Microwave-assisted fast cyclohexane oxygenation catalyzed by iron-substituted polyoxotungstates

Article abstract of DOI:10.1002/adsc.200505111

Under microwave irradiation, iron-substituted polyoxotungstates (Fe-POMs) catalyze cyclohexane oxygenation to A/K oil with 90-95% selectivity, unmatched turnover frequencies (40-400 h^-1), and multi-turnover regime (>1000 TON). Such a rapid reaction protocol allowed the screening of so-called inorganic Fe-synzymes with 1-4 nuclearity, including a family of Krebs-type isostructural complexes. Product distribution, kinetic analysis, mechanistic probes and fitting calculations, are consistent with a radical chain oxidation propagated by Fe-catalysed decomposition of organic peroxides which ultimately depends on the redox-properties and structural arrangement of the iron moiety within the POM cage.

Full text of DOI:10.1002/adsc.200505111

COMMUNICATIONS
DOI: 10.1002/adsc.200505111
Microwave-Assisted Fast Cyclohexane Oxygenation Catalyzed by
Iron-Substituted Polyoxotungstates
a,
a
a
b
Marcella Bonchio, * Mauro Carraro, Gianfranco Scorrano, Ulrich Kortz
a
Istituto CNR per la Tecnologia delle Membrane (ITM-CNR), Universit a` di Padova, via Marzolo 1, 35131 Padova, Italy
Fax: (þ39)-(0)49-827-5239, e-mail: marcella.bonchio@unipd.it
b
School of Engineering and Science, International University Bremen, P. O. Box 750 561, 28725 Bremen, Germany
E-mail: u.kortz@iu-bremen.de
Received: March 14, 2005; Accepted: August 31, 2005
[
1]
Abstract: Under microwave irradiation, iron-substi-
tuted polyoxotungstates (Fe-POMs) catalyze cyclo-
hexane oxygenation to A/K oil with 90–95% selec-
around 80%. A formidable challenge in both industri-
al and academic research is the invention of new cata-
lysts performing efficiently with 1 atm O . In this direc-
2
À1
tivity, unmatched turnover frequencies (40–400 h ),
tion, benchmark literature results can be ranked on the
basis of the total turnover number (TON) and the turn-
over frequency (TOF) exhibited (Table 1). Inspection of
these data shows a promising perspective for catalytic
oxygenation with transition metal-substituted polyoxo-
metalates (TMS-POMs). Indeed, polyoxometalates
provide an inorganic coordination framework where a
redox-active metal center (Fe, Mn, Ru, or others) is
bound by MO octahedra with M being generally
Mo(VI), W(VI).
Noteworthy, some Fe-POMs display structural analo-
gy to the active site of natural oxygenation enzymes
Figure 1). The synzyme rationale has been recently
and multi-turnover regime (>1000 TON). Such a
rapid reaction protocol allowed the screening of
so-called inorganic Fe-synzymes with 1–4 nuclearity,
including a family of Krebs-type isostructural com-
plexes. Product distribution, kinetic analysis, mecha-
nistic probes and fitting calculations, are consistent
with a radical chain oxidation propagated by Fe-cat-
alysed decomposition of organic peroxides which ul-
timately depends on the redox-properties and struc-
tural arrangement of the iron moiety within the
POM cage.
6
[
7]
(
Keywords: Aerobic oxidation; CÀH bond activation;
homogeneous catalysis; iron; microwave irradiation;
polyoxometalates
invoked to address the superior performance of g-
6
À
[(FeOH ) SiW O ]
in hydrocarbon oxygenation
2
2
10 36
(
4þx)À
compared to homologues [(FeOH ) SiW(12–x)O(40–x)
]
2
x
[3,8]
where x¼1–3. However, the severe mechanistic com-
plexity associated with metal-mediated aerobic oxidations
hampers the understanding of the catalyst role and fate so
that, in most cases, a claimed adherence to bioinspired
The selective oxidation of cyclohexane to a mixture of guidelines remains doubtful.
cyclohexanol and cyclohexanone (K/A oil) is of consid- Herein, we report on the screening of selected Fe -
III
erable importance in the production of adipic acid (AA) POMcatalystsforcyclohexaneoxygenationbyarapidre-
finally used in the manufacture of polyamide fibers (ny- action protocol under microwave (MW) induced dielec-
lon 6,6). The current industrial process employs homo- tric heating. Such a novel approach takes advantage of
geneous cobalt salts, molecular oxygen at 8–9 atm and the extreme robustnessand polyanionic natureof the cat-
[
9]
[10]
temperatures above 150–1608C to obtain cyclohexane alyst to elicit MWabsorption. Our results include: (i)
conversions in the range 3–6% and a K/A oil selectivity the discovery of hits operating at unprecedented peak ef-
Table 1. Benchmark systems for aerobic oxidation of cyclohexane with 1 atm O2.
À1
Catalyst
Conv. [%] (Time [h])
TON (TOF [h ])
Ref.
6
À
[2]
[3]
g-[(MnOH ) SiW O ]
6.4 (96)
1.1 (96)
65 (20)
36 (9)
789 (8.2)
135 (1.4)
130 (6.5)
121 (13)
90 (15)
2
2
10 38
6
À
g-[(FeOH ) SiW O ]
2
2
10 38
[
[
[
4]
5]
6]
Mn(acac) /NHPI
2
Fe(II)-picolinic acid
Co(acac) /NHPI
45 (6)
2
Adv. Synth. Catal. 2005, 347, 1909 – 1912
ꢀ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
1909

COMMUNICATIONS
Marcella Bonchio et al.
FeSiW performs with unmatched TOFs (entries 4–
1
1
[
16]
6). Moreover FT-IR spectra confirmthat the recov-
ered catalyst preserves the original structure, at variance
[
17]
with the organic porphyrin analogue (entry 7). In the
case under study, converging lines of evidence delineate
a radical chain mechanism, namely: (i) steady accumula-
tion of the autoxidation product CHHP on lowering the
catalyst loading (entries 4–6); (ii) acceleration by tert-
butyl hydroperoxide (TBHP) (entry 8), (iii) inhibition
by the radical scavenger 3,5-di-tert-butyl-4-hydroxyto-
luene (BHT) (entry 9), (iv) a biphasic kinetic trace,
with a slow initiation phase followed by a fast zero-order
regime, indicating an autocatalytic behavior (Fig-
Figure 1. Structure of representative iron-substituted polyox-
otungstates.
[
18]
ure 2).
À1
ficiency withTOFin the range 40–400 h and total turn-
III
over number>1000, (ii) examination of Fe -POMs, in-
cluding a series of Krebs-type isostructural complexes,
and (iii) the drawing of a general reaction scheme, based
on mechanistic probes and fitting the oxidation kinetics.
III
Fe -POMs of nuclearities 1–4 were obtained by met-
alation of lacunary polyoxotungstates, yielding [a-
5
À
6À
10 38
Fe(H O)SiW O ] (FeSiW ), [g-Fe (H O) SiW O ]
2
11 39
11
2
2
2
7
À
(
Fe SiW ), [a-Fe (H O) SiW O ] (Fe SiW ) and [b-
2
10
3
2
3
9
37
3
9
nÀ
IV
IV
Fe (H O) (XW O ) ] (Fe X W with X¼Se , Te ;
4
2
10
III
9
33
2
4
2
18
III
[11–14]
As , Sb , and n¼4, 6).
These complexes were
used as tetrahexylammonium (THA) salts in 1,2-di-
chloroethane (DCE). Data in Table 2 are meant to ver-
ify the catalyst activity and selectivity after 150–250 min
MW irradiation. Cyclohexyl hydroperoxide (CHHP),
cyclohexanol (A) and cyclohexanone (K) account for
[
15]
9
0–95% selectivity.
Control experiments indicate that in the absence of
catalyst or MW exposure, the oxidation is negligible or Figure 2. Experimental (points) and calculated (see
extremely slow (entries 1–3). Under MW irradiation, Scheme 1) curves for run 6 in Table 1.
III
[a]
Table 2. Cyclohexane oxygenation by mononuclear Fe catalysts.
[
b]
[c]
[d]
[e]
À1
Entry
Catalyst [mmol]
Additive [mmol]
Time [min]
Conv. [%]
Selectivity
TON
TOF [h ]
CHHP:A:K
1
2
3
4
5
6
7
8
9
–
–
–
–
–
–
–
–
250
150
5520
250
250
250
150
150
150
0.15
0.07
1.89
2.95
2.74
2.42
1.38
2.40
<0.02
59: 39: 2
n.d.
–
–
[
f]
FeSiW11 (0.75)[
n.d.
229
358
656
1159
167
291
n.d.
n.d.
4.7
141
207
439
72
g]
FeSiW11 (0.75)
FeSiW11 (0.75)
FeSiW11 (0.38)
FeSiW11 (0.19)
0: 60: 40
0: 66: 34
5: 57: 38
11: 60: 29
2: 66: 32
5: 64: 31
n.d.
[
h]
FeTMP (0.75)
FeSiW11 (0.75)
FeSiW11 (0.75)
TBHP (20)
BHT (300)
147
n.d.
[
a]
DCE 0.8 ml, cyclohexane 9.1 mmol, O 1 atm, T¼1208C, applied power of MW irradiation¼240 W. FeTMP¼iron (III)
2
tetrakis(5,10,15,20-mesityl)porphyrin.
[
[
[
[
[
[
[
b]
c]
d]
e]
f]
%
of substrate conversion.
Product distribution.
Total turnover number (TON): products (mmol)/catalyst (mmol).
Highest turnover frequency calculated as TON per hour on the linear kinetic phase.
Conventionally heated at 1208C.
g]
h]
Conventionally heated at 838C.
90% catalyst decomposition determined by UV-Vis analysis at 418 nm.
1910
asc.wiley-vch.de
ꢀ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2005, 347, 1909 – 1912

Microwave-Assisted Fast Cyclohexane Oxygenation
COMMUNICATIONS
III
[a]
Table 3. Cyclohexane oxygenation by Fe -POM catalysts.
[a]
[b]
[c]
[d]
[e]
À1
#
Catalyst
Conv. [%]
Selectivity CHHP:A:K
TON
TOF [h ]
1
2
3
4
5
6
7
FeSiW11
Fe SiW
1.42
0.54
0.37
0.30
0.51
0.57
0.96
6: 67: 28
73: 7: 20
76: 6: 18
88: 0: 12
66: 25: 9
45: 35: 20
49: 33: 18
172
66
45
36
62
141
30
38
23
34
37
70
2
10
9
Fe SiW
3
Fe Se W
4
2
18
18
Fe Te W
4
2
Fe Sb W
69
116
4
2
18
Fe As W
18
4
2
[
[
a]
DCE 0.8 mL, cyclohexane 9.1 mmol, catalyst 0.75 mmol, O 1 atm, T¼1208C, MW power 240 W for 150 min.
2
b–e]
See footnotes in Table 2.
Therefore, the overall mechanistic hypothesis in-
[
19]
cludes steps 1–8, (Scheme 1). In particular, because
the TBHP-induced acceleration results fromsuppress-
ing the initial lag-time while not affecting the propaga-
[20]
tion rate (cf. TOFs in entries 4 and 8), the main role
of the iron catalyst may be envisaged in the propagation
chain, promoting an efficient decomposition of CHHP
through Haber–Weiss reductive and oxidative steps
[
20–22]
(
Scheme 1, steps 4 and 5).
Further support to such
a scenario comes from the 2:1 A/K selectivity, which
cannot be simply explained by the uncatalyzed Russel-
type decomposition of the organic peroxide accounting
[
21,23]
for a selectivity ratio close to 1 (Scheme 1, step 7).
This model turns out to fit the experimental kinetics
(
5
Figure 2) and identifies the relevant catalytic steps (4,
in Scheme 1) mainly within the propagation chain.
[
24]
The iron-mediated mechanism is also probed by the se-
lectivity of adamantane oxidation, yielding a C-3/C-2 ra-
[25,26]
tio of 9.5.
this value falls within the expected range
for a radical reaction (1–25) and is similar to those ex-
[
27]
hibited by analogous porphyrin catalysts (6–10). Al-
though it is obviously problematic to encode such a mul- Scheme 1. Kinetic model of cyclohexane oxygenation cata-
lyzed by Fe(III)-POMs.
ti-step mechanism by TON and TOF macro-parameters,
which strongly depend on the catalyst loading (entries
III
4
–6 in Table 2), they provide straightforward guidelines progress to look at the interaction between Fe -POMs
for process optimization.
with model organic peroxides in order to gain insight
A different behavior can be traced for substituted into the nature of postulated peroxo-iron intermediates
POMs with higher iron nuclearity, leading to CHHP ac- and their evolution to products through homolytic or
cumulation and to increase of ketone (entry 1 vs. entries heterolytic pathways. Our final aimis to obtain struc-
2
–7 in Table 3). Both features indicate a depressed per- ture-activity correlations so to implement the design of
oxide decomposition via homolytic pathways, favoring new selective catalytic systems and exploit the rich and
[
18,28]
heterolytic ketonization (Scheme 1, step 8).
This versatile chemistry of polyoxometalates.
latter reaction is known to proceed through a peroxome-
tal intermediate whose nature and stability may depend
on the structural arrangement of the iron moiety and on
Experimental Section
[
29]
the POM tungsten framework. Nevertheless, along
the series of the Krebs-type isostructural complexes,
the overall oxidation efficiency can be ascribed to the re-
The iron substituted polyoxotungstates were prepared follow-
[11–14]
ing literature procedures.
ion exchange in water.
THA salts were obtained by cat-
[
14]
dox properties of the iron moiety. In particular, the
higher TOF process is given by Fe As W which, to-
4
2
18
gether with the isocharged Fe Sb W species, displays
a well behaved cathodic wave at a higher reduction po-
4
2
18
General Oxidation Procedure
tential, thus suggesting that the catalyst is also partici- DCE (0.8 mL), Fe-POM (0.75–0.19 mmol) and cyclohexane
[
14]
pating in the Haber–Weiss chemistry. Studies are in (0.98 mL) were mixed, under a dioxygen atmosphere, in a 10-
Adv. Synth. Catal. 2005, 347, 1909 – 1912
ꢀ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
1911

COMMUNICATIONS
Marcella Bonchio et al.
mL closed teflon reactor (HPR-1000/10S, Milestone) equipped
with temperature and pressure control units and irradiated in-
side the cavity of a MW Ethos-1600 labstation (Milestone) at
103–112; c) A. de la Hoz, A. Diaz-Ortiz, A. Moreno,
Chem. Soc. Rev. 2005, 34, 164–178.
[11] F. Zonnevijlle, C. M. Tourn e´ , G. F. Tourn e´ , Inorg. Chem.
1982, 21, 2751–2757.
12] C. Nozaki, I. Kiyoto, Y. Minai, M. Misono, N. Mizuno,
Inorg. Chem. 1999, 38, 5724–5729.
2
40 W, with Tbulk ¼1208C and Pbulk ffi 2 atm. In the system un-
der examination, higher irradiation powers caused heating and
pressure peaks which are detrimental for obtaining a reprodu-
cible heating profile. The reaction was sampled (50 mL) every
[
[
13] J. Liu, F. Ort e´ ga, P. Sethuraman, D. E. Katsoulis, C. E.
50 minutes always restoring the dioxygen atmosphere. The
Costello, M. T. Pope, J. Chem. Soc. Dalton Trans. 1992,
samples were analyzed by GLC and GLC-MS. Peroxide con-
tent was determined using the triphenylphosphine quencher
1
901–1906.
[
14] U. Kortz, M. G. Savelieff, B. S. Bassil, B. Keita, L. Nadjo,
Inorg. Chem. 2002, 41, 783–789.
15] Carboxylic acids were mainly observed as overoxidation
products while significant chlorinated products were de-
tected only under conventional heating likely originating
fromautoxidation pathways involving the solvent.
[30]
method, carboxylic acids were revealed by silylation with
BSTFA before GLC-MS analysis. Fe-POM stabilities were as-
sessed by FT-IR after precipitation fromthe reaction mixture
and washing with diethyl ether. Kinetic parameters were ob-
tained by non-linear experimental data fitting with the Scientist
Micromath software.
[
[
16] CH CN completely inhibits catalysis, thus explaining a
3
[8]
previously reported observation.
[
17] Fe-TMP-catalyzed oxidation occurs with no induction
time likely due to formation of organic radicals from li-
gand degradation.
18] R. A. Sheldon, J. K. Kochi, Metal-Catalyzed Oxidations
of Organic Compounds, Academic Press, New York,
Acknowledgements
Financial support from the Italian National Council of Research
is gratefully acknowledged.
[
[
1
981.
19] R. Pohorecki, J. Baldyga, W. Moniuk, W Podg o´ rska, A.
References and Notes
Zdr o´ jkowski, P. T. Wierzchowski, Chem. Eng. Sci. 2001,
5
6, 1285–1291, and references cited therein.
[
[
[
[
[
[
[
1] K. Weissermel, H. J. Arpe, Industrial Organic Chemistry,
nd edn., VCH, Weinheim, 1993.
2] T. Hayashi, A. Kishida, N. Mizuno, Chem. Commun.
000, 381–382.
3] C. Nozaki, M. Misono, N. Mizuno, Chem. Lett. 1998,
263–1264.
4] T. Iwahama, K. Syojyo, S. Sakaguchi, Y. Ishii, Org. Proc-
ess Res. Dev. 1998, 2, 255–260.
5] D. H. R. Barton, T. Li, J. MacKinnon, Chem. Commun.
[20] a) M. Bonchio, V. Conte, F. Di Furia, G. Modena, J. Org.
Chem. 1989, 54, 4368–4371; b) M. Bonchio, V. Conte, F.
Di Furia, G. Modena, S. Moro, J. O. Edwards, Inorg.
Chem. 1994, 33, 1631–1637.
[21] A. S. Goldstein, R. S. Drago, Inorg. Chem. 1991, 30,
4506–4510.
2
2
1
[22] MW-induced super-heating is likely to promote the unca-
talyzed initiation of the autoxidation.
[23] G. A. Russel, J. Am. Chem. Soc. 1957, 79, 3871–3877.
À8
À1
8
1
997, 557–558.
[24] Calculated kinetic constants k ¼7.0ꢀ10 (s ); k ¼10
1
2
À1
À1 À1
III
À8
À1
6] Y. Ishii, T. Iwahama, S. Sakaguchi, K. Nakayama, Y.
Nishiyama, J. Org. Chem. 1996, 61, 4520–4526.
7] a) B. J. Wallar, J. D. Lipscomb, Chem. Rev. 1996, 96,
(s ); k ¼3.8 (L mol
s
); k ꢀ[Fe ]¼10 (s ); k ꢀ
3
4
7
5
II
À4
À1
À1 À1
8
[Fe ]¼2.8ꢀ10 (s ); k ¼10 (L mol
s
); k ¼10 (L
6
7
À1 À1
III
mol
s
); k ꢀ[Fe ]¼0.
8
2
625–2658; b) M. Costas, M. P. Mehn, M. P. Jensen, L.
[25] The selectivity parameter is defined as the relative reac-
tivity of tertiary to secondary CH groups calculated per
H atom. Under the conditions explored (see footnotes
in Table 1), adamantane (0.15 mmol) is oxidized, in
chlorobenzene, to 1-adamantanol, 2-adamantanol and
adamantanone in a relative ratio 76:16:8.
Que Jr., Chem. Rev. 2004, 104, 939–986; c) T. J. McMur-
ry, J. T. Groves, in: Metalloporphyrin Models for Cyto-
chrome P-450, (Ed.: P. R Ortiz de Montellano), Plenum
Press, New York, 1986, pp. 1–28.
[8] a) Y. Nishiyama, Y. Nakagawa, N. Mizuno, Angew.
Chem. 2001, 113, 3751–3753; Angew. Chem. Int. Ed.
[26] M. Bonchio, G. Scorrano, P. Toniolo, V. Artero, A.
Proust, V. Conte, Adv. Synth. Catal. 2002, 344, 841–
844, and references cited therein.
[27] a) Y. Ishii, S. Sakaguchi, T. Iwahama, Adv. Synth. Catal.
2001, 343, 393–427; b) F. Minisci, F. Fontana, S. Araneo,
F. Recupero, S. Banfi, S. Quici, J. Am. Chem. Soc. 1995,
117, 226–232.
2
001, 40, 3639–3641; b) N. Mizuno, C. Nozaki, T. Hirose,
M. Tateishi, M. Iwamoto, J. Mol. Catal. A: Chem. 1997,
17, 159–168; c) N. Mizuno, M. Hashimoto, Y. Sumida,
1
Y. Nakagawa, K. Kamata, in: Polyoxometalate Chemistry
for Nano-Composite Design, Kluwer Academic/Plenum,
New York, pp. 197–203, 2002; d) N. Mizuno, Catal. Surv.
Japan 2000, 4, 149–154.
[28] J. D. Chen, J. Dakka, R. A. Sheldon, Appl. Catal. A: Gen.
1994, 108, L1-L6.
[
9] C. L. Hill , C. M. Prosser-McCartha, Coord. Chem. Rev.
1
995, 43, 407–455.
[29] J. Kim, R. G. Harrison, C. Kim, L. Que Jr., J. Am. Chem.
Soc. 1996, 118, 4373–4379.
[
10] M. Larhed, C. Moberg, A. Hallberg, Acc. Chem. Res.
2
002, 35, 717–727; b) A. Loupy CR Chimie 2004, 7,
[30] G. B. Shulꢁpin, CR Chim. 2003, 6, 163–178.
1912
asc.wiley-vch.de
ꢀ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2005, 347, 1909 – 1912