Direct synthesis of adipic acid by mono-persulfate oxidation of cyclohexane, cyclohexanone or cyclohexanol catalyzed by water-soluble transition-metal complexes

Article abstract of DOI:10.1039/b102499j

A catalytic system consisting of water-soluble metal sulfophthalocyanines (MPcS) or various ruthenium complexes and mono-persulfate as the oxidant was effective in the oxidation of cyclohexanone, cyclohexanol and cyclohexane to adipic acid with different yields and selectivity. Oxidations were conducted at room temperature and under atmospheric pressure in aqueous media (or, in the case of cyclohexane, in a water-neat substrate double phase). The oxidation of cyclohexanol involved step-by-step formation of cyclohexanone, ε-caprolactone and 6-hydroxyhexanoic acid, all of which have been identified in the reaction mixtures; in selected cases moderate over-oxidation of adipic acid to glutaric and succinic acid was also observed. Various MPcS catalysts were examined (M=Fe, Co, Ni, Cu and Ru), and the ruthenium derivative exhibited the best performances in terms of rate and selectivity. Mono-persulfate was found to be a more convenient oxidizing reagent than hydrogen peroxide; related patterns were observed when H₂O₂ was used, however extended dismutation of the oxidant limited the overall yields. Cyclohexane underwent slow oxidation when reacted with persulfate (water-substrate double phase) in the presence of the water-soluble metal catalysts; adipic acid was selectively produced (95%) in the presence of RuPcS catalyst with yields as high as 21% (48 h). The catalytic performance of simpler ruthenium derivatives, such as [RuCl₂(DMSO)₄] (RuDMS) and K₅[Ru(H₂O)P₁₁O₃₉] (RuPW), was also examined for comparison purposes. A kinetic scheme for cyclohexane oxidation is proposed.

Full text of DOI:10.1039/b102499j

View Article Online / Journal Homepage / Table of Contents for this issue
Direct synthesis of adipic acid by mono-persulfate oxidation of
cyclohexane, cyclohexanone or cyclohexanol catalyzed by
water-soluble transition-metal complexes
Nicola dÏAlessandro,a Lolita Liberatore,a Lucia Tonucci,a Antonino Morvillob and
Mario Bressan*a
a Dipartimento di Scienze, Universita
E-mail: bressan=sci.unich.it; Fax: ]39 085 453 7545
b Dipartimento di Chimica Inorganica, Universita di Padova, V ia Marzolo 1, 35131 Padova, Italy
`
““G. dÏAnnunzioÏÏ, V iale Pindaro 42, 65127 Pescara, Italy.
`
Received (in Strasbourg, France) 16th March 2001, Accepted 2nd July 2001
First published as an Advance Article on the web 11th September 2001
A catalytic system consisting of water-soluble metal sulfophthalocyanines (MPcS) or various ruthenium complexes
and mono-persulfate as the oxidant was e†ective in the oxidation of cyclohexanone, cyclohexanol and cyclohexane
to adipic acid with di†erent yields and selectivity. Oxidations were conducted at room temperature and under
atmospheric pressure in aqueous media (or, in the case of cyclohexane, in a waterÈneat substrate double phase).
The oxidation of cyclohexanol involved step-by-step formation of cyclohexanone, e-caprolactone and
6-hydroxyhexanoic acid, all of which have been identiÐed in the reaction mixtures; in selected cases moderate
over-oxidation of adipic acid to glutaric and succinic acid was also observed. Various MPcS catalysts were
examined (M \ Fe, Co, Ni, Cu and Ru), and the ruthenium derivative exhibited the best performances in terms of
rate and selectivity. Mono-persulfate was found to be a more convenient oxidizing reagent than hydrogen peroxide;
related patterns were observed when H O was used, however extended dismutation of the oxidant limited the
2
2
overall yields. Cyclohexane underwent slow oxidation when reacted with persulfate (waterÈsubstrate double phase)
in the presence of the water-soluble metal catalysts; adipic acid was selectively produced (95%) in the presence of
RuPcS catalyst with yields as high as 21% (48 h). The catalytic performance of simpler ruthenium derivatives, such
as [RuCl (DMSO) ] (RuDMS) and K [Ru(H O)P
kinetic scheme for cyclohexane oxidation is proposed.
O
] (RuPW), was also examined for comparison purposes. A
2
4
5
2
11 39
Adipic acid is an important building block for a variety of
commercially useful products. Currently, adipic acid is manu-
factured mainly by a two-step process involving stoichiomet-
ric nitric acid oxidation of a cyclohexanolÈcyclohexanone
mixture, which is aerobically generated from cyclohexane
using a homogeneous cobalt-based catalyst. Research has par-
ticularly focused on tuning alternative synthetic routes to
obtain cleaner technology and compliance with increasingly
stringent environmental laws.1 In this context the major issue
concerning adipic acid synthesis is the replacement of the
toxic reactant nitric acid, which generates nitrous oxide, which
in turn is associated with important environmental problems.2
The scientiÐc challenge is to devise a one-step synthesis oper-
ating at room temperature in an environmentally friendly
solvent system, such as water, and using green co-oxidants,
such as oxygen (air) or hydrogen peroxide. The substrate of
choice is cyclohexane, but much attention has been paid to
the development of one-step catalytic oxidation of cyclo-
hexanol and/or cyclohexanone with alternate oxidants, such
as alkylhydroperoxides and hydrogen peroxide. Recently,
metal catalysts were proposed for the oxidation of cyclo-
hexane using molecular oxygen, in the presence of homoge-
neous ruthenium and iron catalysts,3 or of heterogeneous
counterparts, such as cobalt molecular sieves4 and alumino-
phosphate molecular sieves containing iron(III) ions.5 A recent
persulfate of a number of organic substrates.7,8 More recently,
we achieved a substantial improvement by using the water-
soluble ruthenium(II) derivative RuPcS (where PcS is tetra-
sodium 2,3-tetrasulfophthalocyaninate) to e†ectively promote
the oxidation of chlorophenols,9 chloro-oleÐns10 and
alcohols11 in aqueous media and in the presence of either
mono-persulfate or hydrogen peroxide. Related metal sulfo-
phthalocyanines (M \ Fe, Mn) have been successfully used by
Meunier et al. for the oxidation of chlorophenols,12
chloroanilines13 and catechols14 with persulfate or hydrogen
peroxide in aqueous acetonitrile, whereas various metal
phthalocyanines, encapsulated in zeolites15 or other inorganic
supports,16 have also been successfully tested for the direct
oxidation of cyclohexane. We present here a series of metal
sulfophthalocyanines and other simple ruthenium complexes,
among them a Keggin-type heteropolyoxotungstate, that cata-
lyze the one-pot oxidation of cyclohexanone, cyclohexanol or
cyclohexane to adipic acid using mono-persulfate or hydrogen
peroxide as the oxidizing sources, under very mild conditions
and in an entirely aqueous phase. Polyoxometallates were
recently reported to catalyze the oxidation of cyclohexane by
molecular oxygen17 and hydrogen peroxide.18
Experimental
RuPcS and CoPcS were prepared by template synthesis start-
improvement of the SbF -catalyzed oxidation of cyclo-
3
hexanone to e-caprolactone using hydrogen peroxide as an
ing from RuCl É 3H O or CoSO É 7H O, sodium 4-
3
2
4
2
oxidant by supporting the catalyst on mesoporous silica has
been reported.6
sulfophthalate and urea, following early general pro-
cedures for the synthesis of metal sulfophthalocyanines;19
cis-[RuCl M(CH ) SON ] (RuDMS)20 and K [Ru(H O)-
Over the last few years we have been interested in using
ruthenium complexes as catalysts for the oxidation with mono-
2
3 2
4
5
2
PW
O
]
(RuPW)21 were prepared using published
11 39
DOI: 10.1039/b102499j
New J. Chem., 2001, 25, 1319È1324
1319
This journal is ( The Royal Society of Chemistry and the Centre National de la Recherche ScientiÐque 2001

View Article Online
amount of D O to the reaction mixtures; water suppression
2
was carried out by a presaturation sequence using a compos-
ite pulse (zgcppr Bruker sequence). A co-axial capillary tube
containing a 30 mM water (D O) solution of 3-(trimethylsilyl)-
2
propionic-2,2,3,3-d acid sodium salt (TSP) was used as refer-
4
ence. The identity of each product and their quantitation were
conÐrmed by comparison of the position and intensity of suit-
able signals after adding measured amounts of the pure com-
pounds to the reaction mixtures. Average reproducibility of
quantitative measurements was within 5%.
Results and discussion
procedures. Organic substrates, chromatographic and spectro-
scopic standards and FePcS, NiPcS and CuPcS complexes
were obtained from Aldrich.
Oxidation of cyclohexanone and cyclohexanol
Persulfate oxidations of cyclohexanone took place at room
temperature in a completely aqueous phase containing an
excess of the peroxidic oxidant (Oxone}), typically 0.5 N (pH
ca. 2), and the substrate (50 mM). The MPcS catalysts
(M \ Fe, Co, Ni, Cu and Ru) were added at the 0.2È2 mol%
level with respect to the substrate (Table 1). It should be recal-
led that mono-persulfate alone is able to easily perform the
uncatalyzed BaeyerÈVilliger oxidation of cyclohexanone to
the corresponding lactone; in the reaction conditions quanti-
tative transformation of cyclohexanone to e-caprolactone
(and/or its hydrolysis product, 6-hydroxyhexanoic acid) was
achieved within minutes. Some of the MPcS complexes inves-
tigated were able to further oxidize the alcoholic function of
6-hydroxyhexanoic acid to adipic acid, with various rates and
selectivity. With FePcS (1 mM) conversion into adipic acid
was indeed observed, although slow (34% yield after 24 h).
The result was nevertheless interesting, since ““nakedÏÏ iron
ions, for example iron(III) sulfate, are inactive for the oxidation
of the lactone. CoPcS was more active, leading to adipic acid
in a 62% yield, but glutaric and succinic acid were also
formed, arising from signiÐcant over-oxidation of adipic acid.
Adipic and succinic acid oxidation by the metal
sulfophthalocyanineÈpersulfate catalytic systems (Table 2)
suggested that the rate of the oxidative decarboxylation of the
dicarboxylic acids followed the trend: glutaric acid [ adipic
acid [ succinic acid. In the presence of RuPcS (1 mM), cyclo-
hexanone was quantitatively transformed into adipic acid
within a few hours and with a selectivity close to 90%; glu-
A procedure for the catalytic oxidations is as follows. (a) A
water solution (10 ml) containing cyclohexanone or cyclo-
hexanol (50 mM) and the metal catalysts (0.1È1 mM) was
stirred magnetically in a vial together with commercial
Oxone}, corresponding to a 0.5 N concentration of active
oxygen as KHSO , or an equivalent amount of hydrogen per-
5
oxide, as determined by iodometric titration. (b) Neat cyclo-
hexane (0.5 ml) was added to 2.5 ml of a water solution
containing the metal catalysts (1È5 mM) and KHSO (0.5 N)
5
and the double-phase mixture stirred magnetically in a vial.
The reactions were carried out at 20 ¡C and were not a†ected
by the presence of air.
Organic analyses were performed on an HP 6890 GLC
instrument equipped with a FID, using a 30 m HP-5 capillary
column (0.32 mm i.d.; 0.25 Ðlm thick) with the injection port
kept at 250 ¡C (carrier gas: He) on aliquots withdrawn with a
microsyringe from the aqueous reaction mixtures either as
such or diluted 1 : 10 with acetone. Commercial products were
used for all gas chromatographic standardizations. The reac-
tion mixtures were treated by standard procedures with a
10 : 1 excess of 2-methyl-1-butanol to analyze the dicarboxylic
acids formed as their isobutyl esters. The identity of each
product was conÐrmed by comparison of the fragmentation
patterns in the mass spectra obtained with a MD 800 Fisons
mass spectrometer operating in the electron ionization mode
at 70 eV. The reactions were also followed by 1H NMR on a
Bruker Avance 300 MHz instrument, by adding a small
Table 1 Oxidation of cyclohexanone by metal catalysts and KHSO a
5
Product selectivityc/mol%
[cat]/
Yield w.r.t. the
Catalyst
mM
Time/h
oxidantb/equiv.%
A
B
C
D
E
Noned
Èe
3f
10
È
100
È
È
È
FePcS
0.1
3
10
10
10
34
8
92
100
100
66
È
È
È
34
È
È
È
tr
È
È
È
È
2
4
3
4
È
È
È
1
2
CoPcS
RuPcS
RuPW
0.1
1
3
4
3
4
10
30
33
68
2
È
2
98
75
65
21
È
25
33
62
È
È
È
6
È
È
È
11
2
2
È
0.1
1
3
4
3
4
10
45
64
72
7
93
43
6
È
55
89
90
È
È
tr
È
2
5
2
2
È
È
È
tr
10
1g
3
92
È
54
14
32
a Conditions: catalyst cyclohexanone 50 mM, and KHSO 0.5 N, in water (non-bu†ered solutions, pH ca. 2); 20 ¡C. Yields determined by 1H
5
NMR and GC. b Consumed (calc.) over available oxidant (equiv./equiv.). Equivalents of oxidant necessary for the formation of each mol of
product from cyclohexanone were calculated to be: e-caprolactone and 6-hydroxyhexanoic acid (2); adipic acid (6); glutaric acid (12); succinic
acid (18). In all reported cases, substrate conversions are quantitative. c Product abbreviation: A, e-caprolactone; B, 6-hydroxyhexanoic acid; C,
adipic acid; D, glutaric acid; E, succinic acid. d Same results with NiPcS, CuPcS and Fe (SO ) (1 mM). e Substrate 31 mM. f Reaction no longer
2
4 3
running. g Substrate 31 mM; KHSO 0.36 N.
5
1320
New J. Chem., 2001, 25, 1319È1324

View Article Online
Table 2 Oxidation of adipic acid by MPcS catalysts and KHSO a
5
Products/mol%b
Substrate
Catalyst
Time/h
6
conv. (%)
D
E
FePcSc
64
64
2
2
2
4
12
2
88
78
CoPcS
RuPcS
6
4
42
57
6
4
23
42
19
81
a Conditions: catalyst 1 mM, substrate 50 mM, and KHSO5
0.6 N, in
water (non-bu†ered solutions, pH ca. 2); 20 ¡C. Yields by 1H NMR.
b Product abbreviation: D, glutaric acid; E, succinic acid. c Oxidation
of succinic acid gave 9% conversion after 24 h.
Fig. 1 Time course for cyclohexanol oxidation. Reaction conditions:
RuPcS catalyst (1 mM), cyclohexanol (45 mM), KHSO (0.5 N in
5
water) 20 ¡C. Cyclohexanol (ÈKÈ); cyclohexanone (È=È); e-
caprolactone (È…È); 6-hydroxyhexanoic acid (È]È); adipic acid
(
- - K - -); glutaric acid (- -= - -); succinic acid (- - … - -).
taric and succinic acid were the only other products detected
in the Ðnal reaction mixture. By lowering the catalyst concen-
tration to 0.1 mM, the reaction deÐnitely slowed down and
the reaction mixtures at 24 h showed formation of adipic acid
the early stages of the reaction and slow formation of the
over-oxidation products of adipic acid (glutaric and succinic
acids) at longer times. With 0.1 mM RuPcS, only a 45% con-
version was achieved after 24 h. Kinetic isotope e†ects for the
(
55%), with only a trace of succinic acid, but signiÐcant
amounts of unreacted e-caprolactone or 6-hydroxyhexanoic
acid. A simple water-soluble derivative of ruthenium, RuPW,
tested for comparison purposes, also e†ected the oxidation of
cyclohexanone, but the reaction was rather unselective,
involving massive over-oxidation (ca. 50%) of the formed
adipic acid.
The good results obtained for the cyclohexanone oxidation
prompted us to test the persulfate oxidation of cyclohexanol
in the presence of FePcS, CoPcS and RuPcS. NiPcS and
CuPcS, which were unable to promote the oxidation of e-
caprolactone or 6-hydroxyhexanoic acid and were also found
to be poor catalysts for the oxidation of cyclohexanol (Table
oxidation of cyclohexanol-d
sured, but with both RuPcS and FePcS at 0.1 mM concentra-
tion, almost undetectable di†erences were found (r /r \ 1.1
for RuPcS and 1.2 for FePcS). The observed absence of detect-
able KIEs strongly indicates a concerted oxygen transfer from
the oxidizing metal species to the alcoholic substrate, likely
coordinated to the metal.22
Hydrogen peroxide oxidations of cyclohexanol and cyclo-
hexanone were conducted at room temperature in an aqueous
phase containing the substrates (50 mM) and an excess of
hydrogen peroxide (ca. 10 equiv. mol~1), both in acidic (pH
ca. 2) and in neutral media, with the MPcS catalysts (M \ Fe,
Ru) added in at 2 mol% with respect to the substrates. Typical
conversions, oxidation rates and product distributions are
reported in Table 4. Contrary to the experiments with per-
sulfate, where yields to the oxidant were generally close to the
stoichiometric requirement, extensive dismutation of H O
vs. cyclohexanol were mea-
1
2
H D
3). In the presence of 1 mM FePcS, cyclohexanol was rapidly
and quantitatively converted into 6-hydroxyhexanoic acid,
which remained however the overwhelmingly major product
even after 24 h; CoPcS behaved better, leading to a 37% yield
of adipic acid (24 h). The reaction in the presence of 1 mM
RuPcS was very fast and complete disappearance of the sub-
strate occurred within minutes, with quantitative formation
2
2
was observed, even in the acidic media, thus considerably lim-
iting the yields of the reactions; cyclohexanol was oxidized
only into cyclohexanone and in very low yields. FePcS caused
the least amount of decomposition of H O (1 and 0.1 mM
(
ca. 90%) of adipic acid after a few hours. The time course for
the oxidation of cyclohexanol (Fig. 1), which was followed by
H NMR (Fig. 2), clearly showed the formation and disap-
pearance of all the expected reaction intermediates
cyclohexanone, e-caprolactone, 6-hydroxyhexanoic acid) in
2
2
1
solutions of FePcS induced dismutation of 0.25 M solutions of
H O with half-times of 12 and 6 h, respectively). However,
2
2
(
the overall conversions of cyclohexanone by FePcS and
Table 3 Oxidation of cyclohexanol by MPcS catalysts and KHSO a
5
Substrate
Product selectivityc/mol%
[
cat]/
conv.
(%)
Yield w.r.t. the
oxidantb/equiv.%
Catalyst
FePcS
mM
Time/h
1
A
B
C
D
E
1
100
35
36
6
È
42
È
52
100
È
tr
È
È
24
CoPcS
1
1
4
34
100
10
49
35
È
64
È
1
63
È
37
È
trd
2
NiPcS
CuPcS
RuPcS
1
1
1
24
24
20
5
5
2
65
È
35
tr
È
È
È
È
100
È
0.5
75
100
12
22
87
4
3
È
30
È
59
È
70
38
È
È
È
86
È
65
È
14d
È
È
2
2
4
0
.1
1
4
45
27
È
35
a Conditions: catalyst, cyclohexanol, 45 mM, and KHSO , 0.5 N, in water (non-bu†ered solutions, pH ca. 2); 20 ¡C. Yields determined by 1H
5
NMR and GC. In the absence of catalysts, 1% conversion of the substrate to cyclohexanone within 24 h. b Consumed (calc.) over available
oxidant (equiv./equiv.). Equivalents of oxidant necessary for the formation of each mol of product from cyclohexanol were calculated to be:
cyclohexanone (2), e-caprolactone and 6-hydroxyhexanoic acid (4); adipic acid (8); succinic acid (20). c Product abbreviation: A, cyclohexanone;
B, e-caprolactone; C, 6-hydroxyhexanoic acid; D, adipic acid; E, succinic acid. d Also traces of glutaric acid.
New J. Chem., 2001, 25, 1319È1324
1321

View Article Online
Table 4 Oxidation of cyclohexanone and cyclohexanol by MPcS
catalysts and H O a
2
2
Product selectivityb,c/mol%
Substrate
conv. (%)
Catalyst
Cyclohexanone
A
B
C
FePcS
21d
6
1
99
99
1
1
15
8e
00f
RuPcS
10
50
5
1
Cyclohexanol
FePcS
RuPcS
16d
5
100
100
a Conditions: catalyst 1 mM, substrate (cyclohexanone 42 mM or
cyclohexanol 45 mM) and H O 0.5 N, in water (non-bu†ered solu-
2
2
tions, pH ca. 7, unless otherwise stated); 20 ¡C, unless otherwise
stated. Yields at 24 h, determined by 1H NMR and GC. No detect-
able substrate conversion within 24 h in the absence of catalyst or in
the presence of NiPcS, CuPcs, RuPW, RuDMS or Fe (SO ) (1 mM)
2
4 3
or RuPcS or FePcS catalysts (0.1 mM). b CO , CO and other VOCs
2
are also present in the gas phase. c Product abbreviation: A, 6-
hydroxyhexanoic acid; B, adipic acid; C, cyclohexanone. d Oxidant
still present. e AcidiÐed (pH ca. 2) by adding H SO . f At 60 ¡C.
2
4
Fig. 2 1H NMR spectra during cyclohexanol oxidation (for reaction
conditions, see Fig. 1): a: t \ 0 min; b: t \ 60 min; c: t \ 180 min;
d: t \ 24 h; A \ e-caprolactone, B \ 6-hydroxyhexanoic acid, C \
adipic acid, D \ glutaric acid, E \ succinic acid, F \ cyclohexanone,
I \ TSP (reference in a co-axial capillary tube).
hexane to adipic acid. The oxidations were conducted in a
double phase system, with the catalysts and persulfate dis-
solved in the aqueous phase and the organic phase consisting
of the neat substrate, in large excess to the oxidant. Addition
of conventional phase-transfer agents, such as tri-
methylcetylammonium hydrogen sulfate, was useless, because
of fast oxidation of the organic chain. Nevertheless, oxidation
of cyclohexane also took place in the absence of phase-
transfer agents, although slowly (days) and in the presence of
high concentrations (1È5 mM) of the catalysts (Table 5).
FePcS and CoPcS were poor catalysts, with small overall con-
version of cyclohexane. Moreover, FePcS led to the formation
of 6-hydroxyhexanoic acid only, whereas CoPcS promoted an
e†ective over-oxidation of the formed adipic acid. With
RuPcS (1 mM) were comparable, likely due to an intrinsically
higher catalytic activity of RuPcS. In all cases yields in adipic
acid were deÐnitely small. Higher temperature (60 ¡C) sped up
the reactions, leading to higher conversions of the substrate,
but with low yields of adipic acid.
Oxidation of cyclohexane
The promising results obtained with persulfate prompted us
to test the possibility of a direct, one-step oxidation of cyclo-
Table 5 Oxidation of cyclohexane by metal catalysts and KHSO a
5
Productsc/mM (in the aqueous phase)
[cat]/
Yield w.r.t. the
Catalyst
FePcS
mM
Time/h
24
oxidantb/equiv.%
A
B
C
D
5
18
20
15
17
È
È
È
È
È
È
48
CoPcS
RuPcS
5
5
24
6
7
È
È
1.9
2.5
tr
tr
0.5
0.5
48
24
8
17
22
8
tr
È
È
2.4
1.3
4.0
8.5
10
2.8
3.4
È
È
tr
È
tr
È
tr
0.4
tr
4
7
8
2
1
5
24
48
9
0.2
RuPW
24
28
53
100
8
9
12
È
È
È
4.5
3.4
2.5
11
18
25
1.5
2.4
3.6
È
È
6
È
È
È
1.4
3.8
7.0
È
È
0.3
4
7
8
2d
1
5
24
4
7
8
2
RuDMS
24
8e
41
53
È
È
12.5
13
tr
3.0
3.7
4.0
4
a Conditions: 2.5 mL of water (non-bu†ered, pH ca. 2) containing the catalyst and persulfate 0.5 N, stirred with 0.5 mL of cyclohexane. Yields
determined by 1H NMR. b Consumed (calcd) over available oxidant. Equivalents of oxidant necessary for the formation of each mol of product
from cyclohexane were calculated to be: 6-hydroxyhexanoic acid (6); adipic acid (10); glutaric acid (16); succinic acid (22). c Product abbrevia-
tion: A, 6-hydroxyhexanoic acid; B, adipic acid; C, glutaric acid; D, succinic acid. d On addition of a further 5 mequiv. of fresh oxidant to the
reaction mixture and reaction for a further 72 h, the following products are detected: adipic acid, 25; glutaric acid, 6; succinic acid, 45 mM.
e Reaction no longer running after 48 h.
1322
New J. Chem., 2001, 25, 1319È1324

View Article Online
RuPcS progressive and selective accumulation of adipic acid
was observed in the aqueous phase, accounting for a ca. 20%
yield after 72 h (calculated with respect to the oxidant), with
the reaction still running; only smaller amounts (5%) of glu-
taric and succinic acid were detected in the reaction mixtures.
Two other ruthenium derivatives, RuDMS and RuPW, both
soluble in water, were tested for comparison purposes; adipic
acid was produced in both cases, although with lower selec-
tivity. Moreover, in the presence of RuDMS the reaction
stopped at ca. 50% of the available oxidant, likely because of
poisoning of the catalyst (vide supra); on the contrary, RuPW
was active for several days, up to quantitative consumption of
the available oxidant.
Fate of the catalyst
Scheme 1 Proposed reaction pathway for the oxidation of cyclo-
In all reported cases, the metal complexes used underwent
rapid degradation in the oxidizing reaction media, giving rise
to new species, likely the catalytically active ones; the fact that
di†erent catalytic behaviors were exhibited (for example, by
the various ruthenium derivatives) strongly indicates that deg-
radation did not lead to the same, totally uncomplexed
aqueous metal species. Only RuDMS behaved as a true pre-
hexane.
hydroxyhexanoic acid) and Ðnally the catalytic oxidation of
6-hydroxyhexanoic acid to adipic acid. In the case of RuPcS,
fast and quantitative conversions of cyclohexanone or cyclo-
hexanol were observed at catalyst loadings down to 0.2%,
with selectivities in adipic acid of up to 90%. RuPcS is also
active and equally selective for the direct oxidation of cyclo-
hexane to adipic acid, whereas RuPW and RuDMS shows
poorer selectivity. Oxidation of 6-hydroxyhexanoic acid
appears to be the crucial step, which was e†ectively promoted
only by CoPcS and the ruthenium derivatives. However, since
oxidation of cyclohexanol to cyclohexanone by CoPcS was
scarcely e†ective and major problems arose with RuPW and
RuDMS, because of e†ective decarboxylation of the formed
adipic acid, RuPcS turns out to be the pre-catalyst of choice
for the overall process, leading to adipic acid when starting
from both cyclohexanol and cyclohexane with acceptable
selectivity (Fig. 3).
As the authors of a very recent review claim in the title
itself, ““cyclohexane oxidation continues to be a challengeÏÏ,31
a survey of the recent literature on liquid phase oxidations of
cyclohexane to adipic acid did indeed show conÑicting results.
With molecular oxygen moderate conversions, low selectivity
in adipic acid and, often, severe reaction conditions were gen-
erally reported;5,6,32 when milder conditions were used, only
cyclohexanol or cyclohexanone were detected as oxidation
products.33,34 Persulfate oxidation of cyclohexane or cyclo-
hexanol, promoted by manganese porphyrins, only led, and
after long reaction times (up to 160 h), to e-caprolactone;35 on
the contrary, oxidations e†ected by hydrogen peroxide (in the
presence of heteropolytungstates18) or by TBHP (over cerium-
containing silicates36) produced cyclohexanol and cyclo-
hexanone (or cyclohexyl hydroperoxide). The good selectivity
in adipic acid reported for the present RuPcS pre-catalyst and
cursor of a “nakedÏ metal species, likely RuO ; in the presence
4
of persulfate the diagnostic 1H NMR signals of the RuDMS
complex23 were completely replaced within 30 min by a sharp
singlet at 3.05 ppm, assigned to dimethylsulfone, a poor metal-
coordinating ligand.
The fate of the RuPW catalyst was less clear-cut; 31P NMR
spectra (measured under the same reaction conditions, i.e. at
pH ca. 2 for H SO ) showed complete disappearance of the
2
4
diagnostic signal of the compound, at [11.4 ppm, within 1 h
upon addition of persulfate. A new signal appeared as a
singlet at [12.8 ppm, whereas no other species, apart from
moderate amounts of H PO (singlet at ]1.1 ppm), were
3
4
detected after 48 h reaction. The nature of the new species is
elusive, although the above data should rule out massive frac-
tionation of the heteropolyoxometallate moiety, at least to the
level described in the literature for simpler Keggin derivatives;
polyoxometallates are indeed generally considered to be very
sensitive to peroxolytic degradation,24,25 but signiÐcant and
recent exceptions have been reported.26h28 The catalytically
active solutions containing the “RuPWÏ compound could be
reused in the experiments of cyclohexane oxidation, but at
longer reaction times succinic, and not adipic acid, was the
dominant product.
RuPcS (and the other MPcS derivatives used) also behaved
as a pre-catalyst; we have already reported10 that peroxidic
oxidants rapidly and irreversibly turn the deeply colored com-
pounds into yellow species, with complete disappearance of
the diagnostic absorptions due to the p-conjugated aromatic
system of the phthalocyanine ring (Q band). This change is
accompanied by the formation of ammonia, roughly in a 1 : 1
molar ratio with the complex. This unexpected feature sug-
gests that selective cleavage at one of the nitrogen atoms of
the phthalocyanine ring occurs, strongly reminiscent of the
celebrated oxidative degradation of heme to verdoheme and
biliverdin29 and also observed in relevant iron porphyrin
model compounds.30 Detailed investigations are under way,
aiming to deÐne the nature of the species resulting upon deg-
radation of the phthalocyanine ring.
Conclusion
The oxidation pathway for the present catalytic cyclohexane
oxidation consists of the conventional multistep process
(
Scheme 1), involving two catalytic oxidations (Ðrst to cyclo-
hexanol and then to cyclohexanone), followed by two
uncatalyzed steps (BaeyerÈVilliger oxidation of cyclohexanone
to e-caprolactone and acid hydrolysis of the latter to 6-
Fig. 3 Bar chart comparing the product selectivity (mol%) upon per-
sulfate oxidation of cyclohexane with various metal catalysts (48 h
reaction).
New J. Chem., 2001, 25, 1319È1324
1323

View Article Online
the mild conditions used may allow one to envisage its use as
an alternative possibility for the direct synthesis of adipic acid.
18 M. M. Q. Simoes, C. M. M. Conceicao, J. A. F. Gamelas,
P. M. D. N. Domingues, A. M. V. Cavaleiro, J. A. S. Cavaleiro,
A. J. V. Ferrer-Correia and R. A. W. Johnstone, J. Mol. Catal. A:
Chem., 1999, 144, 461.
19
0
J. H. Weber and D. H. Busch, Inorg. Chem., 1965, 4, 469.
I. P. Evans, A. Spencer and G. Wilkinson, J. Chem. Soc., Dalton
T rans., 1973, 204.
Acknowledgements
We thank Mr A. Ravazzolo, Centro CNR Composti di
Coordinazione, Padova, for technical assistance. We acknow-
2
21
22
23
24
M. Bressan, L. Forti, F. GhelÐ and A. Morvillo, J. Mol. Catal.,
1
993, 79, 85.
ledge the support of the Ministero dellÏUniversita
MURST-PRIN 1998).

, Italy
R. Mello, L. Cassidei, M. Fiorentino, C. Fusco, W. Hummer, V.
Jager and R. Curci, J. Am. Chem. Soc., 1991, 113, 2205.
E. Alessio, G. Mestroni, G. Nardin, W. M. Attia, M. Calligaris,
G. Sava and S. Zorzet, Inorg. Chem., 1988, 27, 4099.
C. L. Hill and C. M. Prosser-McCartha, Coord. Chem. Rev., 1995,
143, 407.
(
References
1
2
M. Freemantle, Chem. Eng. News, 2000, 78, 7.
K. D. Draths and J. W. Forst, in Green Chemistry, ed. P. T.
Anastas and T. C. Williamson, Oxford University Press, Oxford,
UK, 1998, p. 58.
25 L. Salles, C. Aubry, R. Thouvenot, F. Robert, C. Doromieux-
Morin, G. Gottard, H. Ledon, Y. Jeannin and J.-M. Bregeault,
Inorg. Chem., 1994, 33, 871.
26 R. Neumann and A. M. Khenkin, J. Mol. Catal. A: Chem., 1996,
114, 169.
27 W. P. Griffith, N. Morley-Smith, H. I. S. Nogueira, A. G. F.
Shoair, M. Suriaatmaja, A. J. P. White and D. J. William, J.
Organometal. Chem., 2000, 607, 146.
28 J. Server-Carrio, J. Bas-Serra, M. E. Gonzalez-Nunez, A. Garcia-
Gastaldi, G. B. Jameson, L. C. W. Baker and R. Acerete, J. Am.
Chem. Soc., 1999, 121, 977.
3
C. Xu, J. Oyloe and D. E. Richardson, Abstracts of the 218th
Meeting of the American Chemical Society, ACS, Washington,
DC, 1999, p. 569-INOR.
4
5
6
7
8
9
0
1
2
I. Belkhir, A. Germain, F. Fajula and E. Fache, J. Chem. Soc.,
Faraday T rans., 1998, 94, 1761.
M. Dugal, G. Sankar, R. Raja and J. M. Thomas, Angew. Chem.,
Int. Ed., 2000, 39, 310 and references therein.
A. Lambert, D. J. Macquarrie, G. Carr and J. H. Clark, New J.
Chem., 2000, 24, 485.
29 L. Avila, H.-W. Huang, J. C. Rodriguez, P. Moenne-Loccoz and
M. Rivera, J. Am. Chem. Soc., 2000, 122, 7618 and references
therein.
M. Bressan, L. Forti and A. Morvillo, New J. Chem., 1995, 19,
9
51.
M. Bressan, L. Forti and A. Morvillo, J. Chem. Soc., Chem.
Commun., 1994, 253.
30 T. N. St. Claire and A. L. Balch, Inorg. Chem., 1999, 38, 684 and
references therein.
M. Bressan, N. dÏAlessandro, L. Liberatore and A. Morvillo,
Coord. Chem. Rev., 1999, 185–186, 385.
31 U. Schuchardt, D. Cardoso, R. Sercheli, R. Pereira, R. S. de Cruz,
M. C. Guerreiro, D. Mandelli, E. V. Spinace and E. L. Fires,
Appl. Catal. A, 2001, 211, 1.
32 M. Besson, A. Blackburn, P. Gallezot, O. Kozynchenko, A.
Pigamo and S. Tennison, T op. Catal., 2000, 13, 253.
33 K. Teramura, T. Tanaka, T. Yamamoto and T. Funabiki, J. Mol.
Catal. A: Chem., 2001, 165, 299.
34 N. Perkas, Y. Q. Wang, Y. Koltypin, A. Gedanken and S. Chan-
drasekaran, Chem. Commun., 2001, 998.
35 L. Cammarota, S. Campestrini, M. Carrieri and F. Di Furia, J.
Mol. Catal. A: Chem., 1999, 137, 155.
36 E. L. Pires, U. Arnold and U. Schuchardt, J. Mol. Catal. A:
Chem., 2001, 169, 157 and references therein.
1
1
1
M. Bressan, N. Celli, N. dÏAlessandro, L. Liberatore, A. Morvillo
and L. Tonucci, J. Organomet. Chem., 2000, 593–594, 416.
N. dÏAlessandro, L. Liberatore, L. Tonucci, A. Morvillo and M.
Bressan, J. Mol. Catal. A: Chem., 2001, in press.
A. Hadash, A. Sorokin, A. Rabion and B. Meunier, New J.
Chem., 1998, 22, 45.
1
1
3
4
A. Hadash and B. Meunier, Eur. J. Inorg. Chem., 1999, 2319.
A. Sorokin, L. Fraisse, A. Rabion and B. Meunier, J. Mol. Catal.
A: Chem., 1997, 117, 103.
1
1
5
6
R. Raja and P. Ratnasamy, Catal. L ett., 1997, 48, 1.
E. Armengol, A. Corma, V. Fornes, H. Garcia and J. Primo,
Appl. Catal. A, 1999, 181, 305.
1
7
C. Nozaki, M. Misono and N. Mizuno, Chem. L ett., 1998, 1263.
1324
New J. Chem., 2001, 25, 1319È1324