Fluorous biphasic catalytic oxidation of sulfides by molecular oxygen/2,2-dimethylpropanal

Article abstract of DOI:10.1002/1099-0690(200101)2001:1<181::AID-EJOC181>3.0.CO;2-3

The use of perfluoroalkyl-substituted cobalt complexes as catalysts for the oxidation of alkyl aryl sulfides with molecular oxygen/2,2-dimethypropanal has been studied in a fluorous organic biphasic system. The addition of very small amounts of a Co¹¹-tetraarylporphyrin (Co-4) led to increased substrate conversions (67-100%). Sulfoxide was generally obtained as the main product, together with variable quantities of sulfone (0-100%), depending on the nature of the substrate. A perfluoroalkyl-substituted Co¹¹-phthalocyanine (Co-6) proved to be less efficient with regard to substrate conversion (40-78%), but afforded sulfoxides selectively. Although the mechanism has not been investigated in detail, the reaction probably proceeds through a free-radical oxidative process, initiated by the Co¹¹ complexes. The attempts at recycling the catalysts through phase separation were partly ineffectual owing to their instability under the reaction conditions, which is more pronounced in the case of Co-6.

Full text of DOI:10.1002/1099-0690(200101)2001:1<181::AID-EJOC181>3.0.CO;2-3

FULL PAPER
Fluorous Biphasic Catalytic Oxidation of Sulfides by
Molecular Oxygen/2,2-Dimethylpropanal
Stefano Colonna,[a] Nicoletta Gaggero,[a] Fernando Montanari,[b] Gianluca Pozzi,*[b] and
Silvio Quici[b]
Dedicated to the memory of Professor Shigeru Oae
Keywords: Cobalt / Biphasic catalysis / Porphyrinoids / Phthalocyanines / Oxidations
The use of perfluoroalkyl-substituted cobalt complexes as
6) proved to be less efficient with regard to substrate conver-
catalysts for the oxidation of alkyl aryl sulfides with molecu- sion (40−78%), but afforded sulfoxides selectively. Although
lar oxygen/2,2-dimethypropanal has been studied in
fluorous organic biphasic system. The addition of very small
amounts of a CoII−tetraarylporphyrin (Co-4) led to increased
a
the mechanism has not been investigated in detail, the reac-
tion probably proceeds through a free-radical oxidative pro-
cess, initiated by the CoII complexes. The attempts at recyc-
substrate conversions (67−100%). Sulfoxide was generally ling the catalysts through phase separation were partly inef-
obtained as the main product, together with variable quantit-
ies of sulfone (0−100%), depending on the nature of the sub-
strate. A perfluoroalkyl-substituted CoII−phthalocyanine (Co-
fectual owing to their instability under the reaction condi-
tions, which is more pronounced in the case of Co-6.
Introduction
systems, both in the absence[6] and in the presence[7] of a
reducing agent such as a branched alkyl aldehyde. More
´
´
recently, Begue and his group discovered the beneficial role
played by hexafluoro-2-propanol in the stoichiometric ox-
idation of sulfides to sulfoxides with aqueous 30% hydrogen
peroxide.[8] High yields and complete selectivities for sulfox-
ides were obtained under mild reaction conditions. The use
of solid Mn(OAc)3·2H2O as a catalyst for the oxidation of
sulfides with dioxygen/2,2-dimethypropanal in perfluoro-2-
butyltetrahydrofuran was also examined, but in this case
no particular solvent effect was observed.[8a] The positive
features expected from the combination of the FB approach
with the use of oxygen as primary oxidant were possibly
hidden by the heterogeneous nature of the catalyst. Our
long-standing interest in the biomimetic oxidation of sulf-
ides prompted us to investigate the use of the cobalt com-
plexes of perfluoroalkylated tetradentate N-ligands (tetraar-
ylporphyrins and phthalocyanines) as homogeneous FB
catalysts for this reaction. We describe here the results ob-
tained with two of these complexes, as well as a new effect-
ive approach to the synthesis of tetraarylporphyrin 4
(Scheme 1), previously prepared by a rather cumbersome
procedure.[3a]
The concept of Fluorous Biphasic (FB) catalysis is cur-
rently attracting considerable interest, in view of the poten-
tial advantages over classical homogeneous catalytic sys-
tems.[1] The easy separation and recycling of the perfluorin-
ated organometallic complexes used in FB catalytic systems
are perhaps the most appealing features of this technique.
The peculiar physico-chemical properties of the perfluoro-
carbons used as reaction media also suggest new intriguing
possibilities, e.g. in terms of unexpected selectivities due to
the unique solvation environment.[2] In the past, we took
advantage of the chemical inertness of perfluorocarbons
and of their exceptionally high affinity for dioxygen in the
FB catalytic oxidation of hydrocarbons.[3] Another import-
ant oxidative process ideally suited for FB modification is
the conversion of sulfides to sulfoxides or sulfones. This re-
action has been thoroughly investigated for many years and
the use of several oxidizing agents combined with classic
catalytic as well as enzymatic systems have been pro-
posed.[4,5] For obvious reasons, those systems relying on the
use of molecular oxygen or air would be most attractive,
provided that acceptable reactivities and selectivities were
obtained. The combination of perfluorinated solvents and
catalysts has thus been explored by a few groups that briefly
reported the aerobic oxidation of sulfides in FB catalytic
Results and Discussion
Synthesis of the Catalysts
[a]
`
Istituto di Chimica Organica, Facolta di Farmacia,
Tetraarylporphyrins are versatile ligands able to complex
almost all transition-metal cations. Such complexes have
been used as catalysts for a variety of biomimetic oxidation
reactions, including the selective conversion of sulfides into
sulfoxides or sulfones.[9] Three years ago we achieved the
via Venezian 21, 20133 Milano, Italy
[b]
Centro CNR Sintesi
Organici,
via Golgi 19, 20133 Milano, Italy
Fax: (internat.) ϩ 39-02/266-3354
E-mail: gianluca.pozzi@unimi.it
e Stereochimica di Speciali Sistemi
Eur. J. Org. Chem. 2001, 181Ϫ186
WILEY-VCH Verlag GmbH, 69451 Weinheim, 2001
1434Ϫ193X/01/0101Ϫ0181 $ 17.50ϩ.50/0
181

S. Colonna, N. Gaggero, F. Montanari, G. Pozzi, S. Quici
FULL PAPER
Transition-metal complexes of phthalocyanines are an
important group of macrocyclic compounds endowed with
properties close to those of metalloporphyrins.[12] A perflu-
orocarbon-soluble CoIIϪphthalocyanine obtained through
direct free-radical perfluoroalkylation of CoIIϪphthalo-
cyanine with C8F17I was found to catalyze the aerobic FB
oxidation of hydrocarbons and sulfides in the absence of a
reducing agent.[6] In this case, however, conversions were
low and the precise structure of the ligand (i.e. the number
and the location of the perfluoroalkyl substituents on the
macrocycle) was not reported. An alternative synthetic
pathway leading to the better-defined perfluorocarbon-sol-
uble CoIIϪphthalocyanine 6 is reported in Scheme 2.[13]
Complex Co-6 was easily obtained when 1,2-dicyano-4-(n-
perfluorooctyl)benzene (5) was heated with finely divided
cobalt metal in a tightly closed vessel, according to a clas-
sical procedure for the synthesis of metallophthalocyanines
(Scheme 2).[14] The presence of a single C8F17 substituent
on each of the four benzene rings of the phthalocyanine is
ensured by the choice of the perfluoroalkylated precursor.
Scheme 1
synthesis of the highly fluorinated ligand 4 and of the cor- We did not try to determine the substitution pattern of Co-
responding CoII complex which are both selectively soluble 6 and the fluorinated chains are therefore shown at ran-
in perfluorocarbons.[3a] Unfortunately, the preparation of 4 dom positions.
was complicated by the number of low-yielding synthetic
steps required, and the scope of this approach was therefore
considerably restricted.
The absence of 4 (or Co-4) in the organic phase of vari-
ous FB systems was checked taking advantage of the char-
acteristic, intense UV/Vis absorption shown by porphyrins
and their metal complexes (Soret band, λ ϭ 400Ϫ425 nm,
ε in the order of 104Ϫ105 dm3 molϪ1 cmϪ1). Phthalocyan-
ines also strongly absorb in the UV/Vis region, so traces of
these compounds can be easily detected in solution.
As repeatedly highlighted by Curran and co-workers,
partly fluorinated solvents such as α,α,α-trifluorotoluene
can have a beneficial influence on the course of reactions
involving highly fluorinated reagents.[10] It was thus decided
Scheme 2
to undertake the synthesis of 4 through Lewis acid cata-
lyzed condensation of pyrrole and 3,5-bis(n-perfluorooctyl)-
benzaldehyde (3), using a partly fluorinated solvent as the
The fluorine load of Co-6 (F ϭ 57.6%) is lower than that
reaction medium (Scheme 1). Aldehyde 3 was readily pre- required to make other oxidation catalysts (e.g.
pared by the copper-catalyzed coupling reaction of methyl MnIIIϪtetraarylporphyrins or MnIIIϪsalen complexes)
3,5-dibromobenzoate with C8F17I (92% yield), followed by preferentially soluble in perfluorocarbons (F Ͼ 60%).[3]
reduction of the perfluoroalkylated ester 1 to benzyl alcohol Nevertheless, Co-6 easily dissolves in n-perfluorooctane,
2 (90% yield) and selective oxidation (73% yield), according giving deep-blue solutions, whereas it is completely insol-
to a procedure that we developed for a related salicylal- uble in other organic solvents such as acetone and alcohols,
dehyde.[3d]
as estimated by visual tests. Diethyl ether is an exception,
Preliminary tests run in common organic solvents but the solubility of Co-6 in this solvent is much lower than
showed that condensation of aldehyde 3 and pyrrole fails in perfluorooctane (1.3 mg mLϪ1 versus Ͼ 15 mg mLϪ1).
to give porphyrin 4 under a variety of conditions. The re- The partition coefficient of Co-6 in n-perfluorooctane/
placement of CH2Cl2 with CFCl2CFCl2 allowed for the re- CH2Cl2 (1:1, v/v) is 25.6 at 20 °C, as indicated by a decrease
action to proceed as desired, under conditions otherwise in the intensity of the characteristic UV/Vis absorption
similar to those used by Lindsey and co-workers in the syn- band at 654 nm after stirring a solution of Co-6 in n-per-
thesis of a number of tetraarylporphyrins.[11] Porphyrin 4 fluorooctane (2.86 ϫ 10Ϫ4 ) with CH2Cl2 for 2 h. Taking
was obtained in 23% yield after purification by column only the solubility of Co-6 in CH2Cl2 (which is just
chromatography on silica gel with CFCl2CFCl2 as the elu- 0.04 mg mLϪ1) into account, a partition coefficient higher
ent. The cobalt complex of 4 was then prepared as previ- than 366 would be expected. The marginal leaching of the
ously described.[3a]
complex into the organic phase of the FB system can thus
182
Eur. J. Org. Chem. 2001, 181Ϫ186

Catalytic Oxidation of Sulfides
FULL PAPER
be mainly ascribed to the low, but not negligible, mutual
miscibility of the two solvents.[15]
The phthalocyanine complex Co-6 is less effective than
Co-4 at S/C ϭ 1000, giving a substrate conversion of 40%
The remarkably high fluorous affinity of Co-6 for perflu- in 4 h (Table 1, Entry 4). However, only sulfoxide was ob-
orocarbons is probably related to its electroneutrality, which tained, in agreement with the reported relationship between
is in contrast with the cationic nature of incomplete conversion and selectivity in solid/liquid FB sys-
MnIIIϪtetraarylporphyrins and MnIIIϪsalen complexes. In- tems.[8a] On this time scale, only small amounts of substrate
deed, the very low dielectric constants of perfluorocarbons were oxidized in FB blank experiments, as was also found
diminishes the solubilization of charged complexes, even by other authors investigating homogeneous reactions run
though being highly fluorinated. The present finding also in dichloroethane.[17]
confirms the tentative nature of all predictions of the solu-
Rather unexpectedly, catalyst recycling according to the
bility behaviour of a molecule in perfluorocarbons, when well-established liquid/liquid FB protocol turned out to be
only the fluorine load is taken into account.[3d]
unsuccessful. The fluorous phase, which was recovered by
simply decanting from the organic solution containing the
products, was added to a fresh CH2Cl2 solution of sulfide
and treated with O2/2,2-dimethypropanal. As shown in
Table 1, in the case of Co-6, the substrate conversion de-
creased markedly without any significant variation in the se-
lectivity when the fluorous layer was reused (Entry 5). Leach-
ing of Co-6 into the organic phase could not explain this
drop of activity in view of the high partition coefficient of
this complex in n-perfluorooctane/CH2Cl2. However, de-
gradation of CoIIϪphthalocyanines into several unidentified
fragments was noticed in oxidation reactions proceeding un-
der free-radical conditions.[18] When occurring in an FB sys-
tem, a similar degradation would obviously prevent the re-
cycling of the fluorous phase, as indeed we observed. The
complete oxidation of methyl p-tolyl sulfide was still achieved
in reactions catalyzed by Co-4 (Entries 2 and 3), but the
sulfone/sulfoxide ratio increased significantly after the first
recycling of the fluorous phase, and only sulfone was ob-
tained in the subsequent recycling. The change in selectivity
was paralleled with the disappearance of the characteristic
split UV/Vis absorption band of Co-4 (λmax ϭ 405 and
427 nm) and by the emergence of a new band at 438 nm
(assigned to [CoIII-4]ϩ)[19] that also disappeared after the
third run. These observations indicated that the nature of the
catalytic species involved in the oxidation reaction was no
longer the same in the recycling steps. This was not found in
the epoxidation of alkenes with O2/aldehyde; in this case, the
FB conditions enabled us to recycle Co-4 as well as other
fluorous complexes employed as catalysts for the reaction.[3]
A series of para-substituted methyl phenyl sulfides were
oxidized in the presence of the two complexes under the FB
conditions described above. Data shown in Table 2 con-
firmed that Co-4 is a more active catalyst than Co-6, which
in turn promotes the selective oxidation of most substrates
to sulfoxides, with the remarkable exception of methyl p-
nitrophenyl sulfide (Entry 8, Table 2).
FB Oxidation of Sulfides
The catalytic oxidation of methyl p-tolyl sulfide (1 mmol)
with molecular oxygen (1 atm) in the presence of 2,2-di-
methylpropanal (3 mmol) was chosen as a model reaction in
order to test the feasibility of the FB approach. Reactions
catalyzed by Co-4 or Co-6 were carried out in the dark at
20 °C under O2, with the catalyst dissolved in the fluorous
phase and the substrate dissolved in CH2Cl2. 2,2-Dimethyl-
propanal was added in small portions to the vigorously
stirred biphasic mixture until the aldehyde/starting sulfide
molar ratio was equal to 3, and the reaction progress was
1
followed by H NMR spectroscopy of samples drawn from
the organic layer. The porphyrin complex Co-4 effectively
catalyzes the oxidation of methyl p-tolyl sulfide and the re-
action was complete in 4 h, even with a very high substrate/
catalyst ratio (S/C ϭ 1000, Table 1, Entry 1). This figure
compares favourably with the substrate/catalyst ratio (S/
C ϭ 35) of a preceding example of liquid/liquid FB oxida-
tion of sulfides with O2/aldehyde,[7] and with the S/C ϭ
20Ϫ60 commonly reported for similar homogeneous cata-
lytic reactions.[9,16] The selectivity for sulfoxide was higher
´
´
than that observed by Begue and co-workers in the oxida-
tion of ethyl phenyl sulfide catalyzed by solid
Mn(OAc)3·2H2O in perfluoro-2-butyltetrahydrofuran,[8a]
but partial formation of sulfone could not be avoided.
Table 1. Catalytic oxidation of methyl p-tolyl sulfide with O2/2,2-
dimethylpropanal under fluorous two-phase conditions
Entry Catalyst Conversion (%) Selectivity[a]
Sulfoxide (%) Sulfone (%)
1
Co-4
Co-4
Co-4
Co-6
Co-6
Co-6
Co-6
100
100
100
40
94
6
2[b]
3[c]
4
42
58
100
Ϫ
3
Ϫ
The influence of para substituents on the selectivity of
reactions catalyzed by Co-4 was more evident, even though
a definite trend could not be discerned. The formation of
sulfone was enhanced by the strong electron-donating sub-
stituent ϪOCH3 (Entry 9, Table 2), but the strong electron-
withdrawing substituent ϪNO2 behaved similarly (Entry 7,
Table 2). Furthermore, the formation of sulfoxide was fa-
voured by halogen substituents (Entries 3 and 5, Table 2)
rather than by ϪH (Entry 1, Table 2) or ϪCH3 (Entry 1,
Table 1).
100
5[d]
6[e]
7[f]
11
97
27
6
100
100
Ϫ
Ϫ
[a]
Reaction conditions: see Experimental Section; conversion and
selectivity (yield/conversion) were determined by 1H NMR spectro-
[b]
scopy. Ϫ
Reaction run with the fluorous layer recovered from
[c]
Entry 1. Ϫ Reaction run with the fluorous layer recovered from
[d]
Entry 2. Ϫ Reaction run with the fluorous layer recovered from
[e]
Entry 4. Ϫ Reaction run in the presence of 1 mol-% of 4-tert-
[f]
butylcatechol. Ϫ Reaction run in the presence of 5 mol-% of 4-
tert-butylcatechol.
Eur. J. Org. Chem. 2001, 181Ϫ186
183

S. Colonna, N. Gaggero, F. Montanari, G. Pozzi, S. Quici
Table 2. Catalytic oxidation of para-substituted methyl phenyl sulfides with O2/2,2-dimethylpropanal under fluorous two-phase conditions
FULL PAPER
Entry
Catalyst
Substrate
Conversion (%)[a]
Selectivity[a]
Sulfoxide (%)
Sulfone (%)
1
2
3
4
5
6
7
8
9
10
Co-4
Co-6
Co-4
Co-6
Co-4
Co-6
Co-4
Co-6
Co-4
Co-6
PhSCH3
82
90
10
Ϫ
Ϫ
Ϫ
5
Ϫ
90
85
100
Ϫ
PhSCH3
68
100
100
100
95
p-ClPhSCH3
p-ClPhSCH3
p-FPhSCH3
p-FPhSCH3
p-NO2PhSCH3
p-NO2PhSCH3
p-OCH3PhSCH3
p-OCH3PhSCH3
67
58
100
49
100
100
100
59
100
10
15
Ϫ
100
[a] Reaction conditions: see Experimental Section; conversion and selectivity (yield/conversion) were determined by 1H NMR spectroscopy.
These results are opposite to those obtained with biomi-
metic catalytic systems for which a heterolytic mechanism
Conclusions
is now generally accepted. Indeed, microsomal cytochrome
P-450 and metallotetraarylporphyrins promote the conver-
sion of sulfides into sulfoxides with oxygen donors such as
O2/NADPH, PhIO and H2O2.[9c,20,21] In the presence of ex-
cess oxidizing agent, sulfoxides are subsequently converted
into sulfones, and therefore mixtures of products are usually
obtained. Electron-donating substituents increase the rate
of oxidation of para-substituted methyl phenyl sulfoxides
and an inverse linear relationship between the one-electron
oxidation potential of the substrate and log(Vmax) has been
found for microsomal cytochrome P-450 catalyzed oxida-
tions with O2/NADPH.[20b] The involvement of sulfur-
centred radical cations, generated by the interaction of the
substrate with a catalytic high-valent oxoiron species was
first suggested,[20] but more recently convincing evidence for
The limited examples of catalytic aerobic oxidation of
sulfides under fluorous biphasic conditions reported up to
now indicate that this new methodology offers a viable ac-
cess to sulfoxides and sulfones. We have started to explore
the use of perfluoroalkylated metallotetraarylporphyrins
and -phthalocyanines, and a first comparison can now be
made with the other known FB catalysts. The cobalt(II)
complex of porphyrin 4 gave good results in terms of select-
ivity for sulfoxide in comparison with solid
Mn(OAc)3·2H2O,[8a] being only slightly inferior to the
nickel(II) complex of a perfluorinated 1,3-diketone reported
by Knochel and co-workers.[7] Co-4 was active at an S/C
ratio of 1000, which is the highest value reported for a FB
system. Although less active than Co-4, the phthalocyanine
complex Co-6 showed complete selectivity for sulfoxide, ex-
cept in the oxidation of methyl 4-nitrophenyl sulfide. Des-
pite the good chemoselectivity and efficiency observed, the
two catalysts investigated here cannot withstand the free-
radical conditions under which the FB oxidation of sulfides
probably proceeds. Further modification of the peripheral
structure of the perfluoroalkylated macrocycles is thus re-
quired to increase their robustness and to enhance the re-
coverability of the FB catalysts.
the direct oxygen transfer from
a
high-valent
oxoironϪporphyrin radical cation to the neutral sulfide has
been obtained.[9c,21]
The selectivity data presented here are better explained
in terms of a homolytic mechanism in which the metal com-
plexes simply speed up the formation of acyl and peroxyacyl
radicals that are then responsible for the oxidation of the
substrate.[22] This view is also supported by the inhibiting
effect of the free-radical scavenger 4-tert-butylcatechol on
the Co-6 catalyzed oxidation of methyl p-tolyl sulfide
(Table 1, Entries 6 and 7) and by the decrease in the
stability of the catalyst relative to that observed in the epox-
idation of alkenes with the same oxidizing system.[3] As
Experimental Section
General Remarks: Solvents were purified by standard methods and
found in the present work, metalloporphyrins and -phthalo- dried if necessary, except perfluorooctane, which was used as re-
ceived. 2,2-Dimethypropanal was distilled prior to use and all the
other commercially available reagents were used as received. Com-
plex Co-4 was prepared as described in ref.[3a] Ϫ TLC was carried
out on silica gel 60 F254. Ϫ Column chromatography (CC) was
carried out on silica gel SI 60, mesh size 0.040Ϫ0.063 mm (Merck,
Darmstadt, Germany) or neutral alumina 90 Activity I, mesh size
0.063Ϫ0.200 mm (Merck, Darmstadt, Germany). Ϫ Melting points
(uncorrected) were determined with a Büchi SMP-20 capillary
melting-point apparatus. Ϫ UV/Vis spectra were measured using a
Lambda 6 PerkinϪElmer spectrometer. Ϫ 1H NMR (300 MHz),
13C NMR (75.4 MHz) and 19F NMR (282 MHz) spectra were re-
cyanines are prone to oxidative decomposition under free-
radical conditions.[23,24] On the other hand, homolytic and
heterolytic pathways coexist in the epoxidation of alkenes
with O2/aldehyde and the correct choice of reaction condi-
tions can increase the relative importance of the latter.[25]
For instance, the catalytic FB enantioselective epoxidation
of indene with O2/2,2-dimethylpropanal was achieved in the
presence of MnIIIϪsalen complexes,[3d] thus pointing to the
preponderance of a heterolytic reaction mechanism possibly
involving (acylperoxo)metal complexes.[26]
184
Eur. J. Org. Chem. 2001, 181Ϫ186

Catalytic Oxidation of Sulfides
FULL PAPER
corded with a Bruker XL 300 spectrometer with tetramethylsilane
(δ ϭ 0), CDCl3 (δ ϭ 77) and CFCl3 (δ ϭ 0) as internal standard
respectively, unless otherwise indicated. Ϫ Elemental analyses:
Redox S.n.C. (Cologno Monzese, Italy) and Departmental Service
of Microanalysis (University of Milano).
Tetraarylporphyrin 4: A solution of aldehyde 3 (0.94 g, 1 mmol),
pyrrole (0.07 g, 1 mmol) and BF3·Et2O (0.04 g, 0.3 mmol) in
CF2ClCFCl2 (100 mL), was stirred at room temp. for 7 h. After
addition of 2,3-dichloro-5,6-dicyano-1,4-benzoquinone (DDQ,
0.19 g, 0.75 mmol), the reaction mixture was stirred for 2 h. Et3N
(0.5 mL) was then added, and the solvent evaporated. The residue
was washed with MeOH (20 mL), CH2Cl2 (20 mL) and hexane
(20 mL) and purified by column chromatography (silica gel,
CF2ClCFCl2) affording the title compound 4 as a dark purple solid
(0.11 g, 20%) which is insoluble in organic solvents. Ϫ 1H NMR
(CF2ClCFCl2, ext. ref. [D6]benzene): δ ϭ Ϫ2.85 (s, 2 H), 8.36 (s, 4
Perfluoroalkyl Ester 1: Activated copper bronze (1.90 g,
30 mmol)[27] was added to a solution of methyl 3,5-dibromobenzo-
ate (1.47 g, 5 mmol) in DMF (10 mL). The stirred suspension was
heated at 130 °C and purged with N2. n-Perfluorooctyl iodide
(3.96 mL, 15 mmol) was added dropwise to the stirred suspension
over 15 min and the reaction mixture was stirred overnight at 130
°C. After cooling to room temp., the mixture was diluted with H2O
(20 mL) and Et2O (50 mL) and filtered through a Celite plug. The
solid residue was washed with Et2O (3 ϫ 30 mL). The aqueous
phase was extracted with Et2O (20 mL). The combined ether layers
were washed with brine (30 mL) and dried with Na2SO4. The solv-
ent was evaporated affording crude 1 (4.82 g, 99%) as a pale yellow
solid (m.p. 59Ϫ61 °C) sufficiently pure for further reactions. Crys-
tallization from MeOH gave pure 1 (4.47 g, 92%) as a white solid
(m.p. 63Ϫ64 °C). Ϫ 1H NMR (CDCl3): δ ϭ 4.00 (s, 3 H), 7.97 (br.
s, 1 H), 8.48 (br. s, 2 H). Ϫ 13C NMR (CDCl3) δ ϭ 53.1, 105Ϫ120
(m, C8F17), 129.4 (t, JCϪF ϭ 6 Hz), 130.7 (t, JCϪF ϭ 24 Hz), 131.5
(t, JCϪF ϭ 6 Hz), 132.2, 164.3. Ϫ 19F NMR (CDCl3): δ ϭ Ϫ81.4
(t, J ϭ 10 Hz, 3 F), Ϫ111.7 (t, J ϭ 14 Hz, 2 F), Ϫ121.7 (br. s, 2
F), Ϫ122.4 (br. s, 6 F), Ϫ123.3 (br. s, 2 F), Ϫ126.7 (br. s, 2 F).
Ϫ C24H6F34O2 (972.11): calcd. C 29.65, H 0.62; found C 29.62,
H 0.66.
H), 8.72 (s, 8 H), 8.77 (s, 8 H). ؊ UV/Vis (CF2ClCFCl2): λmax
ϭ
413 nm (ε ϭ 7.2 ϫ 104 dm3 molϪ1 cmϪ1). Ϫ MS (FABϩ) for
C108H22F136N4; m/z (%): 3959 (100). Ϫ C108H22F136N4 (3958.58):
calcd. C 32.76, H 0.56, N 1.41; found C 32.80, H 0.59, N 1.36.
(Perfluoroalkyl)phthalonitrile 5: Activated copper bronze (0.94 g,
15 mmol)[27] was added to a solution of 1,2-dicyano-4-iodobenzene
(1.02 g, 4 mmol)[13] in dry DMSO (6 mL) in a Schlenk vessel. The
stirred suspension was warmed to 110 °C and purged with N2. n-
Perfluorooctyl iodide (2.46 g, 4.5 mmol) was added dropwise to the
stirred suspension over 10 min. After 3 h, the suspension was co-
oled to room temp., and H2O (20 mL) and Et2O (50 mL) were ad-
ded. The solid was removed by filtration through a Celite plug and
washed with Et2O (3 ϫ 15 mL). The aqueous phase was extracted
with Et2O (20 mL). The combined ether layers were washed with
brine (30 mL) and dried with Na2SO4. The solvent was evaporated
and the residue purified by column chromatography (light petro-
leum ether/Et2O, 4:1), affording the title compound as a white solid
(0.98 g, 45%). ؊ M. p. 112Ϫ113 °C. Ϫ 1H NMR (CDCl3): δ ϭ
7.95Ϫ7.99 (m, 2 H), 8.01Ϫ8.03 (m, 1 H). Ϫ 13C NMR (CDCl3):
δ ϭ 105Ϫ120 (m), 131.9 (t, J ϭ 6 Hz), 132.4 (t, J ϭ 6 Hz), 134.4,
134.7 (t, J ϭ 22 Hz). Ϫ 19F NMR (CDCl3): δ ϭ Ϫ81.2 (t, J ϭ
10 Hz, 3 F), Ϫ112.3 (t, J ϭ 14 Hz, 2 F), Ϫ121.6 (br. s, 4 F), Ϫ122.3
(br. s, 4 F), Ϫ123.2 (br. s, 2 F), Ϫ126.6 (br. s, 2 F). Ϫ C16H3F17N2
(546.11): calcd. C 35.19, H 0.55, N 5.13; found C 35.23, H 0.60,
N 5.13.
Benzyl Alcohol 2: A solution of crude ester 1 (4.86 g, 5 mmol) in
dry Et2O (70 mL) was added to a suspension of LiAlH4 (0.26 g,
7.5 mmol) in dry Et2O (10 mL). This mixture was stirred at room
temp. under N2 for 8 h, before EtOAc (1 mL) was added. After
15 min, H2SO4 (10%, 1 mL) was carefully dropped into the stirred
mixture, which was stirred for a further 20 min. The upper organic
phase was removed and the aqueous phase was extracted with Et2O
(3 ϫ 20 mL). The combined ether layers were washed with brine
(30 mL) and dried with Na2SO4. Evaporation of the solvent af-
forded a pale yellow solid (m.p. 96Ϫ98 °C) that was recrystallized
from hexane to give pure 2 (4.25 g, 90%) as a white solid (m.p.
CoII Complex of Phthalocyanine 6: 1,2-Dicyano-4-(n-perfluorooc-
tyl)benzene (5) (2.18 g, 4 mmol) and finely divided cobalt metal
(1.47 g, 2.5 mmol) were placed in a tightly closed, heavy-wall Pyrex
tube and slowly heated to 180 °C. After 4 h, the dark, melted mix-
ture was heated at 250 °C for 12 h. After cooling to room temp.
the crude mixture was extracted with n-perfluorooctane (50 mL).
The deep-blue solution was washed with toluene (3 ϫ 20 mL) and
CH2Cl2 (20 mL), and concentrated to dryness. The solid residue
was warmed to 130 °C under vacuum, thus eliminating volatile by-
products. Co-6 (0.27 g, 12%) was obtained as a blue solid. Ϫ UV/
Vis (CF2ClCFCl2): λmax ϭ 654 nm (ε ϭ 4.0 ϫ 104 dm3 molϪ1
cmϪ1), 614 nm (ε ϭ 3.0 ϫ 104 dm3 molϪ1 cmϪ1), 318 nm (ε ϭ 4.7
ϫ 104 dm3 molϪ1 cmϪ1). Ϫ MS (FABϩ) for C64H22F68N8; m/z (%):
2244 (100). Ϫ C64H12CoF68N8 (2243.38): calcd. C 34.27, H 0.54,
N 4.99; found C 34.44, H 0.81, N 4.75.
1
101Ϫ102 °C). Ϫ H NMR (CDCl3): δ ϭ 1.92 (br. s, 1 H), 4.88 (s,
2 H), 7.74 (br. s, 1 H), 7.84 (br. s, 2 H). Ϫ 13C NMR (CDCl3): δ ϭ
64.1, 105Ϫ120 (m, C8F17), 124.9 (t, JCϪF ϭ 6 Hz), 128.7 (t, JCϪF ϭ
6 Hz), 130.8 (t, JCϪF ϭ 23 Hz), 143.5. Ϫ 19F NMR (CDCl3): δ ϭ
Ϫ81.3 (t, J ϭ 10 Hz, 3 F), Ϫ111.1 (t, J ϭ 13 Hz, 2 F), Ϫ121.7 (br.
s, 2 F), Ϫ122.4 (br. s, 6 F), Ϫ123.3 (br. s, 2 F), Ϫ126.7 (br. s, 2
F). Ϫ C23H6F34O (944.10): calcd. C 29.26, H 0.64; found C 29.31,
H 0.65.
Benzaldehyde 3: MnO2 (4.87 g, 56 mmol) was added to a solution
of alcohol 2 (2.63 g, 5 mmol) in boiling toluene (100 mL). The
stirred mixture was heated at reflux for 12 h in a DeanϪStark ap-
paratus. The solid was filtered off, washed with Et2O (100 mL) and
the combined liquid phase was concentrated under reduced pres-
sure. The solid residue was purified by column chromatography
(silica gel, light petroleum ether/Et2O, 95:5), affording pure 3
General Procedure for the Aerobic Epoxidation of Sulfides Under
Fluorous Biphasic Conditions: Reactions were carried out in the
dark in a 20-mL Schlenk vessel connected to a gas burette charged
(2.22 g, 73%) as a colourless solid (m.p. 75Ϫ76 °C). Ϫ 1H NMR with O2 (250 mL). The Schlenk vessel was placed in a thermoregul-
(CDCl3): δ ϭ 8.05 (br. s, 1 H), 8.32 (br. s, 2 H), 10.15 (s, 1 H). Ϫ ated bath maintained at 20 Ϯ 0.2 °C and was then charged with:
13C NMR (CDCl3): δ ϭ 105Ϫ118 (m, C8F17), 130.8 (t, JCϪF
ϭ
(i) a CH2Cl2 solution of the substrate (0.5 , 2 mL); (ii) a solution
6 Hz), 131.6 (t, JCϪF ϭ 6 Hz), 132.0 (t, JCϪF ϭ 24 Hz), 137.7, of the catalyst in n-perfluorooctane (2.9 ϫ 10Ϫ4 , 3.5 mL). The
189.2. Ϫ 19F NMR (CDCl3): δ ϭ Ϫ81.4 (t, J ϭ 10 Hz, 3 F), Ϫ111.6
(t, J ϭ 14 Hz, 2 F), Ϫ121.5 (br. s, 2 F), Ϫ122.3 (br. s, 6 F), Ϫ123.2
two-phase mixture was magnetically stirred at 1300 Ϯ 50 rpm in
order to ensure optimum contact between the organic and the
(br. s, 2 F), Ϫ126.6 (br. s, 2 F). Ϫ C23H4F34O (942.08): calcd. C fluorous phase. 2,2-Dimethylpropanal (0.3 mL, 3 mmol) was added
29.32, H 0.43; found C 29.30, H 0.46.
to the mixture over 20 min. After 4 h, the fluorous layer was reco-
Eur. J. Org. Chem. 2001, 181Ϫ186
185

S. Colonna, N. Gaggero, F. Montanari, G. Pozzi, S. Quici
FULL PAPER
in catalytic oxidations (Ed. R. A. Sheldon), Marcel Dekker,
vered, washed with CH2Cl2 (2 ϫ 1 mL) and reused in further runs.
The combined CH2Cl2 layers were washed with aqueous NaHCO3
(5%, 1 mL), brine (1 mL) and dried (MgSO4). The solvent was
evaporated under reduced pressure, affording a residue containing
only the starting product plus sulfoxide and sulfone in variable
[9c]
New York, 1994. Ϫ
A. Marques, M. di Matteo, M.-F. Ru-
asse, Can. J. Chem. 1998, 76, 770Ϫ775 and references therein
cited.
D. P. Curran, Angew. Chem. 1998, 110, 1230Ϫ1255; Angew.
Chem. Int. Ed. 1998, 37, 1175Ϫ1196 and references cited
therein.
J. S. Lindsey, I. C. Schreiman, H. C. Hsu, P. C. Kearney, A.
M. Marguerettaz, J. Org. Chem. 1987, 52, 827Ϫ836.
Phthalocyanines Ϫ Properties and Applications (Eds. C. C. Lez-
noff, A. B. P. Lever), VCH, Weinheim, 1993.
T. M. Keller, J. R. Griffith, J. Fluorine Chem. 1978, 12, 73Ϫ77.
R. P. Linstead, P. A. Barrett, C. S. Dent, J. Chem. Soc. 1936,
1719Ϫ1737.
[10]
1
amounts, as determined by H NMR spectroscopy.
[11]
[12]
Acknowledgments
The support of the COST Action D12 ‘‘Fluorous medium: a tool
for environmentally compatible oxidation processes’’ is gratefully
acknowledged.
[13]
[14]
[15]
[16]
[17]
[18]
[19]
L. P. Barthel-Rosa, J. A. Gladysz, Coord. Chem. Rev. 1999,
192, 587Ϫ605.
M. M. dell’Anna, P. Mastrorilli, C. F. Nobile, J. Mol. Catal.
A, Chem. 1996, 108, 57Ϫ62 and references cited therein.
G. Song, F. Wang, H. Zhang, X. Lu, C. Whang, Synth. Com-
mun. 1998, 28, 2783Ϫ2787.
[1] [1a]
[1b]
´
´
I. T. Horvath, J. Rabai, Science 1994, 266, 72Ϫ75;
I .T.
[1c]
´
Horvath, Acc. Chem. Res. 1998, 31, 641Ϫ650. Ϫ
R. Fish,
Chem. Eur J. 1999, 5, 1677Ϫ1680 Ϫ [1d]E. De Wolf, G. van
[1e]
Koten, B.-J. Deelman, Chem. Soc. Rev. 1999, 28, 37Ϫ41. Ϫ
Y. Gaigneaux, R. Maggi, P. Ruiz, B. Delmon, J. Mol. Catal.
A, Chem. 1996, 109, 67Ϫ74.
M. Cavazzini, F. Montanari, G. Pozzi, S. Quici, J. Fluorine
Chem. 1999, 94, 183Ϫ193.
A. Wolberg, J. Manassen, J. Am. Chem. Soc. 1970, 92,
[2]
J. A. Gladysz, Science 1994, 266, 55Ϫ56.
2982Ϫ2991.
[3] [3a]
[20] [20a]
G. Pozzi, F. Montanari, S. Quici, Chem. Commun. 1997,
Y. Watanabe, T. Numata, S. Oae, Tetrahedron Lett. 1980,
[3b]
[20b]
69Ϫ70. Ϫ
G. Pozzi, M. Cavazzini, S. Quici, S. Fontana,
21, 3685Ϫ3688. Ϫ
hedron Lett. 1982, 23, 533Ϫ536. Ϫ
K. Fujimori, Tetrahedron Lett. 1982, 23, 1189Ϫ1192.
[21] [21a]
S. Oae, Y. Watanabe, T. Iyanagi, Tetra-
Tetrahedron Lett. 1997, 38, 7605Ϫ7608. Ϫ [3c] G. Pozzi, F. Cin-
[20c]
S. Oae, Y. Watanabe,
ato, F. Montanari, S. Quici, Chem. Commun. 1998, 877Ϫ878.
[3d]
Ϫ
G. Pozzi, M. Cavazzini, F. Cinato, F. Montanari, S.
E. Baciocchi, O. Lanzalunga, F. Marconi, Tetrahedron
Quici, Eur. J. Org. Chem. 1999, 1947Ϫ1955.
[21b]
Lett. 1994, 35, 9771Ϫ9774. Ϫ
E. Baciocchi, M. Ioele, O.
[4]
[5]
J. Drabowicz, P. Kielbasinski, M. Mikolajczik, The Chemistry
Lanzalunga, M. Malandrucco, S. Steenken, J. Am. Chem. Soc.
[21c]
of Sulfones and Sulfoxides (Eds.: S. Patai, Z. Rappoport, C. J.
1996, 118, 8973Ϫ8974. Ϫ
E. Baciocchi, O. Lanzalunga, B.
M. Stirling), Wiley, New York, 1988, p. 233.
Pirozzi, Tetrahedron 1997, 53, 12287Ϫ12298.
W. Nam, H. J. Kim, S. H. Kim, R. Y. N. Ho, J. S. Valentine,
Inorg. Chem. 1996, 35, 1045Ϫ1049.
[5a]
[22]
[23]
[24]
[25]
[26]
[27]
For recent developments see:
S. Colonna, N. Gaggero, C.
[5b]
Richelmi, P. Pasta, Trends Biotech. 1999, 17, 163Ϫ168. Ϫ
S. Colonna, N. Gaggero, P. Pasta, G. Ottolina, Chem. Commun.
1996, 2303Ϫ2307.
I. T. Horvath, J. Rabai (Exxon Comp.), U.S. Pat. 5,463,082,
1995; Chem. Abstr. 1995, 123, 87349a.
F. Montanari, S. Banfi, G. Pozzi, S. Quici, Rev. Heteroatom
Chem. 1992, 6, 94Ϫ141 and references cited therein.
A. Pentilla, C. R. Boyle, M. L. Salin, Biochem. Biophys. Res.
Commun. 1996, 226, 135Ϫ139.
J. Haber, T. Mlodnicka, J. Mol. Cat. 1992, 74, 131Ϫ141 and
references cited therein.
T. Mukaiyama, T. Yamada, Bull. Chem. Soc. Jpn. 1995, 68,
17Ϫ35 and references cited therein.
G. Pozzi, I. Colombani, M. Miglioli, F. Montanari, S. Quici,
[6]
[7]
´
´
I. Klement, H. Lütjens, P. Knochel, Angew. Chem. 1997, 109,
1603Ϫ1605; Angew. Chem. Int. Ed. Engl. 1997, 36, 1454Ϫ1456.
[8] [8a]
´
´
K. S. Ravikumar, J.-P. Begue, D. Bonnet-Delpon, Tetrahed-
[8b]
ron Lett. 1998, 39, 3141Ϫ3144. Ϫ
K. S. Ravikumar, Y. M.
´
´
Zhang, J.-P. Begue, D. Bonnet-Delpon, Eur. J. Org. Chem.
1998, 2937Ϫ2940.
Tetrahedron 1997, 53, 6145Ϫ6162.
[9] [9a]
Metalloporphyrin-catalyzed oxidations (Eds. F. Montanari,
Received June 6, 2000
[9b]
L. Casella), Kluwer, Dordrecht, 1994. Ϫ
Metalloporphyrins
[O00295]
186
Eur. J. Org. Chem. 2001, 181Ϫ186