Synthesis, characterization, and catalytic activity of anionic iron(III) porphyrins intercalated into layered double hydroxides

Article abstract of DOI:10.1016/j.jcat.2008.04.026

The first generation free-base anionic porphyrin [H₂(TSPP)]^4-, its iron(III) porphyrin [Fe(TSPP)]^3-, and the second generation anionic complexes [Fe(TDFSPP)]^3- and [Fe(TDCSPP)]^3- were intercalated into the layered double hydroxide Zn_nAl-LDH (n = 2, 4 or 5) by coprecipitation at constant pH. The materials were characterized by X-ray powder diffraction, UV/visible spectroscopy in glycerin mull, attenuated total reflectance Fourier transform infrared spectroscopy, and electron paramagnetic resonance. Results revealed that the coprecipitation method led to intercalation of the free-base porphyrin and the iron(III) porphyrins between the Zn_nAl-LDH layers. The materials were used as catalysts in the oxidation of cyclooctene, cyclohexene, and cyclohexane by iodosylbenzene. The catalytic activity of [Fe(TDCSPP)]-Zn₂Al-LDH was higher than that of the homogeneous [Fe(TDCSPP)]^3-, but the opposite effect was observed in the case of [Fe(TDFSPP)]-Zn₂Al-LDH. Although [Fe(TDFSPP)]^3- and [Fe(TDCSPP)]^3- are structurally similar, their intercalation into Zn₂Al-LDH likely results in different chemical environments, leading to distinct catalytic activities.

Full text of DOI:10.1016/j.jcat.2008.04.026

Journal of Catalysis 257 (2008) 233–243
Contents lists available at ScienceDirect
Journal of Catalysis
www.elsevier.com/locate/jcat
Synthesis, characterization, and catalytic activity of anionic iron(III) porphyrins
intercalated into layered double hydroxides
Matilte Halma a, Kelly Aparecida Dias de Freitas Castro a, Christine Taviot-Gueho c, Vanessa Prévot c, Claude Forano c,
Fernando Wypych b, Shirley Nakagaki a,∗
a
Laboratory of Bioinorganic Chemistry and Catalysis, Department of Chemistry, Federal University of Paraná, UFPR, P.O. Box 19081, Curitiba, Paraná 81531-990, Brazil
Solid State Chemistry Laboratory, LQES, Department of Chemistry, Federal University of Paraná, UFPR, P.O. Box 19081, Curitiba, Paraná 81531-990, Brazil
Laboratoire des Matériaux Inorganiques, UMR CNRS 6002, Université Blaise Pascal, 63177 Aubière Cedex, France
b
c
a r t i c l e i n f o
a b s t r a c t
The first generation free-base anionic porphyrin [H2(TSPP)]4−, its iron(III) porphyrin [Fe(TSPP)]3−, and
the second generation anionic complexes [Fe(TDFSPP)]3− and [Fe(TDCSPP)]3− were intercalated into the
layered double hydroxide ZnnAl-LDH (n = 2, 4 or 5) by coprecipitation at constant pH. The materials
were characterized by X-ray powder diffraction, UV/visible spectroscopy in glycerin mull, attenuated
total reflectance Fourier transform infrared spectroscopy, and electron paramagnetic resonance. Results
revealed that the coprecipitation method led to intercalation of the free-base porphyrin and the iron(III)
porphyrins between the ZnnAl-LDH layers. The materials were used as catalysts in the oxidation of
cyclooctene, cyclohexene, and cyclohexane by iodosylbenzene. The catalytic activity of [Fe(TDCSPP)]-
Zn2Al-LDH was higher than that of the homogeneous [Fe(TDCSPP)]3−, but the opposite effect was
observed in the case of [Fe(TDFSPP)]-Zn2Al-LDH. Although [Fe(TDFSPP)]3− and [Fe(TDCSPP)]3− are
structurally similar, their intercalation into Zn2Al-LDH likely results in different chemical environments,
leading to distinct catalytic activities.
Article history:
Received 7 March 2008
Revised 29 April 2008
Accepted 30 April 2008
Available online 13 June 2008
Keywords:
Intercalation
Porphyrins
Layered double hydroxides
Catalysis
Oxidation
© 2008 Elsevier Inc. All rights reserved.
1. Introduction
strate and the active catalytic species, which would not occur in
homogeneous medium [1,7–9,22,23,27]. Immobilization prevents
metalloporphyrin aggregation and/or bimolecular self-destruction,
thus avoiding deactivation of the metalloporphyrin catalytically ac-
tive species. In addition, metalloporphyrin immobilization facili-
tates catalyst recovery and reuse [18].
Bioinspired oxidation catalysts based on metalloporphyrins have
been extensively studied over the past 3 decades. These com-
pounds are well known for their ability to catalyze the oxida-
tion of various substrates under mild conditions, including inert
molecules such as alkanes. For this reason, they are considered
mimics of the cytochrome P450-dependent monooxygenases [1–5].
Metalloporphyrins can be used in homogeneous solution or immo-
bilized on solid matrices, such as organic amorphous polymers and
crystalline inorganic materials. Examples of the latter supports in-
clude silica [6–9], fibrous chrysotile [2], zeolite [10–12], and clays
from the smectite group (montmorillonite) [13,14] (among others
[15–25]). Immobilization of metalloporphyrins carrying electron-
withdrawing substituents, the so-called “second generation” met-
alloporphyrins [26], on inorganic supports leads to efficient and
selective catalysts for hydrocarbon oxidation. The inorganic sup-
port confers shape-selectivity to the metalloporphyrin, thus pro-
moting a special environment for the reaction between the sub-
Layered double hydroxides (LDHs) have high surface area and
good anion-exchange and expansion properties, which make them
promising matrices for the immobilization of anionic metallo-
porphyrins. Metalloporphyrins can either intercalate between the
layers or bind to the surface of the support [17,19]. In this
work, the second-generation iron(III) porphyrins [Fe(TDFSPP)]3−
and [Fe(TDCSPP)]3− (Fig. 1) were intercalated in the interlayer
space of ZnnAl-LDH by coprecipitation at constant pH. The cat-
alytic activity of [Fe(TDFSPP)]-ZnnAl-LDH and [Fe(TDCSPP)]-ZnnAl-
LDH in the oxidation of cyclooctene, cyclohexane, and cyclohex-
ene by iodosylbenzene also was investigated, and the results
were compared with those obtained with other iron(III) porphyrin
systems described in the literature. [H2(TSPP)]-ZnnAl-LDH and
[Fe(TSPP)]-ZnnAl-LDH were also prepared for comparison purposes.
[H2(TSPP)]-ZnnAl-LDH and [Fe(TSPP)]-ZnnAl-LDH were subjected to
hydrothermal treatment, which led to improved crystallinity and
allowed for deeper structural characterization.
Corresponding author. Fax: +55 41 3361 3186.
*
E-mail address: shirley@quimica.ufpr.br (S. Nakagaki).
0021-9517/$ – see front matter © 2008 Elsevier Inc. All rights reserved.
doi:10.1016/j.jcat.2008.04.026

234
M. Halma et al. / Journal of Catalysis 257 (2008) 233–243
Fig. 1. Schematic representation of iron(III) porphyrins.
2. Experimental
Table 1
Conditions of the porphyrins/LDH coprecipitation at constant pH
[M2+ + M3+
and [NaOH]
(mol/L)
]
[Porphyrin] or
[iron(III) porphyrin]
(mol/L)
V of salts Time
2.1. Materials
Solid
(mL)
(min)
All solvents and reagents were commercial grade (Aldrich,
Merck, Fluka), unless stated otherwise. Authentic samples of the
alcohols, ketones, and epoxides that could be produced in the oxi-
dation reactions were purchased at their highest commercial purity
grade (Aldrich) and used as received. The substrates cyclooctene,
[H2(TSPP)]-Zn2Al-LDH
[H2(TSPP)]-Zn4Al-LDH
[H2(TSPP)]-Zn5Al-LDH
[Fe(TSPP)]-Zn2Al-LDH
[Fe(TSPP)]-Zn4Al-LDH
0.100
0.125
0.125
0.100
0.125
0.012
0.014
0.012
0.017
0.013
0.004
0.003
2.57
2.54
3.32
2.48
2.47
0.70
0.62
371
369
480
360
328
310
292
◦
cyclohexene, and cyclohexane were stored at 5 C and purged with
[Fe(TDFSPP)]-Zn2Al-LDH 0.100
[Fe(TDCSPP)]-Zn2Al-LDH 0.100
argon before use. After the experiments, all reagents were dis-
carded in an appropriate container for later treatment and reuse
when possible, or for final disposal.
by coprecipitation at constant pH, adapted for the preparation of
small quantities of material. Toward this end, an aqueous solution
containing AlCl3 and ZnCl2 and another aqueous solution contain-
ing NaOH were simultaneously added to a container (see Table 1
for details on the various AlCl3, ZnCl2 and NaOH concentrations
used in the procedure). The volume of added NaOH was moni-
tored via a pH electrode immersed in the reagent solution, so as
to maintain the pH of the coprecipitation reaction at 7.5 0.2. The
average flow rate of the solution containing AlCl3 and ZnCl2 was
obtained by dividing the transferred solution volume by the time.
The mixture containing NaOH, AlCl3, and ZnCl2 was then slowly
added to the free-base porphyrin or to a FeP solution (see Table 1
for details on concentration) under nitrogen atmosphere. The con-
centration of the free-base porphyrin or the FeP in the solution
corresponded to fourfold the number of mol required for the ex-
change reaction stoichiometry. Addition of the NaOH, AlCl3, and
ZnCl2 solution to the free-base porphyrin or FeP was completed
within 6 h, and the mixture was left to age at room temperature
for 2 days.
After the intercalation reaction, [H2(TSPP)]-ZnnAl-LDH, [Fe-
(TSPP)]-ZnnAl-LDH, [Fe(TDFSPP)]-Zn2Al-LDH, and [Fe(TDCSPP)]-
Zn2Al-LDH were washed with deionized water and recovered
by centrifugation. The combined washing solutions were stored
and analyzed by UV–vis spectroscopy, to quantify any free-base
porphyrin or FeP that might have leached from the matrix dur-
ing the dispersion-centrifugation process. The obtained green
solids were air-dried, and the following [H2(TSPP)]/solid and
2.1.1. Porphyrins
Reagent-grade free-base porphyrin H2(TSPP) [5,10,15,20-tetra-
kis(4-sulfonatophenyl)porphyrin] was purchased from Aldrich and
used without purification. UV/vis data H2(TSPP) (deionized wa-
−1
ter): 421 nm (ε = 2.7 × 105 L mol cm−1). The second-generation
free-base porphyrins H2(TDFSPP); [5,10,15,20-tetrakis(2,6-difluoro-
3-sulfonatophenyl)porphyrin] and H2(TDCSPP); [5,10,15,20-tetrakis-
(2,6-dichoro-3-sulfonatophenyl)porphyrin]; were synthesized and
purified as described previously. The anionic tetra charges were
omitted for all free-base porphyrins for simplification (Fig. 1)
[25,28,29].
2.1.2. Iron(III) porphyrins
Iron(III) porphyrin (FePs) were obtained by iron insertion into
the free-base porphyrin ligands using ferrous chloride tetrahydrate
in dimethylformamide (DMF), as described by Adler and Longo
[30,31]. Purification of the FePs was performed by column chro-
matography on Sephadex, using deionized water as an eluent. The
products were characterized by UV–vis and EPR spectroscopy. Data
were consistent with those of the compound expected after the
iron insertion reaction. UV/vis data: [Fe(TSPP)Cl] (deionized water)
−1
394 nm (ε = 2.4 × 104 L mol cm−1), [Fe(TDFSPP)Cl] (deionized
−1
water) 394 nm (ε = 3.7 × 104 L mol cm−1), and [Fe(TDCSPP)Cl]
−1
(deionized water) 390 nm (ε = 1.7 × 104 L mol cm−1).
2.1.3. Iodosylbenzene (PhIO)
PhIO was obtained through hydrolysis of iodosylbenzene diac-
etate, as described previously [32,33]. Purity was checked periodi-
cally by iodometric assay.
FeP/solid loadings were obtained: [H2(TSPP)]-Zn2Al-LDH = 8.8 ×
−4
10
mol free-base porphyrin/g solid; [H2(TSPP)]-Zn4Al-LDH = 1.1
−3
× 10
mol free-base porphyrin/g solid; [H2(TSPP)]-Zn5Al-LDH =
−4
7.7 × 10
mol free-base porphyrin/g solid; [Fe(TSPP)]-Zn2Al-LDH
2.2. Intercalation of [H2(TSPP)], [Fe(TSPP)], [Fe(TDFSPP)], and
[Fe(TDCSPP)] into ZnnAl-LDH
−4
= 8.9 × 10
mol iron(III) porphyrin/g solid; [Fe(TSPP)]-Zn4Al-
−4
LDH = 7.9 × 10
mol iron(III) porphyrin/g solid; [Fe(TDFSPP)]-
−3
Zn2Al-LDH = 1.1 × 10
mol iron(III) porphyrin/g solid, and [Fe-
The solids [H2(TSPP)]-ZnnAl-LDH, [Fe(TSPP)]-ZnnAl-LDH, [Fe-
(TDFSPP)]-ZnnAl-LDH, and [Fe(TDCSPP)]-ZnnAl-LDH were prepared
−3
(TDCSPP)]-Zn2Al-LDH = 1.2 × 10
mol iron(III) porphyrin/g solid.

M. Halma et al. / Journal of Catalysis 257 (2008) 233–243
235
To improve the crystallinity of the materials prepared as above
and to gain further insight into their structural arrangement, the
[H2(TSPP)]-Zn2Al-LDH, [H2(TSPP)]-Zn4Al-LDH, [Fe(TSPP)]-Zn2Al-
LDH, and [Fe(TSPP)]-Zn4Al-LDH samples were submitted to hydro-
thermal treatment. The resulting products were denoted [H2-
(TSPP)]-Zn2Al-LDHHyd, [H2(TSPP)]-Zn4Al-LDHHyd, [Fe(TSPP)]-Zn2Al-
LDHHyd, and [Fe(TSPP)]-Zn4Al-LDHHyd, respectively. Toward this
end, 50 mg of the coprecipitated phases were suspended in 25
mL of deionized water in a Teflon inner vessel (30 mL) placed in
a stainless steel outer autoclave, and the solids were kept there at
◦
120 C for 72 h under autogeneous pressure.
2.3. Oxidation of cyclooctene, cyclohexene and cyclohexane by PhIO
catalyzed by anionic FePs in homogeneous medium or intercalated into
ZnnAl-LDH
The oxidation reactions were carried out in a thermostatic
glass reactor (2 mL) equipped with a magnetic stirrer, placed
inside a dark chamber [15,18]. In the case of the ZnnAl-LDH-
intercalated FePs (heterogeneous catalysis), the solid catalyst and
iodosylbenzene (in an FeP/PhIO molar ratio 1:10) were suspended
in dichlorometane/acetonitrile 1:1 v/v mixture (0.350 mL) and de-
gassed with argon for 15 min. The substrate (cyclohexane, cy-
clooctene, or cyclohexene) was then added to the glass reactor
at an FeP/substrate molar ratio of 1:1000, and the oxidation re-
action was performed for 1 h under magnetic stirring. Sodium
sulfite was added to eliminate any excess iodosylbenzene and
to quench the reaction. The reaction products were separated
from the solid catalyst (heterogeneous catalysis) by exhaustive
washing and centrifugation of the FeP-ZnnAl-LDH catalyst with a
dichlorometane/acetonitrile 1:1 v/v mixture. The combined wash-
ing solutions were analyzed by capillary gas chromatography. The
products were identified by comparing their retention times with
those of authentic samples. Product yields were determined by
the internal standard method. Control reactions were carried out
with the chlorine-intercalated ZnnAl-LDH containing no FeP and
with a solution blank setup with no solid. A similar procedure was
adopted when iron(III) porphyrins were used as homogeneous cat-
alyst.
Fig. 2. X-ray powder diffraction patterns of [Zn2Al-Cl] LDH (a), [H2(TSPP)]-Zn4Al-
LDH before (b) and after the hydrothermal treatment (c) and [H2(TSPP)]-Zn2Al-LDH
before (d) and after the hydrothermal treatment (e). Ticks and labels give the Bragg
reflections in the space group R-3m and peaks noted with an asterisk correspond to
ZnO.
2.4. Instruments
X-ray powder diffraction patterns (XRPD) were recorded on a
Philips X-Pert Pro diffractometer equipped with an X-Celerator de-
tector using CuKα1/α2 radiation. Data were collected between
◦
◦
Fig. 3. X-ray powder diffraction patterns in the 2θ range 2 –70 of [Fe(TSPP)]-Zn4Al-
LDH before (a) and after the hydrothermal treatment (b) and [Fe(TSPP)]-Zn2Al-LDH
(c) and after hydrothermal treatment (d). Ticks give the Bragg reflections in the
space group R-3m and peaks noted with an asterisk correspond to ZnO.
◦
◦
◦
2
and 70 (2θ, with a step size of 0.03 and a counting time
of 100 s/step). The cell parameters were determined from peak
profile analysis using the Fullprof program (Full Pattern Match-
ing) [34]. Before the analyses, the samples were mixed with 10 wt%
of a polycrystalline silicon standard, to obtain accurate unit cell
parameters for the ZnnAl-LDH-containing free-base porphyrin or
FeP. The Lorentzian component of the TCH pseudo-Voigt profile
function was modeled using a linear combination of spherical har-
monics to take into account the anisotropic broadening due to
anisotropic size effects [34,35]. A Rietveld refinement was car-
ried out on the [Zn2Al-Cl] reference sample (LDH containing no
intercalated complexes) starting from the model given in the liter-
ature [36].
−1
NICOLET 5700 equipment in the 4000–400 cm range, with a res-
−1
olution of 8 cm
and accumulation of 120 scans. Products from
the catalytic oxidation reactions were identified using a Shimadzu
CG-14B gas chromatograph equipped with a flame ionization de-
tector and a DB-WAX capillary column (J & W Scientific).
3. Results and discussion
3.1. Characterization of the ZnnAl-LDH-intercalated FeP catalysts
UV–visible spectra (UV/vis) were obtained with a NICOLET evo-
lution 500 Diode Array spectrophotometer, using solid samples
dispersed in glycerol mulls placed between two quartz plates.
Electron paramagnetic resonance (EPR) of the supported metallo-
porphyrin systems in the solid state was performed on a Bruker
ESP 300E spectrometer operating at the X-band (approximately
9.5 GHz), at 77 K, using liquid N2. Attenuated total reflectance
Fourier transform infrared (ATR/FTIR) spectra were recorded on a
3.1.1. X-ray powder diffraction
All of the materials prepared in this work displayed XRPD pat-
terns typical of ZnnAl-LDH materials (Figs. 2–4), yet the diffraction
patterns were broad compared with those for the [Zn2Al-Cl] pris-
tine material containing no free-base porphyrin or FeP. This is
likely due to the simultaneous effects of small coherent domain
size and structural disorder following porphyrin intercalation. The

236
M. Halma et al. / Journal of Catalysis 257 (2008) 233–243
Fig. 4. Results of the profile analysis of XRPD patterns for [Fe(TDCSPP)]-Zn2Al-LDH (a) and [Fe(TDFSPP)]-Zn2Al-LDH (b): experimental X-ray diffraction (dots) and calculated
(line).
Table 2
Refined cell parameters and microstructural data for porphyrin-LDH materials
ꢀ
LDH phases
Cell parameters (Å)
In-plane (110) and
out–of-plane (001)
coherence lengths (Å)
R
calc/Rtheo
So (Å2/e) = 3a2/2x
Zn2Al-Cl
a = 3.07526(2)
c = 23.2459(8); d003 = 7.749
a = 3.077(2)
1950/460
115/75
[H2(TSPP)]-Zn2Al-LDH
[H2(TSPP)]-Zn4Al-LDH
[Fe(TSPP)]-Zn2Al-LDH
[Fe(TSPP)]-Zn4Al-LDH
[H2(TSPP)]-Zn2Al-LDHHyd
[H2(TSPP)]-Zn4Al-LDHHyd
[Fe(TSPP)]-Zn2Al-LDHHyd
[Fe(TSPP)]-Zn4Al-LDHHyd
[Fe(TDCSPP)]-Zn2Al-LDH
[Fe(TDFSPP)]-Zn2Al-LDH
c = 66.9(3); d003 = 22.3
a = 3.082(1)
2.07/2.0
2.29/4.0
1.82/2.0
1.98/4.0
1.53/2
25.15
27.06
23.07
24.37
20.55
20.50
20.78
20.78
c = 64.5(1); d003 = 21.5
a = 3.071(3)
–
–
c = 68.5(2); d003 = 22.8
a = 3.075(2)
–
c = 69.3(2); d003 = 23.1
a = 3.0620(2)
680/505
c = 68.98(4); d003 = 22.99
a = 3.0618(7)
–
–
–
–
–
c = 68.99(1); d003 = 22.99
a = 3.063(1)
1.52/4
c = 68.86(6); d003 = 22.95
a = 3.063(1)
1.56/2
c = 69.23(5); d003 = 23.07
a = 3.073(2)
1.56/4
c = 64.9(1); d003 = 21.6
a = 3.069(2)
c = 60.49(5); d003 = 20.16
lower signal-to-noise ratio seen in the Fe(TSPP)-containing samples
may be attributed in part to iron fluorescence. Diffraction peaks
are indexed as in the case of [Zn2Al-Cl], on a hexagonal unit cell
with the space group R-3m. The cell parameters gathered in Ta-
ble 2 were obtained from profile peak analysis.
The intercalation of porphyrin complexes into the LDH led
to strong enlargement of the basal distance d003, as seen from
the displacement of the 003 reflection toward low 2θ angles.
The d003 values of 20–23 Å suggest a perpendicular arrangement
of the porphyrin ring with respect to the hydroxide layer (dis-
cussed below). As evidenced from the position of the reflection
markers at the bottom of the XRPD patterns, the [H2(TSPP)]-Zn2Al-
LDH, [H2(TSPP)]-Zn4Al-LDH, [Fe(TSPP)]-Zn2Al-LDH, and [Fe(TSPP)]-
Zn4Al-LDH samples are single-phase materials (Figs. 2 and 3),
whereas [Fe(TDFSPP)]-Zn2Al-LDH and [Fe(TDCSPP)]-Zn2Al-LDH are
not pure samples (Fig. 4). The impurities in the latter con-
sist of Zn2Al-LDH phases intercalated with inorganic anions. For
[Fe(TDCSPP)]-Zn2Al-LDH, the basal distance d003 ∼ 7.7 Å can be
ascribed to either chloride (competition with chloride metallic
salts) or carbonate anions (contamination from atmospheric CO2).
As for the [Fe(TDFSPP)]-Zn2Al-LDH sample, the additional diffrac-
tion peaks may be attributed to the [ZnRAl-SO4] phase with
d003 ∼ 8.6 Å; however, the source of sulfate anions is question-
able and could lie on the porphyrin sulfonation process. Using a
correlation between the cell parameter a and the Zn/Al molar ratio
R established elsewhere for the [ZnRAl-Cl] series, an R value close
to 2 was obtained in all cases, even for samples prepared with a
Zn/Al ratio of 4. Therefore, it can be inferred that the intercalated
porphyrins are partly responsible for the Zn/Al ratio, that is, the
charge density of the hydroxide layers.
To improve the crystallinity of the samples and gain bet-
ter insight into the porphyrin-ZnnAl-LDH structures, the single

M. Halma et al. / Journal of Catalysis 257 (2008) 233–243
237
Fig. 5. Results of the analysis of the XRPD profile of Rietveld refinement [Zn2Al-Cl] (a), [H2(TSPP)]-Zn2Al-LDH before (b) and after hydrothermal treatment (c). Experimental
X-ray diffraction (dots) and calculated (line). The insets show the average apparent shape of the crystallite coherent domains.
phases [H2(TSPP)]-Zn2Al-LDH, [H2(TSPP)]-Zn4Al-LDH, [Fe(TSPP)]-
Zn2Al-LDH, and [Fe(TSPP)]-Zn4Al-LDH were subjected to a post-
synthesis hydrothermal treatment. The narrowing of the diffrac-
tion peaks suggests a net improvement in the crystallinity of the
[H2(TPPS)]-containing materials (Figs. 2 and 3). The results of the
full-pattern matching refinements of [H2(TSPP)]-Zn2Al-LDH and
[H2(TSPP)]-Zn2Al-LDHHyd and the Rietveld refinement of [Zn2Al-Cl]
are given in Fig. 5.
Good profile refinements were obtained by assuming that the
anisotropic broadening is due to size effects, which allowed deter-
mination of both the size and the average form of the coherent
domains. It was possible to reconstruct the apparent sizes along
different directions from the refined profile coefficients. Hexago-
nal platelet coherent domains were obtained, in total agreement
with the hexagonal platelet-like morphology of ZnnAl-LDH mate-
rials. The size of the coherent domains of [H2(TSPP)]-Zn2Al-LDH

238
M. Halma et al. / Journal of Catalysis 257 (2008) 233–243
The sulfonate groups led to maxima at the outer parts of the
interlayer space, that is, d = 4.0 and 19.0 Å. This position is co-
herent with the existence of hydrogen-bonding interactions be-
tween these sulfonate groups and the hydroxylated layers. The
central maximum is attributed to the four inner nitrogen atoms
of the porphyrin ring, whereas the intermediate peaks of lower
intensity spaced approximately 7.6 Å from each other arise from
the opposite phenyl groups. With such an orientation, the por-
phyrin molecule displayed a surface area of ∼17.3 Å2 per unit
charge, largely satisfied by the host lattice for
a Zn/Al ratio
−
of 2 (24.8 Å2/e ), and in far better accordance with R = 1.5
−
(20.3 Å2/e ). In the latter case, a slightly inclined orientation of
the porphyrin, achieved by increasing the surface required by the
porphyrin, should lead to exact matching between the Zn2Al-LDH
host and the porphyrin guest, thus causing an optimization of the
electrostatic interactions. Yet the 1-D plot does not allow us to
distinguish between a perpendicular or a slightly oriented arrange-
ment.
A similar perpendicular orientation is proposed for all the
other materials, including those containing FeP. The slight varia-
tions observed in the d003 values must be first ascribed to the
different crystallinity of these materials and/or different inter-
layer water contents. Nevertheless, in [Fe(TDFSPP)]-Zn2Al-LDH and
[Fe(TDCSPP)]-Zn2Al-LDH, the substitution of the two ortho hydro-
gen atoms on the mesophenyl group by fluorine or chlorine cer-
tainly had an effect on gallery height. In the particular case of the
chlorine-substituted derivative [Fe(TDCSPP)]-Zn2Al-LDH, the d003
of 21.6 Å suggests that intramolecular steric hindrance forced the
rotation of both the phenyl and the sulfonate groups, leading to
rearrangement of the porphyrin anions to optimize the guest–host
interactions.
Fig. 6. One-dimensional electron density ρ(z) projection along the c-axis for [H2-
(TSPP)]-Zn2Al-LDHhyd and the corresponding structural model.
(in-plane dimension of 115 Å and thickness of 75 Å) was found
to be 6-fold larger after the solid underwent hydrothermal treat-
ment (in-plane dimension of 680 Å and thickness of 450 Å), and
the same effect on particle size can be assumed. For [Zn2Al-Cl],
the size of the coherent domains (195/45 nm) is associated with
an average particle size of 500 nm (i.e., in the submicrometric
range). The broad features displayed by the materials containing
FeP suggest the presence of very small particles, certainly in the
nanometer range (<100 nm). On the other hand, the Zn/Al ratio
deduced from the value of the cell parameter a was found to de-
crease and converge to 1.5 on hydrothermal treatment for all of the
samples. The simultaneous formation of ZnO was observed in the
XRPD pattern, and more of this oxide was present in the Zn4Al-
samples than in the Zn2Al-samples. Although further characteri-
zation is needed to clarify this phenomenon, this result strongly
suggests that the hydrothermal process proceeds via a dissolution,
precipitation, and crystallization mechanism. Under autogeneous
3.1.2. UV–visible absorption spectra
The presence of FeP in the ZnnAl-LDH matrix also was con-
firmed by UV–visible spectra of the solids in glycerin mull (Fig. 7).
Two sets of absorption bands are seen in the UV–visible spectra
of porphyrins. The Soret band is characterized by a large absorp-
tion coefficient and lies in the 400–450 nm range. A second set of
weaker bands, the Q bands, occurs between 450 and 700 nm [17].
Our measurements suggested that no demetallation (characterized
by a blue-shift of the Soret band associated with a significant
amount of free-base porphyrin [22,24]) or significant exchange of
the Fe(III) ion with the support occurring during the catalyst in-
tercalation carried out in this work [2,18]. The Soret peaks of the
intercalated phases [Figs. 7b (418 nm), 7c, 7e, and 7g (414 nm)]
were slightly shifted to higher wavelengths compared with those
of the pure FePs [Figs. 7a (408 nm), 7d (410 nm), and 7f (414 nm)].
This behavior can be attributed to steric constraints caused by the
support, which substantially distort the FeP molecule in these sup-
ported catalysts [31]. It also could be an indication that most of
the FePs was intercalated between the ZnnAl-LDH layers [18]. Peak
broadening also was observed in the case of the intercalated FePs
compared with the parent complexes in solution, which also can
be explained by the presence of the different intermolecular inter-
actions.
◦
pressure (∼2 bars), dissociation of water at 120 C is increased,
and the pH of the solution is slightly acidic. Based on the Pourbaix
diagrams of these species, only ZnRAl-LDH with low Zn/Al molar
ratio can be precipitated in these conditions.
The relative large number of basal reflections observed for
[H2(TSPP)]-Zn2Al-LDHHyd (up to seven harmonics), related to the
large size of the intercalated porphyrin, allowed us to probe the
structure of the interlayer space projected on the c-axis via Fourier
transform analysis. The electron density was calculated from the
known structure of the hydroxide layer, assuming a weak con-
tribution from the interlayer portion to the total scattering. The
1-D electron density map of [H2(TSPP)]-Zn2Al-LDHHyd (Fig. 6) ρ(Z)
gave rise to two strong peaks at d = 0 and 23.0 Å, because the hy-
droxide layers contain the metal cations. Five additional peaks of
lower electron density, due to the porphyrin molecule, were ob-
served between the layers. On comparison with the dimensions
of the porphyrin molecule, which is a parallelepiped of about
16.0 × 15.8 × 4.3 Å as obtained from steric energy minimization
using the MM2 method [36], we propose a perpendicular arrange-
ment of the porphyrin against the hydroxide layers, in agreement
with previous studies [17].
3.1.3. Electron paramagnetic resonance
Fig. 8 shows the EPR spectra of all the ZnnAl-LDHs intercalated
with the FePs. They all display a common signal in g = 5.7 (axial
symmetry), typical of a high-spin 5/2 FeP complex [10,15,18]. There
also was a slight distortion from the rhombic symmetry, as demon-
strated by the signal in g = 4.3. The higher intensity of the signal
due to axial symmetry in g = 5.7 suggests intercalation of the FeP
between the layers of the ZnnAl-LDH and also that no demetal-
lation occurred during the immobilization procedure [13,14,18], as

M. Halma et al. / Journal of Catalysis 257 (2008) 233–243
239
Fig. 7. UV–vis spectra of the FePs and FeP-ZnnAl-LDHs in glycerin mull. Fe(TSPP) (a),
[Fe(TSPP)]-Zn4Al-LDH (b), [Fe(TSPP)]-Zn2Al-LDH (c), Fe(TDFSPP) (d), [Fe(TDFSPP)]-
Zn2Al-LDH (e), Fe(TDCSPP) (f), [Fe(TDCSPP)]-Zn2Al-LDH (g).
Fig. 8. EPR spectra of solid samples obtained at 77 K. [Fe(TSPP)]-Zn2Al-LDH (a), [Fe-
(TSPP)]-Zn4Al-LDH (b), [Fe(TDFSPP)]-Zn2Al-LDH (c), [Fe(TDCSPP)]-Zn2Al-LDH (d).
also suggested by the UV–vis data. This signal also indicates that
only slight distortion of the porphyrin ring occurred on intercala-
tion.
3.2.1. Cyclooctene
It is well known that the oxidation of cyclooctene by metallo-
porphyrin/PhIO systems produces epoxide as the oxidation prod-
uct, with no traces of allylic alcohol or ketone. This behavior is
justified by the low stability of the allyl radical intermediate tran-
sient species proposed for substrate conversion into the allylic
products [39]. For this reason, cyclooctene is frequently used as
a diagnostic substrate in catalytic systems involving metallopor-
phyrins. In the present work, we used this alkene to investigate the
efficiency and stability of the immobilized anionic FePs as catalyst
in alkene oxidation by PhIO. This study also provided information
regarding the accessibility of the substrate and the oxidant to the
iron(III) sites on the intercalated catalyst.
Table 3 (runs 1–7) characterizes the epoxidation of cyclooctene
by PhIO catalyzed by the FePs either in solution or immobilized on
ZnnAl-LDH. The reproducibility of the reaction results was excellent
(deviation of ca. 2%).
Control reactions using PhIO in the same reaction conditions
(runs 8–10 in Table 3), in the absence of FePs (run 8) and with
pristine Zn2Al-LDH with no FeP as catalyst (runs 9 and 10), led to
lower yields of oxidation product. This confirms that the catalytic
activity observed in runs 1–7 was actually due to the FeP com-
plexes.
3.1.4. Infrared spectroscopy
Infrared spectroscopy (diffuse reflectance) was performed to
provide information about the main vibrations of the porphyrin
rings and the vibration bands of the ZnnAl-LDH support (spectra
not shown). The porphyrin rings displayed vibration bands in the
region 1600–1370 cm−1, due to νC=C phenyl and νC=N. Bands
−1
were also seen in the 1200–1020 cm
region, typical of the
νsym S − φ and νsym S − φ symmetric and asymmetric vibrations
−1
of the φ − SO3 groups. The bands in the 720–640 cm
region
are attributed to out–of-plane C–H vibrations [17,37,38]. The in-
tercalated porphyrin rings displayed vibration bands in the same
positions as the free porphyrins, indicating weak interactions be-
tween the host and guest partners. Similarly, all of the intercalated
materials displayed the bands typical of the ZnnAl-LDH matrix in
−1
the 430-cm region, attributed to M–O and O–M–O bonds. There-
fore, the layer integrity within the support was maintained despite
the large basal spacing. A slight contamination of the phase with
carbonate anions (likely due to the drying method used in this
work), characterized by the presence of a band in 1365 cm−1, also
was observed.
No FeP leaching from the support was observed when excess
substrate (FeP/substrate molar ratio = 1:1000) and a solvent mix-
ture of CH2Cl2/CH3CN 1:1 (v/v) was used. In addition, good quanti-
tative epoxide yields based on PhIO (as high as 90%) were obtained
within 1 h of reaction in the case of the second-generation FePs
(run 7 in Table 3).
3.2. Investigation of the catalytic activity of the intercalated FePs
The catalytic activity of all the intercalated FePs (heterogeneous
catalysis) and of the parent FePs in solution (homogeneous cataly-
sis) was investigated in the oxidation of cyclooctene, cyclohexene,
and the poorly reactive alkane cyclohexane.
As for the more structurally simple anionic iron(III) porphyrin
Fe(TSPP), a higher epoxide yield was achieved with the homoge-

240
M. Halma et al. / Journal of Catalysis 257 (2008) 233–243
Table 3
and cyclooctene to the iron center. This steric effect should play a
more significant role in the catalysis than the activating electronic
effect of the fluorine atoms, explaining why the epoxide yield did
not increase for Fe(TDFSPP)-Zn2Al-LDH. This behavior is similar to
that previously reported by us for the same FeP immobilized in a
“house of cards” LDH structure [15].
Fe(TDCSPP)-Zn2Al-LDH led to increased cyclooctenoxide yields
(run 7 in Table 3) compared with its homogeneous counterpart
(run 6 in Table 3). This change in behavior can be explained by re-
calling the XRPD data. Because Fe(TDCSPP)Zn2Al-LDH had a larger
basal spacing than Fe(TDFSPP)-Zn2Al-LDH, the two orthochlorine
substituents in each mesophenyl porphyrin group should provide
more space around Fe(TDCSPP), creating a larger cavity. This must
favor the access of the oxidant and the substrate to the active cat-
alytic site, thus promoting higher product yields [15,18,20].
Oxidation of cyclooctene by PhIO catalyzed by iron porphyrins in solution (homo-
geneous catalysis) and immobilized in LDH (intercalated iron porphyrins: Fe(TSPP),
Fe(TDFSPP), and Fe(TDCSPP)) (heterogeneous catalysis)a
Catalyst
Run
Cyclooctene oxidea,b
(%)
FePor/ HDL
TONf
(g/g)c
Fe(TSPP)
1
2
3
4
5
6
7
8
9
64
43
34
79
69
76
90
10
11
16
6.2
3.4
2.4
8.4
8.4
7.8
4.8
[Fe(TSPP)]-Zn2Al-LDH
[Fe(TSPP)]-Zn4Al-LDH
Fe(TDFSPP)
[Fe(TDFSPP)]-Zn2Al-LDH
Fe(TDCSPP)
1.00
0.88
1.38
1.41
[Fe(TDCSPP)]-Zn2Al-LDH
No catalyst solidd
Zn4Al-LDHe
Zn2Al-LDHe
10
a
Conditions: purged argon, cyclooctene/solvent mixture CH2Cl2:CH3CN 1:1 (v/v)
at room temperature. FeP/PhIO/substrate molar ratio = 1:10:1000.
b
3.2.2. Cyclohexane
Yields obtained after 1 h of reaction, based on starting PhIO.
c
Concentration of FeP in the ZnnAl-LDH (g/g) calculated by UV–vis and Lambert–
Cyclohexane is a very useful substrate for investigating the ef-
ficiency of FePs as catalysts for alkane hydroxylation by iodosyl-
benzene (PhIO) [8,13,40,41]. Table 4 presents the results obtained
from cyclohexane hydroxylation by PhIO catalyzed by the interca-
lated FePs studied in this work. Systems using the same catalysts
immobilized on the surface of LDH [6,18] and on LDH layers ar-
ranged in a “house of cards” fashion [15] are also discussed for
comparison.
When 1.1 to 1.4 mmol of substrate was used in the solvent mix-
ture CH3CN/CH2Cl2 1:1 (v/v ratio), no catalyst leaching from the
support occurred, and these proportions lead to selective alcohol
formation within 1 h. PhIO consumption in all the reactions was
monitored by the presence of iodobenzene (PhI) in the gas chro-
matograph.
Differences between the porphyrin generations become more
significant when cyclohexane, a substrate that is harder to oxi-
dize compared with cyclooctene, is used (Table 4). Low catalytic
yields and low alcohol/ketone selectivities are obtained when both
homogeneous and immobilized Fe(TSPP) are used (runs 1–3 in Ta-
ble 4). As for the second-generation anionic FePs (runs 4–7 in
Table 4), the steric and electronic effects of the substituents led
to more selective cyclohexane hydroxylation, and the results were
far better than those obtained with homogeneous Fe(TSPP). Com-
paring the two second-generation iron(III) porphyrins [Fe(TDFSPP)]
and [Fe(TDCSPP)] shows that the steric and electronic effects of
the fluorine and chlorine atoms on the ring structures can ac-
count for the differences in the product yields. The alcohol yields
were higher for the heterogeneous Fe(TDCSPP) (run 7 in Table 4)
than for its homogeneous counterpart (run 6 in Table 4), demon-
strating that catalyst immobilization led to improved catalytic
performance. But the opposite behavior was seen for homoge-
neous Fe(TDFSPP) (run 4 in Table 4) and heterogeneous Fe(TDFSPP)
(run 5 in Table 4). As described above, the larger radii of the
two orthochlorine substituents compared with those of the or-
thofluorine substituents led to a larger basal spacing in the case
of Fe(TDCSPP)-Zn2Al-LDH, making the active site of the latter FeP
better dispersed and more readily available for the reagents com-
pared with Fe(TDFSPP)- Zn2Al-LDH, thus resulting in higher prod-
uct yields.
Beer law.
d
Control reaction performed with PhIO and substrate without solid catalyst or
FeP in solution.
e
Control reaction using ZnnAl-LDH without FeP as catalyst in the same conditions
as the oxidation reaction described in (a).
f
TON = turnover number: mol of product/mol of catalyst.
neous catalyst (run 1) compared with the heterogeneous counter-
part (runs 2 and 3). The techniques that we used to characterize
the heterogeneous catalysts proved that the FeP was confined be-
tween the layers of ZnnAl-LDH. Fe(TSPP) between the ZnnAl-LDH
layers forms high-density pillared molecules, hindering the access
of iodosylbenzene to the iron center compared with situation for
the catalyst in a homogeneous medium. This hinders formation
of the active oxometal species with two oxidation levels above
the resting oxidative state, the so-called oxoiron(IV) porphyrin π-
radical cation (OFeIV
tion. The access of the substrate cyclooctene to the heterogeneous
catalyst also may be impaired compared with the situation for the
homogeneous counterpart.
P
) [40–44], responsible for substrate oxida-
•+
The Zn/Al ratio in the LDH also could be an important factor
contributing to catalytic activity. The catalyst with a Zn/Al ratio of
2:1 (run 2 in Table 3) led to slightly higher epoxide yields com-
pared with that with a Zn/Al ratio of 4:1 (run 3), even when the
yield resulting from the control reaction using ZnnAl-LDH only was
deduced from the final results (runs 9 and 10). The solid with a
Zn/Al ratio of 2:1 had a greater charge and consequently a higher
density of FeP, which should favor the catalytic reaction.
Compared with Fe(TSPP), the second-generation anionic iron(III)
porphyrins Fe(TDFSPP) and Fe(TDCSPP) led to better catalytic per-
formance in solution. This was expected, because it is well known
that the halogen groups in the porphyrin structure protect the
porphyrin ring from self-oxidative destruction by an electronic in-
ductive effect and stabilizes the intermediate catalytically active
species [25].
Intercalation of the second-generation FePs into Zn2Al-LDH re-
sults in interesting catalytic behavior. Zn4Al-LDH was not used in
this study because it was less efficient than Zn2Al-LDH in the case
of Fe(TSPP). Compared with the homogeneous system (run 4 in
Table 3), the intercalated Fe(TDFSPP) (run 5 in Table 3) led to
lower epoxide yields. These results suggest that Fe(TDFSPP) was
intercalated between the Zn2Al-LDH layers, making it less accessi-
ble to the reagents [20]. Because the atomic radii of the hydro-
gen and fluorine atoms differ only slightly, the porphyrin rings
of Fe(TDFSPP) and Fe(TSPP) are expected to be of similar size.
Therefore, it is reasonable to view the arrangement of these two
catalytic species between the LDH layers as similar, and thus to
expect the same steric effect described above for Fe(TSPP) for
Fe(TDFSPP), which should hinder the access of the reactants PhIO
The catalytic yields presented here for the intercalated Fe (TDF-
SPP) were lower than those reported previously for the same
FeP immobilized on the surface of LDHs [6,18]. Compared with
the Fe(TDFSPP) catalyst immobilized on the LDH support in a
“house of cards” arrangement, the catalytic results are virtually the
same [15]. These comparisons suggest that access of reactants to
the catalytic site has an important affect on the catalytic activity
of FePs intercalated between LDH layers [6,18].
It is noteworthy that reactions using the ZnnAl-LDH support
itself without intercalated FeP did not give any hydroxylation

M. Halma et al. / Journal of Catalysis 257 (2008) 233–243
241
Table 4
Oxidation of cyclohexane and cyclohexene by PhIO catalyzed by iron porphyrins in solution (homogeneous catalysis) and immobilized in LDH (intercalated iron porphyrins:
Fe(TSPP), Fe(TDFSPP), and Fe(TDCSPP)) (heterogeneous catalysis)a
Catalyst
Cyclohexanea
Cyclohexenea,e
Run
Ol
One
(% yield)d
TONi
Oxide
(% yield)
One
(% yield)
Ol
(% yield)
TONi
(% yield)c
Fe(TSPP)b
[Fe(TSPP)]-Zn2Al-LDH
[Fe(TSPP)]-Zn4Al-LDH
Fe(TDFSPP)b
[Fe(TDFSPP)]-Zn2Al-LDH
Fe(TDCSPP)b
1
2
3
4
5
6
7
8
9
5
<1
<1
42
21
13
10
8
5
<1
<1
<1
<1
–
0.6
0.7
0.6
3.1
1.3
0.7
1.4
67
10
62
49
37
100
69
48
25
40
20
82
33
45
18
6
9.3
8.8
31
[Fe(TDCSPP)]-Zn2Al-LDH
Zn4Al-LDHf
21
21
18
<1
<1
<1
9.4
8.3
<1
Zn2Al-LDHf
–
–
No catalyst solidg
10
15
Final catalytic resulth
Oxide
One
Ol
(% yield)
(% yield)
(% yield)
[Fe(TDFSPP)]-Zn2Al-LDH [run 5 − (run 9 + run 10)]
Fe(TDFSPP) [run 4 − run 10]
2
67
13
62
–
–
34
9
34
34
85
[Fe(TDCSPP)]-Zn2Al-LDH [run7 − (run 9 + run 10)]
Fe(TDCSPP) [run 6 − run 10]
76
a
Typical reaction conditions: purged argon, catalyst/oxidant/cyclohexane molar ratio = 1:10:1000 mmol; solvent mixture dichloromethane/acetonitrile 1:1 (v/v) (350 μL)
at room temperature (yields obtained after 1 h of reaction based on starting PhIO). It was assumed that two mols of PhIO are necessary for ketone formation.
b
Homogeneous catalysis was carried out under identical conditions, in dichloromethane/acetonitrile 1:1 solvent mixture (v/v).
Total yield of cyclohexanol.
Total yield of cyclohexanone.
c
d
e
Cyclohexene oxidation products: oxide = cyclohexane oxide (epoxide), one = 2-cyclohexen-1-one, ol = 2-cyclohexen-1-ol.
f
Control reaction using ZnnAl-LDH without FeP as catalyst in the same conditions as the oxidation reaction described in (a).
g
Control reaction performed with PhIO and substrate without solid catalyst or FeP in solution.
h
Final catalytic result is the yield % of each product from the cyclohexene oxidation after deducing the contribution of the product yields obtained from the control
reactions (runs 9 and 10).
i
TON = turnover number: mol of products/mol of catalyst.
Table 5
Conversion percentage of cyclohexene in the catalytic process involving iron porphyrins in solution and immobilized in LDH and the autooxidation reaction. Transformation
of the final results in yield % based on PhIO
Catalyst system
Cyclohexene conversion (%)a
Run
Oxide
One
–
0.0042
0.041
Ol
–
0.0063
0.014
[Fe(TDFSPP)]-Zn2Al-LDH + PhIO + substrate + solventb
[Fe(TDFSPP)]-Zn2Al-LDH + substrate + solventc
Substrate + solventd
1
2
3
0.024
–
0.00092
Final conversion resulte
0.023
−0.045f
−0.020
Yield (%)g
[Fe(TDFSPP)]-Zn2Al-LDH + PhIO + substrate + solventh
Catalyst system
1.9
–
–
Cyclohexene conversion (%)a
Run
Oxide
One
Ol
Fe(TDFSPP)+ PhIO + substrate + solventb
Fe(TDFSPP) + substrate + solventc
Substrate + solventd
4
5
3
0.54
0.029
0.00092
0.27
0.88
0.041
0.27
0.38
0.014
Final conversion resulte
0.51
−0.65
−0.12
Yield (%)g
Fe(TDFSPP) + PhIO + substrate + solventh
63
–
–
a
The total mol of cyclohexene used was considered 100% conversion.
Typical condition for run 1 was the same described in Table 3.
Reaction performed without air control, in the same molar concentration and experimental conditions as run 1, in the absence of PhIO.
Reaction performed in the same experimental condition of run 1 without air control, in the absence of PhIO and catalyst.
b
c
d
e
% of cyclohexene conversion after deducing the results from the autooxidation routes [run 1 − (run 2 + run 3) in the case of heterogeneous catalysis and run 4 − (run
5 + run 6) for homogeneous catalysis.
f
Negative results means that the autooxidation process, when there was no control of dioxygen conditions, leads to large and out of control conversion of cyclohexene to
the allylic product.
g
Yield % corrected: the results represent the tabled cyclohexene conversion results (oxide = 0.10%, one = 0.25%, and ol = 0.16%) now converted to the yield % based on
the PhIO used in run 1 (heterogeneous catalysis) or run 4 (homogeneous catalysis), since all the reactions were performed in the same experimental conditions.
h
% of cyclohexene conversion expressed as yields % (based on the PhIO) after all autooxidation processes are deduced.
products. This indicates that hydroxylation of cyclohexane can
be attributed to the presence of the intercalated/adsorbed FeP
only.
3.2.3. Cyclohexene
The products generated from cyclohexene oxidation mediated
by PhIO/FeP systems result from a competition between the C=C

242
M. Halma et al. / Journal of Catalysis 257 (2008) 233–243
and allylic C–H groups on the alkene for the electrophilic ac-
tive species, the ferryl porphyrin π-cation radical [45] generated
in the reaction between FeP and iodosylbenzene. Oxidation of
these groups should produce cyclohexenoxide and/or allylic alcohol
(1-cyclohexen-3-ol) and ketone (1-cyclohexen-3-one), respectively
[39,42,44]. Homogeneous FeP/PhIO systems lead to allylic products
in minor yields compared with epoxide, and the efficiency and se-
lectivity of the catalytic reaction toward the epoxide is controlled
by the reaction conditions (solvent, temperature, inert atmosphere,
and reactants molar ratio) [46], the structure of the porphyrin ring,
and the presence of axial ligands to the iron center [33,37,46,47].
In a preliminary examination, the use of second-generation FePs
in the oxidation of cyclohexene produced epoxide and significant
yields of allylic oxidation products (runs 4–7 in Table 4). However,
the control reactions indicated that the ZnnAl-LDH support alone
contributed greatly to the total product yields (runs 8 and 9 in Ta-
ble 4), and that the reaction performed with PhIO and substrate
only (run 10 in Table 4) also led to significant amounts of prod-
ucts. Therefore, the final product yields obtained after deducting
the yields achieved in the control reactions (presented at the end
of Table 4) should provide a more realistic picture of the catalytic
results obtained with both homogeneous and intercalated FePs.
The product distributions and product yields achieved in the
oxidation reactions catalyzed by the FePs are consistent with the
presence of dioxygen in the reaction medium, especially in the
case of the homogeneous second-generation FePs and the inter-
calated Fe(TDCSPP). Despite the argon purge carried out before
the oxidation reaction, homogeneous catalysis probably promoted
dioxygen solubilization throughout the reaction. Controlling the
presence of dioxygen also is a problem inherent to heterogeneous
catalysis, because this gas can be distributed in voids of small crys-
tals of the ZnnAl-LDH support, making it difficult to ensure the
absence of this gas. The presence of dioxygen in the support was
confirmed by the high yields of allylic products obtained when
only the Zn2Al-LDH support (with no intercalated FeP) was used
as catalyst (run 9 in Table 4). The fact that [Fe(TDFSPP)]-Zn2Al-
LDH led to an epoxide yield of only 2% after correction confirms
the low catalytic activity of this intercalated FeP compared with
Fe(TDCSPP)-Zn2Al-LDH, as was observed for the cyclohexane and
cyclooctene substrates.
tone and 7% alcohol) were obtained. These results corroborate the
results described above for the other two substrates (cyclohexane
and cyclooctene); the heterogeneous catalyst led to lower yields
than those achieved with the homogeneous counterparts, due to
steric constraints on the access of the oxidant and reagent to the
active site of the intercalated catalyst. In homogeneous medium,
the two anionic second-generation FePs displayed similar catalytic
behavior, leading to preferable epoxide formation. This is because
both active species were electron-deficient, due to the presence of
halogen atoms. Despite the fact that we could not control all of
the autooxidation routes so as to know the exact amount of allylic
products that they generated, the approximate final yield % results
given in Table 5 suggest that Fe(TDFSPP) was even more selective
toward the epoxide than Fe(TDCSPP) in the conditions used here.
This may be related to the different electron-deficiency imposed
by the halogen atoms fluorine and chlorine on the structures of
the anionic FeP rings. The presence of these electron-withdrawing
groups in the periphery of the porphyrin rings make the catalytic
species more electrophilic, and thus more reactive, toward the
electron-rich C=C bond. Because the fluorine atoms are more elec-
tronegative than chlorine, the active species generated from the
reaction between Fe(TDFSPP) and PhIO is more reactive toward the
C=C bond, leading to higher selectivity toward the epoxide [46].
In summary, the catalytic results obtained from cyclohexane ox-
idation reactions give evidence that both second-generation FePs
are catalytically active in homogeneous solution and when interca-
lated into ZnnAl-LDH. Interestingly, the catalytic results depend on
the particular features of the porphyrin ring structures. Research
using other substrates, such as linear alkanes, is currently under-
way in our laboratory, as is an investigation of the kinetic behavior
of these catalysts.
4. Conclusion
First- and second-generation tetraanionic porphyrins and
iron(III) porphyrins were successfully intercalated between ZnnAl-
LDH layers for the first time, using coprecipitation reaction at
constant pH. The intercalation process seemed to favor the cat-
alytic activity of Fe(TDCSPP), probably because the supported cat-
alyst was less susceptible to inactivation by molecular aggregation
and/or bimolecular self-destruction than the parent FeP in solution.
The inverse behavior was observed with Fe(TDFSPP), suggesting
that steric hindrance to the access of the oxidant and substrate to
the active site of the intercalated FeP should be considered during
the design of immobilized catalysts, especially when intercalated
molecules are involved.
To gain insight into the contribution of the various oxidation
routes to cyclohexene oxidation, we performed various experi-
ments in the presence of dioxygen and absence of PhIO for both
second-generation FeP systems. The results, expressed as the per-
centage of cyclohexene conversion to the reaction products, are
given in Table 5 for Fe(TDFSPP). The same procedure also was car-
ried out for Fe(TDCSPP), with similar results (not shown).
In fact, the oxidation of cyclohexene in the presence of FeP and
air under magnetic stirring, under the same catalytic reaction con-
ditions shown in Table 4 but without PhIO, resulted in preferable
conversion of cyclohexene to allylic products, which is consistent
with a free-radical autooxidation mechanism (heterogeneous catal-
ysis, run 2; homogeneous catalysis, run 5 in Table 5) mediated by
the FeP in solution or intercalated into ZnnAl-LDH [48]. Most of
the allylic products from cyclohexene conversion were produced
when the substrate and solvent were stirred magnetically in air
for the same period and at the same temperature as for the cat-
alytic reaction (run 3 in Table 5). After considering all contribu-
tions from the autooxidation routes from the catalytic results and
converting cyclohexene conversion % into yield %, only epoxide
yields were obtained for both the homogeneous and heteroge-
neous Fe(TDFSPP) catalysts (homogeneous catalysis, 63% epoxide
yield; heterogeneous catalysis, 1.9% epoxide yield). For the homo-
geneous Fe(TDCSPP), the epoxide yield was 61% and the yield of
allylic products was >50%. For the heterogeneous Fe(TDCSPP), an
epoxide yield of 13% and an allylic product yield of 36% (29% ke-
Acknowledgments
Financial support was provided by the Coordenação de Aper-
feiçoamento de Pessoal de Nível Superior (CAPES) and Conselho
Nacional de Desenvolvimento Científico e Tecnológico (CNPq). The
authors also thank the Fundação Araucária, Fundação da Universi-
dade Federal do Paraná (FUNPAR), Universidade Federal do Paraná
(UFPR), Laboratoire des Matériaux Inorganiques (LMI), Centre Na-
tional de la Recherche Scientifique (CNRS), and Université Blaise
Pascal for support and assistance.
References
[1] D. Mansuy, C. R. Chim. 10 (2007) 1.
[2] S. Nakagaki, K.A.D.F. Castro, G.S. Machado, M. Halma, S.M. Drechsel, F. Wypych,
J. Braz. Chem. Soc. 17 (2006) 1672.
[3] J.D. Harvey, C.J. Zeigler, J. Inorg. Biochem. 100 (2006) 869.
[4] F. Bedioui, Coord. Chem. Rev. 144 (1995) 39.
[5] J. Haber, L. Matachowski, K. Pamin, J. Poltowicz, J. Mol. Catal. A Chem. 198
(2003) 215.

M. Halma et al. / Journal of Catalysis 257 (2008) 233–243
243
[6] M. Halma, A. Bail, F. Wypych, S. Nakagaki, J. Mol. Catal. A Chem. 243 (2006)
[25] S. Nakagaki, F.L. Benedito, F. Wypych, J. Mol. Catal. A Chem. 17 (2004) 121.
[26] D. Dolphin, T.G. Traylor, L.Y. Xie, Acc. Chem. Res. 30 (1997) 251.
[27] T. Hibino, W. Jones, J. Mater. Chem. 11 (2001) 1321.
44.
[7] F.L. Benedito, S. Nakagaki, A.A. Saczk, P.G. Peralta-Zamora, M.C.M. Costa, Appl.
Catal. A Gen. 250 (2003) 1.
[28] H. Turk, W.T. Ford, J. Org. Chem. 511 (1991) 253.
[8] Y. Iamamoto, Y.M. Idemori, S. Nakagaki, J. Mol. Catal. A Chem. 99 (1985) 187.
[9] M.A. Matinez-Lorente, P. Battioni, W. Kleemiss, J.F. Bartoli, D. Mansuy, J. Mol.
Catal. A Chem. 113 (1996) 344.
[10] S. Nakagaki, C.R. Xavier, A.J. Wosniak, A.S. Mangrich, F. Wypych, M. Cantão,
I. Denicoló, L.T. Kubota, Colloids Surf. A 168 (2000) 261.
[11] N. Herron, J. Coord. Chem. Soc. 19 (1998) 25.
[29] J.S. Lyndsey, I.C. Schreiman, H.C. Hsu, P.C. Kearney, A.M. Marguerettaz, J. Org.
Chem. 52 (1987) 827.
[30] A. Adler, F.R. Longo, J. Am. Chem. Soc. 86 (1964) 3145.
[31] A. Adler, F.R. Longo, F. Kampas, J. Kim, Inorg. Nucl. Chem. 32 (1979) 2443.
[32] J.G. Sharefkin, H. Saltzmann, Org. Synth. 43 (1963) 62.
[33] H.J. Lucas, E.R. Kennedy, M.W. Forno, Org. Synth. 43 (1963) 483.
[34] J. Rodriguez-Carvajal, Newslet. Powder Diff. Comm. Int. Un. Crystall. 26 (2001)
12.
[12] P. Battioni, R. Iwanejko, D. Mansuy, T. Mlodnicka, J. Poltowicz, F. Chanchez,
J. Mol. Catal. A Chem. 109 (1996) 91.
[13] L. Barloy, J.P. Lallier, P. Battioni, D. Mansuy, Y. Pitfard, M. Tournoux, J.B. Valim,
W. Jones, New J. Chem. 16 (1992) 71.
[14] A.M. Machado, F. Wypych, S.M. Drechsel, S. Nakagaki, J. Colloid Interface Sci.
254 (2002) 158.
[15] S. Nakagaki, M. Halma, A. Bail, G.G.C. Arízaga, F. Wypych, J. Colloid Interface
Sci. 281 (2005) 417.
[16] F. Wypych, G.A. Budniak, M. Halma, S. Nakagaki, J. Colloid Interface Sci. 264
(2003) 203.
[35] A. Ennadi, A. Legrouri, A. Roy, J.P. Besse, J. Mater. Chem. 10 (2000) 2337.
[36] CS Chem3D Ultra Program—Cambridge Soft Corporation, CS Chem3D Ultra 5.0:
Ultimate Modeling, Visualization and Analysis, Cambridge, 1998.
[37] Z. Chang, D.G. Evans, C. Vial, J. Ghanbaja, V. Prévot, M. Roy, C. Forano, J. Solid
State Chem. 178 (2005) 2766.
[38] K. Nakamoto, Infrared and Raman Spectra of Inorganic and Coordination Com-
pounds, fifth ed., Wiley, New York, 1997, pp. 263, 324.
[39] A.J. Appleton, S. Evans, J.R. Lindsay Smith, J. Chem. Soc. Perkin Trans. 2 (1995)
28.
[40] J.T. Groves, T.E. Nemo, R.S. Meyers, J. Am. Chem. Soc. 101 (1979) 1032.
[41] J.T. Groves, R.C. Haushalter, M. Nakamura, T.E. Nemo, B.J. Evans, J. Am. Chem.
Soc. 103 (1981) 2884.
[42] J.T. Groves, J. Inorg. Biochem. 100 (2006) 434.
[43] T. Kamachi, T. Kouno, W. Nam, K. Yoshizawa, J. Inorg. Biochem. 100 (2006)
751.
[17] S. Bonnet, C. Forano, A. Roy, J.P. Besse, P. Maillard, M. Momenteau, Chem. Mat. 8
(1996) 1962.
[18] M. Halma, F. Wypych, S.M. Dreschel, S. Nakagaki, J. Porph. Phthal. 6 (2002) 502.
[19] Z. Tong, T. Shichi, G. Zhang, K. Takagi, Res. Chem. Intermed. 29 (2003) 335.
[20] F. Wypych, A. Bail, M. Halma, S. Nakagaki, J. Catal. 234 (2005) 431.
[21] S. Nakagaki, G.S. Machado, M. Halma, A.A.S. Marangon, K.A.D.F. Castro, N. Mat-
toso, F. Wypych, J. Catal. 242 (2006) 110.
[22] S.S. Cady, T. Pinnavaia, J. Inorg. Chem. 17 (1978) 1501.
[23] M. Chibwe, L. Ukrainczyk, S.A. Boyd, T.J. Pinnavaia, J. Mol. Catal. A Chem. 249
(1996) 113.
[24] S. Nakagaki, A.R. Ramos, F.L. Benedito, P.G. Peralta-Zamora, A.J.G. Zarbin, J. Mol.
Catal. A Chem. 185 (2002) 203.
[44] J.T. Groves, T.E. Nemo, J. Am. Chem. Soc. 105 (1983) 5786.
[45] J.T. Groves, J. Inorg. Biochem. 110 (2006) 434.
[46] W.J. Song, Y.O. Ryu, R. Song, W. Nam, J. Biol. Inorg. Chem. 10 (2005) 294.
[47] J.A. Smegal, C.L. Hill, J. Am. Chem. Soc. 105 (1983) 3515.
[48] D. Paulson, R. Ullman, R.B. Sloane, J. C. S. Chem. Commun. (1984) 186.