An immobilized imidazolyl manganese porphyrin for the oxidation of olefins

Article abstract of DOI:10.1016/j.molcata.2015.04.021

A new catalytic system based on an immobilized imidazolyl manganese porphyrin for the oxidation of olefins is presented. Merrifield resin (MR) and functionalized silica gel (SG) were chosen as supports. The results indicate that the MR system shows high reaction rates, high efficiency with hydrogen peroxide as oxidant and good recyclability up to four times, without a dramatic loss in the catalytic efficiency. The catalytic behavior seems to be strongly influenced by the immobilization reaction conditions. The oxidation reactions performed for cis-cyclooctene, styrene, cyclohexene and geraniol give the corresponding epoxides, with very high selectivity, when the MR system is used. Some considerations concerning the high efficiency of the MR system are put forward.

Full text of DOI:10.1016/j.molcata.2015.04.021

Accepted Manuscript
Title: An immobilized imidazolyl manganese porphyrin for
the oxidation of olefins
Author: Rodrigo De Paula Isabel C.M.S. Santos M a´ rio M.Q.
Sim o˜ es M.Gra c¸ a P.M.S. Neves Jos e´ A.S. Cavaleiro
PII:
S1381-1169(15)00171-5
DOI:
Reference:
http://dx.doi.org/doi:10.1016/j.molcata.2015.04.021
MOLCAA 9484
To appear in:
Journal of Molecular Catalysis A: Chemical
Received date:
Revised date:
Accepted date:
17-12-2014
28-3-2015
27-4-2015
Please cite this article as: Rodrigo De Paula, Isabel C.M.S.Santos, M a´ rio M.Q.Sim o˜
es, M.Gra c¸ a P.M.S.Neves, Jos e´ A.S.Cavaleiro, An immobilized imidazolyl manganese
porphyrin for the oxidation of olefins, Journal of Molecular Catalysis A: Chemical
http://dx.doi.org/10.1016/j.molcata.2015.04.021
This is a PDF file of an unedited manuscript that has been accepted for publication.
As a service to our customers we are providing this early version of the manuscript.
The manuscript will undergo copyediting, typesetting, and review of the resulting proof
before it is published in its final form. Please note that during the production process
errors may be discovered which could affect the content, and all legal disclaimers that
apply to the journal pertain.

An immobilized imidazolyl manganese porphyrin
for the oxidation of olefins
§
*
Rodrigo De Paula , Isabel C. M. S. Santos, Mário M. Q. Simões, M. Graça P. M. S. Neves , José
*
A. S. Cavaleiro
QOPNA, Department of Chemistry, University of Aveiro, 3810-193 Aveiro, Portugal
*
Corresponding Authors
Professor Dr. M. Graça P. M. S. Neves
Phone: +351 234 370 710; Fax: +351 234 370 084; E-mail: gneves@ua.pt
Professor Dr. José A. S. Cavaleiro
Phone: +351 234 370 717; Fax: +351 234 370 084; E-mail: jcavaleiro@ua.pt
§
Present Address: Universidade Federal do Recôncavo da Bahia – Centro de Formação de Professores, 45.300-000
Amargosa/BA – Brazil, Tel: +55 75 3634 3042, Fax: +55 75 3634 3703); e-mail: rodrigodepaula@ufrb.edu.br
1

Graphical Abstract
2

Highlights
•
•
•
•
A new catalytic system based on an immobilized imidazolyl manganese porphyrin.
Merrifield resin (MR) and functionalized silica gel (SG) as supports.
MR system shows high reaction rates, high efficiency and good recyclability.
Very high selectivity for cyclooctene, styrene, cyclohexene and geraniol epoxides.
3

Abstract
A new catalytic system based on an immobilized imidazolyl manganese porphyrin for the
oxidation of olefins is presented. Merrifield resin (MR) and functionalized silica gel (SG) were
chosen as supports. The results indicate that the MR system shows high reaction rates, high
efficiency with hydrogen peroxide as oxidant and good recyclability up to four times, without a
dramatic loss in the catalytic efficiency. The catalytic behavior seems to be strongly influenced
by the immobilization reaction conditions. The oxidation reactions performed for cis-
cyclooctene, styrene, cyclohexene and geraniol give the corresponding epoxides, with very high
selectivity, when the MR system is used. Some considerations concerning the high efficiency of
the MR system are put forward.
KEYWORDS: manganese(III)porphyrin; tert-butylhydroperoxide; hydrogen peroxide;
Merrifield resin; functionalized silica; oxidation.
4

1
. INTRODUCTION
Metalloporphyrins are widely used as catalysts in oxidation reactions,[1-3] mimicking the natural
functions performed by cytochrome P450 monooxygenase enzymes.[4-7] The natural system is
constituted by an iron porphyrin surrounded by proteins which is capable of selectively catalyze
the oxidation and, in particular, the epoxidation of a wide range of substrates.[8-10] Iron is the
“
workhorse” in many biological processes and, concerning the monooxygenases, iron drives the
oxidation of many substrates by mechanistic routes involving a hypervalent oxo-species.[9,11-
8] Like iron, manganese also forms a high-valent oxo intermediate that can promote the
1
oxidation of many substrates,[19-21] and manganese complexes are often preferred for
biomimetic in vitro catalysis, due to their higher activity.[9,10]
There are a lot of reports concerning metalloporphyrins catalyzed oxidation under homogeneous
conditions.[9,10,21-34] However, the synthesis of metalloporphyrins is a challenging task and
usually a low yielding process. Also the known instability of these catalysts towards oxidative
degradation and the difficulty in their recovery can limit the putative practical applications of
metalloporphyrins as catalysts in both laboratory synthetic chemistry and industrial processes.
On the other hand, immobilization of metalloporphyrin complexes on solid supports can provide
catalysts easier to handle, that may exhibit improved selectivity and activity due to the support
environment.[35-40] To date, different approaches have been developed for the design of
heterogeneous metalloporphyrin catalysts, including: (i) electrostatic binding of charged
porphyrins to counter-charged supports; (ii) intercalation or entrapment of porphyrins between
the layers of clays or within the pores or matrices of solids; (iii) axial anchoring to surface-bound
ligands; and (iv) covalent binding to appropriate supports. Recently, our group was able to
5

develop two immobilized systems, one using a manganese complex of 5,10,15,20-tetrakis(2,6-
dichlorophenyl)porphyrin[41] and the other based on the corresponding chlorin.[42]
Differently from the well-studied meso-phenyl substituted porphyrins, we have recently
published the use of an imidazolium-based tetracationic manganese porphyrin as catalyst under
homogeneous conditions.[12,24] The neutral parent of that catalyst, the manganese complex of
5
,10,15,20-tetrakis(1-methylimidazol-2-yl)-21H,23H-porphyrin [Mn(Porph)] did not exhibit a
great catalytic activity in homogeneous media; however, when linked to a solid support, a great
improvement is observed. This paper deals with the experimental results obtained for the
imidazolyl-based manganese porphyrin covalently bound to two distinct solid supports, the
3
-bromopropylfunctionalized silica gel and the Merrifield resin (Figure 1) in the epoxidation of
cis-cyclooctene. This support selection was based on the fact that the 3-bromopropylsilica and
the Merrifield resin allow, by a simple nucleophilic substitution reaction, to covalently anchor
the metalloporphyrin, instead of other systems where the immobilization lies on support/metal
interactions or through support/macrocycle electrostatic interactions.
The most promising Mn(Porph)-MR heterogeneous catalyst was also tested for styrene,
cyclohexene and geraniol oxidation and the results are also presented here.
2
2
. RESULTS AND DISCUSSION
.1. Synthesis and Characterization
Some previous results showed that the neutral manganese(III) complex of 5,10,15,20-tetrakis(1-
methylimidazol-2-yl)-21H,23H-porphyrin has extremely low catalytic activity.[11] However,
once positively charged, its activity changes dramatically reaching high conversion and high
selectivity to epoxide for many olefins.[24] In this way, we have decided to attach the
6

metalloporphyrin into a solid support, in order to obtain a positively charged and immobilized
material, simultaneously. The different experimental procedures used to immobilize the
Mn(Porph) in the two supports selected, the Merrifield resin (MR) and the 3-
bromopropylfunctionalized silica gel (SG), are summarized in Scheme 1; these conditions were
adapted from a procedure reported by Tangestaninejad et al..[36,43]
N
N
CH
3
Br
CH2
CH2
CH2
silica
N
N
support
CH
3
n
N
N
N
N
Cl
Mn(III)
N
N
n = 1-4
Cl
CH
2
Merrifield
N
3
H C
N
3
H C
SG support
MR support
(CH2)3
silica
=
=
3
H C
N
N
CH
2
Merrifield
Mn(Porph)
Figure 1. Structure of the immobilized manganese porphyrin (Mn(Porph)). The supports are a
functionalized silica (3-bromopropylfunctionalized silica gel) and a Merrifield resin (SG and
MR, respectively).
For the Merrifield resin based heterogeneous catalyst [Mn-(Porph)-MR], two different
conditions were used being the principal difference the temperature of the immobilization. In the
first procedure, designed by MR-Batch-1, the MR was added to a DMF solution of
Mn(TMImP)Cl and the reaction was left at 80 ºC under strong stirring for 2 days in an inert
atmosphere (N ), plus 24 hours at r.t. before filtration. Under these conditions, the experimental
2
loading obtained was 0.75% far from the maximum expected one, 4.78% (w/w) if all the
7

metalloporphyrin added was incorporated into the solid. In the second procedure (MR-Batch-2),
aiming to improve the load of catalyst, the ratio catalyst/MR was slightly increased to 6.6%
(
w/w) and the immobilization was performed at 150 ºC. After 4 hours under N , a control
2
showed that almost no free porphyrin was present, and the reaction was terminated as in the
previous procedure. Under these conditions the porphyrin loading was really improved to 4.8%.
MR-Batch 3 with the same loading of 4.8 % was obtained under similar conditions of MR-
Batch-2 which confirmed the reproducibility of the immobilization conditions.
The immobilization in the 3-bromopropylfunctionalized silica gel (SG) was also performed in
DMF at 150 ºC for 4 hours under N , and the work-up was exactly the same as for the MR
2
heterogeneous materials. For SG-materials, the two batches were prepared using two different
theoretical porphyrin/support ratios – 5.2 % (w/w) and 3.6% (w/w). Interestingly, the best
loading (4.2%, SG-Batch-2) was obtained with the more diluted proportion.
The materials obtained have a light brown colour due to the presence of the metalloporphyrin.
Several spectroscopic techniques were used to characterize the new compounds. All the results
seem to indicate that the porphyrin was incorporated into the new material. For the immobilized
systems, diffuse reflectance (Figure 2-A), FT-IR (Figure 2-B) and SEM (Figure 3) were used to
characterize the materials. The diffuse reflectance spectra show the usual electronic profile for
this kind of materials (Figure 2A). The visible spectra in solution of the Mn(Porph) is
characterized by a strong Soret band at about 455–465 nm and one Q-band between 550 and 650
nm (Figure 2A). The diffuse reflectance spectra of these materials were similar, showing the Q-
band observed in solution spectra but with the relative intensity of the Soret band comparatively
diminished, as it is possible to observe in the case of Mn(Porph)-SG (Figure 2A). In the case of
Mn(Porph)-MR the Soret band is not seen in the spectra, however this is a characteristic of
8

some porphyrins, probably due to the very dark material or to some aggregation. In the same
figure, the UV-Vis spectra of the Mn(Porph) in CH CN solution, of the Mn(Porph) powder and
3
of the supports (MR and SG) are also shown for comparison.
The infrared spectra are shown in Figure 2B, where it can be seen the spectra of MR,
Mn(Porph)-MR, Mn(Porph) and its charged complex (the cationic manganese porphyrin, a N-
methylated catalyst). Unfortunately, the infrared spectra for Mn(Porph)-SG was not useful, due
to the O-H stretch band of silica, which overlaps the other bands.
Scanning electronic microscopy (SEM) images were taken before (A) and after (B-I)
incorporation of the metalloporphyrin (Figure 3), showing some differences on the materials
surface. All the characterization results indicate that the metalloporphyrin used is present in the
solid compounds obtained without noticeable changes.
Mn(Porph) powder
Mn(Porph)-SG
Mn(Porph) in CH CN
3
1.4
1.2
1.0
0.8
0.6
0.4
0.2
0.0
Mn(Porph)-MR
Merrifield Resin (MR)
Merrifield
Resin (MR)
3-bromopropylsilica (SG)
Mn(Porph) - MR
Mn(Porph)
4+
[
Mn(Porph)]
3
50
400
450
500
550
600
650
700
750
800
5
00
1000
1500
2000
ν
2500
3000
3500
4000
λ
,
nm
-1
,
cm
(
A)
(B)
Figure 2. (A) Electronic profile for the Merrifield resin (MR), the 3-bromopropylsilica (SG), the
Mn(Porph) powder, the immobilized materials (diffuse reflectance) and the metalloporphyrin in
CH CN solution (UV-Vis); (B) FT-IR for Merrifield resin (MR), Mn(Porph)-MR, Mn(Porph)
3
4
+
and [Mn(Porph)] (N-methylated Mn(Porph)).
9

Scheme 1. Routes for the preparation of the heterogeneous catalysts; 3-bromopropylfunctionalized silica gel (SG); Merrifield resin (MR).
SG-BATCH 1
MR-BATCH 1
Mn(Porph)-MR
3
-bromopropyl-
Merrifield
Resin
Mn(Porph)-SG
SG-BATCH 2
Mn(Porph)
functionalised silica gel
MR-BATCH 2 and 3
-5
-5
BATCH 1: 16.5 mg (2.29x10 mol) of Mn(Porph), 295.5 mg
of SG, 153 ºC, DMF, N2(g), 4 hours.
BATCH 1: (1) 25.2 mg (3.50x10 mol) of Mn(Porph), 502.6 mg
of MR, 80 ºC, DMF, N2(g), 48 hours; (2) 24 hours under stirring.
-5
-5
BATCH 2: 15,1 mg (2,10x10 mol) of Mn(Porph), 405.5 mg
of SG, 153 ºC, DMF, N2(g), 4 hours.
BATCH 2: 39.7 mg (5.52x10 mol) of Mn(Porph), 563.1 mg
of MR 150 ºC, DMF, N2(g), 4 hours. (BATCH 2)
-
5
BATCH 3: 16.3 mg (2.27x10 mol) of Mn(Porph), 227.8 mg
of MR 150 ºC, DMF, N2(g), 4 hours. (BATCH 3)
10

(A)
(B-I)
(B-II)
(C)
Figure 3. SEM images taken from: (A) MR; (B-I) Mn(Porph)-MR (Batch 3) before and (B-
II) after the catalytic reaction; (C) a zoom on (B-I) image.
2
.2. Catalytic activity for cis-cyclooctene epoxidation reactions
(
C) (B´)
To assess the catalytic properties of the new heterogeneous materials, the oxidation of cis-
cyclooctene was carried out using H O and t-BuOOH as the oxidants. In Table 1 are
2
2
summarized the results obtained with the different Mn(Porph)-SG batches and in Table 2 the
ones obtained with the Mn(Porph)-MR batches. In general the Mn(Porph)-MR systems
shows higher activity than Mn(Porph)-SG counterpart. Although silica containing materials
are frequently used as supports in catalysis, some reports highlight the efficiency of MR in
biomimetic oxidation with metalloporphyrins.[44]
Regarding the catalytic activity and recyclability registered with the Mn(Porph)-SG systems
the results summarized in Table 1 show that the high conversion observed in the first cycle
2 2
suffers a drastic drop with the reuse, independently of the oxidant used, H O or t-BuOOH.
11

The results obtained for Mn(Porph)-SG allow us to conclude that after the addition of 5-6
equivalents of H O , the batch 2 (4.19 % loading) material gives rise to a higher conversion in
2
2
the first (99.2 %) and in the second cycles (52.7 %), against 79.4 % and 2.2 %, respectively,
for batch 1 (1.08 % loading). This can be related to the amount of Mn(Porph) anchored into
the silica materials, since batch 2 presents a nearly 4 times higher loading than batch 1.
2 2
Unfortunately, the efficiency of batch 2 in the presence of H O drops drastically in the third
cycle (entry 7). It is worth to refer that the reactions were stopped when, after two successive
oxidant additions, the substrate conversion remained constant (see experimental section).
Moreover, under homogeneous conditions, the same Mn(Porph) gives rise to 100 % of
conversion after 75-90 min and 2.5-3.0 equivalents of H O , under similar reaction
2
2
conditions, so the somewhat longer reaction times are not surprising under heterogeneous
conditions.
Table 1. Experimental results obtained for cis-cyclooctene oxidation catalyzed by
(
a)
heterogeneous catalysts Mn(Porph)-SG with t-BuOOH or H
O .
2 2
(
b)
Entry
Batch
Oxidant
Oxidant addition
frequency (min)
Oxidant
(equiv.)
Time of
reaction (min)
Conversion
(%)
Cycles
1
2
3
4
5
6
7
Batch 1
Batch 1
Batch 1
Batch 1
Batch 2
Batch 2
Batch 2
---
t-BuOOH
t-BuOOH
30
30
30
30
15
15
15
15
6
6
6
5
5
6
3
6
360
360
360
300
150
180
90
81.2
12.8
79.4
2.2
1
2
H
2
H
2
H
2
H
2
H
2
H
2
O
O
O
O
O
O
2
2
2
2
2
2
1
2
99.2
52.7
4.9
1
2
3
(
c)
8
180
0
---
(
a)
substrate (0.3 mmol), catalyst (50 mg) and co-catalyst (3.3 mmol). Each oxidant addition corresponds to a half
(
b)
of the initial substrate concentration (i.e., 0.15 mmol). All reactions were carried out in CH
3
CN (0.5 mL); the
corresponding epoxide was obtained as the only product; blank reaction, using just the support with co-catalyst
acetic acid).
(c)
(
12

Table 2. Experimental results obtained for cis-cyclooctene oxidation catalyzed by
(
a)
heterogeneous catalysts Mn(Porph)-MR with t-BuOOH or H O .
2
2
(
b)
Entry
Batch
Oxidant
Oxidant addition
frequency (min)
Oxidant
(equiv.)
6
Time of
reaction (min)
Conversion
(%)
Cycles
1
2
3
4
5
6
7
8
9
Batch 1
Batch 1
Batch 1
Batch 1
Batch 1
Batch 1
Batch 2
Batch 2
Batch 2
Batch 2
Batch 3
Batch 3
Batch 3
---
t-BuOOH
t-BuOOH
t-BuOOH
t-BuOOH
30
30
15
15
30
30
15
15
15
15
15
15
15
15
360
360
210
210
360
360
150
180
180
180
150
180
180
180
89.5
53.2
70.4
60.0
96.7
57.0
100
96.4
91.4
68.7
100
100
85.8
0
1
2
6
3
3
.5
.5
1
2
6
6
5
6
6
6
5
6
6
6
H
2
H
2
H
2
H
2
H
2
H
2
H
2
H
2
H
2
H
2
O
O
O
O
O
O
O
O
O
O
2
2
2
2
2
2
2
2
2
2
1
2
1
2
3
10
11
12
13
4
1
2
3
(
c)
14
---
(
a)
substrate (0.3 mmol), catalyst (50 mg) and co-catalyst (3.3 mmol). Each oxidant addition corresponds to a half
(
b)
of the initial substrate concentration (i.e., 0.15 mmol). All reactions were carried out in CH
3
CN (0.5 mL); the
corresponding epoxide was obtained as the only product; blank reaction, using just the support with co-catalyst
acetic acid).
(c)
(
The first experiments performed with Mn(Porph)-MR batch 1 (Table 2) in the presence of
H O (entries 5,6) and t-BuOOH (entries 1-4) show that the best conversion in the first cycle
2
2
was obtained with the first oxygen donor (96.7% vs 89.5%). However, in both cases the
conversions obtained drop significantly to ca. 50% with the catalyst recycle (entries 2,6).
With this batch no improvement on the recycle was observed when the addition of the oxidant
was performed at each 15 min instead of 30 min (entries 3,4 for t-BuOOH; data not shown for
H O ). The differences in the procedures employed in the synthesis of the heterogeneous
2
2
13

catalysts and consequently in the amount of porphyrin immobilized are reflected in the
catalytic efficiency observed during the epoxidation reactions with Mn(Porph)-MR batches 2
and 3.
The results obtained for Mn(Porph)-MR after the addition of 5-6 equivalents of H O are
2
2
always higher than those registered for Mn(Porph)-SG. Conversions between 96.7-100%
were obtained in the first cycle, the Mn(Porph) catalyst loading representing the major factor
of differentiation. Batch 1 (0.75% loading) gives only 57.0% of conversion in the second
cycle, whereas batches 2 (4.83% loading) and 3 (4.80% loading) give rise to 96.4 and 100%,
respectively. So, Mn(Porph)-MR batch 2 exhibits high catalytic efficiency, that is
maintained after being recycled three times; even in the fourth recycle an agreeable
conversion of near 70% was attained (entries 7-10). To ascertain if the reproducibility of that
feature is not related specially with batch 2, the oxidation experiments were repeated with
Mn(Porph)-MR batch 3 with the same metalloporphyrin content. As expected, the efficiency
of the system is maintained when the Mn(Porph)-MR batch 3 was used as catalyst (entries
1
1-13). These results show that, under the same conditions, Mn(Porph)-SG presents lower
performance when compared with the Mn(Porph)-MR materials. These can be attributed to a
higher stability of the covalent linkage established between the Mn(Porph) and the MR
support, since several cycles are viable for the Mn(Porph)-MR material, and no catalyst
leaching seems to occur, as can be seen on Figure 4.
The control experiments (entries 8 in table 1 and 14 in table 2) confirm that, in the absence of
the Mn-porphyrin, no conversion is observed, even using acetic acid (the co-catalyst) in the
reaction medium. Hence, the possibility of forming peroxyacetic acid in situ as the putative
epoxidation reagent can be disregarded.
Using the conditions described in entry 7 (Table 2), additional experiments were run to
evaluate the possibility of bleaching during catalysis reactions. Figure 4 illustrates the
14

bleaching and the recycling assays. Apparently, the decrease in catalytic activity observed
from the first to the fourth cycles using Mn(Porph)-MR (Figure 4) was not due to the
bleaching of the catalyst. Thus, a possible explanation can be attributed to the degradation or
the inactivation of the catalyst, along the catalytic cycles. Scanning electronic microscopy
(SEM) images were taken before and after the first cycle (Figure 3 B-I and 3 B-II), showing
some differences on the materials surface, which can be the reason for the registered decrease
in the catalytic activity.
[
H O ]/[Olefin]
2 2 0
0
1
2
3
4
5
6
100
7
5
0
5
0
Filter off
5
2
cis-cyclooctene oxidation with H O2
2
catalyzed by Mn(Porph)-MR
bleaching
cycle 1
cycle
cycle 2
cycle 4
3
0
30
60
90
120
150
180
Time of reaction, min
Figure 4. Catalyst recycling and bleaching experiments with Mn(Porph)-MR for
cis-cyclooctene oxidation with H . The results shown for cycles 1-4 are related to entries 7-
2
O
2
10 (Table 2).
2.3. Styrene, cyclohexene and geraniol oxidation
Additional reactions were performed to check the catalytic efficiency of the best material
Mn(Porph)-MR towards other olefins such as styrene, cyclohexene and geraniol. The results
obtained in those studies are summarized in Tables 3-5.
It is known that styrene epoxidation is not a clean reaction, since side-products like
15

phenylacetaldehyde and benzaldehyde can be observed in biomimetic oxidations catalyzed by
metalloporphyrins[45,46] and cytochrome P450 enzymes.[47] In this work, high substrate
th
conversion was registered for the first four cycles, decreasing in the 5 cycle, while high
selectivity to the styrene epoxide (ca 90%) was observed up to 5 cycles (Table 3, entries 1-5).
The selectivity for the phenylacetaldehyde (7%) and benzaldehyde (3%) was not altered
during the recycles. Under homogeneous conditions, the tetracationic metalloporphyrin gives
the same products, although the selectivity for the epoxide can be improved by decreasing the
substrate/catalyst ratio (Table 3, entries 6,7).
Regarding cyclohexene oxidation, the high conversion (97-100%) accompanied by high
selectivity for the cyclohexene oxide (ca 90%) were kept almost constant in all the catalyst
recycles. The selectivity for cyclohex-2-en-1-ol and cyclohex-2-en-1-one, resulting from
allylic oxidation (global yield ~9%) did not suffer also great fluctuations during the catalyst
reuse (Table 4, entries 1-4), although the selectivity is higher than that observed under
homogeneous conditions (Table 4, entries 5,6). As it was already mentioned, for this olefin,
the efficiency was kept almost constant in all the catalyst recycles, while in homogeneous
conditions the catalyst stability was lower than 50% at the end of the reaction (from UV/Vis
spectrum; not shown). It is known that cyclohexene auto-oxidation can involve molecular
oxygen participation through a radical pathway,[48] thus more than a single active catalytic
species being present.[49]
As expected, the 6,7-epoxide was the main product obtained for geraniol oxidation, followed
by the 2,3-epoxide isomer and the corresponding 2,3,6,7-diepoxide (Table 5, entries 1-4).
Homogeneous conditions yielded a slightly higher selectivity for 6,7-epoxide than the
heterogeneous conditions (Table 5, entry 5). Curiously, under heterogeneous conditions a
little bit more 2,3-epoxide was obtained, a typical situation found in reactions where the metal
centre undergoes attraction to the hydroxyl group present in the geraniol structure.[50] Such
16

an observation indicates that, under some circumstances, the somewhat analogous protein
matrix microenvironment created by the support - in heme enzymes this is the role of the
protein matrix which surrounds the prosthetic group - can drive some changes on the
selectivity of the reaction.[51] The efficiency observed with this kind of metalloporphyrins
can be due to some structural features which do both effects, the electron withdrawing effect
and the protection of the porphyrin ring. The conversion obtained in the first cycle (ca 70%),
after a fall in the second run into ~50%, maintain this last value almost constant in the
posterior reuses.
17

(a)
Table 3. Styrene oxidation with H
2
O
2
catalyzed by Mn(Porph)-MR following the conditions described in Table 2 (for batch 3, entry 11).
(cationic metalloporphyrin).
In the homogeneous conditions the Mn(Porph) used was [Mn(TDMImP)Cl]I
4
Time of
reaction
Entry
Olefin
Cycle
Selectivity (%)
Conversion (%)
(min)
Styrene oxide
90.9 ± 2.1
90.6 ± 1.4
88.6 ± 2.3
89.4 ± 5.8
88.1 ± 1.7
Phenylacetaldehyde
7.2 ± 3.0
Benzaldehyde
2.3 ± 0.4
1
2
3
4
5
1
2
3
4
92.1
89.8
84.4
78.8
57.9
180
180
180
180
180
7.4 ± 2.2
2.4 ± 0.3
6.9 ± 1.8
4.5 ± 0.8
7.7 ± 4.6
2.9 ± 1.5
5
6.6 ± 2.5
5.3 ± 0.8
Olefin/Catalyst ratio
(
b)
–
homogeneous
(
c)
6
75
> 95
81
traces
18
traces
< 1
100
57
60
60
Styrene
(
d)
7
300
(
a)
reaction conditions (batch 3, Table 2, entry 11): substrate/catalyst/co-catalyst = 0.3 mmol/0.522 µmol/3.325 mmol; CH
3
CN, 0.5 mL. Each oxidant addition
(
b)
corresponds to a half of the initial substrate concentration (i.e., 0.15 mmol). homogeneous conditions: solvent: acetonitrile (2 mL of total volume); concentration
of the catalyst employed was 2.5 x 10 mol.dm ; co-catalyst: acetic acid (0.42 mmol); the oxidant used was 30% (w/w) aqueous H O diluted in acetonitrile
2 2
(
-
4
-3
(
c)
(d)
1:10) and its addition was made gradually each 15 min of reaction. Each H
2 2
O addition corresponds to a half-substrate amount; from reference [24]. similar to
reference [24], except for the Olefin/Catalyst ratio.
18

Table 4. Cyclohexene oxidation with H
2
O
2
catalyzed by Mn(Porph)-MR following the conditions described in Table 2 (for batch 3, entry
(a)
1
1). In the homogeneous conditions the Mn(Porph) used was [Mn(TDMImP)Cl]I (cationic metalloporphyrin).
4
Time of reaction
Entry
Olefin
Cycle
Selectivity (%)
Conversion (%)
(min)
Cyclohexene oxide
90.8 ± 1.9
Cyclohex-2-en-1-ol
3.9 ± 2.4
Cyclohex-2-en-1-one
5.4 ± 3.7
1
2
3
4
1
2
3
100
98.9
98.5
97.4
150
150
150
150
90.7 ± 2.5
3.3 ± 2.1
5.9 ± 3.7
91.0 ± 2.0
3.4 ± 2.2
5.6 ± 3.7
4
91.3 ± 2.0
3.5 ± 2.3
5.2 ± 3.7
Olefin/Catalyst ratio
(
b)
–
homogeneous
(
c)
5
75
100
95
0
3
0
2
100
57
60
60
Cyclohexene
(
d)
6
300
(
a)
reaction conditions (batch 3, Table 2, entry 11): substrate/catalyst/co-catalyst = 0.3 mmol/0.522 µmol/3.325 mmol; CH
3
CN, 0.5 mL Each oxidant addition corresponds
(
b)
to a half of the initial substrate concentration (i.e., 0.15 mmol). homogeneous conditions: solvent: acetonitrile (2 mL of total volume); concentration of the catalyst
employed was 2.5 x 10 mol.dm ; co-catalyst: acetic acid (0.42 mmol); the oxidant used was 30% (w/w) aqueous H O diluted in acetonitrile (1:10) and its addition was
2 2
made gradually each 15 min of reaction. Each H
-
4
-3
(
c)
(d)
2 2
O addition corresponds to a half-substrate amount; from reference [24]. similar to reference [24], except for the
Olefin/Catalyst ratio.
19

(a)
Table 5. Geraniol oxidation with H
2
O
2
catalyzed by Mn(Porph)-MR following the conditions described in Table 2 (for batch 3, entry 11).
(cationic metalloporphyrin).
In the homogeneous conditions the Mn(Porph) used was [Mn(TDMImP)Cl]I
4
Time of reaction
Entry
Olefin
Cycle
Selectivity (%)
Conversion (%)
(min)
2
,3-epoxy
6,7-epoxy
geraniol
2,3,6,7-diepoxy
geraniol
geraniol
1
2
3
4
1
2
3
36.2 ± 1.8
49.2 ± 1.9
48.9 ± 1.7
50.5 ± 1.1
46.5 ± 1.6
14.6 ± 3.2
9.9 ± 2.6
9.7 ± 2.1
14.2 ± 5.6
72.7
53.4
54.0
51.2
180
180
180
180
41.2 ± 1.0
39.8 ± 1.1
39.3 ± 4.7
OH
4
Olefin/Catalyst ratio
–
homogeneous
Geraniol
(
b)
5
300
30
58
12
67
60
(
a)
3
reaction conditions (batch 3, Table 2, entry 11): substrate/catalyst/co-catalyst = 0.3 mmol/0.522 µmol/3.325 mmol; CH CN, 0.5 mL Each oxidant addition
(
b)
corresponds to a half of the initial substrate concentration (i.e., 0.15 mmol). from reference [52].
20

2.4. Efficiency rationale
It is know that the manganese(III) complexes of 5,10,15,20-tetrakis(2,6-
dichlorophenyl)porphyrin [Mn(TDCPP)Cl] and 5,10,15,20-tetrakis(pentafluorophenyl)
porphyrin [Mn(TPFPP)Cl] are two of the most important compounds in porphyrins’
catalytic biomimetic oxidation. Their efficiency is related to the strong electron-
withdrawing character and the protection effect. Figure 5 illustrates the putative
4
+
structural similarities among [Mn(TDCPP)Cl], [Mn(TPFPP)Cl], [Mn(4-TDMImP)Cl]
as well as the immobilized analogue of the latter, where the highlights show the bulky
substituents near the macrocycle.
[
support]
H C
3
+
+
N
N
N
N
N
N
+
+
Mn
Mn
N
N
H C
H C
3
3
III
4+
III
4+
[
Mn (4-TDMImP)Cl]
[Mn (4-TDMImP)Cl]
[support]
Cl
F
F
N
N
+
+
Mn
Mn
F
N
N
Cl
F
F
III
III
Mn (TDCPP)Cl
Mn (TPFPP)Cl
Figure 5. Structural similarities amongst some manganese (III) porphyrin derivatives.
The highlight shows the bulky substituents near the macrocycle.
2
1

One way to assess the robustness of metalloporphyrins is reached by comparing the
UV/Visible spectrum in neutral and in acidic media (e.g., with and without hydrochloric
acid).[53] The differences observed point out the orbital cross susceptibility, which is
due to the substituent effect. From the practical point of view, such spectral differences
are given by intensity, profile and number of Q-bands (mainly Q (0,0)).
x
R
NH
N
N
R
R
HN
R
N
N
H -TPP
R =
R =
H -TMImP R =
2
2
Cl
H C
3
H -TDCPP
2
N
N
N
H -TTz2PP R =
Cl
F
2
F
R =
F
H C
H -TPFPP
3
I-
2
N
N
F
F
R =
H -TDMImP.I
2
4
H C
R =
R =
N
3
H -4TPyP
2
I-
N
H -4TPyP.I
R =
N
CH
3
2
4
H -2TPyP
2
Figure 6. Structures and acronyms of the porphyrin derivatives studied
The results summarized in Table 6 (please refer to Figure 6 for the structures and
acronyms of the porphyrin derivatives) indicate that electron-withdrawing groups
2
2

attached to the porphyrin ring turn the metallic center highly electrophilic. Nonetheless,
ionic substituents avoid the reaction once they change the oxidation state of manganese
hindering the oxygen atom transfer from the active porphyrin species to the substrate.
The results show that H -TPP was stabilized in acidic media, once its conjugated acid
2
2
+
[
H
4
-TPP] exhibited a batochromic shift (∆λ= +16 nm), thus indicating an extension in
conjugation between the phenyl rings and the macrocycle (entries 1,2).
On the other hand, the phenyl derivatives H -TDCPP (entries 3,4) and H
entries 5,6), which are precursors of high stable and efficient catalysts, originated
2
2
-TPFPP
(
hypsochromic shifts indicating that the substituents exhibit a strong electron-
withdrawing effect on the macrocycle and, due to steric effects, no conjugation can be
thought. Dealing with porphyrins bearing heterocyclic units in the meso-positions
2
+
(
entries 7-18)
,
entries 7-8 show the pair H
2
-4TPyP/[H
4
-4TPyP] and entries 9,10 show
2
+
the isomeric pair H
2
-2TPyP/[H -2TPyP] . By comparing the results obtained for these
4
systems, it can be seen that in the first case a stabilization gave a shift of ∆λ= +24 nm,
whereas in the latter case the shift was ∆λ= +17 nm, indicating that steric hindering is
the limiting feature to explain such differences. In the same way, when the analysis is
2
carried out to H -TMImP (entry 11), the hindering effect due to a methyl group attached
to one imidazolyl nitrogen is responsible by a shift of +6 nm (entry 12). The behavior
observed among H -4TPyP, H -2TPyP and H -TMImP reflects the differences in steric
2
2
2
and electronic properties due to the meso-substituent. The imidazolyl group is more
electronegative than the pyridil moiety. But such difference has no correlation with the
ring size, since a great stabilization was observed with the derivative H
2
-TTz2PP
(entries 13,14 for ionic species), which has a triazolyl group attached to the porphyrin
ring. The most interesting results were observed when the analyses were carried out
with the tetracationic derivatives H -TDMImP.I and H -4TMPyP.I . The first
2
4
2
4
2
3

correlation to be assigned is the difference between cationic and its neutral counterpart
for each derivative. In the case of H -4TMPyP.I , a slight shift of +2 nm (entry 18) was
2
4
observed which is completely different when comparing with the neutral parental
porphyrin profile (entry 8). Also the spectrum of compound H -4TMPyP.I is quite
2
4
different from that of the same compound in neutral system (aqueous media; Figure 7).
In the case of H -TDMImP.I , no changes were observed in the electronic spectrum in
2
4
acidic media (Figure 7). Such feature indicates that the inner hydrogens resist to
protonation even with concentrated HCl, thus suggesting that imidazolium groups are
strong electron-withdrawing substituents. The introduction of another methyl group to
the imidazolyl substituent enhances the electron-withdrawing effect as well as the
bulkiness, thus turning difficult the conjugation with the macrocycle. These
observations are intimately related with orbital stabilization (Figure 8). Regarding the
2 2
structural aspects, there is a similarity between H -2TPyP and H -TMImP. These
similarities will be also observed in their cationic forms and in the metallocomplexes. In
this way, some theoretical hypothesis established for the first derivative can be applied,
at a given extension, to the latter compound. Jin and Groves[54] studied the reactivity
4
+
4+
profile between the regioisomers [H
2 2
-4TMPyP] and [H -2TMPyP] and attributed the
4
+
differences to the orbital stabilization. The cationic derivative [H
2
-2TMPyP] resists to
4
+
protonation even at pH=0,[54] such as observed for the [H -TDImP] . This uncommon
2
feature decreases the donation character of A u molecular orbital, which undergoes the
2
influences of meso-substituent groups (as can be seen in Figure 8) and increases the gap
2 g
energy between A u-e molecular orbitals, justifying the reported behavior. In the
metallocomplex form (e.g., manganese complex), the same behavior is expected to take
place.
2
4

4+
[
H -4TMPyP]
517
2
4
+
[
H -4TMPyP] + HCl
2
5
84
554
642
40
H -TDMImP]
6
5
91
4+
506
[
2
4
+
[
H -TDMImP] + HCl
2
539
577
631
450
500
550
600
650
700
λ
, nm
4
+
Figure 7. Electronic spectra of tetracationic derivatives [H -4TMPyP]
and
2
4+
[H2-TDMImP] with and without concentrated HCl.
2
5

Table 6. Electronic parameters of the Q-bands in metalloporphyrins. Intensities and wavelengths recorded in neutral and in acidic media.
∆
λ, nm
Entry
Derivative
-TPP
Solvent
CHCl
+ HCl conc.
CHCl
+ HCl conc.
CHCl
+ HCl conc.
/CH OH (85:15)
OH (85:15) + HCl conc.
CHCl /CH OH (85:15)
/CH OH (85:15) + HCl conc.
CHCl
+ HCl conc.
CHCl
+ HCl conc.
Qy (0;1)
Qy (0;0)
Qx (0;1)
Qx (0;0)
Qy
Qx
1
2
3
4
5
6
7
8
9
H
2
3
514
549
608
588
644
660
657
637
656
641
642
666
647
664
657
663
656
677
631
630
640
642
2
+
[H
4
-TPP]
CHCl
CHCl
CHCl
3
+59
+49
+55
+31
+70
+22
+69
0
+16
H
2
-TDCPP
3
512
505
511
510
512
519
506
539(sh.)
588
587
582
585
586
585
593
577
2
+
[H
4
-TDCPP]
3
-20
-15
+24
+17
+ 6
+21
-1
H
2
-TPFPP
3
534(sh.)
589
2
+
[H
4
-TPFPP]
3
(
a)
H
2
-4TPyP
CHCl
3
3
545
2
+
[H
4
-4TPyP]
CHCl
CHCl
3
/CH
3
616
(
a)
H
2
-2TPyP
3
3
542
2
+
10
11
12
13
14
15
16
17
18
[H
4
-2TPyP]
3
3
612
(
a)
+
H
2
-TMImP
3
546(sh.)
607
2
[H
4
-TMImP]
CHCl
3
(
a)
H
2
-TTz2PP
3
555
2
+
[H
4
-TTz2PP]
CHCl
3
624(sh.)
539
H
2
-TDMImP.I
4
2
H O
2
+ (b)
(c)
(c)
[H
4
-TMImP.I
4
]
H
2
O + HCl conc.
~ 506
539
~ 577
H
2
-4TMPyP.I
4
H
2
O
517
554
584
2
+ (b)
[H
2
-4TMPyP.I
4
]
H
2
O + HCl conc.
591
+37
+2
(
a)
(b)
(c)
protonation of substituents was disregarded; counter-ion charge was considered; no change was recorded.
26

4
+
2
Those observations made for [H -TDMImP] play an important role to explain the high
catalytic efficiency observed in oxidation reactions using the immobilized systems. A
different efficiency for the system Mn(Porph)-SG should be expected; a probable
explanation must be entirely structural, since the linear propyl moiety that links the porphyrin
to the support (silica gel) does not confer protection against oxidative degradation, while the
tolyl group linking the porphyrin to the MR support is bulkier and protects better the
macrocycle from oxidative attack.
NH
N
N
NH
N
N
H -Porph
(Porph)Mn(V)=O
2
NH
NH
eg
)
.
u
.
a(
c2
c1
LUMO+1
LUMO
y
g
r
(Eg)
d π-π *
d π-π * (E )
xz
x
yz
y
g
e
n
E
HOMO
b2 (A1u
)
A1u HOMO
NH
N
N
NH
N
N
HOMO-1
b1 (A2u)
A2u HOMO-1
dxy
HOMO-2
NH
NH
A1u
A2u
Figure 8. Molecular orbital distribution according to Martin Gouterman theory for H -Porph
2
(left)[53,54] and orbital distribution according to DeAngelis and Groves[55] proposal for
elusive species (2-TMPyP)Mn(V)=O (right). The diagram does not account the energy
differences amongst orbitals.
3. CONCLUSIONS
This work presents a new heterogeneous catalytic system based on an imidazolyl manganese
porphyrin, able to selectively oxidize olefins exclusively or mainly to their corresponding
epoxides. The catalytic efficiency seems to be influenced by the synthetic conditions used
during the immobilization reactions. The silica support did not yield a totally satisfactory
heterogeneous oxidation system, even if t-BuOOH in decane is used as the oxygen donor,
27

thus avoiding the water present in aqueous hydrogen peroxide. The conditions studied allow
high reaction performance using an environmentally benign oxidant (H ), the catalyst being
2
O
2
recycled up to four times without appreciable loss in efficiency. The efficiency observed can
be explained by molecular orbital stabilization achieved by the linkage between the
metallocomplex and the support. The lower catalytic efficiency observed for Mn(Porph)-SG
in comparison with Mn(Porph)-MR may be attributed to structural differences, owing to the
support linker.
4. EXPERIMENTAL
All solvents and reagents were used as received without further purification, except for
pyrrole, which was distilled before use. Pyrrole, cis-cyclooctene, styrene, geraniol,
cyclohexene, n-octane, tert-butylhydroperoxide (t-BuOOH), 3-bromopropyl functionalized
silica gel (200-400 mesh, 1.5 mmol/g), Merrifield resin (30-40 mesh, 2.0 mmol/g, 1% cross-
linked) and dimethylformamide were purchased from Aldrich. Propionic acid and acetonitrile
were obtained from Merck whereas aqueous hydrogen peroxide (30% w/w) was obtained
from Riedel-de-Häen. Acetic acid was purchased from PanReac. The Membrane filter N16
(0.2 µm; φ = 50 mm) was used to filter the crude reaction in order to separate the
heterogenized materials during catalyst synthesis.
GC-FID analyses were performed using a Varian 3900 chromatograph, whereas GC–MS
analyses were performed using a Finnigan Trace GC-MS (Thermo Quest CE instruments)
both using helium as the carrier gas (35 cm/s) and equipped with a fused silica capillary 5%
phenyl 95% dimethylpolysiloxane column (30 m x 0.25 mm i.d.; 25 µm film thickness).
1
The H NMR spectra were recorded in a Bruker Avance 300 at 300.13 MHz. CDCl was used
3
as solvent and TMS as the internal reference; chemical shifts are expressed in δ (ppm). The
mass spectrometry analyses were carried out in a 4800 MALDI TOF/TOF Analyzer, Applied
28

Biosystems, with and without matrix. The UV/Visible measurements were recorded in a
double beam spectrophotometer Shimadzu UV-2501 PC. C, H, N elemental analyses were
performed on a Leco CHNS-932 apparatus. Infrared spectra were obtained on a Mattson 7000
spectrophotometer using KBr pellets. Diffuse reflectance spectra were registered on a Jasco
V-560 spectrophotometer, using MgO as reference. Scanning electron microscope images
(SEM) were recorded on a Hitachi S-4100 microscope.
4.1. Porphyrin and manganese(III) complex syntheses
The free-base porphyrin, 5,10,15,20-tetrakis(1-methylimidazol-2-yl)-21H,23H-porphyrin (H2-
TMImP) was obtained as a mixture of atropoisomers from pyrrole and 1-methylimidazole-2-
carboxaldehyde in propionic acid under microwave (MW) irradiation following the method
already described by us.[56] The synthesis of the manganese complex was carried out
according to procedures described elsewhere.[2,24,52]
1
Structural characterization was made by H NMR spectroscopy and by mass spectrometry
1
(
MALDI TOF/TOF) for the free-base porphyrin (H -TMImP). The H NMR results obtained
2
2
for the atropoisomers of H -TMImP are similar to those already known.[12,24,57]
4.2. Manganese porphyrin immobilization
To immobilize the manganese porphyrin onto the solid supports, a method reported by
Tangestaninejad et al. was adopted.[36,43] Depending on the support used, some adjustments
were done. Scheme 1 shows the procedures adopted in each case and the experimental
differences among the batches produced. In a conic flask, and to a DMF solution of the
manganese porphyrin, the Merrifield resin (MR) or the 3-bromopropyl-functionalized silica
gel (SG) were added. The mixture was vigorously stirred at a given temperature for a given
time (see Scheme 1 for details). After cooling, the material was collected by filtration and
29