A novel iron-based catalyst for the biphasic oxidation of benzene to phenol with hydrogen peroxide

Article abstract of DOI:10.1002/1521-3773(20001201)39:23<4321::AID-ANIE4321>3.0.CO;2-5

Improved selectivity and decreased formation of overoxygenated byproducts were achieved by using a MeCN/H₂O biphasic system for the iron-catalyzed oxidation of benzene to phenol with H₂O₂. Under optimized reaction conditions with 5-carboxy-2-methylpyrazine-N-oxide as the ligand for iron [Eq. (1)] a benzene conversion of 8.6% and selectivities based on benzene and H₂O₂ of 97 and 88%, respectively, were obtained.

Full text of DOI:10.1002/1521-3773(20001201)39:23<4321::AID-ANIE4321>3.0.CO;2-5

COMMUNICATIONS
[5] a) S. Sato, M. Wada, Bull. Chem. Soc. Jpn. 1970, 43, 1955 ± 1962;
b) A. V. Haynes, H. G. Drickamer, J. Chem. Phys. 1982, 76, 114 ± 125.
[6] M. I. Gaiduk, V. V. Grigoryants, A. F. Mironov, V. D. Rmyantseva,
V. I. Chissov, G. M. J. Sukhin, Photochem. Photobiol. B 1990, 7, 15 ±
20.
[7] M. P. Oude Wolbers, F. C. J. M. vanVeggel, F. G. A. Peters, E. S. E.
van Beelen, J. W. Hofstraat, F. A. J. Geurts, D. N. Reinhoudt, Chem.
Eur. J. 1998, 4, 772 ± 780.
[8] Y. S. Sohn, D. N. Hendrickson, H. B. Gray, J. Am. Chem. Soc. 1971, 93,
3603 ± 3612.
[9] S. L. Murov, I. Carmichael, G. L. Hug, Handbook of Photochemistry,
2nd ed., Marcel Dekker, New York, 1993.
transfer rate. The energy-transfer rate is further determined
by the so-called spectral overlap of the antenna and the
lanthanide ion, that is, how well the donating and receiving
energy levels match. The energy-transfer rate in [:Yb2-
Ru)1]2‡ will be evenslower thanin[:Nd 2-Ru)1]2‡, because
the antenna triplet state :ET ˆ 17400 cmÀ1)[16] matches the
NdIII 4G5/2 state at 17100 cmÀ1, whereas the overlap with the
only YbIII energy level, the 2F5/2 state which is at 10240 cmÀ1, is
much smaller.
Deoxygenated [D6]DMSO solutions of the ferrocene-
functionalized :Nd)2 and :Yb)2 complexes :10À4 m) exhibit
the typical linelike lanthanide emission upon excitation of the
ferrocene antenna at 320 nm.[17] At room temperature,
[10] a) M. P. Oude Wolbers, F. C. J. M. vanVeggel, J. W. Hofstraat, F. A. J.
Geurts, D. N. Reinhoudt, J. Chem. Soc. Perkin Trans. 2 1997, 2275 ±
2282; b) M. P. Oude Wolbers, F. C. J. M. vanVeggel, B. H. M. Snel-
È
link-Ruel, J. W. Hofstraat, F. A. J. Geurts, D. N. Reinhoudt, J. Chem.
4
sensitized emission at 1060 and 1330 nm :4F3/2 ! I11/2, an d
Soc. Perkin Trans. 2 1998, 2141 ± 2150.
4I13/2 transitions, respectively) is observed for :Nd)2, an d at
[11] S. I. Klink, G. A. Hebbink, L. Grave, F. G. A. Peters, F. C. J. M.
vanVeggel, D. N. Reinhoudt, J. W. Hofstraat, Eur. J. Org. Chem. 2000,
1923 ± 1931.
2
980 nm :2F5/2 ! F7/2 transition) for :Yb)2 :see Figure 3). The
[12] S. I. Klink, L. Grave, D. N. Reinhoudt, F. C. J. M. van Veggel, M. H. V.
Werts, F. A. J. Geurts, J. W. Hofstraat, J. Phys. Chem. A 2000, 104,
5457 ± 5468.
[13] D. L. Dexter, J. Chem. Phys. 1953, 21, 836 ± 850.
[14] Comparison of the antenna luminescence spectrum of [:Nd2-Ru)1]2‡
with the standard [Ru:bpy)3]2‡ complex inDMSO shows that the
emissionmaximum of [:Nd 2-Ru)1]2‡ has shifted to the red: from
3
634 nm to 666 nm. The amide groups stabilize the MLCT state, and
probably the excited electronis localized onthe amide-functionalized
bpy ligand.
[15] V. Balzani, A. Juris, M. Venturi, Chem. Rev. 1996, 96, 759 ± 833.
[16] A. Juris, V. Balzani, F. Barigelleti, S. Campagna, P. Belser, A.
vonZelewsky, Coord. Chem. Rev. 1988, 84, 85 ± 277.
[17] Excitation in the weaker ferrocene absorption band at 440 nm :e ˆ
200mÀ1 cmÀ1) also resulted in sensitized lanthanide luminescence.
Figure 3. Emissionspectra of :Yb) 2 and :Nd)2 in[D 6]DMSO :10À4 m)
uponexcitationat 320 mn . The spectra have beencorrected for the
absorbance of the samples.
excitationspectra of :Nd) 2 and :Yb)2 :not shown), closely
follow the absorption spectrum of the ferrocene antenna and
confirm unambiguously that excitation takes place through
the antenna. Upon excitation at 337 nm the lanthanide
luminescence decay curves could be fitted mono-exponen-
tially with lifetimes of 2.0 Æ 0.2 ms for :Nd)2 and 18.8 Æ 0.8 ms
for :Yb)2.
In conclusion, the transition metal complexes [Ru:bpy)3]2‡
and ferrocene are able to sensitize both NdIII and YbIII.
Moreover, these sensitizers enable excitation with visible
light. The luminescent nature of the [Ru:bpy)3]2‡3MLCT state
has allowed a direct study of the sensitization process. It has
beenestablished for the [:Nd 2-Ru)1]2‡ complex that the
energy transfer takes place in a ªclassical wayº through the
antenna triplet state with a rate of 1.1 Â 106 sÀ1. The sensitiza-
tionprocess of [:Yb 2-Ru)1]2‡ follows the same pathway,
although the energy transfer rate is lower :ꢀ105 sÀ1).
A Novel Iron-Based Catalyst for the Biphasic
Oxidation of Benzene to Phenol with Hydrogen
Peroxide
Daniele Bianchi,* Rossella Bortolo, Roberto Tassinari,
Marco Ricci, and Rodolfo Vignola*
Direct oxidation of benzene to phenol is an important
reactioninorganic chemistry. Often, however, its selectivity is
rather poor since phenol is more reactive toward oxidation
than benzene itself, and substantial formation of over-oxy-
genated by-products :catechol, hydroquinone, benzoqui-
nones, and tars) also occurs.[1]
Similar over-oxidationproblems are efficiently solved in
biological systems by segregating catalysts and products into
Received: June 14, 2000 [Z15273]
[*] Dr. D. Bianchi, Dr. R. Bortolo, R. Tassinari, Dr. M. Ricci
EniChem S.p.A.-Centro Ricerche Novara ªIstituto G. Doneganiº
Via Fauser 4, 28100 Novara :Italy)
[1] M. H. V. Werts, J. W. Hofstraat, F. A. J. Geurts, J. W. Verhoeven,
Chem. Phys. Lett. 1997, 276, 196 ± 201.
[2] Y. Hasegawa, T. Ohkubo, K. Sogabe, Y. Kawamura, Y. Wada, H.
Nakashima, S. Yanagida, Angew. Chem. 2000, 112, 365 ± 368; Angew.
Chem. Int. Ed. 2000, 39, 357 ± 360.
Fax : :‡390)321-447425
E-mail: daniele.bianchi@enichem.it
Dr. R. Vignola
EniTecnologie
[3] L. H. Slooff, A. Polman, M. P. Oude Wolbers, F. C. J. M. van Veggel,
D. N. Reinhoudt, J. W. Hofstraat, J. Appl. Phys. 1998, 83, 497 ± 503.
[4] L. H. Slooff, A. Polman, S. I. Klink, G. A. Hebbink, L. Grave,
F. C. J. M. vanVeggel, D. N. Reinhoudt, J. W. Hofstraat, Opt. Mater.
2000, 14, 101 ± 107.
Via Ramarini 32, 00016 Monterotondo :Italy)
Fax : :‡390)6-90673243
E-mail: rvignola@enitecnologie.eni.it
Angew. Chem. Int. Ed. 2000, 39, No. 23
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 2000
1433-7851/00/3923-4321 $ 17.50+.50/0
4321

COMMUNICATIONS
different environments. The active sites of several oxygenases
[2]
are deeply buried inhydrophobic pockets where lipophilic
substrates are readily oxidized, while the more hydrophilic
reaction products are promptly released into the surrounding
aqueous environment. Likewise, the cellular membranes have
important functions as selective barriers around the cell that
allow the import of components and the export of products.[3]
In the course of industrial research aimed at developing a
new process for the one-step production of phenol, we
wondered whether a biphasic system could mimic these
important features of biological systems. Here we report on a
new method for the selective biphasic oxidation of benzene to
phenol with hydrogen peroxide as oxidant, an iron complex as
catalyst, and trifluoroacetic acid as cocatalyst.
The first aspect that we investigated was the solvent system.
The importance of this component is well illustrated by the
history of hydrocarbonoxidationwith hydrogenperoxide. For
example, water is the solvent in the Fenton system,[4] while
pyridine is characteristic of the Gif chemistry widely studied
by Bartonet al. inthe last tenyears. [5] We found that the use of
anaqueous/organic reactionmedium dramatically affects the
selectivity of the reaction. Screening was carried out with
FeSO4 :a conventional Fenton catalyst) and a series of organic
solvent/water mixtures as reaction media.
Figure 1. Effect of the reaction medium on the oxidation of benzene. The
reaction was carried out under the conditions described in the Exper-
imental Section with FeSO4 as catalyst inabsecne of ligadns. The
composition of the reaction medium was water/solvent/benzene 45/45/10
:v/v/v). Black bar: Fractionof H 2O2 oxidizing benzene to phenol; gray bar:
fractionof H 2O2 oxidizing benzene to dioxygenated products :hydro-
quinone, catechol, and 1,4-benzoquinone). The concentration of over-
oxidation products :tars) was not determined. S ˆ selectivity based on
H2O2.
As shown in Figure 1, a remarkable enhancement of
selectivity was obtained with the biphasic system water/
acetonitrile :1/1) in the presence of benzene. In this medium
the concentration of benzene in the aqueous phase increased
from 0.18% :solubility of benzene in water)[6] to 0.76%, and
the phenol was largely extracted :85%) into the organic
phase. Thus, the biphasic process minimizes over-oxidation by
reducing the contact between phenol and the catalyst, which is
soluble inthe aqueous phase. This hypothesis was confirmed
on comparing the benzene/phenol molar ratios in the aqueous
phase of water/benzene and water/acetonitrile/benzene reac-
tionmixtures, which were 0.25 and 3.7, respectively. Such
enhancement is not possible in the case of solvents that are
too hydrophobic, such as n-octane, which quantitatively
extract benzene from the aqueous phase, or with those that
are too hydrophilic, such as dimethylformamide :DMF),
which do not form a two-phase system.
The identification of a more efficient and selective catalyst,
which is an essential requirement for industrial application,
was the second key aspect of the study. After testing a series of
bidentate N,N, N,O and O,O ligands for
iron:Table 1), 5-carboxy-2-methylpyra-
zine-N-oxide :1; entry 15) was selected
as the most efficient catalyst precursor.
The nature of the ligand plays a crucial
role in determining both the catalyst
reactivity and selectivity. In general,
complexes with N,N ligands :entries 2 ± 5) were almost
inactive, those with O,O ligands :entries 6 ± 8) displayed good
activity but poor selectivity, and those with N,O ligands,
Table 1. Oxidation of benzene: selection of the ligand for iron.[a]
Entry
Ligand
H2O2 conversion
[%]
Selectivity
Selectivity
onH O2[b] [%]
onC H6[c] [%]
2
6
1
2
3
4
5
6
7
8
9
none
2,2'-bipyridine
1,10-phenanthroline
2,6-di:2-pyridyl)pyridine
2,3-di:2-pyridyl)pyrazine
1,2-dihydroxybenzene
1,2-dihydroxynaphthalene
1,2-dihydroxynaphthalene-4-sulfonic acid
8-hydroxyquinoline
77
6
6
0
4
100
100
100
0
38
< 1
< 1
0
< 1
44
25
3
33
n.d.
n.d.
±
n.d.
60
n.d.
n.d.
±
0
10
11
12
13
14
15
pyridine-2-carboxylic acid
pyridine-2,6-dicarboxylic acid
quinoline-2-carboxylic acid
pyrazinecarboxylic acid
pyrazine-2,3-dicarboxylic acid
5-carboxy-2-methylpyrazine-N-oxide :1)
13
2
90
78
90
94
15
< 1
33
70
52
78
n.d.
±
n.d.
81
78
85
[a] Screening was carried out under the reaction conditions described in the Experimental Section with a molar ratio Fe/ligand/CF3COOH of 1/4/1. The
reaction medium was water/acetonitrile/benzene :5/5/1). [b] Selectivity based on H2O2: :moles of phenol produced) Â 100/:moles of converted hydrogen
peroxide). [c] Selectivity based on benzene: :moles of phenol produced) Â 100/:moles of converted benzene). n.d.: not determined.
4322
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 2000
1433-7851/00/3923-4322 $ 17.50+.50/0
Angew. Chem. Int. Ed. 2000, 39, No. 23

COMMUNICATIONS
aqueous solutionof hydrogenperoxide :30% w/w, 1.8 mmol) was added
stepwise over 1 h. At the end of the reaction, the hydrogen peroxide
concentration in the aqueous phase was determined by titration with
potassium permanganate, and the concentration of oxidation products in
the aqueous and organic phases was determined by HPLC analysis.
especially the pyrazinecarboxylic acid derivatives :en-
tries 13 ± 15), gave the best results.
The cocatalyst was selected from a series of acids, including
acetic, phosphoric, sulfuric, p-toluenesulfonic, and trifluoro-
acetic acids. The latter gave the best results and increased the
phenol selectivity :based on H2O2) from 68 :inthe absence of
any acid cocatalyst) to 78%. Conversely, complete inhibition
of the catalytic activity occurred with phosphoric acid.
With FeSO4, ligand 1, and trifluoroacetic acid, in the
selected water/acetonitrile/benzene biphasic system, under
optimized reaction conditions :a further increase in selectivity
was obtained by lowering the temperature from 50 to 358C),
we achieved a benzene conversion of 8.6% with selectivities
of 97% :based on benzene) and 88% :based on H2O2). These
values are higher thanthose reported for other iron-based
catalysts such as the Fenton [7] and Gif systems.[8] Inthe former
case, high selectivities based onH 2O2 :up to 83%) were only
obtained at low benzene conversion :<1%). Inthe latter, a
series of ironcomplexes with pyridien-2-carboxylic acid
derivatives were reported to be completely ineffective in the
oxidation of benzene under Gif reaction conditions.
Catalyst screening: FeSO4 ´ 7H2O :0.06 mmol), CF3COOH :0.06 mmol),
and one of the ligands listed in Table 1 :0.24 mmol) were dissolved in water
:8 mL), and the reaction mixture was stirred for 1 h at 258C. The resulting
solution was added to a mixture of acetonitrile :8 mL) and benzene
:1.6 mL, 18 mmol). The reactions were carried out as described for solvent
screening.
Optimized conditions for benzene oxidation: FeSO4 ´ 7H2O :0.4 mmol),
CF3COOH :0.4 mmol), and 1 :0.4 mmol) were dissolved inwater :109 mL),
and the reaction mixture was stirred for 1 h at 258C. The resulting solution
was added to a mixture of acetonitrile :109 mL) and benzene :22 mL,
248 mmol). The resulting biphasic system was stirred at 800 rpm at 358C,
and an aqueous solution of hydrogen peroxide :30% w/w, 24 mmol) was
progressively added with a peristaltic pump over 4 h. At the end of the
reaction, the concentrations of hydrogen peroxide and of the oxidation
product were measured as described above. Hydrogenperoxide adn
benzene conversions were 95 and 8.4%, respectively, with selectivities to
phenol of 88 and 97% based on hydrogen peroxide and benzene,
respectively. The molar product distributionwas: pheonl 97.1, hydro-
quinone 0.5, catechol 0.6, 1,4-benzoquinone 0.1, biphenyl 0.1, and tars
1.5 %.
It is also noteworthy that in the biphasic system, only a
negligible amount :<1%) of biphenyl was detected among
the secondary products, whereas in the classical Fenton
oxidation, this compound is formed by radical dimerization
of hydroxycyclohexadienyl radicals in a yield of 8 ± 39%.[9]
This different reactivity and selectivity was confirmed in the
oxidationof toluene. Inthe biphasic system, this reaction
afforded a mixture of hydroxylationproducts : o- and p-
cresol) and benzylic oxidation products :benzylic alcohol,
benzaldehyde, benzoic acid) in 76 and 23% overall yield,
respectively :see Experimental Section). Conversely, under
Fenton conditions, the reported product distribution was:
hydroxylation 4%, benzylic oxidation 56%, and radical
Toluene oxidation: The reaction was carried out as described above for
benzene, with water :109 mL), acetonitrile :109 mL), FeSO4 ´ 7H2O
:0.4 mmol), CF3COOH :0.4 mmol), 1 :0.4 mmol), toluene :240 mmol),
and hydrogen peroxide :30% w/w; 24 mmol). After 4 h at 508C the
hydrogen peroxide and toluene conversions were 60 and 3.1%, respec-
tively. The molar product distributionwas: o-cresol 44, p-cresol 32,
2-methyl-1,4-benzoquinone 1, benzyl alcohol 5, benzaldehyde 15, and
benzoic acid 3%.
Received: April 17, 2000
Revised: August 23, 2000 [Z15005]
[1] a) R. A. Sheldon, J. K. Kochi, Metal-Catalyzed Oxidations of Organic
Compounds, Academic Press, New York, 1981, pp. 329 ± 333; b) G.
Goor in Catalytic Oxidations with Hydrogen Peroxide as Oxidant :Ed.:
G. Strukul), Kluver, Dordrecht, 1992, pp. 29 ± 31.
[2] T. L. Poulos, J. Cupp-Vickery, H. Li in Cytochrome P-450. Structure,
Mechanism, and Biochemistry :Ed.: P. R. Ortiz de Montellano), 2nd
ed., Plenum, New York, 1995, p. 132.
[3] B. Poolman, W. N. Konings, Biochim. Biophys. Acta 1993, 1183, 5 ± 39.
[4] a) C. Walling, Acc. Chem. Res. 1975, 8, 125 ± 129; b) C. Walling, Acc.
Chem. Res. 1988, 31, 155 ± 157.
[5] a) D. H. R. Barton, Tetrahedron 1998, 54, 5805 ± 5817; b) D. H. R.
Barton, F. Launay, Tetrahedron 1998, 54, 12699 ± 12706.
[6] J. A. Riddick, W. B. Bunger, T. K. Sakano in Organic Solvents :Ed.: A.
Weissemberg) 4th ed., Wiley, New York, 1986, p. 136.
[7] a) C. Walling, J. Am. Chem. Soc. 1975, 97, 363 ± 367; b) S. Ito, A.
Mitarai, K. Hikino, M. Hirama, K. Sasaki, J. Org. Chem. 1992, 57,
6937 ± 6941.
[9]
dimerizationto bibenzyl 30%.
The catalyst system is also effective inthe oxidationof
alkanes, including a substrate as poorly reactive as methane,
which is mainly converted :708C, 5 Mpa) to formic acid :46%
based onhydrogenperoxide); inthis case, a biphasic system is
not expected to form.
In conclusion, a new iron-based homogeneous catalyst for
the oxidation of benzene to phenol with 97% selectivity was
discovered. Anattractive feature of the process is the
operationina biphasic system, with the quantitative distri-
butionof the catalyst inthe aqueous phase, that canbe
separated and reused in several consecutive cycles as a
ªworking solutionº without any appreciable loss of activity. In
the light of these promising results, the development of a
continuous process, more suitable for industrial application, is
presently under investigation.
[8] D. H. R. Barton, F. Halley, N. Ozbalik, W. Mehl, Tetrahedron Lett.
1989, 30, 6615 ± 6618.
[9] J. R. L. Smith, R. O. C. Norman, J. Chem. Soc. 1963, 2897 ± 2905.
[10] C. Venturello, R. DꢁAloisio :Montedison S.p.A.) D-EP 201934, 1986
[Chem. Abstr. 1987, 106, 67349 w].
Experimental Section
1
was synthesised according to a
literature procedure.[10] All other
chemicals :Aldrich) were used as purchased.
Solvent screening: FeSO4 ´ 7H2O :0.06 mmol) and CF3COOH :0.06 mmol)
were dissolved inwater :8 mL), and the solutionadded to a mixture of an
organic solvent :acetonitrile, propionitrile, tert-butyl alcohol, acetone,
dioxane, dimethylformamide, or n-octane; 8 mL) and benzene :1.6 mL,
18 mmol). The resulting biphasic system was stirred at 508C, and an
Angew. Chem. Int. Ed. 2000, 39, No. 23
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 2000
1433-7851/00/3923-4323 $ 17.50+.50/0
4323