An exploratory and mechanistic study of the defluorination of an (aminofluorophenyl)oxazolidinone: SN1(Ar*) vs. S R+N1(Ar*) mechanism

Article abstract of DOI:10.1039/b812372a

The morpholinofluorophenyloxazolidinone 1 (the antibacterial drug linezolid) is found to undergo reductive defluorination upon irradiation in water (Φ 0.33), in some of the products accompanied by the simultaneous oxidative degradation of the morpholine side chain. In the presence of chloride, iodide and pyrrole, the fluorine is substituted by these groups (with pyrrole, in position 2). The defluorination is less efficient in methanol and mainly leads to reduction (Φ 0.053). These data can be accommodated through two different mechanisms, viz. either C-F bond heterolysis to give a phenyl cation [S_N1(Ar*)], or ionization to give a radical cation [S _R+N1(Ar*)]. Steady-state and time resolved data have been gathered for clarifying this issue. It is found that, indeed, ionization of 1 is efficient and proceeds from the singlet, but leads to no irreversible change. On the contrary, triplet ³1 (lifetime 0.5 μs in MeOH, <0.1 μs in water) fragments and gives the corresponding triplet phenyl cation. The last intermediate explains well the observed hydrogen abstraction both inter- (from the solvent, when this is reducing) and intramolecularly (from the morpholine group), as well as addition to a charged anion or to a neutral π nucleophile such as pyrrole. The rationalization is supported by the study of some related molecules. Thus, the only photochemical reaction from the non fluorinated analogue of linezolid (that ionizes just as 1) is an inefficient degradation of the morpholine chain (Φ 0.001), while a simple model such as N-(2-fluorophenyl)morpholine undergoes photosolvolysis in water and is not trapped by pyrrole.

Full text of DOI:10.1039/b812372a

View Article Online / Journal Homepage / Table of Contents for this issue
PAPER
www.rsc.org/obc | Organic & Biomolecular Chemistry
An exploratory and mechanistic study of the defluorination of an
+
(aminofluorophenyl)oxazolidinone: SN1(Ar*) vs. SR N1(Ar*) mechanism
Elisa Fasani,*a Fedele Tilocca,a Stefano Protti,a Daniele Merlib and Angelo Albini*a
Received 18th July 2008, Accepted 12th September 2008
First published as an Advance Article on the web 28th October 2008
DOI: 10.1039/b812372a
The morpholinofluorophenyloxazolidinone 1 (the antibacterial drug linezolid) is found to undergo
reductive defluorination upon irradiation in water (U 0.33), in some of the products accompanied by
the simultaneous oxidative degradation of the morpholine side chain. In the presence of chloride, iodide
and pyrrole, the fluorine is substituted by these groups (with pyrrole, in position 2). The defluorination
is less efficient in methanol and mainly leads to reduction (U 0.053). These data can be accommodated
through two different mechanisms, viz. either C–F bond heterolysis to give a phenyl cation [SN1(Ar*)],
+
or ionization to give a radical cation [SR N1(Ar*)]. Steady-state and time resolved data have been
gathered for clarifying this issue. It is found that, indeed, ionization of 1 is efficient and proceeds from
the singlet, but leads to no irreversible change. On the contrary, triplet 31 (lifetime 0.5 ms in MeOH,
<0.1 ms in water) fragments and gives the corresponding triplet phenyl cation. The last intermediate
explains well the observed hydrogen abstraction both inter- (from the solvent, when this is reducing) and
intramolecularly (from the morpholine group), as well as addition to a charged anion or to a neutral p
nucleophile such as pyrrole. The rationalization is supported by the study of some related molecules.
Thus, the only photochemical reaction from the non fluorinated analogue of linezolid (that ionizes just
as 1) is an inefficient degradation of the morpholine chain (U 0.001), while a simple model such as
N-(2-fluorophenyl)morpholine undergoes photosolvolysis in water and is not trapped by pyrrole.
Introduction
Photochemistry has an important role in aromatic nucleophilic
substitution1 and, along with catalysis by transition metal
complexes,2 contributes to widen the scope of a class of reactions
that otherwise requires harsh conditions or activating (electron-
withdrawing) substituents. A number of convenient photochem-
ical procedures for obtaining nucleophilic substitutions has been
reported and include reactions characterized by good yields and
mild conditions. This topic is actively pursued, in particular
for the formation of an aryl–carbon bond by substitution of a
phenyl halide or similar reagent by a carbon-based nucleophile.
A peculiarity of photoinduced reactions is that substitution of a
fluorine atom is generally viable also for non activated substrates
(i.e., those not bearing electron-withdrawing substituents).3 In
thermal chemistry, fluorine substitution is common for activated
substrates, but otherwise less frequent, e.g. in catalytic methods,
where the other halides are used more often.4
As for the mechanism, the photochemical substitution may oc-
cur via the excited state analogue of the addition elimination path
typical of the ground state reactions, the SN2(Ar*) mechanism,
path a in Scheme 1. However, a larger number of examples pertain
to the family of the SRN1 reactions, where the key step is cleavage
of the C–X bond from the radical anion (arising by photoinduced
electron transfer) to form the actual reacting intermediate, the
phenyl radical. Due to this characteristic, this method is generally
Scheme 1 Mechanisms of aromatic photosubstitution reactions.
limited to less strongly bonded aryl halides, such as iodides and
bromides (path b).5–7
A different pathway has taken a more extensive role in
recent years, however, with the identification of a number of
reactions that appear to proceed via unimolecular fragmentation
of the (triplet) excited state and to form a phenyl cation as
the intermediate [SN1(Ar*) mechanism, path c].1a,3,8 In this case
a fluorine atom is accessible to substitution. This reaction has
attracted interest both because of the smooth substitution in
chlorides and fluorides and because in this way a phenyl cation
is produced and this intermediate, difficult to access by non
photochemical methods, has demonstrated to be a synthetically
useful electrophilic reagent for the arylation of alkenes, aromatics
and heteroaromatics under mild conditions. This reaction has
been documented for electron-donating substituted aromatics in
strongly polar media, e.g. for chloro- or fluoroanilines, phenols and
(thio)anisoles in alcohols, acetonitrile or aqueous mixtures.3 It has
aDep. Organic Chemistry, University of Pavia, via Taramelli 10, 27100, Pavia,
Italy. E-mail: elisa.fasani@unipv.it, angelo.albini@unipv.it
bDep. General Chemistry, University of Pavia, via Taramelli 12, 27100, Pavia,
Italy
4634 | Org. Biomol. Chem., 2008, 6, 4634–4642
This journal is
The Royal Society of Chemistry 2008
©

View Article Online
been further identified for some terms of an important family of
drugs, that of fluoroquinolones.
Mechanistic and computational investigations have been carried
out and support for the involvement of path c has been obtained
through several pieces of evidence.8 However, the characteristics
of the reaction, at least as presently known, viz. the limitation
to electron rich substrates and highly polar media, suggest a
further possible mechanism, likewise initiated by a unimolecular
step. Thus, it may be that in a polar solvent such easily oxidized
compounds lose an electron, either by photoionization, or by
electron tranfer within an exciplex ([A ◊ ◊ ◊ A]* → A∑+ + A∑-).
The radical cation thus formed adds a nucleophile leading to
a neutral radical intermediate, and the halide is cleaved at this
+
stage [SR N1(Ar*) mechanism, path d]. Such a mechanism has
been invoked in some cases, indeed has been considered to be
characteristic of good donors in water9 and, although it seems in
principle less appealing for the cleavage of the strong C–F bond,
it deserves attention.
Below we report a study on the photochemistry of the antimi-
crobial drug linezolid, which helps in understanding the scope of
nucleophilic aromatic substitution, since both paths c and d are
followed, while on the other hand adds to the continuing effort
from various laboratories for clarifying the photochemistry (and
phototoxicity mechanism) of fluorinated drugs.10
Scheme 2 Products from the photoreactions of fluorophenyloxazolidi-
none 1 in water.
further minor peak, with m/z 217, corresponding to that of 2 less
two hydrogen atoms. This was proposed to be the corresponding
dehydromorpholino derivative (6). Non-identified high molecular
weight products were also present. Omitting degassing did not
affect the product distribution, but the reaction was slowed by a
factor of 3. Carrying out the irradiation in D2O led to no significant
deuterium incorporation in products 2 to 6.
Compound 1 reacted also by irradiation in nitrogen-flushed
methanol, though the process was slower, and gave defluorinated
2 as the main product (70%, see Scheme 3), along with a small
amount ofthe methoxyphenylderivative 7 andtraces of compound
3. A similar experiment in CD3OH gave again 2, which was isolated
and shown to be 100% deuterated at position 3¢ (product d-2,
see Experimental). The reaction course was again similar in air,
although much slower. Preliminary experiments in MeCN (N2 or
air) led to a complex product mixture and discouraged further
examination.
Results
Linezolid pertains to the family of oxazolidinones a highly
promising new class of antimicrobials11 and is the only term of
this series presently licensed for clinical use. The heterocyclic ring
bears a N-phenyl substituent, and the latter in turn has a fluorine
and a dialkylamino group as substituents.
Photochemistry
In the frame of the validation study, (S)-3-[3¢-fluoro-4¢-(N-morp-
holino)phenyl]-5-N-acetamidomethyloxazolidin-2-one (linezolid,
1) has been previously irradiated in parenteral sterile preparations
in citrate buffer, where compounds 3–5 were characterized.12
The present investigation was carried out in neat water and
methanol. Irradiation of 1 (6 ¥ 10-3 M) in nitrogen-flushed water
gave several products. The main one was isolated by chromatog-
raphy and shown to preserve intact the acetylaminomethyloxa-
zolidinone moiety (Oxaz in Scheme 2) of the starting material,
while the fluorine atom on the aniline ring had been substituted
by a hydrogen (compound 2, 30% yield). Two further products
were isolated (3 and 4, together 16%) that had undergone both
ring defluorination and degradation of the morpholine chain to a
N-(2-hydroxyethylamino) group, with the difference that the latter
compound bore also a N-formyl group. These products along with
the corresponding O, N-diformyl derivative 5 (present in a small
amount as indicated by HPLC/MS analysis), but not 2, had been
previously characterized from the above-mentioned irradiation of
1 in citrate buffer. The analysis showed also the presence of a
Scheme 3 Products from the photoreaction of compound 1 in methanol.
The quantum yield (U) of these reactions was measured in
separate low conversion experiments at 280 nm. The values
obtained were 0.33 in water and 0.053 in methanol.
As seen above, the processes occurring with compound 1 were
defluorination and oxidative degradation of the morpholine chain.
In order to explore the relation between the two processes, the
photochemistry of fluorine-free 2 was examined. This compound
was prepared from 4-(N-morpholino)-aniline following the same
approach reported for 1 (see Experimental). Under irradiation,
This journal is
The Royal Society of Chemistry 2008
Org. Biomol. Chem., 2008, 6, 4634–4642 | 4635
©

View Article Online
compound 2 reacted sluggishly both in water and in methanol to
give products 3 and 4 (see Scheme 4). The quantum yield was quite
low (ca. 0.001 in water and 0.0025 in methanol).
Spectroscopy
The UV absorption of linezolid in water and methanol exhibits
a maximum at 252 and 259 nm respectively (log e 4.3) and a
further partially superimposed band (shoulder at ca. 275 nm,
tail extending up to 310 nm). In the fluorine-free analogue 2
these values were slightly (5 nm) red shifted. In fact, the UV
spectra of compounds 1 and 2 are quite similar to that of 2-
fluoroaniline and aniline, respectively, and indeed the absorbing
chromophores are centered on those moieties. As for emission,
compound 2 (lmax 377 nm; UF = 0.073 in water) fluoresced much
more intensively than fluorinated 1 (lmax 377 nm; UF = 0.0004).
Also the phosphorescence in ether–pentane–alcohol glass was
apparent with 2 (lmax 430 nm), while it was barely detectable with
1 (lmax 450 nm, ca. 30 times less intensive). The phosphorescence
lifetime of 1 was likewise much shorter than that of 2 (5 ms vs.
>100 ms in ether–pentane–alcohol glass at 77 K). Comparison
with literature data shows that the aniline fluorescence is at
334 nm (UF = 0.15 in MeCN) and the phosphorescence at
405 nm (in methyltetrahydrofuran glass at 77 K),14 with an efficient
intersystem crossing in solution (UISC ª 0.7); 4-fluoroaniline
fluoresces with a similar efficiency (UF = 0.12 in MeCN).15
Scheme 4 Products from the photoreaction of the defluoro-analogue 2.
The reactions observed with 1 differed from those known for
2-fluoroaniline, for which photosubstitution to give the phenol is
the main process in water, and were similar to those reported for 4-
fluoroaniline (and the corresponding N,N-dimethyl derivatives),
which undergo reduction rather than substitution.13 In order to
understand the scope of the two reactions, we studied on one
hand whether the morpholine side-chain could introduce some
difference with respect to the NH2 or the NMe2 group and on the
other we explored whether substitution reactions different from
solvolysis could take place with 1. Notice that with 4-fluoroaniline
substitution of the fluorine by a halide or by a p nucleophile occur
efficiently.3
Transient spectroscopy
As for the latter point, the effect of some additives on the
photochemistry of compound 1 in water was examined. Thus,
irradiation in 0.2 M sodium iodide solution led to the isolation
of the 3¢-iodo derivative 8 in 60% yield, accompanied by 19% of
2 (Scheme 5) An analogous experiment in 0.2 M sodium chloride
caused a less marked change, since 2 remained the main product
(45%), but HPLC/MS showed the presence of a chlorinated
product, reasonably of structure 9 in ca. 18% yield.
Laser flash photolysis was then used searching for intermediate(s).
Actually, flashing a nitrogen-flushed 1 ¥ 10-4 M solution of 1 in
water caused the appearance of a transient absorption extended
over a large wavelength range (from 270 to over 600 nm, see
Fig. 1a), with a further absorption centered at 700 nm (shown
enlarged in Fig. 1b). When a solution saturated with N2O (a
known trap for free electrons) was flashed, the red end portion
of the transient was eliminated (see inset in Fig. 1b), while the
remaining part was unaffected. This had a lifetime of ca. 70 ms and
underwent only a small quenching (by a few percent) in an oxygen-
equilibrated solution. On the other hand, flashing in methanol
gave similar results, but now a short-lived component could be
identified in the 300 nm region (lifetime ca. 0.5 ms, 2nd order
decay, see Fig. 2), which was barely discernible in the spectrum
in water. This short-wavelength part was completely quenched in
an oxygen-flushed solution, as expected for a triplet.
Most notably, quite similar transients in terms both of
the spectrum shape and of lifetime were observed when the
fluorine-free morpholinophenyloxazolidinone 2 was flashed (not
reported).
Scheme 5 Photochemistry of compound 1 in water: trapping by chloride,
iodide and pyrrole.
Electrochemical measurements
Finally, irradiation of 1 in an aqueous solution containing
0.2 M pyrrole led to a compound identified as the 3¢-(2¢¢-pyrrolyl)
derivative 10 as virtually the only products (Scheme 5, isolated
yield, 66%).
On the other hand, irradiation of N-(2-fluorophenyl)-
morpholine (11) gave the corresponding hydroxyphenyl derivative
12 as the only isolated product. The product distribution remained
unchanged in the presence of 0.2 M (as well as 1 M) pyrrole. In
methanol by far the main product was phenylmorpholine 13. Also
in this case, no change in the product distribution occurred in the
presence of pyrrole.
The detection of free electrons suggested that photoionization
was occurring and support was seeked by cyclic voltammetry.
With an aqueous solution of 1, two anodic waves were detected
(see Fig. 3). The first one corresponded to a reversible, mo-
noelectronic oxidation (E◦ = +830 mV vs. SCE, average from
three measurements at different scanning rates), the latter to an
irreversible wave. This corresponded to a bielectronic (as deduced
from the Ep vs. scanning rate plot) oxidation (E◦ ca. 1100 mV vs.
SCE). A similar examination of fluorine-free 2 (E◦ = 680 mV
vs. SCE) and of fluorophenylmorpholine 11 gave very similar
results.
4636 | Org. Biomol. Chem., 2008, 6, 4634–4642
This journal is
The Royal Society of Chemistry 2008
©

View Article Online
Fig. 1 Difference absorption spectrum of a 1 ¥ 10-4 M solution of 1 in water 0.5 ms after flashing at 266 nm. (a) Spectrum in the 270–620 nm region.
Inset: decay of the absorption at 340 nm. (b) Spectrum in the 650–800 nm region (different scale). Inset: decay of the 675 absorption in the nitrogen
flushed (upper trace) and nitrogen oxide flushed (lower trace) solution.
process from the fluorophenylmorpholine derivative linezolid (1)
in water or methanol. Homolytic cleavage of the C–F bond is
excluded because the energy of both singlet (ES 84 kcal mol-1) and
triplet (ET 75 kcal mol-1) states were too low to make this (EAr-F
ca. 120 kcal mol-1) a viable path. For the same reason, the SRN
mechanism (path b in Scheme 1) is discarded.
1
As for the addition–elimination mechanism SN2(Ar*), path a in
Scheme 1, this can be excluded on the basis of the product distri-
bution, since if this were involved, a nucleophilic solvent such as an
alcohol or water would substitute the fluorine atom. Solvolysis is
indeed the main process with N-(2-fluorophenyl)morpholine (11)
which gives the phenol 12 in water (although reduction to 13 is
the main process in methanol, see Scheme 6), analogously to what
was previously observed with 2-fluoroaniline,16 but contrary to the
case of 1, where no phenol is formed in water and only a minor
amount of the methoxyphenyl derivative is formed in methanol.
Thus, the morpholine group per se does not introduce a difference
in the photochemistry of fluorobenzenes with respect to other
amino substituents. Apparently, the presence of the oxazolidinonyl
group as a further donating substituent determines the peculiar
chemistry observed with 1, for which two rationalizations remain,
viz. either monomolecular fragmentation of the excited state to
yield the phenyl cation or photoionization (paths c and d in
Scheme 1).
Fig. 2 Difference absorption spectrum of a 1 ¥ 10-4 M solution of 1 in
methanol 0.5 ms after flashing at 266 nm: upper trace, nitrogen-flushed so-
lution; lower trace, oxygen-flushed solution. Inset: decay of the absorbance
at 310 nm in a nitrogen (upper trace) and oxygen-saturated (lower trace)
solution.
Fig. 3 Cyclic voltammogram of 1 (1.5 ¥ 10-3 M in water, 1.2 V min-1, KCl
0.1 M as supporting electrolite).
Scheme 6 Photochemistry of 2-fluorophenylmorpholine.
Discussion
There is no doubt that ionization occurs, as conspicuously
shown by the flash photolysis experiments. The characteristics
of the observed transient, extended over most of the visible, that
The above results show that reductive defluorination, accompa-
nied or not by reaction at the morpholine moiety, is the main
This journal is
The Royal Society of Chemistry 2008
Org. Biomol. Chem., 2008, 6, 4634–4642 | 4637
©

View Article Online
is the long lifetime and the minimal oxygen effect, excluded that
this was an excited state or a radical, while were well compatible
with an ionic intermediate. Indeed, the transient spectrum shown
in Fig. 1a was closely similar to that of the radical cations of N,N-
dimethylaniline (Wurster blue) and of N,N,N¢,N¢-tetramethyl-
p-phenylendiamine (two intense bands at 320 nm and at 530–
640 nm).17 This transient can thus be safely attributed to the radical
cation 1∑+ formed through a photoionization process according to
eqn (1).
endoergonic in polar/non protic solvents, but became exergonic
in the presence of water (see eqn (2)) due to the determining
contribution of the formation of the H–F bond. It is thus
understandable that the photoreaction is more efficient in water
than in methanol, since the former is both a more polar (e 80.2
vs. 32.6) and a slightly more acidic solvent. As a result, heterolysis
according to eqn (2) is exclusive in water, while in MeOH solvolysis
contributes to some degree (see below).
Ar-F + hn + H2O → Ar+ + HF + OH-
(2)
1 + hn → 1∑+ + e-
(1)
The triplet path is supported by the marked deceleration of
the photoreaction under air, the quenching of the 31 in methanol
and the chemistry observed, which well corresponds to that of a
phenyl cation in the triplet state, the expected first intermediate
from the cleavage of 31 and presumably the ground state of
this species analogously to the 4-aminophenyl cations13,21 (see
intermediate 14+ in Scheme 7). There is now plentiful indication
of the contrasting reactivity of singlet and triplet phenyl cations,
both from experiments and from computation.22 Thus, triplet
phenyl cations are not quenched by O2 (and indeed the product
distribution is unchanged) and form a complex with, but do
not add to, a neutral n donor such as water because in such
The electron ejected was indeed detected farther to the red with
respect to the previous transient (see Fig. 1b), and the identification
was confirmed by the selective quenching of that part of the
spectrum by N2O. If one assumes that the molar absorptivity of
1∑+ is similar to that of Wurster blue (e565 12 500 mol-1 cm-1),17 then
the quantum yield of formation of this species is ca. 0.25.
The flash photolysis evidence fitted with the electrochemical
results, showing that single electron oxidation of 1 was possible
under mild conditions and involved the aniline moiety (compare
model compound 11 and N-phenylpiperidine that are oxidized
at 0.95 V vs. SCE,18 respectively, while non phenylated oxazo-
lidinones are oxidized at a much more positive potential, ca.
2.4 V vs. SCE).19 The excited state involved in the photoionization
was the singlet, as indicated by the lack of oxygen effect in the
formation of 1∑+. Importantly, fluorine-free 2 gave a transient
absorption closely similar to that observed with 1 both in shape
and in lifetime (and likewise not affected by oxygen). This again
supported the attribution of the radical cation structure to the
transient from 2, just as that from 1, since ionization appeared to
be the only conceivable phenomenon both of the two compounds,
of almost identical oxidation potential, may undergo, with the
further proviso that electron ejection had to be reversible, since
2 was virtually photostable. The above evidence excluded that
these transients were involved in the observed photochemical
defluorination of 1, because otherwise the transient from this
compound should be much shorter-lived than that from 2, as the
former reacts almost 100 times more efficiently than the latter in
water.
5
1
intermediates the charge is dispersed on the ring (p s structure)
and is negligible at C1, contrary to what happens in the singlet
6
0
(p s structure, charge localized at C1).13,21,23 On the other hand,
both singlet and triplet cations react with p-nucleophiles and with
charged n nucleophiles.
Excluding ionization (path d in Scheme 1) left excited state
fragmentation (path c) as the viable mechanism for defluorination.
The conspicuous ionization made it unlikely that this was a
competitive process from the singlet. Moreover, a direct indication
came from flash photolysis in MeOH, where besides the intense
signal of 1∑+, a shorter-lived transient was observed at 280–320 nm.
This was diffusion-controlled quenched by oxygen (see Fig. 2,
inset) and had no analogy in water, where a transient in that
region was barely detected with a short time delay. These pieces of
evidence supported the assignment of the part of the transient
around 300 nm to triplet 13* that fragmented six times more
efficiently in water than in MeOH and thus was much shorter-
lived in the former solvent. This fact, together with the increased
intensity of the radical cation, made the triplet all but undetectable
in the latter solvent. A triplet path was also supported by the
remarkably decreased phosphorescence lifetime observed with 1
in comparison to 2, indicating that a further process involving
the fluorine atom was competing with emission. The behavior
is analogous to that of 4-fluoroaniline, where calculations on
the triplet state20 showed that heterolysis of the C–F bond was
Scheme 7 Mechanism for the intra- and intermolecular reaction of
compound 1.
The photoreactions of linezolid (see Scheme 7) can thus be
attributed to triplet cation 14+. In methanol, the main path is
intermolecular hydrogen abstraction (path c) leading to compound
2, for which the role of hydrogen abstraction from the solvent is
demonstrated by the formation of 3¢-d 2 in CD3OH. Ether 7 is
a minor product and arises via either a minor path from 14+ or
solvolysis directly from the excited state.
4638 | Org. Biomol. Chem., 2008, 6, 4634–4642
This journal is
The Royal Society of Chemistry 2008
©

View Article Online
Another possibility for the cation is intramolecular hydrogen
abstraction from the vicinal alkylamino group. In the present
cases, this kind of process is revealed in non-hydrogen donating
water by formation of products 3 to 6, where the morpholine
group is stepwise oxidized. This process can be rationalized
via deprotonation of the cation to form enamine 6 (path b,
Scheme 7) and subsequent hydrolysis as well as further oxidation
yielding products where a more deep degradation of the side
chain has occurred (such as products 3 to 5). The oxidation
of the morpholine group had been recognized in the above-
mentioned photostability studies on 1,12 but the connection with
defluorination had not been explicitly recognized.
This process is closely analogous to that observed with flu-
orinated heterocycles bearing an alkylamino side-chain, such
as fluoroquinolones orbifloxacin24a and lomefloxacin (when in
the anionic form).24b In that case, reductive defluorination is
accompanied by oxidative degradation (to a various degree) of
the dialkylamino side chain (a piperazine ring). It clearly differs,
on the other hand, from the mere degradation of the morpholine
group in non-fluorinated anilines, as observed here in the case
of 2. The latter process occurs with a much lower quantum
yield (300 times lower than that of 1 in water). This implies
that secondary oxidation of 2 gives no major contribution to the
formation of products 3 to 6 from 1, which are likely formed in a
monophotonic reaction.
The photochemical reaction of 2 is an instance of the oxidation
of alkyl groups in N-alkylanilines, a common process with this
class of compounds.25 This generally involves electron transfer to
an excited state or a photochemically formed radical and is quite
inefficient, e.g. occurring with U ª1 ¥ 10-3 with fluoroquinolones
that do not undergo defluorination, such as ofloxacin10a and
rufloxacin,26 exactly as it is the case with compound 2.
While water and methanol have no effect, a p nucleophile such
as pyrrole is an effective trap and diverts the reactivity of a
strong electrophile such as cation 14+ (path d). Selective arylation
in position 2 of electron rich heterocycles had been previously
observed upon photolysis of 4-chloroaniline in organic solvents.22
Furthermore, the cation is trapped by charged nucleophiles. The
soft iodide is by far a better trap than chloride, again an indication
of the non-localized nature of cation 14+.
families of fluorinated electron-rich (hetero)aromatics (besides
anilines, anisoles27a and indoles)27b that likewise undergo smooth
substitution via unimolecular fragmentation from the triplet.
The peculiar chemistry of the resulting cation (H abstraction or
insertion in a C–H bond, addition to p nucleophiles) might find
some preparative application, e.g. for selective arylation reactions,
as here with pyrrole, with the advantage of a favorable route in an
eco-friendly solvent such as water.
On the other hand, the frequent occurrence of the fluoroaniline
motif in widely used drugs prompts attention to photolability and
phototoxicity of such derivatives.28 Virtually all of the oxazolidi-
none drugs bear an aminofluorophenyl substituent and are ex-
pected to be photolabile, in a similar manner to fluoroquinolones.
Differently to those drugs, oxazolidinones absorb only a part of
the UV-B radiation, but in a preliminary attempt we found that
when using lamps with emission centered at 360 nm, compound
1 (1 ¥ 10-2 M) decomposes at about half of the rate with similar
lamps centered at 310 nm. Thus, one cannot count too much
on the poor absorption of near-UV light and drug preparations
with compound 1 as the active principle should be protected from
light, with attention to the possible phototoxicity,12,29 even if we
are not aware of clinical reports as yet. Furthermore, the arylation
of pyrrole suggests that attack at nucleic acids is a possibility. A
phototoxic effect28 may be understood on this basis and on the
other hand photoactivated drugs based on this structure may be
conceived.
Experimental section
General
H (300 MHz) and C (75.4 MHz) NMR spectra were registered
by means of a Brucker instrument and IR spectra by using a
Perkin Elmer Fourier transform spectrophotometer. Flash silica
gel was used for column chromatographic separation. Reverse
phase HPLC spectrometry was carried out by using a C8 Zorbax
SB column and eluting with MeCN–H2O 3 : 7). The same
set up was used for HPLC/MS experiments. Fluorescence and
phosphorescence spectra were measured by means of a Perkin
Elmer spectrometer. (S)-3-[3¢-Fluoro-4¢-(N-morpholino)phenyl]-
5-(N-acetamidomethyl)-oxazolidin-2-one (linezolid, 1) was pre-
pared according to the published procedure.30 The fluorine-
free analogue 2, the pharmacological activity of which has
been reported,31 was prepared in a similar way from 4-(N)-
morpholinoaniline as indicated below.
To summarize, the observed photochemistry with 1 is strictly
analogous to that of 4-fluoroaniline (reductive defluorination and
fluorine substitution by halides and by p-nucleophiles) and differs
from that of 2-fluoroanilines, as confirmed here for the case
of 2-fluorophenylmorpholine 11, where solvolysis of fluorine is
observed and the conspicuous photoionization has no role in the
defluorination.
(R)-3-(4¢-N-Morpholinophenyl)-5-(hydroxymethyl)-oxazolidin-
2-one. To a solution of 4-(N)-m◦orpholinoaniline32 (4.2 g) in
water–acetone (50 + 100 mL) at 0 C, sodium carbonate (4.2 g)
was added. After 10 min, benzyl choroformate (3.6 mL) was slowly
added and then the mixture was stirred for 4 h at rt. Addition
of ice–water gave 7 g (90% yield) of the benzyl urethane, mp.
130–132 ◦C; elemental analysis C 69.0, H 6.5, N 8.8, C18H20N2O3
requires C 69.21, H 6.45, N 8.97%; IR (nujol) n 3290, 1724 cm-1;
H NMR (CDCl3) d 3.2 (t, 4H, J = 5 Hz), 3.9 (t, 4H, J = 5 Hz),
5.25 (s, 2H), 6.55 (br s, 1H), 6.9 (d, 2H, J = 9 Hz), 7.3–7.5 (m, 5H),
7.5 (d, 2H, J = 9 Hz); m/z 312 (M+). A solution of this material in
anhydrous THF (110 mL) was brought to -78 ◦C under nitrogen
and treated with butyl lithium (16 mL of a 1.7 M solution in
pentane). After 40 min stirring, a solution of (R)-glycidyl butyrate
Conclusion
In conclusion, linezolid is highly photoreactive in water and to a
lesser degree in methanol. This compound offers the possibility
of comparing two mechanisms for aromatic photosubstitution,
viz. SR N1(Ar*) and SN1 (Ar3*). The latter one is indicated by
+
the fact that ionization occurs (from the singlet) but causes no
irreversible decomposition, while the reactions observed (inter-
and intramolecular hydrogen transfer, trapping by halide ions and
by pyrrole) are all compatible with a triplet phenyl cation as the
intermediate. This finding extends what has been found with other
This journal is
The Royal Society of Chemistry 2008
Org. Biomol. Chem., 2008, 6, 4634–4642 | 4639
©

View Article Online
(4.75 mL) in THF (4.75 mL) was added, the mixture stirred for a
further hour and then left overnight at rt. Upon adding saturated
ammonium chloride (90 mL), ethyl acetate (65 mL) and water
(70 mL), two phases formed and were separated. The organic
phase was extracted with 3 ¥ 60 mL ethyl acetate. Washing of the
reunited extracts with saturated NaCl and evaporation gave the
product (2.29 g, 37% yield), mp 156–158 ◦C; elemental analysis
C 60.9, H 6.5, N 9.8, C14H18N2O4 requires C 60.42, H 6.52, N
10.07%; IR (nujol) n 3420, 1700 cm-1; H NMR (CDCl3) d 3.1 (t,
4H, J = 5 Hz), 3.75 (m, 1H), 3.9 (t, 4H, J = 5 Hz), 3.95–4.05 (m,
3H), 4.75 (m, 1H), 6.9 (d, 2H, J = 9 Hz), 7.5 (d, 2H, J = 9 Hz);
m/z 278 (M+).
N-(2-Fluorophenyl)-morpholine (10)33. The compound was
prepared by adapting a method reported for other phenylated
morpholines.34 Bis-(2-chloroethyl)ether (6.5 ml, 0.055 moles)
was dissolved in 100 ml butanol and 2-fluoroaniline (5.35 ml,
0.055 moles) was added while stirring at rt and then the mixture
was refluxed for 48 h. The solution was cooled to rt, shaken over
Na2CO3, filtered and refluxed for a further 48 h. The solution was
cooled, and water (50 ml) and CH2Cl2 (50 ml) were added. Phase
separation, extraction of the aqueous phase with 3 ¥ 30 ml CH2Cl2
and drying and distillation under vacuum gave 3 ml (0.016 moles,
30% yield) of the title compound as a colorless liquid, elemental
analysis C 66.0, H 6.6, N 7.5, C10H12NOF requires C 66.28, H 6.67,
N 7.73%; IR (neat) n 1502, 1120 cm-1; H NMR (CDCl3) d 3.1 (t,
4H, J = 5 Hz), 3.9 (t, 4H, J = 5 Hz), 6.8–7.0 (m, 2H), 7.0–7.1 (m,
2H); m/z 181.
(R)-3-(4¢-N-Morpholinophenyl)-5-(methanesulfonyloxymethyl)-
oxazolidin-2-one. The above compound (3.8 g) was dissolved in
dry dichloromethane (80 mL) and dry triethylamine (3.75 mL)
was added. The solution was brought to 0 ◦C and methanesulfonyl
chloride (1.48 mL) was slowly added while stirring. After further
20 min stirring, the white precipitate was filtered off. Extraction
of the aqueous phase with dichloromethane and evaporation
gave a solid which was reunited with the original precipitate and
recrystallized from acetonitrile–water to yield the product (3.95 g,
Photochemical reactions
General. 6 ¥ 10-3 M Solutions of linezolid were nitrogen
flushed and irradiated under either of the two conditions: (a) in
an immersion well apparatus (125 mL) by means of a medium
pressure mercury arc or (b) in a number of quartz tubes (each
10 mL) by means of 4 external phosphor coated lamps (centre
of emission, 310 nm). The course of the reaction was monitored
by HPLC. When the starting material was consumed, the solvent
was removed by rotary evaporation and the crude product was
purified by flash chromatography on silica gel (cyclohexane–
ethyl acetate). The key data for the identification of the isolated
new photoproducts are reported below. Compound 2 has been
characterized above and compound 3 was recognized on the
basis of the spectroscopic properties identical to those previously
reported.12 For the other products, HPLC/MS data were used for
structure proposals: compounds 4, m/z 338 (M+, 100); 5, m/z
338 (M+, 100); 6, m/z 318 (M + H+, 100%) (for compounds
4–6 compare ref. 12; 9, m/z 338 (M+, 100). The main charac-
teristics of the other photoproducts are reported below. N-(2-
Fluorophenyl)morpholine (10) was irradiated in the same way
(option b above) to form the hydroxy derivative 11 (see below) and
reduced 12 (identical to a commercial sample).
◦
80% yield), mp 165–168 C; elemental analysis C 50.1, H 6.0, N
7.8, C15H20N2O6S requires C 50.55, H 5.66, N 7.86%; IR (nujol) n
1739 cm-1; H NMR [DCON(CD3)2] d 3.1 (s, 3H), 3.2 (t, 4H, J =
5 Hz), 3.8 (t, 4H, J = 5 Hz), 3.9 (dd, 1H, J = 6, 9 Hz), 4.1 (t, 1H,
J = 9), 4.45 (AA¢dq, 2H, J = 4, 12 Hz), 4.9 (m, 1H), 6.95 (d, 2H,
J = 9 Hz), 7.45 (d, 2H, J = 9 Hz); m/z 356 (M+).
(R)-3-(4¢-N-Morpholinophenyl)-5-(azidomethyl)-oxazolidin-2-
one. A solution of the above compound (2 g) and sodium
azide (1.39 g) in anhydrous THF was heated at 75 ◦C for 16 h.
Upon cooling and treating with water (100 mL) and ethyl acetate
(50 mL), two phases formed and were separated. Extraction of the
aqueous phase and evaporation of the collected organics gave the
product (1.28 g, 74% yield); elemental analysis C 55.9, H 5.3, N
22.8, C14H17N5O3 requires C 55.44, H 5.65, N 23.09%; IR (nujol)
n 2114, 1730 cm-1; H NMR (CDCl3) d 3.15 (t, 4H, J = 5 Hz), 3.6
(AA¢ dq, 2H, J = 5, 12 Hz), 3.8 (t, 4H, J = 4 Hz), 3.8–3.9 (m, 1H),
4.1 (t, 1H, J = 9 Hz), 4.8 (m, 1H), 6.95 (d, 2H, J = 9 Hz), 7.5 (m,
2H, J = 9 Hz); m/z 303.
3¢-d (S)-3-(4¢-N -Morpholinophenyl)-5-(N -acetamidomethyl)-
oxazolidin-2-one (d-2). This differed from the non-deuterated
analogue 2 (see above)12 for the following features: H NMR
(CDCl3) d 6.95 (1H rather than 2H); C NMR (CDCl3) d 117
(CD); m/z (%) 320 (M+, 100).
(S)-3-(4¢-N-Morpholinophenyl)-5-(N-acetamidomethyl)-oxazol-
idin-2-one (2). The above azide (2.33 g) in ethyl acetate (330 mL)
was hydrogenated at room temperature and pressure in the
presence of 0.32 g Pd/C. When the reaction was complete (TLC),
the mixture was cooled at 0 ◦C and pyridine (0.65 mL) and
acetic anhydride (2.25 mL) were added. After stirring at 0 ◦C
for 30 min, the mixture was brought to rt, filtered and evaporated.
The residue was chromatographed on a silica gel column (150 g)
eluting with ethyl acetate to give the title compound (1.73 g, 74%
yield), colorless solid, mp 174–177 ◦C, elemental analysis C 60.0,
H 6.6, N 13.0, C16H21N3O4 requires C 60.17, H 6.63, N 13.16%; IR
(nujol) n 3310, 1730 cm-1; H NMR (CDCl3) d 2.05 (s, 3H), 3.1 (t,
4H, J = 5 Hz), 3.5 (ABX, 2H), 3.7–3.85 (m, 2H), 3.9 (t, 4H, J =
5 Hz), 4.05 (t, 1H, J = 9 Hz), 4.75 (m, 1H), 6.05 (br t, 1H), 6.95
(d, 2H, J = 9 Hz), 7.45 (d, 2H, J = 9 Hz); C NMR (CDCl3): d
22.8 (CH3), 43.5 (CH2), 49.3 (CH2), 51.1 (CH2), 68.2 (CH2), 73.7
(CH), 117.6 (CH), 121.8 (CH), 132.4, 150.2, 157.4, 174.3; m/z
319 (M+).
(S)-3-[3¢-Methoxy-4¢-(N-morpholino)phenyl]-5-(N-acetamido-
methyl)-oxazolidin-2-one (7). Oil that solidifies on standing; IR
(nujol) n 1743 cm-1; elemental analysis C 58.8, H 6.7, N 11.9,
C17H23N3O5 requires C 58.44, H 6.64, N 12.03%; H NMR (CDCl3)
d 2.1 (s, 3H), 3.05 (t, 4H J = 5 Hz), 3.7–3.85 (m, 3H), 3.85 (t, 4H
J = 5 Hz), 3.85 (s, 3H), 4.1 (t, 1H J = 9 Hz), 4.7 (m, 1H), 6.15 (br
t, 1H), 6.75 (dd, 2H J = 2, 9 Hz), 6.9 (d, 1H J = 9 Hz), 7.4 (d,
1H J = 2 Hz). C NMR (CDCl3) d 23.0 (CH3), 41.9 (CH2), 47.8
(CH2), 51.1 (2 CH2), 55.5 (CH3), 67.0 (2 CH2), 71.7 (CH), 103.0
(CH), 110.1 (CH), 117.8 (CH), 132.4, 150.2, 157.5, 174.3.
(S)-3-[3¢-Iodo-4¢-(N-morpholino)phenyl]-5-(N-acetamidome-
thyl)-oxazolidin-2-one (8). Colorless solid, mp 200–203 ◦C; IR
(nujol) n 1737 cm-1; elemental analysis C 42.9, H 4.5, N 9.3,
C16H20N3O4I requires C 43.16, H 4.53, N 9.44%; H NMR
4640 | Org. Biomol. Chem., 2008, 6, 4634–4642
This journal is
The Royal Society of Chemistry 2008
©

View Article Online
[(CD3)2SO] d 1.85 (s, 3H), 2.9 (t, 4 H, J = 5 Hz,), 3.3 (ABX,
2H), 3.7 (m, 1H), 3.75 (t, 4H, J = 5 Hz), 4.1 (t, 1H, J = 9 Hz), 4.7
(m, 1H), 7.2 (d, 1H, J = 9 Hz), 7.5 (dd, 1H, J = 2.5, 9 Hz), 8.1
(d, 1H, J = 2.5 Hz), 8.2 (t, 1H, J = 6 Hz); C NMR [(CD3)2SO]
d 22.8 (CH3), 40.4 (CH2),47.7 (CH2), 52.9 (2 CH2), 66.8 (2 CH2),
71.9 (CH), 98.6 (CI), 119.5 (CH), 121.4 (CH), 129.3 (CH), 133.4,
137.8, 152.5, 154.4, 171.0.
Catalysts, Wiley, Chichester, 2004J. Tsuji, in Transition Metal Reagents
and Catalysts, Wiley, Chichester, 2000; D. A. Culkin and J. F. Hartwig,
Acc. Chem. Res., 2003, 36, 234; I. P. Beletskaya and A. V. Cheprakov,
Chem. Rev., 2000, 100, 3009.
5 R. A. Rossi and R. H. De Rossi, Aromatic Substitution by the SRN
Mechanism, Am. Chem. Soc., Washington, 1983.
6 R. A. Rossi and A. B. Penenory, in Handbook of Organic Photochemistry
and Photobiology, ed. W. Horspool and P. S. Song, CRC Press, Boca
Raton, 2nd edn, 2004, ch. 47.
1
7 M. T. Baumgarter, A. B. Pierini and R. A. Rossi, J. Org. Chem., 1999,
(S)-3-[3¢-(2-Pyrrolyl)-4¢-(N-morpholino)phenyl]-5-(N-acetamido-
methyl)-oxazolidin-2-one (10). Colorless solid, mp 56–58 ◦C;
elemental analysis C 62.4, H 6.3, N 14.9, C20H24N4O4 requires
C 62.49, H 6.29, N 14.57%; IR (nujol) n 1740 cm-1; H NMR
(CD3OH) d 1.9 (s, 3H), 2.8 (t, 4H, J = 5 Hz,), 3.5 (ABX, 2H), 3.8
(m, 1H), 3.8 (t, 4H, J = 5 Hz), 4.1 (t, 1H, J = 9 Hz), 4.7 (m, 1H),
6.2 (t, 1H, J = 2 Hz), 6.5 (dd, 1H, J = 2, 3.5 Hz), 6.9 (dd, 1H, J =
2, 3.5 Hz), 7.1 (d, 1H, J = 9 Hz), 7.3 (dd, 1H, J = 3, 9 Hz), 7.6 (d,
1H, J = 3 Hz); C NMR (CD3OH) d 22.8 (CH3), 43.5 (CH2), 49.8
(CH2), 53.8 (2 CH2), 68.4 (2 CH2), 73.7 (CH), 108.5 (CH), 109.8
(CH), 118.0 (CH), 118.8 (CH), 120.1 (CH), 121.1 (CH), 129.4,
131.3, 135.8, 146.7, 157.2, 174.0.
64, 6487.
8 M. Freccero, E. Fasani, M. Mella, I. Manet, S. Monti and A. Albini,
Eur. J. Chem., 2008, 14, 653.
9 J. Cornelisse, G. Lodder and E. Havinga, Rev. Chem. Intermed., 1979,
2, 231; S. Nilson, Acta Chem. Scand., 1973, 27, 329.
10 S. Monti, S. Sortino, E. Fasani and A. Albini, Chem.–Eur. J., 2001, 7,
2185; E. Fasani, F. F. Barberis Negra, M. Mella, S. Monti and A. Albini,
J. Org. Chem., 1999, 64, 5388; A. Albini and S. Monti, Chem. Soc. Rev.,
2003, 32, 238; L. J. Martinez, G. Li and C. F. Chignell, Photochem.
Photobiol., 1997, 65, 599; J. Claridge Navaratnam, Photochem. Photo-
biol., 2000, 72, 283; M. C. Cuquerella, F. Bosca` and M. A. Miranda,
J. Org. Chem., 2004, 69, 7256; G. Condorelli, G. De Guidi, S. Giuffrida,
S. Sortino, R. Chillemi and S. Sciuto, Photochem. Photobiol., 1999, 70,
280; S. Monti, S. Sortino, E. Fasani and A. Albini, Chem.–Eur. J., 2001,
7, 2185.
11 M. R. Barbachyn and C. W. Ford, Angew. Chem., 2003, 115, 2056;
M. R. Barbachyn and C. W. Ford, Angew. Chem., Int. Ed., 2003, 42,
2010; J. F. Barrett, Curr. Opin. Investig. Drugs, 2000, 1, 181.
12 G. E. Martin, R. H. Robins, P. B. Bowman, W. K. Duholke, K. A.
Farley, B. D. Kaluzny, J. E. Guido, S. M. Sims, T. J. Thamann, B. E.
Thompson, T. Nishimura, Y. Noro and T. Tahara, J. Heterocycl. Chem.,
1999, 36, 265.
13 B. Guizzardi, M. Mella, M. Fagnoni, M. Freccero and A. Albini, J. Org.
Chem., 2001, 66, 6353.
14 F. D. Lewis and W. Liu, J. Phys. Chem. A, 1999, 103, 9678; D. A. Tasis,
M. G. Siskos, A. K. Zarkadis, S. Steenken and G. Pistolis, J. Org. Chem.,
2000, 65, 4274.
N-(2-Hydroxyphenyl)-morpholine (11).35 Colorless solid, mp
130 ◦C; elemental analysis C 66.5, H 7.4, N 7.4, C10H13NO2
requires C 67.02, H 7.31, N 7.82%; IR (nujol) n ca. 3000 (br),
1454 cm-1; H NMR (CDCl3) d 3.0 (t, 4H, J = 5 Hz), 3.95 (t, 4H,
J = 5 Hz), 6.9 (dt, 1H, J = 7.5, 1.5), 7.0 (dd, 1H, J = 1.5, 7.5), 7.1
(dt, 1H, J = 1, 7.5), 7.2 (dd, 1H, J = 1, 7.5); C NMR (CDCl3) d
52.7 (2 CH2), 66.5 (2 CH2), 114.2 (CH), 120.1 (CH), 121.3 (CH),
126.6 (CH), 138.6, 151.4.
15 K. Othmen, P. Boule, B. Szczepanik, K. Rotkiewicz and G. Grabner,
J. Phys. Chem. A, 2000, 104, 9525.
Quantum yield measurements
16 K. Othmen and P. Boule, J. Photochem. Photobiol. A: Chem., 2000,
136, 79.
Reaction quantum yields were measured by irradiating 2 mL
samples of 5 ¥ 10-4 M solutions of either 1 or 2 in a quartz
spectrophotometric cuvette on an optical bench. The light source
was a collimated beam from a 100 W high-pressure mercury arc
fitted with an interference filter (transmittance maximum, 280 nm).
The reaction was monitored by HPLC and the consumption
of the starting material (limited to <20%) was determined (a
known volume—20 ml—injection loop was used). The light flux
was measured by ferrioxalate actinometry. Fluorescence quantum
yields were measured by using quinine sulfate as the standard
(UF = 0.54).
17 P. Maruthamuthu, L. Venkatasubramanian and P. Dharmalingam,
J. Chem. Soc., Faraday Trans. 1, 1986, 82, 359; S. Nakayama and K.
Suzuki, Bull. Chem. Soc. Jpn., 1973, 46, 3694; A. Bewick, D. Serve
and T. A. Joslin, J. Electroanal. Chem. Int. Electrochem., 1983, 154,
81; H. Takemura, K. Takehara and M. Ata, Eur. J. Org. Chem., 2004,
4936.
18 K. Le Gall, J. P. Hurvois, T. Renaud, C. Moinet, A. TAllec, P. Uriac, T.
Sinbandhit and L. Toupet, Liebig Ann. Chem., 1997, 2089.
19 Y. Cao, K. Suzuki, T. Tajima and T. Fuchigami, Tetrahedron, 2005, 61,
6855.
20 M. Freccero, M. Fagnoni and A. Albini, J. Am. Chem. Soc., 2003, 125,
13182.
21 M. Aschi and J. N. Harvey, J. Chem. Soc., Perkin Trans. 2, 1999,
1059.
22 M. Mella, P. Coppo, B. Guizzardi, M. Fagnoni, M. Freccero and A.
Albini, J. Org. Chem., 2001, 66, 6344.
Acknowledgements
23 S. M. Gasper, C. Devadoss and G. B. Schuster, J. Am. Chem. Soc.,
1995, 117, 5206.
24 (a) T. Morimura, Y. Nobuhara and H. Matsukura, Chem. Pharm. Bull.,
1997, 45, 373; (b) E. Fasani, M. Mella and A. Albini, Eur. J. Org. Chem.,
2004, 5075.
25 D. Narog, V. Lechowicz, T. Pietryga and A. Sobkoviak, J. Mol. Catal.
A: Chem., 2004, 212, 25 and references cited therein.
26 A. Belvedere, F. Bosca`, M. C. Cuquerella, G. De Guidi and M. A.
Miranda, Photochem. Photobiol., 2002, 76, 252.
Partial support of this work by MURST, Rome, is gratefully
acknowledged.
References
1 M. Fagnoni and A. Albini, in Molecular and Supramolecular Photo-
chemistry, ed. V. Ramamurthy and K. S. Schanze, Taylor & Francis,
Boca Raton, 2006, vol. 14, p. 131; J. Cornelisse and E. Havinga,
Chem. Rev., 1975, 75, 353; J. Cornelisse, in Handbook of Organic
Photochemistry and Photobiology, ed. W. Horspool and P. S. Song,
CRC Press, Boca Raton, 1995, p. 250.
2 U. Mazurek and H. Schwarz, Chem. Commun., 2003, 1321; T. Saeki,
Y. Takashima and K. Tamao, Synlett, 2005, 1771; L. Ackermann, R.
Born, J. H. Spatz and D. Meyer, Angew. Chem., Int. Ed., 2005, 7216.
3 M. Fagnoni and A. Albini, Acc. Chem. Res., 2005, 38, 713.
27 (a) G. Zhang and P. Wan, J. Chem. Soc., Chem. Commun., 1994, 19;
(b) N. C. Yang, A. Huang and D. H. Yang, J. Am. Chem. Soc., 1989,
111, 8060.
28 L. Martinez and C. F. Chignell, J. Photochem. Photobiol. B, 1998, 45,
91.
29 H. Agrawal, K. R. Muhadik, K. R. Paradkar and N. Kaul, Drug Dev.
Ind. Pharm., 2003, 29, 1119.
30 S. J. Brickner, D. K. Hutchinson, M. R. Barbachin, P. R. Manninen,
D. A. Ulanowicz, S. A. Garmon, K. C. Grega, S. K. Hendges, D. S.
4 Transition Metals for Organic Synthesis, ed. M. Beller and C. Bolm,
Wiley-VCH, Weinheim, 1998J. Tsuji, in Palladium Reagents and
This journal is
The Royal Society of Chemistry 2008
Org. Biomol. Chem., 2008, 6, 4634–4642 | 4641
©

View Article Online
Toops, C. W. Ford and G. E. Zurenko, J. Med. Chem., 1996, 39,
673.
31 D. M. Gleave, S. J. Brickner, P. R. Manninen, D. A. Allwine, K. D.
Lovasz, D. C. Rohrer, J. A. Tucker, G. F. Zurenko and C. W. Ford,
Bioorg. Med. Chem. Lett., 1998, 8, 1231.
33 L. J. Goosen, J. Paetzold, O. Briel, A. Rivas-Nass, R. Karch and B.
Kayser, Synlett, 2005, 275.
34 G. E. Martin, R. J. Elgin, J. R. Mathiasen, C. B. Davis, J. M. Kesslick,
W. J. Baldy, R. P. Shank, D. L. DiStefano, C. L. Fedde and M. K. Scott,
J. Med. Chem., 1989, 32, 1052.
32 V. Kinonicek, E. Svatek, J. Holubek and M. Protiva, Collect. Czech.
Chem. Commun., 1986, 51, 937.
35 J. C. Lockhart, A. C. Robson, M. E. Thompson, S. D. Furtado, C. K.
Kaura and A. R. Allan, J. Chem. Soc., Perkin Trans. 1, 1971, 577.
4642 | Org. Biomol. Chem., 2008, 6, 4634–4642
This journal is
The Royal Society of Chemistry 2008
©