Rational design of a reactive yet stable organic-based photocatalyst

Article abstract of DOI:10.1039/b822111c

Zinc perfluoro-fluoroalkyl-phthalocyanine, synthesized in high yield, does not exhibit electron loss, does not aggregate in solution, is photostable and produces ¹O₂ in very high quantum yields. Aerobic photo-oxygenation of an external substrate occurs without catalyst self-oxidation. The encapsulation of a metal center in a refractory organic environment could guide the design of other viable catalysts for oxygenation of substrates either for synthesis or for oxidative destruction of organic or biological molecules, under reaction conditions that include the use of only air and light. The Royal Society of Chemistry 2009.

Full text of DOI:10.1039/b822111c

View Article Online / Journal Homepage / Table of Contents for this issue
Dalton
Transactions
An international journal of inorganic chemistry
www.rsc.org/dalton
Number 7 | 21 February 2009 | Pages 1065–1252
ISSN 1477-9226
COMMUNICATION
Gorun et al.
Long-range solid-state ordering and
high geometric distortions induced in
phthalocyanines by small fluoroalkyl
groups
COMMUNICATION
Gorun et al.
Rational design of a reactive yet stable
organic-based photocatalyst
1477-9226(2009)7;1-V

View Article Online
COMMUNICATION
www.rsc.org/dalton | Dalton Transactions
Rational design of a reactive yet stable organic-based photocatalyst
Robert Gerdes,a Lukasz Lapok,a Olga Tsaryova,a,b Dieter Wo¨hrleb and Sergiu M. Gorun*a
Received 9th December 2008, Accepted 24th December 2008
First published as an Advance Article on the web 8th January 2009
DOI: 10.1039/b822111c
Zinc perfluoro-fluoroalkyl-phthalocyanine, synthesized in
high yield, does not exhibit electron loss, does not aggregate
in solution, is photostable and produces 1O2 in very high
quantum yields. Aerobic photo-oxygenation of an external
substrate occurs without catalyst self-oxidation. The encap-
sulation of a metal center in a refractory organic environ-
ment could guide the design of other viable catalysts for
oxygenation of substrates either for synthesis or for oxidative
destruction of organic or biological molecules, under reaction
conditions that include the use of only air and light.
are well known aromatic macrocyclic sensitizers that produce
singlet oxygen (1Dg 1O2).
Pc’s catalytic efficiency, however, is limited by either molecular
aggregation via p–p interactions that results in a shortening of
their triplet excited state lifetimes and thus inefficient energy
transfer, and/or reaction with 1O2 resulting in degradation of
their C–H bonds. These are many problems for an effective use
of Pc’s as sensitizers in photocatalytic oxygenations of substrates.
Replacement of labile ring C–H bonds by thermodynamically
stronger C–F bonds enhances the relative stability of F16PcM,
(M = metal) but favours aggregation, and therefore could reduce
catalytic efficiency, in addition to exposing the aromatic C–F
bonds to nucleophilic attack.8
Recently we have described that the partial replacement of
aromatic F groups in F16PcM with iso-perfluoroalkyl groups
(Rf) to give octakis-(perfluoro-i-C3F7)-(perfluoro)PcM, F64PcM,
results in fundamental changes in their properties, including
solubility in organic solvents, facile formation of X-ray quality
single-crystals, and energetically depressed frontier orbitals.9–13
We report here that the F → Rf replacement strategy for M =
Zn results in the simultaneous virtual suppression of sensitizer’s
oxidation, enhanced photochemical stability, single-site isolation
(non-aggregation) as well as efficient oxygenation photocatalysis.
This combination of properties allows one to aim the oxidation
catalysis exclusively at an external substrate by excluding the self-
oxidation pathway, an important goal in catalysis.
The efficient transformation of organic substrates requires both
reactive and stable catalytic materials. The incorporation of oxygen
from air in C–H bonds via catalytic and photocatalytic reactions
in solution is attractive from both chemical and environmental
perspectives, but designers of organic-based catalysts face the
dilemma of using a material containing labile C–H bonds for
the transformation of a substrate that also contains C–H bonds.1
In the case of aerobic oxygenations, the problem could be
severe if the oxidizing species are very reactive, such as oxygen-
based free radicals and singlet oxygen. The latter species can be
produced from ground-state triplet dioxygen and light using a
photosensitizer. Amongst materials that activate triplet oxygen
ꢀ
3
-
(
,
3O2), diamagnetic phthalocyanines (Pc), Fig. 1, have
g
attracted considerable academic and industrial interest,2–7 and
H16PcZn (1), F16PcZn (2), and F64PcZn (3), Fig. 1, were prepared
via microwave assisted synthesis.14 The isolated yields for 1, 2,
and 3, were 95, 59 and 64%, respectively.9,16,17 Reaction times are
typically 10 minutes instead of hours, solvents are eliminated and
the purification is facilitated by the reduced level of impurities.
The solution photo-stabilities of complexes 1, 2, and 3 under Ar
or O2, as well as their ability to produce 1O2, differ considerably.18
Thus, illumination of complexes 1, 2 and 3 under Ar leads only
3
to the decomposition of complex 1 (25% in 3 h).15 Under O2, 1
does not aggregate significantly under our reaction conditions, as
evidenced by the lack of split of the UV-Vis Q-band, but it does
1
decompose in 3 h as a result of the O2 it produces.18 Unlike 1,
complex 2 appears stable under an oxygen atmosphere, but its
aggregation in solution results in a short, <1 ms, lifetime of its
triplet excited state, t,21 and thus inefficient production of reactive
1O2. In contrast, complex 3, similar to 1, does not aggregate,
insuring single-site isolation and a higher t = 52 3 ms.9
Fig. 1 Left: structural formulae for Pc complexes 1 (R1 = R2 = H, the
parent molecule), 2 (R1 = R2 = F) and 3 (R1 = CF(CF3)2, R2 = F). Right:
X-ray structure of complex 3: (top) frontal view of the F and CF3 groups
shown as van der Waals spheres; (bottom) side view, with all atoms shown
as van der Waals spheres.
1
The efficient production of O2 is important for catalysis. The
1O2 quantum yields of complexes 1, 2, and 3 are significantly
different, 0.42, 0.10 and 0.81, respectively.19,20 The relatively low
value for 2 reflects its aggregation, rather than halogen effects,
aDepartment of Chemistry & Environmental Science, New Jersey Institute
of Technology, 161 Warren Street, Newark, NJ 07102, USA. E-mail:
gorun@njit.edu
1
deemed to be small.22 The value of the O2 quantum yield for
bInstitute for Organic & Macromolecular Chemistry, University of Bremen,
Leobener Str. NW2, 28355, Bremen, Germany
complex 3 appears to be the highest reported to date for any Pc
metal complex.23
1098 | Dalton Trans., 2009, 1098–1100
This journal is
The Royal Society of Chemistry 2009
©

View Article Online
The above differences could be rationalized on both steric and
electronic grounds. Firstly, we consider that 1 and 2 (assumed
planar) have the thickness of the van der Waals diameter of their
largest atom. The radii of Zn, C, N, H and F are ~1.39, 1.70,
1.55, 1.20, 1.47 A, respectively,24 and thus intermolecular distances
˚
˚
could reach ~3.4 A, consistent with p–p stacking. Complex 3, like
2, could also be expected to aggregate on electronic grounds, but its
˚
maximum thickness is ~7.8 A due to the bulky i-C3F7 groups. In-
˚
termolecular distances between molecules of 3 could reach ~4.2 A
for the most extreme case, i.e. columnar packing with a perfect 45◦
staggering angle, but even in this case the p–p interactions remain
disfavoured and thus single-site isolation is favoured. Moreover,
the higher Lewis acidity of Zn2+ ion in 3·(acetone)2, which results
in a uniquely high PcZn hexacoordination in solid state,25 is
likely to favour axial binding in solution, further hindering p–
p interactions and thus preventing aggregation. Secondly, the
strong stabilization of the frontier orbitals energy induced by
the Rf groups disfavours oxidation via electron transfer. Thus,
the DFT calculated first ionization potentials (IP) values (eV)
for 1 and 2 are 6.44 and 7.70, respectively, i.e. a difference of
~1.2 eV.8 Since 3 was too large, the isopropyl-C3F7 substituents
were modelled by ethyl-C2F5 groups to give ZnPcF48 (complex
3¢), a simplification supported by ZINDO calculations, which
showed that the UV-Vis spectrum predicted for complex 3¢ is
almost identical to that of 3.12 In comparison with 2, a further
increase of ~1 eV to 8.72 eV is computed for the IP of complex
3¢, while the electron affinities (EA) become progressively more
negative, viz. -2.18, -3.52 and -4.74 for 1, 2, and 3¢, respectively.11
The HOMO stabilization is at least partially responsible for the
enhanced photostability.11–13 Consistent with this view, the E1/2
reduction potentials for complexes 1, 2, and 3 increase, rendering
3 the easiest complex to reduce, while preventing its oxidation via
electron transfer.12
The favourable solution photo stability and photosensitizing
efficiency of complex 3, combined with its resistance to degra-
dation by 1O2 are prerequisites for potentially high catalytic
activity. As a model reaction, we studied the oxygenation of an
industrially important molecule (S)-(-)-3,7-dimethyl-6-octen-1-ol
((S)-(-)-Citronellol, a natural acyclic monoterpenoid.26 The pho-
tooxidation of Citronellol (both enantiomers), produces two hy-
droperoxides, Scheme 1 following the Schenck ene-mechanism.27
One of the hydroperoxide is further converted to Rose oxide, a
component of many fragrances.6,13,28–30
Fig. 2 Photocatalytic oxidation of Citronellol.
products extracted with CHCl3, dried over Na2SO4, and analyzed
by GC and GC-MS as per a literature report.31 Only rose oxide
and the diol that cannot be cyclized were observed.30 These
results confirm the quantitative oxygenation of Citronellol with the
exclusive formation of hydroperoxides, as shown in Scheme 1, thus
validating the 1O2 pathway. Control experiments support the above
chemistry. Dark oxidations under O2, photo-reactivity under N2,
and photo-oxidations under O2 in the absence of Citronellol
did not result in any oxygen consumption. The consumption of
3
ground-state O2, Fig. 2, reaches 24.4 0.2 mL for 3, consistent
with the 24.5 mL volume of 1 mmol of O2 expected for the complete
reaction of 1 mmol Citronellol.
In contrast to complex 3, complexes 1 and 2 are less effective
catalysts. The initial reaction rates for complexes 1, 2 and 3
are 20.6, 3.7 and 33.6 mmol O2 min-1, giving normalized O2
consumptions (initial turnover frequencies) of 1236, 222 and
2016 mol O2 h-1 mol-1 Pc, respectively. The photostabilities of
complexes 1, 2 and 3 under reaction conditions, measured by
monitoring the intensities of their Q bands, are remarkably
different too. Complex 2 de-aggregates slowly but incompletely in
ethanol during catalysis, as indicated by an increase of the intensity
of its Q-band. Thus, because of this competing de-aggregation
reaction, the rate of photodecomposition of F16PcZn could not be
determined. For complexes 1 and 3 the k values are 1.9 ¥ 10-3 and
<10-4, respectively. The latter number is only an estimate since the
value was too small to measure accurately. A loss of 13% in the
intensity of the Q-band of complex 1 is noted, but virtually no
change is observed for complex 3 at the end of the reaction.
The RfPc ring-system, of which 3 is the first example, thus
appears suitable for further substitution studies.
In conclusion, we have shown that a perfluoro-fluoroalkyl
substituted phthalocyanine does not exhibit oxidation by electron
loss, does not aggregate in solution, is photostable and produces
1O2, to our knowledge, in the highest quantum yield for any metal
Pc. Taken together, these properties lead to both catalyst stability
and efficient photocatalysis. In essence, complex 3 illustrates
the strategy of hindering the destructive oxidative self-reactivity
pathways of a complex that exhibits a metal center encapsulated
in a stabilized organic environment, while promoting its reactivity
toward external substrates. Unlike related molecules, complex 3
survives the reactive oxygen species it produces. These results
could guide the design of other viable catalysts for oxygenation
Scheme 1 Singlet oxygen conversion of Citronellol to hydroperoxides.
Using complex 3, the products of (S)-(-)-Citronellol oxygena-
tions were analyzed by NMR and GC-MS. The hydroperoxide
1
mixture was identified by H NMR at the end of the reaction,
according to the literature.30 No signal for S-(-)-Citronellol
could be detected, consistent with the stoichiometric oxygen
consumption (Fig. 2). The hydroperoxides were reduced to the
corresponding diols with an aqueous solution of Na2SO3, the
This journal is
The Royal Society of Chemistry 2009
Dalton Trans., 2009, 1098–1100 | 1099
©

View Article Online
of substrates either for productive synthesis or for the oxidative
destruction of organic or biological molecules, using as reagents
only air and light.
95%, vs. the reported yield of 87%.16 Complex 2 was purified by the
same procedure; yields were 59 10%, vs. 45% reported for the non-
microwave assisted procedure.17 In the case of 3 the reaction product
was washed with toluene, purified by column chromatography on silica
gel (acetone:hexane 3:7) and obtained in 64% yield vs. the reported
21% yield of the non-microwave assisted procedure.10 All products,
Acknowledgements
1
10-1g scale, were characterized successfully by IR, H and 19F NMR,
UV-Vis and EI-MS.
15 S. M. Gorun, B. A. Bench, G. Carpenter, M. W. Beggs, J. T. Mague and
H. E. Ensley, J. Fluor. Chem., 1998, 91, 37.
16 D. Villemin, M. Hammadi, Hachemi and N. Bar, Molecules, 2001, 6,
831.
Financial support from the US Army (to SMG) is gratefully
acknowledged.
17 R. W. Boyle, J. Rousseau, S. V. Kudrevich, M. O. K. Obochi and J. E.
Van Lier, Brit. J. Cancer, 1996, 73, 49.
Notes and references
1 S. J. Lippard, Nature, 2002, 416, 587.
18 Ethanol was used for all stability and reactivity studies. The Lambert–
Beer plots of the Q-bands of all complexes were examined and linearity
2 K. M. Kadish, K. M. Smith and R. Guilard, in The Porphyrin
Handbook, New York, Academic Press, 2003, vol. 15–20.
3 C. C. Leznoff and A. B. P. Lever, in Phthalocyanines: Properties and
Applications, VCH Publishers, New York, 1989–1996, vol. 1–4.
4 N. B. McKeown, in Phthalocyanine Materials, Cambridge University
Press, Cambridge, 1998.
1
was verified for 1 and 3. O2 quantum yields, UDPBF, were determined
by monitoring the 1,3-diphenylisobenzofuran (DPBF) absorption at
411 nm. The quantum yield for 1, 0.42, was similar to the previously
reported value of 0.4,19 thus validating our experimental set-up.
Notably, our previous reported value for 3 in acetone was likely an
underestimation20.
5 R. Gerdes and R. Guder, 2003, WLB, Wasser, Luft und Boden, 47, 25.
6 R. Gerdes, O. Bartels, G. Schneider, D. Wo¨hrle and G. Schulz-Ekloff,
Polym. Adv. Technol., 2001, 12, 152.
19 G. Valduga, S. Nonell, E. Reddi, G. Jori and S. E. Braslavsky,
Photochem. Photobiol., 1988, 48, 1.
7 D. Wo¨hrle, M. Kaneko, K. Nagai, O. Suvorova and R. Gerdes in
Environmental Cleaning by Molecular Photocatalysts, Berlin, Springer
Verlag, 2008, vol. 111, p. 236.
8 C. C. Leznoff and J. L. Sosa-Sanchez, Chem. Commun., 2004, 338.
9 B. A. Bench, A. Beveridge, W. M. Sharman, G. J. Diebold, J. E. van
Lier and S. M. Gorun, Angew. Chem., Int. Ed., 2002, 41, 747.
10 B. A Bench, W. W. Brennessel, H.-J. Lee and S. M. Gorun, Angew.
Chem., Int. Ed., 2002, 41, 750.
20 A. C. Beveridge, B. A. Bench, S. M. Gorun and G. Diebold, J. Phys.
Chem. A, 2003, 107, 5138.
21 A. K. Sobbi, D. Wohrle and D. Schlettwein, J. Chem. Soc., Perkin
¨
Trans. 2, 1993, 3, 481.
22 X.-F. Zhang and H.-J. Xu, J. Chem. Soc., Faraday Trans., 1993, 89,
3347.
23 T. Nyokong, Coord. Chem. Rev., 2007, 251, 1707, and references therein.
24 A. Bondi, J. Phys. Chem., 1964, 68, 441.
11 M.-S Liao, J. D. Watts, M.-J. Huang, S. M. Gorun, T. Kar and S.
Scheiner, J. Chem. Theory Comput., 2005, 1, 1201.
12 S. P. Keizer, J. Mack, B. A. Bench, S. M. Gorun and J. Stillman, J. Am.
Chem. Soc., 2003, 125, 7067.
13 H. Shinohara, O. Tsaryova, G. Schnurpfeil and D. Wo¨hrle, J. Pho-
tochem. Photobiol., A, 2006, 184, 50.
14 Commercial reagents and organic solvents were used as received.
Perfluoro-(4,5-di-isopropyl) phthalonitrile, 4, was prepared according
to the literature.15 The synthesis of fluorinated phthalocyanines by
microwave assisted methods appears unreported to date. A microwave
Discover CEM reactor was used for synthesis. Complexes 1, 2 and 3
were prepared by mixing 0.50 mmol of a phthalonitrile with 0.13 mmol
zinc acetate dihydrate, adding two drops of DMF, and heating the
mixture to 200 C in a sealed tube for 10 minutes. Complex 1 was
purified by Soxhlet extraction with acetone, CH2Cl2 and CH3CN,
followed by re-crystallization from pyridine. The yield of 1 was
25 Cambridge Structural Database, August 2008. Notably, only 10% of
the related Zn porphyrins exhibit hexacoordination.
26 Photo-oxidations conditions: 25 ◦C under O2, jacketed 100 mL
glass reactor, 180 mW cm-2 halogen lamp illumination. 50 mL of
20 mM photosensitizer, 182 mL, (1 mmol) Citronellol in ethanol;
(1000:1 substrate/catalyst ratios). Gas consumption was measured
manometrically. For details on the photocatalytic equipment and
reaction conditions, see also ref. 13.
27 G. Ohloff, E. Klein and G. O. Schenck, Angew. Chem., 1961, 73, 578.
28 S. Taneja, V. Sethi, S. Andotra, S. Koul and G. Qazi, Synth. Commun.,
2005, 35, 2297.
29 A. M. Braun, M.-T. Maurette and E. Oliveros, in Photochemical
Technology, Chichester, Wiley, 445, 1991.
30 S. Meyer, D. Tietze, S. Raub, B. Schafer and G. Kreisel, J. Photochem.
Photobiol., A, 2007, 186, 248.
31 H. Dincalp and S. Ic¸li, J. Photochem. Photobiol., A, 2001, 141, 147.
1100 | Dalton Trans., 2009, 1098–1100
This journal is
The Royal Society of Chemistry 2009
©