (4.5%), and
acetophenone (41.1%); yields are based on the peroxides.
Oxid a tion of 2-Meth ylcycloh exa n on e by TBP a n d
TBH. Five mmol of 2-methylcyclohexanone, 1 mmol of TBH,
and 0.5 mmol if TBP in 5 mL of benzene were stirred for 2 h
at 67 °C under N2. An authentic sample of 23 was utilized
for the identification of the reaction product. The direct GLC
analysis of the reaction mixture revealed the presence of 77%
yield of 23. At 45 °C under the same conditions only traces of
23 were formed.
Oxid a tion of Acr ylon itr ile by TBP a n d TBH in THF .
(A) Ten mmol of acrylonitrile, 5 mmol of TBH, and 5 mmol of
J O9700944