Spin trapping of radical intermediates generated by the oxidation of substituted 4-methylphenols

Article abstract of DOI:10.1002/hlca.201000431

A series of substituted 4-methylphenols 1 and 2 was oxidized with PbO ₂ in the presence of nitroso compounds 3-10. The formation of adducts of benzyl radicals with the nitroso spin traps in the reaction mixture was established, suggesting the abstraction of an H-atom from the methyl substituent of 1 or 2. In the consecutive steps, the adducts underwent a further rearrangement to the corresponding nitrones. When the starting phenol contained bulky ^tBu groups in ortho-position (see 2,6-di(tert-butyl)-4- methylphenol (1a)), the stable 2,6-di(tert-butyl)-4R-phenoxy radicals (R=-CH=N⁺(O^-)-X) were detected as the final radical products. The indirect evidence of nitrones in the reaction mixture was performed in one case by the reaction with a RO₂ radicals. Copyright

Full text of DOI:10.1002/hlca.201000431

1260
Helvetica Chimica Acta – Vol. 94 (2011)
Spin Trapping of Radical Intermediates Generated by the Oxidation of
Substituted 4-Methylphenols
by Petr Majzlꢀk, Ladislav Omelka*, Renata Superatova´, and Petra Holubcova´
Faculty of Chemistry, Brno University of Technology, CZ-612 00 Brno, Czech Republic
(omelka@fch.vutbr.cz)
A series of substituted 4-methylphenols 1 and 2 was oxidized with PbO2 in the presence of nitroso
compounds 3 – 10. The formation of adducts of benzyl radicals with the nitroso spin traps in the reaction
mixture was established, suggesting the abstraction of an H-atom from the methyl substituent of 1 or 2. In
the consecutive steps, the adducts underwent a further rearrangement to the corresponding nitrones.
When the starting phenol contained bulky tBu groups in ortho-position (see 2,6-di(tert-butyl)-4-
methylphenol (1a)), the stable 2,6-di(tert-butyl)-4R-phenoxy radicals (R ¼ꢀCH¼Nþ(Oꢀ)ꢀX) were
detected as the final radical products. The indirect evidence of nitrones in the reaction mixture was
.
performed in one case by the reaction with a RO2 radicals.
Introduction. – It is generally known that the oxidation of monohydric phenols with
.
different agents, such as PbO2, MnO2, or RO2 leads to the formation of corresponding
phenoxy radicals. The large number of EPR parameters continuously published in
catalogues during the last decades [1] contributed to the detailed knowledge of this
class of radicals, and seemingly, any new principal information cannot be expected in
this field. The stability of phenoxy radicals substantially depends on the substitution in
the ortho-position. High stability is the typical feature of the phenoxy radicals derived
t
from 2,6-dialkyl-substituted phenols, especially those with bulky Bu substituents
(sterically hindered phenols). For this reason, 2,6-di(tert-butyl)-4R-phenols are
frequently utilized as very efficient antioxidants [2]. Monohydric phenols with 2,6-
dialkyl substituents containing an a-CH bond, as well as phenols with unsubstituted or
partially substituted ortho-position (sterically unhindered phenols) provide on
oxidation unstable phenoxy radicals, which can be detected only by using special
EPR techniques, e.g., flow or spin-trapping methods [3][4]. Although the generation of
phenoxy radicals resulting from the abstraction of the phenolic H-atom is the
dominating radical process during the oxidation, the tendency towards the abstraction
of an H-atom from a Me group in 4-methyl-substituted phenols was also observed. The
benzyl radicals formed were detected by a spin-trapping technique by using nitroso spin
traps [4][5]. The recombination products of benzyl radicals observed in the reaction
mixture of the oxidized phenols also support this mechanism, although other theories
interpreting their formation have been proposed [6]. To understand the specific
behavior of 4-methyl-substituted phenols in more detail, the oxidation of a series of
phenols having this structural requisite was investigated by the spin-trapping method.
Thus H-atom abstraction from a 4-Me group was investigated within the series of
ꢀ 2011 Verlag Helvetica Chimica Acta AG, Zꢁrich

Helvetica Chimica Acta – Vol. 94 (2011)
1261
differently substituted 4-methylphenols 1 and 2 in the presence of substituted nitroso
spin traps 3 – 10.
Results and Discussion. – 2,6-Disubstituted 4-Methylphenols 1. Generally, the
mechanism given in Scheme 1 has to be taken into account in the case of the oxidation
of 2,6-disubstituted 4-methylphenols 1.
Scheme 1
The oxidation of 2,6-di(tert-butyl)-4-methylphenol (1a) in the presence of nitro-
sobenzene (3) afforded a mixture of radical products where the corresponding phenoxy
radical 11a (aH(Me) ¼ 1.094 mT, aH(m) ¼ 0.155 mT) and the benzyl radical 12a,
prevailed, the latter being indirectly detected through the adduct 13a (Ar¼ Ph;
aN(NO) ¼ 1.049 mT, aH(CH2) ¼ 0.565 mT, aH(o,p) ¼ 0.273 mT, aH(m) ¼ 0.092 mT,
Scheme 2).
Scheme 2
The EPR parameters of 13a are similar to those of the spin-trapping adduct of 12a
with 3,5-di(tert-butyl)nitrosobenzene (¼1,3-bis(1,1-dimethylethyl)-5-nitrosobenzene)

1262
Helvetica Chimica Acta – Vol. 94 (2011)
[5]. The concentration ratio between the two radicals 11a and 12a depended on the
experimental conditions, and in the presence of an excess of phenol 1a over the spin-
trap agent 3, the EPR spectrum of the pure benzyl radical adduct 13a was observed
(Fig. 1,a). Nevertheless, gradually, the EPR spectrum changed: One hour after the
preparation of the sample, the EPR spectrum given in Fig. 1,b, was detected, whose
parameters suggest that a secondary phenoxy radical 15a was formed via nitrone 14a
(Ar¼ Ph) in the consecutive steps of the oxidation (g ¼ 2.0049) (Scheme 3). This
statement is evidently supported by the splitting constant of H-atoms (aH ¼ 0.150 mT),
which is the typical value of meta H-atoms in phenoxy radicals. Other dominating
splittings verified by the simulation of the experimental EPR spectra were attributed to
one H- and one N-atom of 15a (Ar¼ Ph) (Table 1). Based on these experimental facts,
it is evident that the secondary phenoxy radicals resulted from the transformation of
the benzyl radical adduct 13a. Moreover, the number of interacting nuclei suggests that
the 4-position in the secondary phenoxy radical 15a (Ar¼ Ph) is occupied by the
nitrone fragment ꢀCH¼Nþ(Oꢀ)ꢀPh, which originates from the conversion of the
.
nitroxide radical moiety ꢀCH2ꢀN(O )ꢀPh in 13a (! intermediate 14a (Ar¼ Ph)).
Fig. 1. a) Experimental and simulated EPR spectrum of nitroxide radical 13a in benzene solution;
b) experimental and simulated EPR spectrum of secondary phenoxy radical 15a (Ar¼ Ph) in benzene
solution

Helvetica Chimica Acta – Vol. 94 (2011)
1263
Scheme 3
Table 1. EPR Parameters of Phenoxy Radicals 15a Prepared by the Oxidation of 2,6-Di(tert-butyl)-4-
methylphenol (1a) with PbO2 in Benzene Solution in the Presence of the Different Aromatic Nitroso
Compounds 3 – 9 or of 2-Methyl-2-nitrosopropane (10)
Nitroso compound Ar-NO Ar
Splitting constants [mT] of 15a
aH(3,5)
aH(CH) aH(N) aH(Ar)
3
4
5
6
7
Ph
0.150
0.150
0.165
0.140
0.290
0.280
0.305
0.280
0.270
–
0.510 0.090 (3 H)
2-MeC6H4
2,4,6-Cl3C6H2
4-PhNHC6H4
0.490
0.485
–
–
0.570 0.095 (2 H)
2,3,5,6-Me4C6H 0.155
0.490
–
–
–
–
–
–
8a)
9b)
10c)
4-Me2NC6H4
2,4,6-tBu3C6H2
tBu
–
–
–
0.170 (1 H), 0.150 (1 H) 0.260
0.505
a) Due to the strong line broadening effect, the individual splitting constant could not be determined.
b) Interpretation of EPR signal is still an open problem. c) Splitting constants taken from [7].
In the course of further experiments, the number of aromatic nitroso compounds as
spin traps was expanded. The analysis of EPR spectra established that the above-
mentioned reaction route is unambiguously valid in the cases of 3 – 8, although the
quality of the EPR signal of phenoxy radical 15a (Ar¼ 4-Me2NC6H4), obtained with 8,
did not allow the determination of the splitting constants (Fig. 2,a; Table 1). The 2,4,6-
tri(tert-butyl)nitrosobenzene (¼1,3,5-tris(1,1-dimethylethyl)-2-nitrosobenzene; 9),
however, afforded an EPR spectrum (Fig. 2,b) whose interpretation is still an open
problem. Probably due to the strong steric effect arising from the presence of bulky tBu
groups in the ortho-positions of the starting nitroso compound 9, the distribution of spin
density in the corresponding phenoxy radical 15a (Ar¼ 2,4,6-tri(tert-butyl)phenyl)
differed from that observed for the previously discussed ones. On the other hand,
neither the adduct 13a nor the phenoxy radical 15a was observed in the trapping
reaction of 1a with 2-methyl-2-nitrosopropane (10). Instead of them, only the nitroxide
.
radical (Me3C)2NO (aN(NO) ¼ 1.555 mT) was detected, as a product of the decom-
position of the spin trap on the surface of PbO2. This fact is surprising because the EPR

1264
Helvetica Chimica Acta – Vol. 94 (2011)
parameters of the phenoxy radical 15a (Ar¼ tBu, Table 1), produced by the oxidation
of the nitrone 14a (synthesized by the reaction of 3,5-di(tert-butyl)-4-hydroxybenzal-
dehyde with N-(tert-butyl)hydroxylamine) with benzoyloxy radicals, are reported in
the literature [7].
Fig. 2. Experimental EPR spectra of radical products produced by the PbO2 oxidation of 2,6-di(tert-
butyl)-4-methylphenol (1a) in the presence of nitroso spin trap a) 8 and b) 9 in benzene solution
The experimental results obtained by PbO2 oxidation of the sterically hindered 2,6-
di(tert-butyl)-4-methylphenol (1a) in benzene solution in the presence of different
nitroso compounds ArꢀNO initiated further investigation in this field. Thus the
tendency towards the abstraction of one H-atom from a 4-Me group was also studied
with the 2,4,6-trisubstituted phenols 1b and 1c. In contrast to 1a, due to the absence of
bulky substituents in ortho-position, the detection of the corresponding phenoxy
radicals cannot be expected in this case. When 2,4,6-trimethylphenol (1b) was
subjected to the oxidation, immediately after the addition of PbO2 to the benzene
solution of 1b and nitrosobenzene (3), a high concentration of benzyl radical adduct
13b (Ar¼ Ph, Table 2) was observed.
Table 2. EPR Parameters of Nitroxide Radicals 13b Prepared by the PbO2 Oxidation of 1b in Benzene in
the Presence of Some Nitroso Spin Traps ArꢀNO
Ar
Splitting constants [mT] of 13b
aN(NO)
aH(CH2)
aH(Ar)
Ph
1.023
1.150
1.261
0.550
0.630
0.950
0.263 (3 H), 0.092 (2 H)
0.190 (5 H), 0.085 (2 H)
0.070 (2 H)
2-MeC6H4
2,4,6-Cl3C6H2
The EPR signal of 13b was stable for more then 24 h without being overlapped by
the signal of another radical. This fact seemingly points out that the suggested reaction
route (unambiguously evidenced for 1a; Scheme 3), based on the conversion of 13b to
nitrone 14b and consecutively to the secondary phenoxy radical 15b, is not valid in this

Helvetica Chimica Acta – Vol. 94 (2011)
1265
case. However, this effect can also be satisfactorily interpreted as a result of a
substantially lower stability of 2,6-dimethyl-4R-phenoxy radicals in comparison with
those derived from 2,6-di(tert-butyl)-4R-phenols. The presence of nitrone 14b in the
reaction mixture was indirectly established by an experiment, where tert-butyl
hydroperoxide was added to the reaction mixture 24 h after the generation of adduct
13b: A high concentration of nitroxide radical 16 with the splitting constants aN(NO) ¼
1.046 mT, aH(CH) ¼ 0.175 mT, aH(o,p) ¼ 0.273 mT, and aH(m) ¼ 0.092 mT (Fig. 3) was
observed in accordance with Scheme 4. The EPR parameters of 16 agreed well with the
Fig. 3. Experimental and simulated EPR spectrum of nitroxide radical 16 in benzene solution
Scheme 4

1266
Helvetica Chimica Acta – Vol. 94 (2011)
.
published data for a RO2 adduct with C,N-diphenylnitrone, reported by Cholvad et al.
[8]. Simultaneously, the authors point out the instability of this adduct, leading to the
.
loss of the CH H-atom in nitroxide radical PhꢀCH(OOR)ꢀN(O )ꢀPh in subsequent
steps of the reaction. The spin adducts 13b were observed also by the PbO2 oxidation of
1b in the presence of other spin traps (Table 2).
The H-atom abstraction from the 4-methyl substituent, unambiguously established
by the behavior of phenols 1a and 1b, was not observed with 2,6-dinitro-4-methylphenol
(1c). In this case, the EPR spectrum of a nitroxide radical with dominating N-splitting
(aN(NO) ¼ 1.155 mT, other splittings unresolved) was registered after the addition of
PbO2 to the benzene solution of 1c in the presence of nitrosobenzene (3). This fact
points out the specific behavior of phenols with electron-accepting ortho-substituents,
which influence the polarity of the CꢀH bonds in the 4-Me group. Consequently, the
tendency towards its homolytic scission, leading in the case of phenols 1a and 1b to the
formation of benzyl radicals, is partially reduced in phenol 1c. The EPR signal of
nitroxide radicals can possibly be ascribed to adducts of generated phenoxy radicals
with nitrosobenzene [4].
Substituted 4-Methylphenols 2. Similarly to the above discussed phenols 1a – 1c, the
simultaneous generation of corresponding phenoxy and benzyl radicals is also expected
in the course of the oxidation with PbO2 of sterically unhindered phenols 2a – 2e. Due
to their instability caused by structural factors (unsubstituted or partially substituted
ortho position), the phenoxy radical cannot be directly detected by the EPR method,
and only indirect evidence by means of the spin-trapping technique is suitable [4]. Due
to the tendency towards the abstraction of an H-atom from the 4-methyl substituent,
the benzyl radicals also reacted with nitroso spin traps and the generated adducts
mostly dominated in the EPR spectrum. With nitrosobenzene (3) as spin trap, the PbO2
oxidation of phenols 2a – 2c in benzene produced the radical adducts 17a – 17c
(Table 3). The 3,4-dimethylphenol (2d) afforded only in CHCl3 solution a high
Table 3. EPR Parameters of Nitroxide Radicals 17 Prepared by the PbO2 Oxidation of 2a – 2d in Benzene
in the Presence of Nitrosobenzene (3)
Phenol
Splitting constants [mT] of 17
aN(NO)
aH(CH2)
aH(phenyl)
2aa)
2b
1.120
1.091
1.049
1.107
0.610
0.600
0.565
0.730
0.290 (3 H), 0.093 (2 H)
0.280 (3 H), 0.095 (2 H)
0.273 (3 H), 0.092 (2 H)
0.279 (3 H), 0.093 (2 H)
2c
2db)
a) Splitting constants taken from [4]. b) Splitting constants in CHCl3.

Helvetica Chimica Acta – Vol. 94 (2011)
1267
concentration of nitroxide radical 17d, in benzene, only the unresolved EPR signal of
the radical adduct of the phenoxy radical (aN(NO) ¼ 1.00 mT) was detected.
Also no adduct of benzyl radicals was observed on PbO2 oxidation of 3,4,5-
trimethylphenol (2e) in benzene in the presence of nitrosobenzene (3). Instead of
them, the EPR signal of hydroxyphenyl adduct 18 was detected. The EPR parameters
of the latter (aN(NO) ¼ 0.980 mT, aH(1 H) ¼ 0.060 mT, aH(o,p) ¼ 0.261 mT, and
aH(m) ¼ 0.085 mT) were similar to those of the species obtained by the spin trapping
of the 3,5-dimethylphenoxy radical with nitrosobenzene [4]. Nevertheless, the PbO2
oxidation of 2e in CHCl3 in the presence of 3 revealed the simultaneous formation of
small amounts of benzyl radical adduct 17e as deduced from the detailed analysis of the
EPR side lines present besides the dominating EPR signal of radical adduct 18.
Conclusions. – The oxidation of the series of 4-methyl substituted phenols with
PbO2 in different solvents showed that besides the formation of phenoxy radicals, also
the abstraction of an H-atom from the 4-methyl substituent leading to the benzyl
radicals has to be taken into account. Theoretically, both radical intermediates can be
evidenced by EPR spectroscopy applying spin-trapping method. The adducts of the
benzyl radicals entered the follow-up reactions, resulting in the formation of nitrones,
which could be further oxidized at the phenolic part of the molecule. Stable secondary
phenoxy radicals produced in this way were formed on PbO2 oxidation of 2,6-di(tert-
butyl)-4-methylphenol in the presence of different aromatic nitroso compounds. Due to
their instability, similar secondary phenoxy radicals could not be detected with other
phenols in our study, the only detectable radical products being adducts of benzyl
radicals with nitroso compounds. Nevertheless, the presence of an intermediate nitrone
was established by the reaction with a peroxy radical.
This work was supported by the Ministry of Education of the Czech Republic under research project
MSM 0021630501.
Experimental Part
General. All chemicals (phenols, nitroso compounds, tert-butyl hydroperoxide, PbO2) were
commercially available and used without further purification. Benzene and CHCl3 of anal. grade purity
(Sigma) were used as the solvent in all experiments. EPR Spectra: EPR spectrometer SpectraNova; at
r.t.; simulation of exper. spectra with the programme Simphonia.
Oxidation of 1 or 2 with PbO2 in the Presence of Nitroso Compounds 3 – 10: General Procedure. In a
5 · 10ꢀ2 m benzene or CHCl3 soln. of the phenol, the corresponding amount of a nitroso compound was
dissolved, adjusting the molar ratio phenol/nitroso compound to 2 :1. To 2 ml of this soln., PbO2 (150 mg)
was added under stirring. The stirring was stopped after 1 min, and after the sedimentation of the solid
phase, 0.3 ml of the reaction mixture was placed into the EPR tube. To study the consecutive
transformations of primary radical products, the sampling of the reaction mixture was performed in
30 min intervals under continuous stirring. To obtain the high quality EPR spectra, the soln. in the tube
was bubbled with N2 for 30 s.
REFERENCES
[1] Landolt-Bçrnstein, ꢂMagnetic Properties of Free Radicalsꢃ, New Series, Vol. 9, Part c2, Springer,
Berlin, 1979; Landolt-Bçrnstein, ꢂMagnetic Properties of Free Radicalsꢃ, New Series, Vol. 17, Part c2,
Springer, Berlin, 1989.

1268
Helvetica Chimica Acta – Vol. 94 (2011)
ˇ
[2] J. Pospꢄsil, Pure Appl. Chem. 1973, 36, 207.
[3] T. J. Stone, W. A. Waters, J. Chem. Soc. 1964, 213.
´ ˇ
´
[4] L. Omelka, J. Kovacova, Magn. Reson. Chem. 1994, 32, 525.
[5] E. I. Zhuzhgov, N. M. Bazhin, M. P. Terpugova, Y. D. Tsvetkov, Bull. Acad. Sci. USSR, Div. Chem.
Sci. (Engl. Transl.) 1974, 23, 2438.
[6] R. Stubbins, F. Sicilio, Tetrahedron 1970, 26, 291.
[7] J. G. Pacifici, H. L. Browning, J. Am. Chem. Soc. 1970, 92, 5231.
ˇ
´ ˇ
[8] V. Cholvad, A. Stasko, A. Tkac, A. L. Buchachenko, L. Malꢄk, Collect. Czech. Chem. Commun. 1981,
46, 823.
Received November 23, 2010