Formation of 2-phenylethanol from styrene in the presence of zeolites and UV
irradiation
a
b
a
Markus Steilemann, John N. Armor and Wolfgang F. H o¨ lderich*
a
Department of Chemical Technology and Heterogeneous Catalysis, University of Technology RWTH Aachen,
Air Products and Chemicals, Inc., 7201 Hamilton Boulevard, Allentown, PA 181951501, USA
b
Received (in Cambridge, UK) 14th December 1998, Accepted 9th March 1999
2
-Phenylethanol is formed via an in situ multistep reaction
products, we identified the substances acting as intermediates in
the reaction. Styrene is oxidized. This step is proved to be made
possible by light below l @ 290 nm and we suspect the
formation of superoxide radical anions and their reaction with
styrene or the styrene radical cations. In the absence of the
zeolite or silica–alumina under otherwise identical conditions
no styrene oxide is found, but benzaldehyde is formed because
of singlet oxygen formation. In the following step phenyl-
acetaldehyde is formed via thermally or photochemically
induced rearrangement of styrene oxide. The next steps are not
clearly identified: possibly a decarbonylation and the formation
of benzyl radicals take place, followed by a further reaction with
formaldehyde or MeOH to give the corresponding alcohol. This
could explain the drastically increased selectivity in the
presence of MeOH. Nevertheless, the MeOH could act as an
electron-donating reductant adsorbed on the surface. In this case
the zeolite and the silica–alumina would show photo-semi-
conducting properties. A two-electron reduction of the phenyl-
acetaldehyde would take place leading to 2-phenylethanol by
in situ protonation. The assumed reaction steps are shown in
Fig. 1.
The use of the deuterium lamp in our experiments increased
the selectivity for 2-phenylethanol to 0.5% in aqueous systems.
Due to the shorter reaction time of 6 h instead of 24 h in these
experiments, the much lower overall irradiation energy, the
specific spectra of both lamps and the absorption spectra of the
organic and inorganic compounds, we assume the required
wavelength to run the reaction to be ca. 210 nm. Although the
spectra of styrene adsorbed on the zeolite shows a new broad
absorption at l ≈ 550 nm, this absorption is not required for the
reaction, as experiments with Pyrex reactors show. Other
spectra with MeOH- or water-loaded zeolites showed no
differences in the range l @ 290 nm that could be assigned to
the creation of special electronic states too. Therefore we expect
by irradiation of styrene in the presence of silica–alumina
compounds such as zeolites in aqueous and methanolic
systems; the first step is presumably an oxidation.
The formation of anti-Markovnikov compounds by direct
functionalization of olefins is of great interest for technical and
synthetic applications. The Brown hydroboration–oxidation
provides a useful tool to obtain the interesting products, but
unfortunately this method is accompanied by disadvantages for
its commercial application.1 It is known that anti-Markovni-
kov addition can be achieved by the use of photosensitizers
under photochemical reaction conditions, involving radical
cations as intermediates.4 It is also well known that zeolites are
excellent hosts for photochemical reactions.6 We tried to
replace these photosensitizers by zeolites due to their known
properties for the generation and stabilization of organic
radicals adsorbed on their surface. Here we present some
unexpected results pointing to an unusual combination of
photochemical and thermal processes.
–3
,5
–8
With regard to the work of Yates and Arnold we choose
9
styrene as a model substance for our investigation. By using
EPR and UV spectroscopy we were able to show the formation
of styrene radical cations on an H-ZSM-5 zeolite. The formation
of this radical cation occurs, either in non-irradiated or in
irradiated samples, equally as well in aqueous as in organic
slurries of the zeolite with styrene. For the reaction we used the
following reaction conditions: a 450 W mercury medium
pressure lamp from Ace Glass, Inc., and quartz tube reactors
with a magnetic stirrer placed beside the immersion lamp, while
the reaction mixture is cooled by a centered cooling finger. In
this reactor a mixture of typically 15 g water, 2 mmol styrene
and 3 g of the zeolite was placed and irradiated for a period of
2
4 h at room temperature. After the reaction the mixture was
centrifuged, the zeolite was separated and the fluid phase was
extracted with Et O, then dried with Na SO and, after
2
2
4
evaporation of the solvent, analysed by gas chromatography.
Instead of the zeolite a commercially available silica–alumina,
Aldrich 34, 335–8 and CdS, ZnS and ZnO were used in other
experiments. A 30 W D lamp from Heraeus was also used as
2
the irradiation source. In all cases the conversion was higher
than 99%. In water the selectivity for 2-phenylethanol was
0
.06% for H-ZSM-5 and 0.04% for SiO
2 2 3
–Al O ; for CdS, ZnS,
ZnO or SiO it was not detectable. But for H-ZSM-5 in MeOH
2
the selectivity for 2-phenylethanol rises to 19.4%; methoxystyr-
ene was not detected. Therefore we do not believe MeOH
participates directly in this reaction and we regard this result as
a proof for our proposed reaction mechanism (discussed below).
Other by-products such as olefins, oligomers or alkylated
styrene compounds were also not observed.
Although in the case of water as reaction medium the
selectivity is negligible and more than 90% of the styrene reacts
to form polymeric products, we tried to determine if direct
addition of water in the anti-Markovnikov direction is the
reason for the formation of 2-phenylethanol. By mass spectros-
copy and cross checking experiments with the identified
Fig. 1 Assumed reaction scheme for the formation of 2-phenylethanol from
styrene.
Chem. Commun., 1999, 697–698
697