Recyclable catalyst reservoir: Oxidation of alcohols mediated by noncovalently supported bis(imidazolium)-tagged 2,2,6,6-tetramethylpiperidine 1-oxyl

Article abstract of DOI:10.1002/cctc.201300234

Bis(imidazolium)-tagged 2,2,6,6-tetramethylpiperidine 1-oxyl (TEMPO) catalysts were adsorbed on different supports such as silica gel, silica gel modified with highly cross-linked polymeric imidazolium networks, and highly cross-linked polymeric imidazolium networks entrapping magnetic particles. These systems provided a convenient tool for the oxidation of both primary and secondary alcohols working as recyclable reservoirs for the bis(imidazolium)-tagged TEMPO catalysts. By using EPR spectroscopy it was demonstrated that the catalyst was released as the corresponding oxoammonium salt in the solution during the recycling step, thus promoting the oxidative process in a homogeneous fashion. After solvent removal, the catalyst was readsorbed on the support allowing an easy recovery and recycle of the catalytic material up to 13 consecutive cycles with no loss in activity. The bis(imidazolium)-tagged TEMPO catalyst could be used in only 1mol% both for the oxidation of benzylic and aliphatic alcohols. The catalytic material was highly recyclable if used on silica or imidazolium-modified silica gel in 10mol% loading. Loading could be scaled down to 1mol% and the catalyst proved to be recyclable up to 8cycles only with imidazolium-modified silica gel. Such a "catalyst-sponge-like" system permits to combine the benefits of homogeneous and heterogeneous catalysis. Catch and release: Why spend money on covalent immobilization or ionic liquids with bulky cations? A bis(imidazolium)-modified 2,2,6,6-tetramethylpiperidine 1-oxyl catalyst can be adsorbed on an imidazolium-modified support, then released in the presence of a proper organic solvent and recaptured at the end of the reaction on removal of the solvent.

Full text of DOI:10.1002/cctc.201300234

CHEMCATCHEM
FULL PAPERS
FULL PAPERS
Catch and release: Why spend money
on covalent immobilization or ionic liq-
uids with bulky cations? A bis(imidazoli-
um)-modified 2,2,6,6-tetramethylpiperi-
dine 1-oxyl catalyst can be adsorbed on
an imidazolium-modified support, then
released in the presence of a proper
organic solvent and recaptured at the
end of the reaction on removal of the
solvent.
H. A. Beejapur, F. Giacalone, R. Noto,
P. Franchi, M. Lucarini, M. Gruttadauria*
&& – &&
Recyclable Catalyst Reservoir:
Oxidation of Alcohols Mediated by
Noncovalently Supported
Bis(imidazolium)-Tagged 2,2,6,6-
Tetramethylpiperidine 1-Oxyl
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemCatChem 0000, 00, 1 – 10
&
1
&
ÞÞ
These are not the final page numbers!

CHEMCATCHEM
FULL PAPERS
www.chemcatchem.org
DOI: 10.1002/cctc.201300234
Recyclable Catalyst Reservoir: Oxidation of Alcohols
Mediated by Noncovalently Supported Bis(imidazolium)-
Tagged 2,2,6,6-Tetramethylpiperidine 1-Oxyl
Hazi Ahmad Beejapur,[a] Francesco Giacalone,[a] Renato Noto,[a] Paola Franchi,[b]
Marco Lucarini,[b] and Michelangelo Gruttadauria*[a]
Dedicated to Professor Maurizio Prato on the occasion of his 60th birthday
Bis(imidazolium)-tagged 2,2,6,6-tetramethylpiperidine 1-oxyl
(TEMPO) catalysts were adsorbed on different supports such as
silica gel, silica gel modified with highly cross-linked polymeric
imidazolium networks, and highly cross-linked polymeric im-
idazolium networks entrapping magnetic particles. These sys-
tems provided a convenient tool for the oxidation of both pri-
mary and secondary alcohols working as recyclable reservoirs
for the bis(imidazolium)-tagged TEMPO catalysts. By using EPR
spectroscopy it was demonstrated that the catalyst was re-
leased as the corresponding oxoammonium salt in the solution
during the recycling step, thus promoting the oxidative pro-
cess in a homogeneous fashion. After solvent removal, the cat-
alyst was readsorbed on the support allowing an easy recovery
and recycle of the catalytic material up to 13 consecutive
cycles with no loss in activity. The bis(imidazolium)-tagged
TEMPO catalyst could be used in only 1 mol% both for the oxi-
dation of benzylic and aliphatic alcohols. The catalytic material
was highly recyclable if used on silica or imidazolium-modified
silica gel in 10 mol% loading. Loading could be scaled down
to 1 mol% and the catalyst proved to be recyclable up to
8 cycles only with imidazolium-modified silica gel. Such a “cata-
lyst-sponge-like” system permits to combine the benefits of
homogeneous and heterogeneous catalysis.
Introduction
Oxidation of primary and secondary alcohols into their corre-
sponding carbonyl compounds is one of the most important
processes both at laboratory and industrial level because of its
strategic importance in organic synthesis.[1] Indeed this conver-
sion, if performed in the absence of metal-based catalysts, may
find application for the synthesis of fine chemicals such as fra-
grances or food additives.[2] In this light, the use of metal-free
catalysts for selective oxidation of alcohols is very appealing,
and to the date the most exploited compounds for this pur-
pose are the 2,2,6,6-tetramethylpiperidine 1-oxyl (TEMPO) de-
rivatives.[3] This family of stable nitroxyl radicals presents low
toxicity along with an intriguing reversible redox behavior.[4]
TEMPO may be used catalytically in combination with stoichio-
metric amounts of a co-oxidant that regenerates the active
TEMPO species during the reaction. Among others, hypervalent
iodine(III) reagent [bis(acetoxy)iodo]benzene (BAIB) has been
found very useful for this purpose.[5] BAIB is a mild organic oxi-
dant unable to oxidize TEMPO directly, but capable to regener-
ate TEMPO from TEMPOH species. Although BAIB is used in
stoichiometric amounts, some procedures for its recovery and
regeneration have been recently reported.[6] Nevertheless,
whichever oxidant is used, separation of TEMPO from the
products remains a key issue. In addition, TEMPO is considera-
bly expensive and, consequently, its separation and reuse after
the oxidation reaction is highly desirable. To this aim, several
groups designed a number of different strategies in which
TEMPO has been chemically anchored onto various supports
such as silica,[6c,7] magnetic nanoparticles,[8] soluble[9] and in-
soluble organic polymers,[10] and also saponite[11] and nano-
tubes.[12] However, an inherent disadvantage of the covalent
immobilization strategy relies in the generally decreased activi-
ty of the modified catalysts owing to the limited diffusion rates
featuring the heterogeneous nature of the process.[13] On the
other hand, as an alternative approach, the highly specialized
TEMPO moiety has been functionalized with perfluoroalkyl
groups[5d,14] (fluorous tags) or has been incorporated into task-
specific ionic liquids[15] (IL tags) with the aim to be employed
in homogeneous or biphasic systems. Unfortunately, their
wider application is limited by the high costs of fluorinated
systems and ionic liquid if used as solvents or for separation
purposes, along with the problematic stemming from the high
[a] H. A. Beejapur, Dr. F. Giacalone, Prof. R. Noto, Prof. M. Gruttadauria
Dipartimento di Scienze e Tecnologie Biologiche
Chimiche e Farmaceutiche (STEBICEF)
Universitꢀ di Palermo
Viale delle Scienze s/n, Ed. 17, I-90128 Palermo (Italy)
Fax: (+39)091596825
E-mail: michelangelo.gruttadauria@unipa.it
[b] Dr. P. Franchi, Prof. M. Lucarini
Department of Chemistry “G. Ciamician”
University of Bologna
Via San Giacomo 11, I-40126 Bologna (Italy)
Supporting information for this article is available on the WWW under
http://dx.doi.org/10.1002/cctc.201300234.
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemCatChem 0000, 00, 1 – 10
&
2
&
ÞÞ
These are not the final page numbers!

CHEMCATCHEM
FULL PAPERS
www.chemcatchem.org
viscosity of IL that may lead both to mass- and heat-
transfer limitations.
In the case of IL, to minimize the amount of uti-
lized ionic liquids, the supported ionic liquid phase
(SILP) approach has been developed.[16] This ap-
proach involves the presence of a support material
that is modified with adsorbed layers of ionic liquid
serving as the reaction phase in which the homoge-
neous catalyst is confined. An evolution of SILP in-
volves the use of covalently attached ionic liquid
moieties affording
a supported ionic-liquid-like
phase (SILLP).[17] In some cases, additional ionic liquid
can be adsorbed on the SILLP giving rise to the for-
mation of several layers of free ionic liquid on the
support even if some leaching of the IL may occur.
Recently we introduced a new approach in order to
circumvent these problems and, in the meantime, to
reduce the costs. Hence, a new class of multilayered
covalently supported IL phase (mlc-SILP) materials
have been reported by covalently linking and, at the
meantime, polymerizing suitably functionalized bis-
vinylimidazolium salts. A similar approach has been
also applied for the preparation of magnetic particles
entrapped into highly cross-linked imidazolium
salts.[18] These novel materials displayed not only
good catalytic activity in the cycloaddition of carbon dioxide
to epoxides to yield cyclic carbonates,[18,19] but also excellent
activity promoting the Suzuki coupling if Pd nanoparticles are
immobilized in the polymerized imidazolium salt network.[20] Fi-
nally, mlc-SILP approach has been recently applied for the non-
covalent immobilization of an IL-tagged organocatalyst.[21,22]
This system has been successfully employed in the asymmetric
aldol reaction for 15 cycles without loss of activity and selectiv-
ity. This approach is based on the desorption–adsorption route
during the reaction and work-up steps. Initially, the catalyst is
adsorbed on the surface of an imidazolium-modified silica gel
support (Scheme 1).
Scheme 1. Function of the IL-tagged catalyst adsorbed on the polymerized imidazolium
network.
Results and Discussion
Catalyst preparation and catalytic activity
With the aim to increase the affinity of a nitroxide catalyst to-
wards an imidazolium-modified silica gel, we introduced a bis-
imidazolium tag in the 4-hydroxy-TEMPO moiety. Such modifi-
cation has been achieved through a simple two-steps synthesis
starting from the 1,3,5-tris(bromomethyl)benzene compounds
1a,b as reported in Scheme 2. As support, the imidazolium-
modified silica gel 4[21] and magnetic particles entrapped into
highly cross-linked imidazolium salts (material 5,[18] Figure 1)
were used. In addition, unmodified silica gel was employed. As
a first attempt we tested catalyst 3a or 3b in high catalytic
amount (10 mol%)[15a,b] in the presence of BAIB (1.1 equiv.) in
dichloromethane at room temperature for the oxidation of
In order to increase the affinity between the surface of
modified silica gel and catalyst, the latter is modified with an
imidazolium tag. During the reaction, the catalyst is dissolved
in the reaction medium. If the reaction is complete, solvent is
removed causing the readsorption of the catalyst, whereas the
final product is extracted by using a proper solvent that does
not dissolve the adsorbed catalyst. In this sense, the catalytic
material (support+catalyst) acts as a catalyst reservoir. Herein
we describe a similar approach for the oxidation of alcohols
mediated by imidazolium-tagged TEMPO catalysts adsorbed
on imidazolium-modified supports and unmodified silica gel.
The reversible immobilization of a soluble catalyst onto an
insoluble support may provide the combination of the benefits
of heterogeneous and homogeneous catalysis. In this regard,
the easy separation of the catalyst from the reaction mixture
and his reuse are associated to higher reaction rates, character-
istic of homogeneous catalysis.[23,24]
Scheme 2. Reaction conditions: a) 4-Hydroxy-TEMPO, Bu4NBr, toluene, 50%
NaOH, 708C, 2 h, 2a: 60% yield, 2b: 63% yield; b) methylimidazole, CHCl3,
608C, overnight, 3a: 89% yield; 3b: quantitative yield.
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemCatChem 0000, 00, 1 – 10
&
3
&
ÞÞ
These are not the final page numbers!

CHEMCATCHEM
FULL PAPERS
www.chemcatchem.org
without affecting the yield and the recyclability
(Table 1, entry 1).
Less reactive substrates were also tested, namely
2-decanol, 1-octanol, and 1-heptanol. The corre-
sponding carbonyl compounds were obtained in
high yields (Table 1, entries 2–4). Because of the in-
teresting results obtained by employing only
1 mol% of catalyst 3a, further cycles were conduct-
ed. 1-Phenylethanol was used in cycles 2–4 (entry 5)
and primary benzylic alcohols, such as p-methoxy
and p-nitrobenzylic alcohols were employed in the
5th and 6th cycle (entries 6–7). Excellent yields were
observed. Then, less reactive substrates, such as ali-
phatic alcohols, were tested in the 7th and 8th
cycle. 1-Boc-4-hydroxypiperidine was converted into
the corresponding ketone in 53% yield in 2 h,
Figure 1. Imidazolium-modified supports used. Highly cross-linked imidazolium-modified
silica gel 4 and highly cross-linked polymeric imidazolium networks entrapping magnetic
particles 5.
1-phenylethanol. Both catalysts gave a quantitative yield of the
corresponding ketone (Scheme 3).
whereas 1-decanol did not react after 6 h (entries 8–9). The re-
cycling study was duplicated by using 1-phenylethanol in each
cycle. Again the yield decreased after 8 cycles (entries 10–13).
The scope of this methodology was further explored by per-
forming a complete recycling study in which different sub-
Table 1. Oxidation reactions with supported catalyst 3a/4 and recycling
studies.[a]
Scheme 3. Reaction conditions: Catalyst 3a or 3b (10 mol%), alcohol
(0.82 mmol), BAIB (1.1 equiv.), CH2Cl2 (2 mL), RT, 0.5 h.
At the end of the reaction, support 4 was added and the sol-
vent (CH2Cl2) removed under reduced pressure. The solid resi-
due was washed with diethyl ether and filtered. The solid,
which contained catalyst 3a or 3b adsorbed on support 4
(3a/4 or 3b/4, 25 wt% of 3a or 3b with respect to 4) was em-
ployed for further 9 cycles. No loss in activity was observed
(Scheme 4). In contrast, when catalyst 3a or 3b was recovered
in the absence of the support, recovery was not quantitative
and yield decreased after 3 cycles. Considering the good re-
sults obtained, the ratio 3/4 was not optimized.
Although catalysts appear to be highly recyclable, the possi-
bility of reducing their loading has been considered. Then,
1-phenylethanol was oxidized by using 5, 2, and 1 mol% of
catalyst 3a. The ratio 3a/4 has been maintained (25 wt% of
3a with respect to 4). Up to four cycles were conducted in
each case without loss in activity (yield>95%). Very interest-
ingly, the catalytic loading could be scaled down to 1 mol%
Entry
1
Product
Cycle
1
t [h]
Yield[b] [%]
0.5
>95
2
1
22
91
3
4
1
1
2
94
95
0.5
5
6
2–4
5
0.5
0.5
>95
>95
7
6
0.5
>95
8
9
7
8
2
6
53
–
10
1–5
0.5
>95
11
12
13
6
7
8
0.5
0.5
0.5
90
75
25
[a] Catalyst 3a (1 mol%) and support 4 (loading 25 wt%, 21.2 mg); CH2Cl2
(2 mL); alcohol (0.82 mmol) and BAIB (1.1 equiv.). [b] By 1H NMR spectros-
copy.
Scheme 4. Reaction conditions: Catalyst 3a (10 mol%)+4 (212 mg) or 3b
(10 mol%)+4 (200 mg), alcohol (0.82 mmol), BAIB (1.1 equiv.), CH2Cl2 (2 mL),
RT, 0.5 h.
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemCatChem 0000, 00, 1 – 10
&
4
&
ÞÞ
These are not the final page numbers!

CHEMCATCHEM
FULL PAPERS
www.chemcatchem.org
strates, benzylic primary and secondary alcohols, primary allylic
and aliphatic alcohols, and cycloalkanols were oxidized by re-
using the same catalyst sample recovered after each reaction.
As both more reactive and less reactive alcohols were em-
ployed in consecutive cycles, we performed this investigation
using catalysts 3a or 3b in 10 mol%.
The results are reported in Table 2. To prove the high reus-
ability of the catalytic systems in the 11th and 12th cycle, the
same type of alcohol oxidized as in the 1st cycle, that is, 1-phe-
nylethanol, was employed. Both catalytic systems, either bis-
imidazolium-derivative TEMPO 3a or 3b adsorbed on imidazo-
lium-modified silica 4 were used. In all cases the carbonyl com-
pounds were obtained in excellent yields. The high yields ob-
tained up to ten cycles are a proof of the high reusability of
the catalytic systems. This point is confirmed in the last two
cycles, in which again 1-phenylethanol was used. Only a minor
decrease in yield was observed.
To prove if catalysts are dissolved in the solvent under the
reaction conditions adopted, the catalytic system 3b/4 was
suspended in CH2Cl2 (by using the same concentration as em-
ployed in the reactions) and then filtered. After filtration, ap-
proximately 90% of catalyst 3b was recovered, demonstrating
that the catalyst was released during the oxidation reaction
and readsorbed at the end of the reaction, if dichloromethane
is removed under reduced pressure.
We also repeated the same reaction procedure with 1-phe-
nylethanol by employing a solvent like diethyl ether, in which
the nitroxide is completely insoluble. Under this condition the
system should be considered as a truly heterogeneous system.
Reaction time was longer (3 h) and conversion was not quanti-
tative (87% yield, Scheme 5). It
is known that the dielectric con-
stant of the medium influences
the yield,[25] however, in this
Table 2. Oxidation reactions with supported catalysts 3a/4 and 3b/4 and recycling studies.[a]
case the increased reaction time
and decreased yield can be as-
cribed also to the heteroge-
neous nature of the catalyst.
Then, we tested the recycling
properties of our catalytic
system by employing a double
amount of support 4 with re-
spect to the previous systems in
the presence of catalyst 3b
(12.5 wt% on 4, Table 2, last two
Entry Substrate
Product
Cycle Catalyst 3a/4
Catalyst 3b/4
25 wt%
Catalyst 3b/4
12.5 wt%
25 wt%
t [h] Yield[b] [%] t [h] Yield[b] [%] t [h] Yield[b] [%]
1
2
1
2
0.5
0.5
>95
>95
0.5
0.5
>95
>95
1
1
60
86
columns).
A comparison with
data previously discussed indi-
cated that in the presence of
a
larger amount of support,
3
3
0.5
>95
0.5
>95
1
93
longer reaction times and not
quantitative conversions were
obtained. These results are in
agreement with the fact that
the higher amount of support
used imparts a more heteroge-
neous character to the system,
in which catalyst 3b is parti-
tioned between the support
and the solvent.
4
5
6
7
4
5
6
7
0.5
5
>95
>95
>95
92
0.5
5
>95
>95
>95
92
15
16
3
70
84
75
70
3
3
0.5
0.5
2
8
8
9
0.5
0.5
0.5
>95
>95
>95
0.5
0.5
0.5
>95
>95
>95
4
1
87
>95
90
However, such
a
catalytic
9
system still performs excellent
recyclability, as indicated by the
product yield in the 13th cycle,
which was even slightly higher
than in the first cycle.
10
11
10
11
15
3
To explore the nature of the
support on the outcome of the
reaction, we also employed
magnetic particles entrapped
into highly cross-linked imidazo-
0.5
0.5
92
92
84
12
13
12
13
3
1
87
68
[a] Catalyst 3a or 3b (10 mol%) and support 4 (loading 25 wt%: 212 mg for catalyst 3a, 200 mg for catalyst
3b; loading 12.5 wt%: 400 mg for catalyst 3b); CH2Cl2 (2 mL); alcohol (0.82 mmol) and BAIB (1.1 equiv.). [b] By
1H NMR spectroscopy.
lium
salts
(material
5,
Figure 1).[18]
The magnetic
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemCatChem 0000, 00, 1 – 10
&
5
&
ÞÞ
These are not the final page numbers!

CHEMCATCHEM
FULL PAPERS
www.chemcatchem.org
were observed in the latter two cycles. Such behavior could be
ascribed to the very low surface area of material 5,[18] which
may cause loss of catalyst 3b upon recycling.
Finally, unmodified silica gel was used as the support. Cata-
lyst 3a was used in 1 or 10 mol% (Table 4). If catalyst 3a was
used in 1 mol%, yields of acetophenone dropped dramatically
after only 3 cycles (Table 4, entries 1, 2). In this case, support 4
proved to be better than silica and the 3a/4 system was clear-
ly more recyclable (see Table 1).
Scheme 5. Oxidation of 1-phenylethanol with 3a/4 in diethyl ether.
nature of such a support can be useful in the recovery proce-
dure. Catalyst 3b was adsorbed on support 5 and used in sev-
eral oxidation cycles as reported in Table 3. In this case, after
evaporation of CH2Cl2 and addition of diethyl ether the materi-
al was easily recovered by decantation after exposition to an
external magnet.
Table 4. Oxidation reactions with supported catalyst 3a/SiO2 and recy-
cling studies.[a]
Data reported in Table 3 show that in many cases yields
were high or excellent, although slightly longer reaction times
than in the 3b/4 system were needed (see Table 2). However,
the 3b/5 system appeared less recyclable because lower yields
Entry
1
Product
Loading [mol%]
1
Cycle
1–2
t [h]
Yield[b] [%]
0.5
>95
Table 3. Oxidation reactions with magnetic-particles-supported catalyst
3b/5 and recycling studies.
2
3
1
10
3
1–5
0.5
0.5
<10
>95
4
5
6
10
10
10
6–7
0.5
94
>95
95
8
3
9
3
Entry
1
Product
Cycle
1
t [h]
Yield [%]
[a] Catalyst 3a (1 or 10 mol%) and support SiO2 (loading 25 wt%:
21.2 mg or 212 mg); CH2Cl2 (2 mL); alcohol (0.82 mmol) and BAIB
1
(1.1 equiv.). [b] By H NMR spectroscopy.
1
>95
2
3
2
3
1
3
>95
>95
On the other hand, 3a in 10 mol% loading on silica worked
very well for 9 cycles without loss in activity (Table 4, entries 3–
6). High yields were observed even with a secondary and a pri-
mary aliphatic alcohol (Table 4, entries 5–6).
The results reported herein can be compared with those ob-
tained by using similar IL–TEMPO catalysts and stoichiometric
amounts of co-catalysts such as BAIB[15a] or nBuNHSO5 in IL at
608C.[15b] However, in these cases, IL–TEMPO catalysts were
used only in 10 mol% loading. Stoichiometric amount of co-
catalyst may be avoided by using CuCl2 and oxygen as the
final oxidant. In these cases IL–TEMPO catalysts were used in
5 mol% at 40–1008C and only benzylic alcohols were oxi-
dized.[15c,g,j] The same catalyst loading (5 mol%) was used for
the IL–TEMPO/NaNO2/CO2/H2O at 1008C, a method mainly de-
veloped for the oxidation of benzylic alcohols.[15e] However, no
recycling studies were described. The amount of IL–TEMPO
could be scaled down to 1 mol% if used with NaOCl/KBr
system in IL/H2O at 08C, but only 2 recycles were described.[15f]
4
5
6
7
4
5
6
7
24
48
3
70
70
>95
90
3
8
8
9
12
2
>95
>95
85
9
10
10
2
11
12
11
12
1
1
81
75
Characterization
Support 4 had a surface area of 128 m2 gÀ1 and a cumulative
pore volume of 0.2 cm3 gÀ1. Catalytic materials, fresh and used,
were characterized by SEM measurements and EPR analysis. In
Figure 2, SEM images of support 4 are reported. From these
[a] Catalyst 3b (10 mol%) and support 5 (loading 25 wt%: 200 mg);
CH2Cl2 (2 mL); alcohol (0.82 mmol) and BAIB (1.1 equiv.). [b] By 1H NMR
spectroscopy.
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemCatChem 0000, 00, 1 – 10
&
6
&
ÞÞ
These are not the final page numbers!

CHEMCATCHEM
FULL PAPERS
www.chemcatchem.org
Figure 2. SEM images of support 4. Pictures b) and d) are magnifications of
a) and c), respectively. Scale bars: a) 50, b) 5, c) 40, and d) 5 um.
Figure 3. SEM images of freshly prepared 3a/4 at different magnifications.
Scale bars: a) 5, b) 10, c) 40, and d) 400 mm.
images it can be seen that such material is a physical mixture
of silica-supported imidazolium salt and the polymer of the
starting bis(vinylimidazolium) salt. After adsorption of the bis-
(imidazolium)-modified TEMPO catalyst, the surface of the sup-
port appeared as covered by a thin film (Figure 3). However,
after 12 cycles the morphology of the material appeared to be
similar to that of the starting material (Figure 4).
EPR measurements
Figure 4. SEM images of 3a/4 after 12 cycles at different magnifications.
Scale bars: a) 5 and b) 40 mm.
The EPR spectrum of nitroxide 3a (0.1 mm) in CH2Cl2 at room
temperature is reported in Figure 5. The expected three-line
spectrum was characterized by the coupling of the unpaired
electron with a nitrogen nucleus with a hyperfine splitting con-
stant, aN, of 15.78 G (g=2.0060).
In Figure 6a the EPR spectrum of the powder catalytic
system 3a/4 is reported instead. To see if in the reaction condi-
tions catalysts were dissolved in the solvent, the catalytic
system 3a/4 was suspended in CH2Cl2 (5 mgmLÀ1 correspond-
ing to 1.25 mgmLÀ1 of nitroxide) and then filtered. After filtra-
tion, we recorded the EPR spectra of both the solid residue
(Figure 6b) and the filtrate (Figure 6c). Quantitative analysis,
performed by comparison of a doubly integrated spectrum to
a
spectrum of a solution of 2,2-diphenyl-1-picrylhydrazyl
Figure 5. EPR spectrum of 3a (0.1 mm) in CH2Cl2 at RT.
(DPPH) of known concentration by using the signal from
a ruby crystal as internal standard, indicated that approximate-
ly 96% of catalyst 3a was recovered in the solution. Only
a negligible amount of the nitroxide (<2%) was instead pres-
ent in the solid residue after the extraction procedure. Similar
results were obtained with 3b/4 system. These results con-
firmed that in the presence of CH2Cl2 both catalysts were re-
leased during the oxidation reaction and readsorbed at the
end of the reaction, if dichloromethane was removed under re-
duced pressure.
We repeated the same EPR experiments on the catalytic sys-
tems 3a/4 or 3b/4 by using Et2O as the solvent. In this case,
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemCatChem 0000, 00, 1 – 10
&
7
&
ÞÞ
These are not the final page numbers!

CHEMCATCHEM
FULL PAPERS
www.chemcatchem.org
that the presence of AcOH, which can be generated by ligand
exchange of PhI(OAc)2 with the alcohol, results in the dismuta-
tion of TEMPO to an oxoammonium salt and hydroxylamine,
the latter being oxidized to TEMPO by BAIB, resulting in the
generation of AcOH (Scheme 6).
To check if the nitroxide was recovered in the oxoammoni-
um form during the different catalytic cycles, we tried to
reduce the oxoammonium salt electrochemically to the corre-
sponding nitroxide by applying negative current to the EPR
cell containing an acetonitrile (ACN) solution of the filtrate ob-
tained after filtration of a suspension of the catalytic systems
employed for 12 cycles (Figure 7).
Actually, an increase of the EPR signals attributable to the ni-
troxide with time was observed upon applying negative cur-
rent to the EPR cell. In this case, the electron uptake led to the
conversion of the oxoammonium salt to the corresponding ni-
troxide species. These evidences can be summarized in the
mechanism proposed in Scheme 6, in which the fresh catalyst
(1st cycle) is released in solution as a nitroxide species, where-
as in the successive cycles the catalyst is adsorbed/released as
an oxoammonium salt.
Figure 6. EPR spectra of 3a/4 (5 mgmLÀ1) before (a) and after solvent extrac-
tion with CH2Cl2 (b,c) or Et2O (d,e).
the solid residue exhibited a strong EPR signal at 298 K, the in-
tensity of which was comparable to that of the starting sample
indicating that the catalyst was recovered in a quantitative
amount. On the other hand, no EPR signal was observed in the
filtrate.
Conclusions
In conclusion, we have developed a recyclable catalytic reser-
voir for alcohol oxidation based on an imidazolium-tagged
TEMPO catalyst adsorbed on an imidazolium-modified support
or on unmodified silica gel. The imidazolium-tagged TEMPO
catalysts were easily obtained with a simple two-steps synthe-
sis. These systems, during the oxidation of alcohols, released in
solution the imidazolium-tagged TEMPO catalyst thus promot-
Notably, the EPR signal of the nitroxide in the solid residue
was characterized by an additional broad signal (indicated by
asterisks) besides the signal already present in the starting
solid sample. We assigned the new broad signal to a new as-
sembly of the catalyst in the support that resulted in a strong
dipolar interaction between paramagnetic centers. Although it
was impossible to obtain a clear
picture of this aggregation, the
new EPR spectrum shape ob-
served after Et2O removal was
consistent with a significant in-
teraction between the support
and the catalyst. Actually, simple
dissolution of the catalyst
powder in Et2O and removal of
the solvent afforded an EPR
spectrum that did not contain
the spectral features observed
in the presence of 4.
To check the loss of catalyst
during the oxidation process we
recorded the EPR spectrum of
the catalytic systems 3a/4 or
3b/4 after 12 cycles of activity
in the powder form. Despite the
good activity shown by the cat-
alytic system after 12 cycles, the
amount of nitroxide still present
in the system was negligible in
both cases.
Previous investigations on the
reaction mechanism[5a] suggest
Scheme 6. Proposed adsorption–desorption equilibrium and reaction mechanism.
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemCatChem 0000, 00, 1 – 10
&
8
&
ÞÞ
These are not the final page numbers!

CHEMCATCHEM
FULL PAPERS
www.chemcatchem.org
cation in concentrated sulfuric acid (g=2.002583). Nitroxide con-
centrations were measured with respect to a solution of DPPH of
known concentration by using the signal from a ruby crystal as an
internal standard.[25] The electrochemical cell was home-made and
consisted of an EPR flat cell (Wilmad WG-810) equipped with
a 25ꢁ5ꢁ0.2 mm platinum gauze (cathode) and a platinum wire
(anode). The current was supplied and controlled by an AMEL 2051
general-purpose potentiostat. In a typical experiment, the cell was
filled with an acetonitrile solution obtained after filtration of a sus-
pension of the catalytic systems employed for 12 cycles in ACN
(5 mgmLÀ1) containing tetrabutylammonium perchlorate (ꢀ0.1m)
as supporting electrolyte. After thoroughly purging the solution
with N2, spectra were recorded at a potential settings of À3.0 V.[26]
Figure 7. EPR spectra of 3a recorded at different times in ACN, containing
Bu4NClO4 (ꢀ0.1m) as supporting electrolyte, at a potential setting of À3.0 V.
The starting solution was obtained after filtration of a suspension of the cat-
alytic systems 3a/4 employed for 12 cycles in ACN.
Synthesis of TEMPO derivatives 2a and 2b
To a solution of 4-hydroxy-TEMPO (2.32 mmol, 400 mg) in toluene
(4 mL), Bu4NBr (0.124 mmol, 40 mg) was added followed by 1,3,5-
tri(bromomethyl)mesitylene 1a (4.64 mmol, 1.85 g) and NaOH 50%
(8 mL). The mixture was heated at 708C for 2 h. After cooling at RT,
toluene and water were added. The aqueous layer was separated
and extracted with ethyl acetate (3ꢁ), then organic layers were
mixed and evaporated. The residue was purified by column chro-
matography by using dichloromethane/petroleum ether 7:3 to re-
cover the unreacted tribromo compound 1a, whereas the desired
compound 2a was eluted with dichloromethane (60% yield).[27]
ing the process in a homogeneous fashion. After solvent re-
moval, the catalyst was readsorbed allowing an easy recovery
and recycle of the catalytic material up to 13 consecutive
cycles with no loss in activity. Such adsorption–desorption
equilibrium has been unequivocally proven by EPR experi-
ments, which further confirmed that the catalyst was in its ox-
oammonium salt form after the catalytic cycle.
The catalytic system could be employed in low catalytic
amount (1 mol%), and resulted to be more recyclable if ad-
sorbed on imidazolium-modified silica gel. Higher loading
(10 mol%) both on silica gel or imidazolium-modified silica gel
gave highly recyclable materials. Thanks to the high recyclabili-
ty of the system, the use of such high amounts does not repre-
sent, in our opinion, a limitation of the method.
Compound 2b was prepared by following the same procedure
(63% yield).
Synthesis of catalysts 3a and 3b
In summary, our system appears to be more recyclable with
respect to other IL–TEMPO catalysts reported in the literature,
although improvements are desirable. Further studies are in
progress to avoid the use of stoichiometric amounts of the
co-oxidant.
A solution of compound 2a (1.13 mmol, 0.55 g) and methylimid-
azole (2.49 mmol, 0.197 mL) in CHCl3 (13 mL) was heated at 608C
overnight. After cooling at RT, the solvent was removed and the
solid residue was carefully washed with hot diethyl ether and fil-
tered to give compound 3a as a light orange solid (89% yield).
TEMPO catalyst 3b was obtained as reddish viscous oil (99% yield).
Experimental Section
Alcohol oxidation
General
General procedure: Catalyst 3a or 3b (10 mol%) were dissolved in
methanol (4 mL) in a round-bottom flask and support 4 (loading
25 wt%: 212 mg for catalyst 3a, 200 mg for catalyst 3b; loading
12.5 wt%: 424 mg for catalyst 3a, 400 mg for catalyst 3b) was
added. The solvent was removed under reduced pressure and
a powder 3a/4 or 3b/4 was obtained. Then, dichloromethane
(2 mL) was added and, to this suspension, the appropriate alcohol
(0.82 mmol) and BAIB (1.1 equiv.) were added. The mixture was
stirred at RT for the time indicated in the tables, then the solvent
was removed under reduced pressure and the residue washed
with diethyl ether and filtered. The catalytic material was dried for
a few minutes and then reused for the next cycle.
Chemicals and solvents were purchased from commercial suppliers
or purified by standard techniques. For thin-layer chromatography
(TLC), silica gel plates (Merck 60 F254) were used and compounds
were visualized by irradiation with UV light and/or by treatment
with a solution of KMnO4. Flash chromatography was performed
by using Macherey–Nagel silica gel (0.04–0.063 mm). Light petro-
1
leum refers to the fraction with the boiling range 40–608C. H and
13C NMR spectra were recorded with a Bruker 300 MHz spectrome-
ter. SEM images were recorded by using an instrument with 20 kV
operating voltage. Support 4 was prepared as described in the lit-
erature.[21] Silica gel used as a support has a surface area of
750 m2 gÀ1 and a cumulative pore volume of 0.68 cm3 gÀ1 and is
commercially available (Aldrich).
Acknowledgements
EPR experiments
Financial support from the University of Palermo (funds for se-
lected topics), University of Bologna and the Italian MIUR within
the national project “Catalizzatori, metodologie e processi inno-
vativi per il regio e stereocontrollo delle sintesi organiche” and
COST Action CM0905 ORCA are gratefully acknowledged.
EPR spectra were recorded at RT using an ELEXYS E500 spectrome-
ter equipped with an NMR gaussmeter for the calibration of the
magnetic field and a frequency counter for the determination of g
factors that were corrected against that of the perylene radical
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemCatChem 0000, 00, 1 – 10
&
9
&
ÞÞ
These are not the final page numbers!

CHEMCATCHEM
FULL PAPERS
www.chemcatchem.org
pp. 1377–1386; b) D. E. De Vos, B. F. Sels, P. A. Jacobs, Adv. Catal. 2001,
46, 1–87; c) J. A. Gladysz, Pure Appl. Chem. 2001, 73, 1319–1324.
[14] a) A. Gheorghe, E. Cuevas-YaÇez, J. Horn, W. Bannwarth, B. Narsaiah, O.
Reiser, Synlett 2006, 2767–2770; b) A. Gheorghe, T. Chinnusamy, E.
Cuevas-YaÇez, P. Hilgers, O. Reiser, Org. Lett. 2008, 10, 4171–4174; c) T.
Chinnusamy, O. Reiser, ChemSusChem 2010, 3, 1040–1042.
[15] a) A. Fall, M. Sene, M. Gaye, G. Gꢅmez, Y. Fall, Tetrahedron Lett. 2010, 51,
4501–4504; b) C. Zhu, L. Ji, Y. Wie, Catal. Commun. 2010, 11, 1017–
1020; c) L. Liu, L.-Y. Ji, Y.-Y. Wei, Monatsh. Chem. 2008, 139, 901–903;
d) C.-X. Miao, L.-N. He, J.-Q. Wang, J.-L. Wang, Adv. Synth. Catal. 2009,
351, 2209–2216; e) C.-X. Miao, L.-N. He, J.-L. Wang, F. Wu, J. Org. Chem.
2010, 75, 257–260; f) X.-E. Wu, L. Ma, M.-X. Ding, L.-X. Gao, Synlett
2005, 607–610; g) L. Liu, J. Ma, J. Zhang, J. Huang, S. Li, Z. Tong, Can. J.
Keywords: alcohols · aldehydes · EPR spectroscopy · ketones ·
oxidation · supported catalysts
[1] a) M. Hudlicky, Oxidations in Organic Synthesis, American Chemical Soci-
ety, Washington, DC, 1990, pp. 114–149; b) I. W. C. E. Arends, R. A. Shel-
don in Modern Oxidation Methods (Ed.: J.-E. Bꢂckvall), Wiley-VCH, Wein-
heim, 2005, Chap. 4; c) S. Jammi, T. Punniyamurthy, G. B. W. L. Ligthart,
R. H. Meijer, J. v. Buijtenen, J. Meuldijk, J. A. J. M. Vekemans, L. A. Hul-
shof, K. Kittigowittana, M. Pohmakotr, V. Reutrakul, C. Kuhakarn in
Regio- and Stereo- Controlled Oxidations and Reductions, Vol. 5 (Eds.:
S. M. Roberts, J. Whittall), Wiley, Chichester, 2007, Chap. 5.
[2] a) R. A. Sheldon, I. W. C. E. Arends, A. Dijksman, Catal. Today 2000, 57,
157–166; b) M. Musawir, P. N. Davey, G. Kelly, I. V. Kozhenikov, Chem.
Commun. 2003, 1414–1415.
´
Chem. 2011, 89, 68–71; h) S.-S. Wang, Z. Popovic, H.-H. Wu, Y. Liu,
ChemCatChem 2011, 3, 1208–1213; i) J. Zhu, P.-c. Wang, M. Lu, RSC Adv.
2012, 2, 8265–8268; j) L. Liu, J. Ma, J. Xia, L. Li, C. Li, X. Zhang, J. Gong,
Z. Tong, Catal. Commun. 2011, 12, 323–326.
[3] a) M. Dagonneau, E. S. Kagan, V. I. Mikhailov, E. G. Rozantsev, V. D.
Sholle, Synthesis 1984, 895–916; b) P. L. Anelli, C. Biffi, F. Montanari, S.
Quici, J. Org. Chem. 1987, 52, 2559–2562; c) P. L. Anelli, C. Biffi, F. Mon-
tanari, S. Quici, J. Org. Chem. 1989, 54, 2970–2972; d) P. L. Anelli, F.
Montanari, S. Quici, Org. Synth. 1990, 69, 212–219; e) L. Tebben, A.
Studer, Angew. Chem. 2011, 123, 5138–5174; Angew. Chem. Int. Ed.
2011, 50, 5034–5068; f) T. Vogler, A. Studer, Synthesis 2008, 1979–1993.
[4] A. E. J. de Nooy, A. C. Besemer, H. van Bekkum, Synthesis 1996, 35,
1153–1176.
[5] a) A. De Mico, R. Margarita, L. Parlanti, A. Vescovi, G. Piancatelli, J. Org.
Chem. 1997, 62, 6974–6977; b) T. M. Hansen, G. J. Florence, P. Lugo-
Mas, J. Chen, J. N. Abrams, C. J. Forsyth, Tetrahedron Lett. 2003, 44, 57–
59; c) J. B. Epp, T. S. Widlanski, J. Org. Chem. 1999, 64, 293–295; d) O.
Holczknecht, M. Cavazzini, S. Quici, I. Shepperson, G. Pozzi, Adv. Synth.
Catal. 2005, 347, 677–688; e) J.-M. Vatꢃle, Tetrahedron Lett. 2006, 47,
715–718; f) J. S. Yadav, N. N. Yadav, T. S. Rao, B. V. S. Reddy, A. Al Khazi-
m Al Ghamdi, Eur. J. Org. Chem. 2011, 4603–4608.
[6] a) W. Qian, E. Jin, W. Bao, Y. Zhang, Angew. Chem. 2005, 117, 974–977;
Angew. Chem. Int. Ed. 2005, 44, 952–955; b) C. Zhu, A. Yoshimura, Y.
Wei, V. N. Nemykin, V. V. Zhdankin, Tetrahedron Lett. 2012, 53, 1438–
1444; c) T. Dohi, K. Fukushima, T. Kamitanaka, K. Morimoto, N. Takenaga,
Y. Kita, Green Chem. 2012, 14, 1493–1501.
[7] a) C. Bolm, T. Fey, Chem. Commun. 1999, 1795–1796; b) R. Ciriminna, J.
Blum, D. Avnir, M. Pagliaro, Chem. Commun. 2000, 1441–1442; c) T. Fey,
H. Fischer, S. Bachmann, K. Albert, C. Bolm, J. Org. Chem. 2001, 66,
8154–8159; d) B. Karimi, A. Biglari, J. H. Clark, V. Budarin, Angew. Chem.
2007, 119, 7348–7351; Angew. Chem. Int. Ed. 2007, 46, 7210–7213; e) B.
Karimi, E. Badreh, Org. Biomol. Chem. 2011, 9, 4194–4198.
[8] a) A. Schꢂtz, R. N. Grass, W. J. Stark, O. Reiser, Chem. Eur. J. 2008, 14,
8262–8266; b) A. K. Tucker-Schwartz, R. L. Garrell, Chem. Eur. J. 2010,
16, 12718–12726; c) B. Karimi, E. Farhangi, Chem. Eur. J. 2011, 17,
6056–6060.
[9] a) M. Benaglia, A. Puglisi, O. Holczknecht, S. Quici, G. Pozzi, Tetrahedron
2005, 61, 12058–12064; b) C.-X. Miao, L.-N. He, J.-Q. Wang, J. Gao, Syn-
lett 2009, 3291–3294.
[10] a) A. Dijksman, I. W. C. E. Arends, R. A. Sheldon, Chem. Commun. 2000,
271–272; b) C. Tanyeli, A. Gꢄmꢄs¸, Tetrahedron Lett. 2003, 44, 1639–
1642; c) M. Gilhespy, M. Lok, X. Baucherel, Catal. Today 2006, 117, 114–
119.
[11] C. Rçben, A. Studer, W. L. Hemme, H. Eckert, Synlett 2010, 1110 – 1114.
[12] a) Y. Wang, X. Song, S. Shao, H. Zhong, F. Lin, RSC Adv. 2012, 2, 7693–
7698; b) J.-P. Lindner, C. Rçben, A. Studer, M. Stasiak, R. Ronge, A. Grein-
er, H.-J. Wendorff, Angew. Chem. 2009, 121, 9035–9038; Angew. Chem.
Int. Ed. 2009, 48, 8874–8877.
[16] a) C. P. Mehnert, R. A. Cook, N. C. Dispenziere, M. Afeworki, J. Am. Chem.
Soc. 2002, 124, 12932–12933; b) C. P. Mehnert, Chem. Eur. J. 2005, 11,
50–56.
[17] a) M. Gruttadauria, S. Riela, C. Aprile, P. Lo Meo, F. D’Anna, R. Noto, Adv.
Synth. Catal. 2006, 348, 82–92; b) C. Aprile, F. Giacalone, M. Gruttadau-
ria, A. Mossuto Marculescu, R. Noto, J. D. Revell, H. Wennemers, Green
Chem. 2007, 9, 1328–1334; c) V. Sans, N. Karbass, M. I. Burguete, V.
CompaÇ, E. Garcꢆa-Verdugo, S. V. Luis, M. Pawlak, Chem. Eur. J. 2011, 17,
1894–1906.
[18] P. Agrigento, M. J. Beier, J. T. N. Knijnenburg, A. Baiker, M. Gruttadauria,
J. Mater. Chem. 2012, 22, 20728–20735.
[19] C. Aprile, F. Giacalone, P. Agrigento, L. F. Liotta, J. A. Martens, P. P. Pescar-
mona, M. Gruttadauria, ChemSusChem 2011, 4, 1830–1837.
[20] a) M. Gruttadauria, L. F. Liotta, A. M. P. Salvo, F. Giacalone, V. La Parola, C.
Aprile, R. Noto, Adv. Synth. Catal. 2011, 353, 2119–2130; b) M. Grutta-
dauria, L. A. Bivona, P. Lo Meo, S. Riela, R. Noto, Eur. J. Org. Chem. 2012,
2635–2642.
[21] E. Montroni, M. Lombardo, A. Quintavalla, C. Trombini, M. Gruttadauria,
F. Giacalone, ChemCatChem 2012, 4, 1000–1006.
[22] For a review on non-covalent immobilization of asymmetric organoca-
talysis see: L. Zhang, S. Luo, J.-P. Cheng, Catal. Sci. Technol. 2011, 1,
507–516.
[23] For a recent example of catalyst recycling via non-covalent adsorption
see: A. M. Kluwer, C. Simons, Q. Knijnenburg, J. I. van der Vlugt, B. de
Bruinc, J. N. H. Reek, Dalton Trans. 2013, 42, 3609–3616.
[24] For a flow system based on non-covalent catalyst immobilization see: Y.
Suzuki, P. Laurino, D. T. McQuade, P. H. Seeberger, Helv. Chim. Acta 2012,
95, 2578–2588.
[25] a) L. Valgimigli, M. Lucarini, G. F. Pedulli, K. U. Ingold, J. Am. Chem. Soc.
1997, 119, 8095–8096; b) V. Mugnaini, C. Punta, R. Liantonio, P. Metran-
golo, F. Recupero, G. Resnati, G. F. Pedulli, M. Lucarini, Tetrahedron Lett.
2006, 47, 3265–3269.
[26] a) A. Alberti, M. Benaglia, P. Hapiot, A. Hudson, G. Le Coustumer, D. Mac-
ciantelli, S. Masson, J. Chem. Soc. Perkin Trans. 2 2000, 1908–1913; b) C.
Boga, M. Calvaresi, P. Franchi, M. Lucarini, S. Fazzini, D. Spinelli, D. Tonel-
li, Org. Biomol. Chem. 2012, 10, 7986–7995.
[27] W. Zhang, W. Zhang, N. Zhou, J. Zhu, Z. Cheng, X. Zhu, J. Polym. Sci.
Part A 2009, 47, 6304–6315.
Received: March 27, 2013
[13] a) G. Oehme in Comprehensive Asymmetric Catalysis I–III, Vol. 3 (Eds.:
E. N. Jacobsen, A. Pfaltz, H. Yamamoto), Springer, Berlin, Germany, 1999,
Published online on && &&, 0000
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemCatChem 0000, 00, 1 – 10
&10
&
ÞÞ
These are not the final page numbers!