The effect of the p-nitro group on the chemistry of phenylnitrene. A study via intramolecular trapping

Article abstract of DOI:10.1039/a903863i

The photodecomposition of a pyrazolyl substituted p-nitrophenyl azide has been studied as an intramolecular model for the reactivity of the parent azide, largely used for photochemical labelling. The singlet nitrene is trapped by intramolecular cyclization onto the pyrazole nitrogen as well as by the low yield addition of intermolecular nucleophiles (EtOH, Et₂NH). Ring expansion to a didehydroazepine is absent. The triplet nitrene abstracts hydrogen (intermolecularly) only slightly more efficiently than the non-nitrated derivative, while it is rather efficiently reduced via electron transfer in the presence of amines. Hydrogen abstraction is efficient for the excited triplet nitrenes, as revealed by an intramolecular reaction.

Full text of DOI:10.1039/a903863i

View Article Online / Journal Homepage / Table of Contents for this issue
The effect of the p-nitro group on the chemistry of phenylnitrene.
A study via intramolecular trapping
Angelo Albini, Gianfranco Bettinetti* and Giovanna Minoli
Dipartimento di Chimica Organica, Università, v. Taramelli 10, 27100 Pavia, Italy.
E-mail: albini@chifis.unipv.it
Received (in Cambridge, UK) 14th May 1999, Accepted 27th September 1999
The photodecomposition of a pyrazolyl substituted p-nitrophenyl azide has been studied as an intramolecular model
for the reactivity of the parent azide, largely used for photochemical labelling. The singlet nitrene is trapped by intra-
molecular cyclization onto the pyrazole nitrogen as well as by the low yield addition of intermolecular nucleophiles
(EtOH, Et2NH). Ring expansion to a didehydroazepine is absent. The triplet nitrene abstracts hydrogen (inter-
molecularly) only slightly more efficiently than the non-nitrated derivative, while it is rather efficiently reduced via
electron transfer in the presence of amines. Hydrogen abstraction is efficient for the excited triplet nitrenes, as
revealed by an intramolecular reaction.
Aryl azides have long been proved to be useful reagents for
photochemical labelling.1,2 Suitably substituted derivatives can
be irradiated at a sufficiently long wavelength, thereby avoiding
damage of the receptor. Aryl nitrenes are generated under these
conditions and have been shown to be both highly reactive and
selective. Their use for labelling and for other applications
has preceded the definition of the mechanism, which is still a
subject of investigation. It is now recognised that four inter-
mediates are involved in the photodecomposition of phenyl
azide and its derivatives, viz. singlet and triplet nitrene as well as
the cyclic isomers, the benzoazirine and didehydroazepine.3–7
However, the poor yield of isolated products obtained in many
cases from the irradiation in solution of aryl azides and the
considerably different sensitivity of the intermediates to the dif-
ferent spectroscopic techniques may hinder the determination
of the main process under a given set of conditions.
facile.16–18 Herein we report a study on the corresponding
5-nitro derivative.
Results
Room temperature photolysis
The photochemistry of 1-(2-azido-5-nitrophenyl)-3,5-dimethyl-
1H-pyrazole (1) was studied in acetonitrile and in ethanol. In
the former solvent, the main product from the decomposition
of a 1 × 10Ϫ4 M solution of 1 was the heterocycle 2 accom-
panied by a small amount of the amine 3 and of the pyrazolo-
quinoxaline 4 (see Scheme 1 and Table 1). When a more
The most commonly used phenyl azides for photochemical
labelling are nitro derivatives and this has led to extensive
investigation of such compounds. Early studies proved that the
lowest state of the nitrene is a triplet8 and investigated the for-
mation of transients in a matrix.9,10 These were complemented
by several spectroscopic and product studies.11–15 The main
conclusions are that triplet reactions, in particular dimerization
to the azo compound, dominate. Hydrogen abstraction is in-
effective, except with amines where an electron transfer path is
involved. Contrary to most phenyl azides, no trapping product
from the didehydroazepine is obtained from the 4-nitro-
and very little from the 3-nitro derivative, possibly due to
the short lifetime of this intermediate.11,12 Likewise, no trapping
of singlet nitrene has been observed, despite the expectation
that the substituent increases the electrophilicity of this
intermediate, except for a low (9%) yield of hydrazine when
irradiating in neat diethylamine.
It is in a way disappointing that useful reagents for labelling
give poor yields of trapping products in solution, and that their
chemistry is dominated by triplet coupling to the azo com-
pound, the typical reaction of stable triplet nitrenes. Several
groups, including our own, have used intramolecular models
for determining the reactivity of nitrenes,16–19 and it may be that
the results obtained from such compounds are indicative
of the chemistry occurring in the complex with the receptor.
Phenyl nitrenes bearing 3,5-dimethylpyrazolyl as a substituent
in position 2 are useful models, since both electrophilic and
radical based intramolecular reactions are expected to be
Scheme 1
concentrated solution (4 × 10Ϫ3 M) was photolysed with a
more powerful lamp the azo derivative 5 became an important
product, but the yield of product 2 was not significantly
diminished.
Carrying out the photodecomposition in ethanol led to
similar results, again with formation of the azo compound only
in the higher concentration experiment, except for the fact that
an additional product formed in low yield was detected, but not
J. Chem. Soc., Perkin Trans. 2, 1999, 2803–2807
This journal is © The Royal Society of Chemistry 1999
2803

View Article Online
Table 1 Products from the irradiation of azide 1 at room temperature
% Yield of the products
Solvent
MeCN
[1]
Lamp Additive
2
3
4
5
6
7
1 × 10Ϫ4
4 × 10Ϫ3
1 × 10Ϫ4
4 × 10Ϫ3
1 × 10Ϫ4
1 × 10Ϫ4
4 × 10Ϫ3
4 × 10Ϫ3
1 × 10Ϫ4
37
41
38
5
32
45
32
4
3
4
38
36
15
12
42
31
6
2
4
a
b
a
27
EtOH, 1 M
b
a
a
b
b
a
MK 9 × 10Ϫ3 M
DEA 0.1 M
EtOH
ca. 5
ca. 5
11
15
MK 9 × 10Ϫ3 M
DEA 0.1 M
36
ca. 5
ca. 1
a Low-pressure mercury arc (15 W) inserted below the cell. b Focused high-pressure mercury arc (200 W).
Table 2 Products from the irradiation of azide 1 in a glassy ethanol
matrix at 90 K
% Yield of the products
Conditions
2
3
5
8
fast heating
20
66
15
30
15
slow heating
tr
2nd irr., >450 nm
50
Fig. 1 Irradiation of azide 1 (1 × 104 M) in glassy ethanol at 90 K.
Inset: spectrum after 8 min irradiation.
isolated. HPLC/mass examination suggested that this was the
O-ethylhydroxylamine 6. This product was not detected in
MeCN containing 1 M ethanol.
Sensitization and trapping experiments
The decomposition of the azide was also performed in the
presence of Michler’s ketone as the sensitizer. As is clear from
Table 1, amine 3 was by far the main product under these
conditions, with only a small amount of heterocycle 2 in
acetonitrile and none of it (or of product 6) in ethanol.
Direct photolysis in the presence of 0.1 M diethylamine
(DEA) in acetonitrile caused a large increase of the yield
of amine 3, the disappearance of quinoxaline 4 and only a
minimal change in the yield of 2. In ethanol the yield of amine
was likewise increased, while the yields of compounds 2 and 6
were little affected and HPLC/mass analysis revealed a trace of
an additional compound formed under these conditions, with
a mass spectrum compatible with the structure of hydrazine 7.
Fig. 2 Evolution at 100 K of the spectrum obtained as in Fig. 1. The
initial spectrum has been obtained after 8 min irradiation at 90 K,
followed by 3 h at 90 K and 1.5 h at 95 K then at 100 K. The following
spectra have been taken after a 30 min interval each. Inset a: final
spectrum. Inset b: amine 3 (1 × 104 M) in glassy ethanol at 90 K.
When the temperature was maintained at 90 K the above
absorption was fairly stable (a 10% decrease in 3 h). Raising
the temperature to 95 K caused a red shift of the longest
wavelength absorption to 578 nm and a further raise at 100 K
led both to a further red shift and to a measurable, though
still slow, decomposition (kobs = 2.4 × 10Ϫ3 minϪ1). A new
spectrum developed with an intensive maximum at 375 nm and
very little absorption in the visible (Fig. 2). As can be seen from
the inset, this was quite close to the spectrum of amine 3 in
ethanol glass, and indeed chemical analysis showed that by far
the main product under these conditions was the amine (except
for a trace of 2, explaining the absorption in the visible).
Photolysis in a glassy matrix
The photochemical decomposition of azide 1 was then per-
formed in ethanol glass at 90 K. New absorption bands
developed in the visible region (Fig. 1) finally giving a spectrum
with maxima at 565 and 316 nm (see the inset). Conservation
of the isosbestic points indicated that no secondary photo-
reactions occurred up to >70% azide conversion. This spectrum
disappeared upon melting the matrix. Analysis of the photolyte
showed that under these conditions the main product was the
azo compound 5 accompanied by the amine 3 (Table 2).
In another experiment the azide was first decomposed by
irradiation at 254 nm in the glass and then a second irradiation
was carried out at λ > 455 nm. Under these conditions the
above spectrum was converted into a new spectrum with
2804
J. Chem. Soc., Perkin Trans. 2, 1999, 2803–2807

View Article Online
of intermolecular reactions must be the short lifetime of the
singlet nitrene, rather than its intrinsic reactivity.
Consistent with this hypothesis, the typical reaction of singlet
phenylnitrene, viz. the rearrangement to the dehydroazepine,
is conspicuously absent with this derivative; should it be
formed, it would be trapped by nucleophiles such as amines
as is consistently observed with other azides,3–7,21 including
pyrazolyl derivatives18b (Scheme 2). Likewise, no aminoazepine
Fig. 3 Irradiation (λ > 455 nm) of the intermediate obtained as in Fig.
1 at 90 K. Each spectrum was taken after 12 min irradiation. Inset a:
after 120 min irradiation and equilibration at 115 K (allowing the
decomposition of the small amount of residual triplet). Inset b: spec-
trum of the HPLC fraction identified as dihydroquinoxaline 8.
maxima at 386, 317 and 294 nm, again maintaining the isosbes-
tic points (Fig. 3). A long irradiation time was required and
after 120 min the conversion of the azide had reached 83% as
judged from the residual absorption at 565 nm. Thawing the
matrix and HPLC analysis of the sample showed that besides
minor amounts of the amine and the azo derivative, a main
product was formed which had a UV spectrum quite similar to
that observed in the matrix after double irradiation (see inset).
This could be identified by HPLC/mass spectroscopic analysis
as the dihydropyrazoloquinoxaline 8. On long standing in solu-
tion (several months), this product underwent spontaneous
oxidation to quinoxaline 4.
Discussion
The presence of a nucleophilic group (the pyrazole nitrogen)
in azide 1 allows for intramolecular trapping of the singlet
state of the corresponding nitrene. In fact, heterocycle 2 is the
main product (yields ranging from 30 to 50%) from the photo-
decomposition of 1 in both MeCN and EtOH.
The success of intramolecular trapping contrasts with the
poor yield of the corresponding intermolecular reaction. Even
in the presence of an electron-withdrawing substituent such as
the nitro group which makes the nitrene a stronger electrophile,
intermolecular trapping remains a minor process. This occurs
with a yield below 10%, viz. in the same order as that observed
with p-nitrophenylnitrene, which, as mentioned in the intro-
duction, gives 9% of the hydrazine in neat DEA (this nitrene
can be trapped by aromatics, however).20 It should be noted
that other substituents allow a high yield of hydrazine to be
obtained by irradiation in the presence of amines. This is the
case for p-cyanophenyl azide,11 where apparently electronic
activation is operating, as well as with some 2,6-disubstituted
phenylnitrenes, apparently due to a steric, not an electronic,
effect since the result is independent from the nature of the
substituent (same result with 2,6-dimethyl and 2,6-difluoro
derivatives).5 With the present nitrene, even a weak nucleophile
such as ethanol is a good enough trap if used as the solvent (not
at a lower concentration, up to 1 M in MeCN). The observed
trapping is unspecific: in 0.1 M DEA in ethanol the main
adduct is hydroxylamine 6 arising from addition of the solvent,
with a much lower amount of hydrazine 7 incorporating the
better nucleophile DEA. Thus, the main cause of the poor yield
Scheme 2
was obtained in a previous study with the non-pyrazole bearing
p-nitrophenylnitrene.12 On the contrary, other electron-with-
drawing substituted (e.g. polyfluoro) phenylnitrenes undergo
ring enlargement as the main process (except, as mentioned
above, with o,oЈ-disubstituted derivatives) and the correspon-
ding aminoazepines are the main products in the presence of
DEA.21
Thus, the simplest hypothesis is that ISC to the triplet state
is faster in this case, and limits all other competing processes
from the singlet except for the highly favoured cyclization to
the heterocycle 2. The ground triplet state of phenylnitrenes
is known to be largely stabilised with respect to the singlet,
but ISC is relatively slow (recently measured as 3 × 106 sϪ1
for parent phenylnitrene).22 The data on the nitro substituted
nitrene suggest that ISC is faster. There is evidence that
polarization increases the rate of ISC in biradicaloid states.23,24
In the present case, this may also be due to the fact that the
triplet state is higher in energy, making ISC over a smaller gap
faster. Support for this suggestion is the fact that the reactivity
J. Chem. Soc., Perkin Trans. 2, 1999, 2803–2807
2805

View Article Online
of the triplet state is increased. In a hydrogen-donating solvent
such as ethanol this is revealed by the significant yield of amine
at room temperature. As one may expect, in an inert solvent and
with a higher absorbed light flux the main reaction from the
triplet state remains dimerization to the azo compound (27% in
MeCN).
However, electronic excitation of the triplet state makes
hydrogen abstraction effective. It has been suggested that the
lowest excited state of triplet arylnitrenes corresponds to a π→n
transition;26 thus, the excited triplet can be likened to an iminyl
radical and is expected to be a much better hydrogen abstractor.
In the matrix such a process obviously takes place intra-
molecularly and leads to the dihydroquinoxaline 8.
Reduction of the nitrene to amine 3 is effective under two
conditions. The first one is triplet photosensitization of the
azide decomposition. As shown in Table 1, Michler’s ketone
(MK, chosen because it allows selective irradiation) sensiti-
zation generates the triplet nitrene and this gives the amine
(presumably via reduction by MK, see below) with a strong
decrease in or elimination of the singlet nitrene products.
The second condition involves electron transfer. The experi-
ments conducted in the presence of DEA show that the yield
of singlet products is marginally affected (except for the
formation of the trace of hydrazine 7) and the yield of amine
3 is greatly increased. This fits with the previous suggestion by
Schuster that the p-nitrophenylnitrene triplet state is a powerful
electron acceptor,12 and strengthens the evidence that a selective
quenching of the triplet state is involved in the formation of the
amine. The electron transfer–proton transfer path is a more
effective mechanism for triplet state reduction than hydrogen
abstraction (compare the yield of 3 in the presence of DEA and
in neat ethanol). This suggests that Michler’s ketone acts both
as sensitizer and as reducing agent and this causes the higher
yield of amine than of azo compound.
The situation changes at low temperature. First of all,
the triplet can be observed in a glassy matrix. The spectrum
measured at 90 K shows the characteristic intense absorption in
the visible (565 nm, log ε ca. 3.4), which is red shifted with
respect to the non-nitrated pyrazolylphenylnitrene (510 nm,
log ε 3.3).18 A similar spectrum has been observed in several
simple phenyl azides,25 although Harder et al. observed very
little visible absorption during the photolysis of p-nitrophenyl
azide in methyltetrahydrofuran glass.15
In conclusion, the data obtained with a model system
demonstrate both singlet and triplet nitrene reactions through
product studies. This reveals the possibility of trapping the
singlet through an intramolecular reaction, while confirming
that intermolecular trapping in solution is a marginal process.
Likewise, nitro substitution only marginally increases the
radical reactivity of the triplet. The utility of nitrophenyl
azides for photochemical labelling of biomolecules is related to
better complexation with the receptors, which would make the
reaction more similar to the present model - rather than to an
intrinsic change in the reactivity of the nitrene.
Experimental
General
Acetonitrile and 95% ethanol were spectroscopic grade
solvents. Azide 1 was prepared and purified as previously
reported.18c Photoproducts 2 to 5 were likewise previously
reported.18c The photoreactions were monitored by HPLC.
A Jasco PU 980 instrument with UV-975 detector was used,
with a 25 cm × 4.6 mm Merck Purospher RP-18 LiChroCART
250-4 column (and a Purospher RP-18 LiChroCART 4-4
precolumn). Various water–acetonitrile mixtures were used
as the eluent. HPLC/mass experiments (on a Finnigan LCQ
instrument) were used for the detection of trace products (see
below).
Room temperature irradiations
As expected from work with other phenylnitrenes,4,6,15,18 only
triplet chemistry takes place under these conditions, since the
thermal barrier for accessing reactions on the singlet surface
cannot be overcome. Heating of the matrix after irradiation
diminishes the viscosity18 and reproduces a situation similar to
that studied by Schuster by laser photolysis of p-nitrophenyl
azide12 with a relatively high triplet nitrene concentration.
Under this condition dimerization to the azo compound is
the main reaction. On the other hand, maintaining the glass at
95–100 K for a long period of time allows hydrogen abstraction
to take place before the matrix softens to a sufficient degree,
and thus coupling of the triplet is unimportant. The reaction
occurs as a pseudo first order process and cleanly leads to
the amine as shown in Fig. 2, obviously through hydrogen
abstraction from the solvent. Harder et al. reported an inter-
mediate spectrum attributed to the iminyl radical in their matrix
study of p-nitrophenyl azide.15 However, in the present case we
have no evidence for the building up of a significant concen-
tration of an intermediate after the nitrene, although this may
escape detection if, as appears to be the case in the study by
Harder, it has a spectrum quite close to that of the amine. The
nitro group somewhat increases the hydrogen atom abstraction
rate, as judged by the fact that under matrix conditions the
amine is practically the only product obtained, while the azo
still competes (amine–azo ratio ca. 2) with the non-nitrated
pyrazolylphenylnitrene under the same conditions.18b
1 × 10Ϫ4 M Solutions of azide 1 in a 1 cm optical path cell
(2 mL) were irradiated by means of a bifilar low-pressure
mercury arc inserted below the cell (Helios Italquartz 15 W).
More concentrated solutions (4 × 10Ϫ3 M) were irradiated by
means of a focused high-pressure mercury arc (Osram HBO
200W/2; in the MK sensitized experiments a cutoff filter
with λtr > 350 nm was inserted). The product distribution as
determined by HPLC is reported in Table 1. Apart from the
previously reported products 2–5, in the experiments in ethanol
HPLC revealed a further product, with a mass spectrum com-
patible with the structure of 1-(2-ethoxyamino-5-nitrophenyl)-
3,5-dimethyl-1H-pyrazole (6). Tr 13.2 min (1 to 1 MeCN–H2O
mixture, 0.5 ml minϪ1). Mass spectrum: m/z 277 (Mϩ ϩ 1),
231 (base peak), 185 (Mϩ Ϫ EtOH, ϪNO2). In the irradiation
in EtOH with 0.1 M DEA, a further trace peak was detected,
with a spectrum compatible with the structure of 1-[2-(2,2-
diethylhydrazino)-5-nitrophenyl]-3,5-dimethyl-1H-pyrazole (7).
Tr 18.5 min (1 to 1 MeCN–H2O mixture, 0.5 ml minϪ1). Mass
spectrum: m/z 304 (Mϩ ϩ 1), 231 (base peak), 185
(Mϩ Ϫ Et2NH, ϪNO2).
Low temperature experiments
1 × 10Ϫ4 M Solutions of the azide 1 in EtOH (2 mL) in a 1 cm
optical path quartz cell with a quartz to glass graded seal were
degassed by means of four freeze–degas–thaw cycles and
sealed. The cell was inserted into an Oxford DN 1704 liquid
nitrogen cryostat fitted with a calibrated ITC4 temperature
controller and placed in a UV-vis Kontron Uvikon 941
spectrophotometer and irradiated from the bottom as above.
Irradiation was discontinued when taking the spectra. In a
typical experiment, the solution was equilibrated for 30 min
at 90 K and then irradiated for 8 min. After this time the
temperature was either gradually raised (see text and Fig. 2)
In contrast to the various intermolecular channels observed
for triplet nitrene both at high and low temperatures, the
seemingly attractive intramolecular hydrogen abstraction
reaction from the pyrazole methyl group is unimportant. The
formation of quinoxaline is only a minor process. Apparently,
the increase in the radical reactivity of the nitrene induced by
the nitro group is not sufficient to make this reaction channel
effective.
2806
J. Chem. Soc., Perkin Trans. 2, 1999, 2803–2807

View Article Online
7 B. Iddon, O. Meth-Cohn, E. F. V. Scriven, H. Suschitzky and
or allowed to quickly reach room temperature (within 30 min
in the latter case). The solution was analyzed by HPLC.
P. T. Gallagher, Angew. Chem., Int. Ed. Engl., 1979, 18, 900.
8 E. Wasserman, Prog. Phys. Org. Chem., 1971, 8, 319.
9 S. P. Efimov, V. A. Smirnov and A. V. Pochinok, Khim. Vysok.
Energ., 1983, 17, 445.
10 V. A. Smirnov and S. B. Brichkin, Chem. Phys. Lett., 1982, 87, 548.
11 R. A. Odum and A. M. Aaronson, J. Am. Chem. Soc., 1969, 91,
5680.
12 T. Y. Liang and G. B. Schuster, J. Am. Chem. Soc., 1987, 109, 7803.
13 M. J. Torres, J. Zayas and M. S. Platz, Tetrahedron Lett., 1986, 27,
791.
14 L. Dähne, J. Bendig and R. Stößer, J. Prakt. Chem., 1992, 334, 707.
15 T. Harder, R. Stößer, P. Wessig and J. Bendig, J. Photochem.
Photobiol., 1997, 103, 105.
16 S. E. Carrol, B. Nay, E. F. V. Scriven, H. Suschitzky and D. R.
Thomas, Tetrahedron Lett., 1977, 18, 3175.
17 A. Albini, G. Bettinetti and G. Minoli, Chem. Lett., 1981, 301.
18 (a) A. Albini, G. Bettinetti and G. Minoli, J. Am. Chem. Soc., 1997,
119, 7308; (b) A. Albini, G. Bettinetti and G. Minoli, J. Am. Chem.
Soc., 1999, 121, 3104; (c) A. Albini, G. Bettinetti and G. Minoli,
Heterocycles., 1995, 40, 597.
19 H. Tomioka, N. Ichikawa and K. Kamatsu, J. Am. Chem. Soc.,
1993, 115, 8621.
20 R. A. Abramovitch, S. R. Challand and E. F. V. Scriven, J. Org.
Chem., 1972, 37, 2705.
21 E. Leyva and R. Sagredo, Tetrahedron, 1998, 54, 7367.
22 N. P. Gritsan, Z. Zhu, C. M. Hadad and M. S. Platz, J. Am. Chem.
Soc., 1999, 121, 1202.
23 J. Michl, J. Am. Chem. Soc., 1996, 118, 3568.
24 F. Kita, W. M. Nau, W. Adam and J. Wirz, J. Am. Chem. Soc., 1995,
117, 8670.
25 E. Leyva, M. S. Platz, G. Persy and J. Wirz, J. Am. Chem. Soc., 1986,
108, 3783.
26 S. J. Kim, T. P. Hamilton and H. F. Schaefer, J. Am. Chem. Soc.,
1992, 114, 5349.
In double irradiation experiments, samples irradiated at 254
nm as above were further irradiated by means of a focused
high-pressure mercury arc (Osram 200 W/2) passed through a
cutoff filter (λtr > 455 nm). The beam reached the cell through a
side opening perpendicular to the analyzing beam. The sample
was then rapidly heated to room temperature and analyzed as
above. In such experiments a new peak was detected in HPLC,
which exhibited a mass spectrum compatible with the structure
of 2-methyl-8-nitro-4,5-dihydropyrazolo[1,5-a]quinoxaline (8).
Tr 12.0 min (1 to 1 MeCN–H2O mixture). Mass spectrum:
m/z 231 (Mϩ ϩ 1), 214 (Mϩ Ϫ NH3), 185 (base peak,
Mϩ Ϫ NO2), 168 (Mϩ Ϫ NH3 Ϫ NO2). Examination of
a
solution of 8 kept in the dark for 12 months showed that this
was completely transformed into the rearomatized derivative 4.
Acknowledgements
Partial support of this work by CNR, Rome, and MURST,
Rome, is gratefully acknowledged. A referee is thanked for
pointing to polar effects on ISC.
References
1 J. R. Knowles, Acc. Chem. Res., 1972, 5, 152.
2 H. Bailey, Photogenerated reagents in Biochemistry and Molecular
Biology, Elsevier, New York, 1983.
3 M. S. Platz, E. Leyva and K. Heider, Org. Photochem., 1991, 11,
367.
4 G. B. Schuster and M. S. Platz, Adv. Photochem., 1992, 17, 69.
5 C. R. Kemnitz, W. L. Karney and W. T. Borden, J. Am. Chem. Soc.,
1998, 120, 3499.
6 M. S. Platz, Acc. Chem. Res., 1995, 28, 487.
Paper 9/03863I
J. Chem. Soc., Perkin Trans. 2, 1999, 2803–2807
2807