Monolayers as models for supported catalysts: Zirconium phosphonate films containing manganese(III) porphyrins

Article abstract of DOI:10.1021/ja017317z

Organized monolayer films of a manganese tetraphenylporphyrin have been prepared and used as supported oxidation catalysts. Manganese 5,10,15,20- tetrakis(tetrafluorophenyl-4′-octadecyloxyphosphonic acid) porphyrin (1) has been immobilized as a monolayer film by a combination of Langmuir-Blodgett (LB) and self-assembled monolayer techniques that use zirconium phosphonate linkages to bind the molecule to the surface. Analysis by FTIR, XPS, UV - vis and polarized optical spectroscopy show that the films consist of noninteracting molecules effectively anchored and oriented nearly parallel to the surface. The monolayer films are stable to the solvent and temperature conditions needed to explore organic oxidations. The activity of films of 1 toward the epoxidation of cyclooctene using iodosylbenzene as the oxidant was compared to that of Manganese 5,10,15,20- tetrakis(pentafluorophenyl) porphyrin (2) and 1 under equivalent homogeneous conditions. The immobilized porphyrin 1 shows an enhanced activity relative to either homogeneous reaction. The main difference between 1 and 2 is the four alkyl phosphonate arms in 1 designed to incorporate the porphyrin within the films. The increased activity of immobilized 1 is a combination of the porphyrin structure, which prohibits the formation of μ-oxo dimers even in solution, and a change in conformation when anchored to the surface. The study demonstrates that careful monolayer studies can provide useful models for the design and study of supported molecular catalyst systems.

Full text of DOI:10.1021/ja017317z

Published on Web 03/30/2002
Monolayers as Models for Supported Catalysts: Zirconium
Phosphonate Films Containing Manganese(III) Porphyrins
Isa O. Ben´ıtez,§ Bruno Bujoli,*,† Laurent J. Camus,† Christine M. Lee,§
Fabrice Odobel,*,† and Daniel R. Talham*,§
Contribution from the Department of Chemistry, UniVersity of Florida,
GainesVille, Florida, 32611-7200, and Laboratoire de Synthe`se Organique,
Faculte´ des Sciences et des Techniques, BP 92208, 44322 Nantes Cedex 03, France
Received October 17, 2001. Revised Manuscript Received January 30, 2002
Abstract: Organized monolayer films of a manganese tetraphenylporphyrin have been prepared and used
as supported oxidation catalysts. Manganese 5,10,15,20-tetrakis(tetrafluorophenyl-4′-octadecyloxyphos-
phonic acid) porphyrin (1) has been immobilized as a monolayer film by a combination of Langmuir-
Blodgett (LB) and self-assembled monolayer techniques that use zirconium phosphonate linkages to bind
the molecule to the surface. Analysis by FTIR, XPS, UV-vis and polarized optical spectroscopy show that
the films consist of noninteracting molecules effectively anchored and oriented nearly parallel to the surface.
The monolayer films are stable to the solvent and temperature conditions needed to explore organic
oxidations. The activity of films of 1 toward the epoxidation of cyclooctene using iodosylbenzene as the
oxidant was compared to that of Manganese 5,10,15,20-tetrakis(pentafluorophenyl) porphyrin (2) and 1
under equivalent homogeneous conditions. The immobilized porphyrin 1 shows an enhanced activity relative
to either homogeneous reaction. The main difference between 1 and 2 is the four alkyl phosphonate arms
in 1 designed to incorporate the porphyrin within the films. The increased activity of immobilized 1 is a
combination of the porphyrin structure, which prohibits the formation of µ-oxo dimers even in solution, and
a change in conformation when anchored to the surface. The study demonstrates that careful monolayer
studies can provide useful models for the design and study of supported molecular catalyst systems.
Introduction
catalysts can mirror their homogeneous behavior or the process
can alter a catalyst’s activity perhaps even changing the
There are several potential advantages to be derived from
confining molecular catalysts at surfaces to form heterogeneous
systems including easier product purification and catalyst
recovery. Common strategies of immobilizing molecular cata-
lysts include incorporating them into organic polymers,1-10 high
surface area carbon,11,12 and inorganic frameworks such as
zeolites,13-15 silica,16-21 or other metal oxides.22,23 Immobilized
mechanism whereby it operates. Detailed understanding of
catalytic behavior at interfaces is clearly desirable, but studies
of supported catalysts can be riddled with uncertainties. Typical
high-surface area supports are often not uniform, giving variation
in catalyst binding sites, aggregation, and orientation with
respect to the surface. It is often even difficult to determine if
the structure of the molecular catalyst has remained intact upon
adsorbing it to a surface. If catalytic activity is observed with
a supported system, it is frequently impossible to know what
species is responsible for the catalysis. Therefore, procedures
for studying molecular catalysts at a uniform and reproducible
interface could enhance the understanding of their behavior in
heterogeneous environments.
* To whom correspondence should be addressed.
§ University of Florida.
† Faculte´ des Sciences et des Techniques.
(1) Dumont, W.; Poulin, J.; Dang, T.; Kagan, H. B. J. Am. Chem. Soc. 1973,
95, 8295.
(2) Takaishi, N.; Imai, H.; Bertelo, C. A.; Stille, J. K. J. Am. Chem. Soc. 1976,
98, 5400.
(3) Cooke, P. R.; Lindsay Smith, J. R. J. Chem. Soc., Perkin Trans. 1 1994,
1913.
(4) Bayston, D. J.; Fraser, J. L.; Ashton, M. R.; Baxter, A. D.; Polywka, M. E.
C.; Moses, E. J. Org. Chem. 1998, 63, 3137.
(15) Sabater, M. J.; Garcia, S.; Alvaro, M.; Garcia, H.; Scaiano, J. C. J. Am.
Chem. Soc. 1998, 120, 8521.
(5) Nozaki, K.; Itol, Y.; Shibahara, F.; Shirakawa, E.; Ohta, T.et al. J. Am.
Chem. Soc. 1998, 120, 4051.
(16) Neumann, R.; Cohen, M. Angew. Chem., Int. Ed. Engl. 1997, 36, 1738.
(17) Rao, Y. V. S.; De Vos, D. E.; Bein, T.; Jacobs, P. A. J. J. Chem. Soc.,
Chem. Commun. 1997, 4, 355.
(6) Pu, L. Tetrahedron: Asymmetry 1998, 9, 1457.
(7) Canali, L.; Sherrington, D. C. Chem. Soc. ReV. 1999, 28, 85.
(8) ter Halle, R.; Colasson, B.; Schulz, E.; Spagnol, M.; Lemaire, M.
Tetrahedron Lett. 2000, 41, 643.
(18) Lindner, E.; Jager, A.; Auer, F.; Wegner, P.; Mayer, H. A. et al. Chem.
Mater. 1998, 10, 217.
(9) Clapham, B.; Reger, T. S.; Janda, K. D. Tetrahedron 2001, 57, 4637.
(10) Astruc, D.; Chardac, F. Chem. ReV. 2001, 101, 2991.
(11) Petrosius, S. C.; Drago, R. S.; Young, V.; Grunewald, G. C. J. Am. Chem.
Soc. 1993, 115, 6131.
(19) De Vos, D. E.; de Wildeman, S.; Sels, B. F.; Grobet, P. J.; Jacobs, P. A.
Angew. Chem., Int. Ed. 1999, 38, 980.
(20) Sandee, A. J.; van der Veen, L. A.; Reek, J. N. H.; Kamer, P. C. J.; Lutz,
M. et al. Angew. Chem., Int. Ed. 1999, 38, 3231.
(12) Drago, R. S.; Jurczyk, K.; Singh, D. J.; Young, V. J. Mol. Catal. A.:
EnViron. 1995, 6, 155.
(21) Hesemann, P.; Moreau, J. J. E. Tetrahedron: Asymmetry 2000, 11, 2183.
(22) Scott, S. L.; Basset, J. M.; Niccolai, G. P.; Santini, C. C.; Candy, J. P. et
al. New J. Chem. 1994, 18, 115.
(23) Maillet, C.; Janvier, P.; Pipelier, M.; Praveen, T.; Andres, Y.; Bujoli, B.
Chem. Mater. 2001, 13, 2879.
(13) Maschmeyer, T.; Rey, F.; Sankar, G.; Thomas, J. M. Nature 1995, 378,
159.
(14) Sutra, P.; Brunel, D. J. Chem. Soc., Chem. Commun. 1996, 21, 2485.
9
10.1021/ja017317z CCC: $22.00 © 2002 American Chemical Society
J. AM. CHEM. SOC. 2002, 124, 4363-4370
4363

A R T I C L E S
Ben´ıtez et al.
Scheme 1. Three-step Deposition Procedure for the Formation of
Films of 1 (The Film Is Present on Both Sides of the Slide)
In this study, we take advantage of a well-understood surface
modification process24-30 to compare the activity of a monolayer
film of a surface-confined manganese porphyrin oxidation
catalyst with its homogeneous behavior. Through careful
molecular design and controlled deposition, monolayers with
known surface coverage, molecular orientation, and mode of
binding to the interface are obtained. Using this controlled
surface for the example of alkene epoxidation, we have observed
increased activity with the surface-confined catalyst relative to
homogeneous reactions under the same conditions. In contrast
to many supported catalysts, all of the catalytic complexes
present in the monolayer films are fully accessible, thus allowing
accurate determination of the ratio of active site to substrate
that is necessary to evaluate the effect of immobilization on
catalytic activity. The monolayer films also allow careful
monitoring of the state of the catalyst during the course of the
reaction so that different behavior of the homogeneous and
supported systems can be quantified. In addition to providing
insight into the specific example of immobilized manganese
porphyrins, the results demonstrate the utility of studying well-
characterized molecular monolayers as models for supported
molecular catalyst systems.
Methods for immobilizing metalloporphyrins in monolayer
films frequently involve coordination bonds to the axial sites
on the metal center.31,32 However, under the conditions needed
to catalyze organic transformations, a covalent linkage through
the macrocycle ligand is preferred to free up the metal center
to participate in the reaction and to guard against desorption
that may result from labile coordination bonds. Our method of
preparing monolayers of covalently confined molecules has been
described previously, and is based on zirconium phosphonate
linkages in combination with Langmuir-Blodgett and “self-
assembled monolayer” adsorption.24-30 It involves the initial
formation a Langmuir-Blodgett monolayer of octadecylphos-
phonic acid (ODPA) on a hydrophobic support that after
adsorption of a layer of Zr4+ ions provides a well-defined surface
for the subsequent deposition of molecules containing phos-
phonic acid groups (Scheme 1).24,25 The thin films are modeled
after layered solid-state metal phosphonates, where organic
layers separate continuous inorganic layers.33-39 The practice
of sequential adsorption of organophosphonates and tetravalent
metal ions has frequently been used to build-up multilayered
films.38-42 However, the present study employs a single
deposition cycle to afford a controlled monolayer. Films formed
in this way are stable to solvents and reaction conditions
commonly used for many organic transformations,30 making it
possible to use them in studies of supported catalysis.
Manganese 5,10,15,20-tetrakis(tetrafluorophenyl-4′-octade-
cyloxyphosphonic acid) porphyrin, 1, was prepared for this
study. The fluorinated tetraphenylporphyrin has four alkylphos-
phonic acid arms intended for binding and orienting the
porphyrin at the surface. Previous studies on the Pd2+ 29 and
free-base analogues43 of 1 have shown that the ligand structure
indeed leads to monolayer coverage with the porphyrin mac-
rocycle oriented parallel to the surface. Another important
feature is that the molecules do not aggregate on the sur-
face.29 The phosphonate groups bind strongly to the zirconated
LB layer to reach a high surface coverage of noninteracting
molecules.
Natural metalloporphyrins are known oxidation catalysts, and
there has been a great deal of investigation of synthetic
analogues44-55 since the early reports that an iron porphyrin
(24) Byrd, H.; Pike, J. K.; Talham, D. R. Chem. Mater. 1993, 5, 709.
(25) Byrd, H.; Whipps, S.; Pike, J. K.; Ma, J.; Nagler, S. E.; Talham, D. R. J.
Am. Chem. Soc. 1994, 116, 295.
(26) Fanucci, G. E.; Seip, C. T.; Petruska, M. A.; Ravaine, S.; Nixon, C. M.;
Talham, D. R. Thin Solid Films 1998, 327-329, 331.
(27) Petruska, M. A.; Fanucci, G. E.; Talham, D. R. Chem. Mater. 1998, 10,
177.
(40) Katz, H. E. Chem. Mater. 1994, 6, 2227.
(41) Obrien, J. T.; Zeppenfeld, A. C.; Richmond, G. L.; Page, C. J. Langmuir
1994, 10, 4657.
(28) Petruska, M. A.; Fanucci, G. E.; Talham, D. R. Thin Solid Films 1998,
329, 131.
(42) Neff, G. A.; Helfrich, M. R.; Clifton, M. C.; Page, C. J. Chem. Mater.
2000, 12, 2393.
(43) Lee, C. M. N. Ph.D. Thesis, University of Florida: Gainesville, 2000.
(44) Groves, J. T.; Nemo, T. E.; Myers, R. S. J. Am. Chem. Soc. 1979, 101,
1033.
(45) Tabushi, I.; Koga, N. Tetrahedron Lett. 1979, 3681.
(46) Chang, C. K.; Ebina, F. J. Chem. Soc., Chem. Commun. 1981, 778.
(47) Mansuy, D.; Battioni, P.; Renaud, J. P. J. Chem. Soc., Chem. Commun.
1984, 1255.
(48) Traylor, P. S.; Dolphin, D.; Traylor, T. G. J. Chem. Soc., Chem. Commun.
1984, 253.
(49) Renaud, J.-P.; Battioni, P.; Bartoli, J. F.; Mansuy, D. J. Chem. Soc., Chem.
Commun. 1985, 888.
(50) De Poorter, B.; Ricci, M.; Meunier, B. Tetrahedron Lett. 1985, 26, 4459.
(51) De Poorter, B.; Meunier, B. J. Chem. Soc., Perkin Trans. 2 1985, 1735.
(52) Traylor, T. G.; Tsuchiya, S. Inorg. Chem. 1987, 26, 1338.
(53) Collman, J. P.; Zhang, X.; Hembre, R. T.; Brauman, J. I. J. Am. Chem.
Soc. 1990, 112, 5356.
(54) Lyons, L. E.; Ellis, E.; Myers, H. K. J. Catal. 1995, 155.
(55) Porhiel, E.; Bondon, A.; Leroy, J. Eur. J. Inorg. Chem. 2000, 1097.
(29) Nixon, C. M.; Claire, K. L.; Odobel, F.; Bujoli, B.; Talham, D. R. Chem.
Mater. 1999, 11, 965.
(30) Petruska, M. A., Talham, D. R. Langmuir 2000, 16, 5123.
(31) Collman, J. P.; Ennis, M. S.; Offord, D. A.; Chng, L. L.; Griffin, J. H.
Inorg. Chem. 1996, 35, 1751.
(32) Ashkenasy, G.; Kalyuzhny, G.; Libman, J.; Rubinstein, I.; Shanzer, A.
Angew. Chem., Int. Ed. 1999, 38, 1257.
(33) Cao, G.; Lee, H.; Lynch, V. M.; Mallouk, T. E. Inorg. Chem. 1988, 27,
2781.
(34) Bujoli, B.; Palvadeau, P.; Rouxel, J. Chem. Mater. 1990, 2, 582.
(35) Wang, R.-C.; Zhang, Y.; Hu, H.; Frausto, R. R.; Clearfield, A. Chem. Mater.
1992, 4, 864.
(36) Bujoli, B.; Pena, O.; Palvadeau, P.; Le Bideau, J.; Payen, C.; Rouxel, J.
Chem. Mater. 1993, 5, 583.
(37) Poojary, M. D.; Hu, H.-L.; Campbell, I., F. L.; Clearfield, A. Acta
Crystallogr. B 1993, 49, 996.
(38) Cao, G.; Hong, H.-G.; Mallouk, T. E. Acc. Chem. Res. 1992, 25, 420.
(39) Thompson, M. E. Chem. Mater. 1994, 6, 1168.
9
4364 J. AM. CHEM. SOC. VOL. 124, NO. 16, 2002

Zirconium Phosphonate Films
A R T I C L E S
solution. The principal difference is that 2 deactivates through
the formation of µ-oxo dimers. On the surface, molecule 1
cannot dimerize, but in solution the alkylphosphonic acid arms
also protect the metalloporphyrin core from dimerization.
However, 1 still shows increased catalytic activity in the
monolayer relative to in solution. In both cases, the molecule
oxidatively degrades with time, but the monolayer deactivates
more slowly, leading to a higher number of turnovers. Confining
the bulky molecule to the surface alters the conformation of
the porphyrin, changing the steric environment around the metal
center. This feature along with the density of the catalytic sites
resulting from the monolayer deposition technique may also play
a role in the observed increase in activity.
Experimental Section
Materials. Reagents were obtained from commercial sources and
used as received unless indicated. Bicyclohexyl 99%, purchased from
Acros (Pittsburgh, PA), was purified by percolation through basic
alumina. Teflon (fluoro(ethylene-propylene)) sheets were purchased
from American Durafilm (Holliston, MA). Single-crystal silicon wafers
(1 0 0) were purchased from Semiconductor Processing Company
(Boston, MA). Germanium attenuated total-reflectance (ATR) crystals
(45°, 50 mm × 10 mm × 3 mm) were purchased from Wilmad Glass
(Buena, NJ).
Manganese 5,10,15,20-tetrakis(tetrafluorophenyl-4′-octadecyloxy-
phosphonic acid) porphyrin (1). Synthetic procedures are included
in the Supporting Information.
Manganese 5,10,15,20-tetrakis(pentafluorophenyl)-porphyrin (2).
This compound was prepared according to published methods.65,66
Substrate Preparation. Glass microscope slides were used as
substrates for UV-vis and catalysis studies. These were cleaned by
the RCA procedure67 and blown dry with a nitrogen stream. To make
them hydrophobic, they were immersed in a 5 mM solution of OTS in
bicyclohexyl for 2 min and dipped in toluene for a few seconds. The
slides were dried with a nitrogen stream and the process repeated.68
Single-crystal silicon wafers (1 0 0), were cut using a diamond glass
cutter to 25 mm × 15 mm × 0.8 mm for XPS substrates. The substrates
were cleaned using the RCA procedure then sonicated for 15 min each
in methanol, 50/50 by volume methanol/chloroform, and chloroform.
The substrates were then sonicated in a 2% OTS solution in hexadecane
and chloroform (50/50 by volume) for 2 h. Finally, the substrates were
sonicated for 15 min each in chloroform, 50/50 by volume methanol/
chloroform, and methanol.69 Germanium attenuated total-reflectance
(ATR) crystals were washed with hot chloroform by placing them in
the thimble of a Soxhlet extractor. They were made hydrophobic by
soaking them in the above OTS solution for an hour, and then rinsed
again with hot chloroform.
with iodosylbenzene could oxidize a number of alkenes and
alkanes.44 The ability of porphyrins to efficiently serve as
oxidation catalysts unfortunately leaves them as potential
substrates. In addition to oxidative degradation of the ligand,
metalloporphyrins deactivate through oxidation of the metal
center to form µ-oxo dimers. One way of avoiding deactivation
is through immobilization of the chromophore.56 Attempts at
immobilizing porphyrin catalysts within supports such as organic
polymers,3,57 silica,3,57-60 zeolites,61,62 and metal phospho-
nates63,64 have also been reported.
The monolayer studies reported here represent an attempt to
better understand the behavior of surface-confined porphyrins.
The activity of a monolayer of porphyrin 1 was carefully
compared to its homogeneous reaction and that of catalyst 2,
an equivalent molecule without the alkylphosphonic acid arms.
Catalyst 2 is manganese 5,10,15,20-tetrakis(pentafluorophenyl)
porphyrin. Catalyst 2 was chosen for its ease of preparation
and because its geometry is likely to mimic the one found in 1
once the alkylphosphonate chains are employed in the im-
mobilization. We find that the films of 1 offer higher yields
and have a longer lifetime than either of the catalysts 1 or 2 in
Film Formation. KSV Instruments (Stratford, CT) Langmuir-
Blodgett systems KSV 2000 and 3000 were used in combination with
a homemade, double barrier Teflon trough for the LB film preparation.
A platinum or filter paper Wilhelmy plate, suspended from a KSV
microbalance, measured the surface pressure. Subphases were prepared
with pure water, with a resistivity of 17.8-18.0 MΩ cm-1 produced
from a Barnstead NANOpure (Boston, MA) purification system.
Zirconium phosphonate films were prepared using a three-step
deposition technique (Scheme 1).24,25 The aqueous subphase was 2.6
(56) Lindsay Smith, J. R. In Metalloporphyrins in Catalytic Oxidations; Sheldon,
R. A., Ed.; Marcel Decker, Inc: New York, 1994; p 325.
(57) Evans, S. J.; Lindsay Smith, J. R. J. Chem. Soc., Perkin Trans. 2 2001,
174.
(58) Gilmartin, C.; Lindsay Smith, J. J. Chem. Soc., Perkin Trans. 2 1995, 243.
(59) Martinez-Lorente, M. A., Battioni, P., Kleemiss, W., Bartoli, J. F., Mansuy,
D. J. Mol. Catal. 1996, 113, 343.
(60) Doro, F. G.; Lindsay Smith, J. R.; G., F. A.; Assis, M. A. J. Mol. Catal.
2000, 164, 97.
mM in CaCl2 adjusted to pH 7.8 with a 0.10 g/L KOH solution.26
A
glass sample vial was placed in the subphase in the well of the trough.
ODPA was spread from a 0.3 mg/mL chloroform solution and
(61) Liu, C. J.; Yu, W. Y. L., S. G.; Che, C. M. J. Org. Chem. 1998, 63, 7364.
(62) Battioni, P.; Iwanejko, R.; Mansuy, D.; Mlodnicka, T.; Poltowicz, J.;
Sanchez, F. J. Mol. Catal. 1996, 109, 91.
(65) Lindsey, J.; Wagner, R. J. Org. Chem. 1989, 54, 828.
(66) Jones, R. D.; Summerville, D. A.; Basolo, F. J. Am. Chem. Soc. 1978,
100, 4416.
(63) Deniaud, D.; Schollorn, B.; Mansuy, D.; Rouxel, J.; Battioni, P.; Bujoli,
B. Chem. Mater. 1995, 7, 995.
(67) Kern, W. J. Electrochem. Soc. 1990, 137, 1887.
(64) Deniaud, D.; Spyroulias, G. A.; Bartoli, J. F.; Battioni, P.; Mansuy, D.et
al. New J. Chem. 1998, 22, 901.
(68) Frydman, E.; Cohen, H.; Maoz, R.; Sagiv, J. Langmuir 1997, 13, 5089.
(69) Advincula, R. Ph.D. Thesis, University of Florida: Gainesville, 1994.
9
J. AM. CHEM. SOC. VOL. 124, NO. 16, 2002 4365

A R T I C L E S
Ben´ıtez et al.
Scheme 2. Synthesis of the Alkyl Spacer 9
compressed at 10 mm/min on the water surface to a target surface
pressure of 20 mN/m. The hydrophobic substrate was dipped down
through the monolayer surface and into the sample vial at 8 mm/min,
transferring the ODPA layer. The substrate and the vial were then
removed from the trough and an amount of zirconyl chloride was added
to the vial to make the solution ca. 3 mM in Zr4+. After at least 36 h
in the zirconium solution, the hydrophilic substrate was removed from
the vial and rinsed with water.
To deposit the porphyrin layer, the hydrophilic substrate was
submerged in a solution of 1 at about 10-5 M in 9/1 ethanol/water,
and the porphyrin was allowed to adsorb for 2 h. The substrate was
then rinsed with methylene chloride or chloroform for 30 min to remove
any physisorbed chromophores. UV-vis and ATR-IR were used to
monitor the completeness of the deposition.
Film Characterization. Methods for attenuated total-reflectance
FTIR, X-ray photoelectron spectroscopy, and transmittance UV-vis
experiments were the same as we have reported in the past.25,29
Catalytic Oxidations. Iodosylbenzene (PhIO) was synthesized from
the diacetate precursor using NaOH.70 Its purity was confirmed using
catalyst 2 in a test reaction of the epoxidation of cyclooctene by PhIO
where the published results were obtained.63
The PhIO compound is only slightly soluble in carbon tetrachloride,
so 27.5 mg were diluted in 20 mL and sonicated for 30 min. After
sonicating, 13 µL of cyclohexanone, the internal standard, and 130 µL
of cyclooctene were added. The mixture was stirred for about 1 min
and solvent was added to complete 25 mL of solution. Samples of 1
mL of this mixture were used for the “blank,” “homogeneous,” and
“heterogeneous” runs in Teflon bags. The heterogeneous bag contained
a microscope slide (34 mm × 26 mm × 1 mm) with the adsorbed
catalyst, while the blank and homogeneous reactions contained a clean
slide of the same dimensions. In all reactions, catalyst:oxidant:substrate
ratios were near 1:5000:40 000. For the homogeneous reactions, 10
µL of a 0.1-mM (1 nmol) solution of 1 or 2 were added, while coverage
on the glass slides was measured as 1.0 ( 0.2 nmol for the
heterogeneous reactions. For the reactions with ethylphosphonic acid
(EtPO3H2), 5.2 µL of 1.9 × 10-4 M or 1.9 × 10-5 M EtPO3H2 were
added for a 2:acid ratio of 1:10 or 1:1.
The Teflon bags were made by sealing sheets with a Technaseal
(Livermore, CA) thermal impulse heat sealer. The sealed bags, with
dimensions of approximately 40 mm × 30 mm, were placed in a closed
beaker and tumbled with a mechanical arm at ca. 100 rpm.
Product concentrations were determined relative to cyclohexanone,
which was used as the internal standard. The gas chromatograph was
a Hewlett-Packard 5890 Series II equipped with a hydrogen flame
ionization detector. A 0.4-µL portion of the reaction solution was
injected onto the 25-m, 0.33-µm film thickness, 0.20-mm ID HP-5
column (Cross-linked 5% PH ME Siloxane). The column was held at
80 °C for 1 min and then ramped at 5 °C/min to a temperature of 95
°C. A second ramp of 70 °C/min was applied to a temperature of 200
°C for 1 min. To avoid evaporation of the solvent, the syringe hole
was sealed after each sampling with a self-adhesive label of a size
similar to the bag. The average solvent loss for each reaction was less
than 6%.
Figure 1. Optical spectra of 1 as (a) 1 × 10-6 M in CH2Cl2 (b) monolayer
film after deposition from ethanol/water and (c) the same film rinsed in
methylene chloride for 30 min. Soret bands near 460-464 nm are assigned
to phosphonate coordinated to the Mn3+ while the peak position at 478 nm
is attributed to Cl- coordination.
group now replaces the tetrahydropyrannyl protecting group of
the alcohol used previously. The benzyl moiety makes this
spacer easy to characterize by TLC (with UV light) and one
step is saved in the synthesis of the spacer, since it can be
cleaved during the palladium catalyzed hydrogenation of the
alkyne group (Scheme 2).
After preparing the substituted porphyrin (Supporting Infor-
mation), the usual harsh conditions for manganese insertion
(Mn(OAc)2 in refluxing DMF) were modified in order to prevent
phosphonate ester hydrolysis. Metalation was preferably per-
formed in DMF at 50 °C using manganese iodide and sodium
acetate and gave the manganese porphyrin in 84% yield. Finally,
the phosphonic ester groups were hydrolyzed in quantitative
yield with bromotrimethylsilane in DMF using triethylamine
as a base to avoid demetalation caused by the formation of
hydrogen bromide.
Optical Properties of the Porphyrins. The optical spectrum
of 1 in a 1 × 10-6 M methylene chloride solution shows a Soret
band at 460 nm (ꢀ ) 2.9 × 104 M-1 cm-1) (Figure 1a). This
peak position is attributed to intramolecular phosphonate binding
to the Mn3+ ion based on studies of the parent compound, 2.
Chloroform solutions of 2 in the concentration range 10-5 to
10-8 M consistently exhibit the Soret band near 476 nm (ꢀ )
4.5 × 104 M-1 cm-1). Optical spectra reveal no tendency for
the chromophore 2 to aggregate in these dilute solutions. The
porphyrin is axially bound by chloride ligands either from the
porphyrin synthesis or present in low concentration from the
solvent.71 Purposeful addition of chloride does not change the
optical spectrum. However, when 1 × 10-4 M ethylphosphonic
After completion of the reaction, films were continuously rinsed with
chloroform in a Soxhlet extractor for at least 1 h to remove any
remaining PhIO. Optical spectra of the slides were then recorded.
Results
Porphyrin Synthesis. The key reaction in the preparation
of the porphyrin 1 is an aromatic nucleophilic substitution
between the 5,10,15,20-tetrakis(pentafluorophenyl) porphyrin
with the alcohol spacer 9 (Scheme 2). The preparation of
compound 9 follows our previous work29 describing the
immobilization of palladium porphyrins, except that a benzyl
(70) Saltzman, H.; Sharefkin, J. G. Org. Synth. 43, 60.
9
4366 J. AM. CHEM. SOC. VOL. 124, NO. 16, 2002

Zirconium Phosphonate Films
A R T I C L E S
according to a procedure described by Mo¨bius.72 Optical spectra
were recorded using polarized light at an incident angle of 45°.
The ratio of s- and p-polarized absorbance for the band at 464
nm corresponds to the porphyrin having an orientation where
the ligand plane is tilted from the substrate, normally in the
range of 72° to 76°. No in-plane anisotropy was observed.
Therefore, the porphyrin molecules are arranged nearly parallel
to the surface, as was found previously for the Pd2+ and free-
base analogues of 1.29 The slight deviation from laying exactly
parallel to the surface may be a consequence of the altered
molecular geometry caused by one of the alkylphosphonate arms
coordinating the metal center.
The porphyrin coverage in the films was measured by etching
the metal phosphonate layer off of the glass slides and measuring
the optical spectra of the porphyrin in the resulting solutions.
The films do not wash off with water or organic solvents because
of the insolubility of the zirconium phosphonate network.
However, zirconium phosphonates are soluble in HF, so films
prepared on glass coverslips were rinsed off with a solution of
50% HF and water. Transmittance optical spectra of the glass
slide after rinsing revealed that porphyrin 1 was entirely
removed. Application of Beer’s law to the washings indicates
a coverage of (6 ( 1) × 10-11 mol/cm2 (an average of four
films). The measured coverage can be compared with that
observed for a Langmuir monolayer of 1 on water. A compres-
sion isotherm of 1 (Supporting Information) on a water subphase
at room temperature shows an increase in surface pressure near
a mean molecular area of 250 Å2/molecule. This value can be
taken as the surface area per molecule in the absence of
constraints on the organization of the alkyl substituents and
corresponds to a coverage of 6.6 × 10-11 moles/cm2. Therefore,
the observed coverage for the metal phosphonate layers is
consistent with a model whereby the porphyrin molecules adsorb
from solution to give a complete monolayer of noninteracting
chromophores laying approximately parallel to the surface.
To show that 1 does not desorb from the film under the
conditions used in the catalysis studies, a slide with a monolayer
of porphyrin was stirred in 1 mL of CCl4 for 24 h. Afterward,
the optical spectrum of the solvent did not show the presence
of 1. For the slide, a decrease of less than 10% in the area of
the Soret band was observed.
Figure 2. Optical spectra of (a) 2.4 × 10-6 M solution of 2 in chloroform
(b) solution of 2.4 × 10-6 M of 2 with 1 × 10-4 M ethylphosphonic acid
and 1 × 10-4 M triisobutylamine in chloroform.
acid (EtPO3H2) and 1 × 10-4 M triisobutylamine were added,
the Soret band shifts to 462 nm (Figure 2). The amine is
necessary to deprotonate the phosphonic acid and it was found
not to bind the porphyrin in the absence of EtPO3H2. The peaks
at 462 nm for 2 and at 460 for 1 are therefore assigned to axial
coordination of the Mn3+ center by phosphonate.
Self-Assembled Monolayers. Monolayers of 1 were prepared
by adsorbing the porphyrin from solution onto zirconated LB
monolayers of octadecylphosphonic acid according to Scheme
1. A key to the process is that the LB layer generated after step
2 provides a flat, uniform surface allowing detailed analysis of
the subsequently adsorbed porphyrin layer. The deposited
porphyrin layer is clearly observed in optical spectroscopy. Films
of 1 deposited from ethanol/water (9/1 mixture) show a λmax at
ca. 463-464 nm (Figure 1b), which corresponds to the peak in
solutions of 1 and can be attributed to nonaggregated metal-
loporphyrins with coordinated phosphonate. Evidently, not all
of the phosphonate groups on each porphyrin bind the surface.
When self-assembled films of 1 are rinsed in methylene chloride
or chloroform, the red shoulder associated with a change in the
ligand environment grows in significantly, indicating that
chloride can displace phosphonate ligation (Figure 1c). However,
as the films dry, the chloride ion is displaced by the phosphonate
group.
Additional evidence for porphyrin adsorption is seen in FTIR
and XPS studies. An ATR-FTIR spectrum of a monolayer of
1, referenced to the zirconated LB layer, clearly shows the
presence of the alkyl chains associated with the four alkylphos-
phonic acid substituents (Supporting Information). Bands at
2926 and 2855 cm-1 corresponding to the νa(CH2) and νs(CH2)
modes, respectively, are consistent with a disordered, liquidlike
arrangement of the alkyl substituents. The size of the porphyrin
prohibits the alkyl substituents from organizing to any extent.
A complex pattern centered near 1050 cm-1, due to the
phosphonate P-O stretches, is also observed. An XPS survey
scan of the film (Supporting Information) shows the presence
of Mn, F, O, and N, corresponding to the porphyrin, as well as
Zr, P, and O from the metal phosphonate network.
Catalysis. We chose to investigate the catalytic properties
of the films using the epoxidation of cis-cyclooctene with
iodosylbenzene (PhIO) as an oxygen donor (1).
The monolayer films were deposited to cover 17.7 cm2,
resulting in approximately 1 nmol of catalyst on each slide. The
small amount of catalyst imposed the condition of a small
solvent volume, chosen to be 1 mL in this study. Significantly
larger volumes would lead to low epoxide concentration and
inaccurate detection. Thermally sealed Teflon bags, formed in
the shape of the slides, served as reaction vessels for the small-
volume reactions. Similar bags have been used by other groups
for solvatothermal solid-state synthesis.73,74 We adapted the
The orientation of the porphyrin in the films can be
determined from polarized absorbance of the Soret band,
(72) Mo¨bius, D.; Orrit, M.; H., G.; Meyer, H. Thin Solid Films 1985, 132, 41.
(73) Harrison, W. T. A.; Nenof, T. M.; Gier, T. E.; Stucky, G. D. Inorg. Chem.
1993, 32, 2437.
(71) Chloride coordination is reduced or eliminated by rinsing the film with
solvent filtered trough basic alumina.
9
J. AM. CHEM. SOC. VOL. 124, NO. 16, 2002 4367

A R T I C L E S
Ben´ıtez et al.
Figure 3. Extent of the epoxidation of cyclooctene with PhIO oxidant over
72 h with catalysts: films of 1 ((), 2 in solution (O), and no catalyst (×).
Figure 4. Optical spectra of (a) the supernatant solution of an oxidation
with a monolayer film of 1 after 72 h, (b) a homogeneous reaction of 1
after 7 h, (c) a film of 1 after 11 h of reaction, and (d) a film of 1 immersed
in the oxidative conditions after 10 min of tumbling.
Table 1. Epoxidation of Cyclooctene by PhIO and Catalyzed by
Manganese Porphyrins in Monolayer Films and in Solutiona
catalyst
reaction time (hr)
yield of epoxided
turnover numbere
monolayer of 1
1 homogeneousb
2 homogeneousb
2 homogeneousb with
1 nmol EtPO3H2
2 homogeneousb with
10 nmol EtPO3H2
blankc
8
7
8
8
20 ( 6%
4.9%
3.6 ( 0.5%
4.1%
893 ( 268
245
180 ( 25
205
7.5
6.1%
305
8
72
72
72
0.5 ( 0.2%
30 ( 7%
monolayer of 1
2 homogeneousb
blankc
1340 ( 350
750 ( 100
15 ( 2%
0.9 ( 0.3%
a Molar ratios of catalyst:PhIO:cyclooctene were 1:5000:40000 in CCl4.
Uncertainties are based on standard deviations from at least three runs. b
1
nmol of catalyst used. c Conditions identical to homogeneous reactions but
no catalyst used. d Calculated using on PhIO as the limiting reagent.
e Calculated as moles of cyclooctene oxide produced over moles of catalyst.
approach to prepare reactors to hold the glass slides. The bags
allow a small solvent volume to wet the slides and eliminate
solvent evaporation during the course of the reaction.
Figure 5. Optical spectrum of homogeneous reaction of 2 after (a) 1 h
and (b) 72 h of tumbling.
show no dissolved porphyrin (Figure 4a). For comparison, the
spectrum of the solution of a homogeneous reaction of 1 shows
similar degradation of the porphyrin, but at an even faster rate
(Figure 4b).
The extent of epoxidation as a function of time in the presence
of the films, and for homogeneous reactions with 1 or 2, is
reported in Figure 3 and Table 1. The yields are calculated
relative to the limiting reagent, which is the oxidant in each
case. Small variations in the amount of catalyst present from
one slide to another were accounted for by normalizing the result
for the amount of catalyst, determined from the area under the
Soret band. Each point is an average of at least three runs and
representative error bars, calculated at the 95% confidence level,
are included. Figure 3 also shows the epoxide yield in the
absence of catalyst (blank reaction). Homogeneous reactions
containing 1 nmol of 1, 1 nmol of 2 with 1 nmol of EtPO3H2,
and 1 nmol of 2 with 10 nmol EtPO3H2 were also monitored
over time and their yields and turnovers after 7 to 8 h are also
included in Table 1.
Characterization of the Catalyst Systems. Heterogeneous
and Homogeneous Reactions with 1. The optical spectra of
the films used in oxidations were recorded after they were rinsed
to remove adsorbed PhIO. The spectra have the same bands as
before use, with a Soret band consistently found near 464 nm
(Figure 4c), although the intensity is decreased by 70 ( 10%
after 11 h (average of 4 films) and 99% after 72 h (average of
3 films). Optical spectra of the supernatant reaction solutions
As the spectra in Figure 4a-c were taken after catalysis runs
and after the slides have been removed from the reaction
conditions, it is possible that they do not represent the state of
the film during the course of the reaction. Therefore, a slightly
different experimental setup was used to monitor the porphyrin
film under the reaction conditions. A film of 1 was prepared
on a thin glass coverslip and placed in an optical cuvette
containing cyclooctene and PhIO in the same concentrations
used in the quantitative runs. The cuvette was tumbled by a
rotating arm and then stopped to record the electronic spectra
of the film while still immersed in the reaction conditions. The
spectrum, shown in Figure 4d, is similar to the ex situ spectrum
and no µ-oxo or other higher oxidation state porphyrin species
are observed.
Homogeneous Reactions with 2. The optical spectra of the
homogeneous reaction solutions of 2 showed at all times the
Soret band shifted to 422 nm with a shoulder at 500 nm (Figure
5). This species has been identified in related porphyrins as a
Mn(IV) µ-oxo dimer.75 The intensity of the Soret band does
(74) Norquist, A. J.; Welk, M. E.; Stern, C. L.; Poeppelmeier, K. R. Chem.
Mater. 2000, 12, 1905.
(75) Smegal, J. A.; Schardt, B. C.; Hill, C. L. J. Am. Chem. Soc. 1983, 105,
1983.
9
4368 J. AM. CHEM. SOC. VOL. 124, NO. 16, 2002

Zirconium Phosphonate Films
A R T I C L E S
Scheme 3. Representation of the Orientation of the Porphyrin 1 in
the Monolayer Films
perseveres over a period of 72 h (Figure 3). Epoxidation
continues over this time, but the rate of conversion is low.
Inclusion of the sterically hindering alkyl phosphonate chains
on 1 prevents the formation of µ-oxo dimers and a band at ca.
420 nm, typical of these species,75 is not seen in optical spectra
of the reaction mixture. On the other hand, the molecule 1 has
been completely destroyed after 7 h as evidenced by the absence
of any Soret band at all (Figure 4b). The two homogeneous
catalysts generate epoxide at similar rates early on in the
reaction, perhaps via the same mechanism, but then deactivate
differently. Compound 1 is destroyed by the reaction conditions,
while 2 forms the less active µ-oxo dimer.
Films of 1 versus the Homogeneous Reactions. The films
of 1 show a significant improvement in catalyst turnovers over
the first few hours relative to the homogeneous reactions (Figure
3). The turnover numbers are five times higher for the films
relative to either 1 or 2 in solution after 8 h. Similar to the
homogeneous reaction, the immobilized catalyst 1 does not form
µ-oxo derivatives (Figures 4c and 4d), as its structure combined
with surface immobilization prevents dimer formation. Never-
theless, the films of 1 degrade during the reaction and there is
no increase in the epoxide yield after about 30 h. The
degradation mechanism is probably the same as in the homo-
geneous reaction of 1, as optical spectra of the reaction solutions
are similar (Figures 4a and 4b). However, immobilization offers
some protection toward porphyrin degradation since 30% of the
porphyrin is still present on the film after 11 h, as compared to
the homogeneous reaction with 1 where no porphyrin is present
in solution after 7 h. In addition, the remaining film continues
to catalyze with the same rate of turnover (corrected for the
amount of remaining porphyrin) as the starting film. This result,
plus the comparison with the homogeneous reactions of 2,
clearly shows that immobilization of the catalyst, and not only
its modified structure, have an important influence on its activity.
not change even after 72 h of reaction, suggesting that this
species is stable under the oxidative reaction conditions. When
EtPO3H2 is added to a solution of 2 in a 1:1 ratio, the optical
spectrum taken after 8 h is similar to that of the reaction without
the acid. On the other hand, a ten-fold excess of the acid causes
a significant inhibition of the µ-oxo dimer with a Soret band at
478 nm.
Discussion
Analysis of how the catalyst 1 is organized in the zirconium
phosphonate films is consistent with a near monolayer coverage
of noninteracting porphyrins, with their macrocycle plane
oriented almost parallel to the substrate. A combination of
optical spectroscopy, XPS, and FTIR spectroscopy confirms that
1 adsorbs to the surface, and the orientation of the molecule is
supported by measurements of surface coverage together with
dichroism of the optical spectra. This result was expected since
the position of the phosphonic acids on the periphery of the
chromophore was designed to allow the porphyrin to stand like
a table on the surface, and a similar monolayer arrangement
was found for the Pd2+ analogue of 1.29 However, not all of
the phosphonate arms of 1 bind to the zirconated surface, as
the Soret band at 464 nm (Figure 1c) indicates that at least one
arm participates in axial binding of the Mn3+ center. Therefore,
our view of the porphyrin on the surface is similar to that
depicted in Scheme 3, with molecules chemisorbed to the
surface such that the porphyrin plane tilts slightly from parallel.
The catalytic activity reported in Table 1 and in Figure 3
clearly shows that the film catalyzes alkene epoxidation,
demonstrating that monolayers prepared in this way can be used
to investigate supported catalysis. To better understand the
monolayer catalyst and its activity, homogeneous studies were
performed using 1 and the related compound 2 under conditions
similar to those used in the heterogeneous studies. The
homogeneous reaction conditions were not chosen to optimize
reaction yields, but rather to allow useful comparisons of the
catalyst behavior. Therefore, homogeneous reactions used the
same amount of catalyst, substrate, oxidant, and solvent as were
used with the monolayer reactions. It is for this reason that
reaction yields reported here are lower than those of other studies
using similar porphyrins but with larger catalyst:substrate
ratios.3,58,60,63,64
Effect of Phosphonate. Another set of control reactions was
performed to investigate the role of phosphonate. Ethylphos-
phonic acid was added to a homogeneous reaction with 2 to
investigate the potential role of the phosphonate as a cocatalyst
in the films. A molar ratio of 1:1 2:EtPO3H2 gives turnovers
comparable to the reaction without the acid (Table 1). Increasing
the amount of phosphonic acid to a molar ratio of 1:10, however,
gives a slightly higher turnover. Optical spectra of these reaction
solutions show that the 1:1 ratio does not affect µ-oxo dimer
formation, while a significant inhibition is observed when a ten-
fold excess of the phosphonic acid is included. The phosphonic
acid appears to stabilize the porphyrin in the oxidation condi-
tions, but the activity is still significantly less than for the
immobilized porphyrin.
Monolayers as Models for Supported Catalysts. While
monolayers on flat surfaces have limited utility as practical
catalysts due to the difficulty of achieving large surface areas,
the present study shows that they can be useful tools in the
study of heterogeneous reactions. An important feature is that
a flat surface, prepared under controlled conditions, allows for
uniform adsorption. Deposition of the catalyst then results in
monolayers that contain essentially a single type of catalytic
site. This feature of monolayers is important when attempting
to correlate structure with properties. Chemical and structural
characterization is easier if only a single species is present,
which then allows catalytic activity to be correlated with a
Comparing Homogeneous Reactions of 1 and 2. The
activity of catalysts 1 and 2 are very similar after 7 to 8 h (Table
1) resulting in ca. 200 turnovers. However, the mechanisms of
deactivation of the molecules are very different. Porphyrin 2
forms a µ-oxo dimer shortly after the reaction starts and
9
J. AM. CHEM. SOC. VOL. 124, NO. 16, 2002 4369

A R T I C L E S
Ben´ıtez et al.
specific, characterized site. These points are in contrast with
the situation for high surface area supports, where different
surface features and a variety of pore shapes and sizes lead to
a dispersion of catalyst environments. Activity must then be
assigned to an ensemble of sites, making it difficult to correlate
behavior with surface structure. The monolayer films can then
serve as models for specific types of molecular sites on high
surface-area supports. Changes in molecular design and deposi-
tion procedures can be used to explore different molecular
arrangements.
those of the individual molecules. For example, with respect to
the present system, the deposition process might be extended
to include surface-bound ligands to axially coordinate the
porphyrin 1. The axial coordination would complement covalent
linkage, but might also be used to direct surface adsorption. In
addition, axial ligands as cofactors can improve the catalytic
performance of the porphyrin and can also extend the array of
useful oxidants. Efforts to explore the possibilities of using
monolayers to build multicomponent catalytic assemblies are
underway.
It should also be noted that new phenomena might arise from
the ability to organize molecular catalysts into controlled
assemblies. The recent example of To¨llner et al. showed that
an LB film of an amphiphilic rhodium bipyridine complex
presented a largely improved catalytic efficiency with respect
to the same complex in solution.76 The authors attributed the
increase to the influence of the assembly of sites at the interface.
In the case of the monolayer films of 1, the observed increase
in activity can be attributed to a combination of enhanced
catalyst lifetime and the altered conformation that the molecule
adopts upon adsorption. However, effects due to the local
concentration of sites cannot be ruled out.
Acknowledgment. This work was partially supported by a
CNRS/NSF Cooperative Research grant, INT-9981579 (D.R.T.)
and 9157 (B.B.). Partial support from the NSF, CHE-9618750
(D.R.T.) is also acknowledged. Anne Robert (Laboratoire de
Chimie de Coordination at Toulouse University) is greatly
acknowledged for fruitful discussions regarding catalytic condi-
tions for the epoxidation.
Supporting Information Available: Synthetic procedures and
full characterization of all new molecules as well as compression
isotherm of 1, ATR-IR spectrum and XPS spectrum of a film
of 1 has been provided for this journal article (PDF). This
material is available free of charge via the Internet at
http://pubs.acs.org.
Finally, monolayer deposition techniques provide special
opportunities for arranging molecular components into “su-
pramolecular assemblies” at solid supports.77 Organized struc-
tures can present different or enhanced properties compared to
JA017317Z
(76) To¨llner, K.; Popovitz-Biro, R.; Lahav, M.; Milstein, D. Science 1997, 278,
2100.
(77) Lehn, J.-M. Supramolecular Chemistry; New York, 1995.
9
4370 J. AM. CHEM. SOC. VOL. 124, NO. 16, 2002