Photochemical generation and lifetimes in water of p-aryloxy- and p-alkoxyphenylnitrenium ions

Article abstract of DOI:10.1039/a807567k

This paper describes product and flash photolysis studies following irradiation in aqueous solution of 4X-C₆H₄N₃ [X = MeO (12a), EtO (12b), PrⁱO (12c), Bu^tO (12d), C₆H₅O (12e), 4-MeOC₆H₄O (12f), F, Cl] and 4-methoxy-l naphthyl azide (15). p-Benzoquinone (or 1,4-naphthoquinone) is observed as a product, in yields of 70-90% with 12a-d, 15, 40% with 12e, 26% with 4-F and 15% with 4-Cl. The quinone arises by a pathway whereby the initially-formed singlet arylnitrene is quenched by protonation by a solvent water molecule to form a nitrenium ion. Hydration of this cation at the para position leads through a hemiacetal (or halohydrin) to the quinone imine, whose hydrolysis results in the final quinone product. Three kinetic processes are observed, the nitrenium hydration on the us time scale, the hemiacetal breakdown on the ms time scale, and the imine hydrolysis on the minutes time scale. The nitrenium ions have lifetimes in aqueous solution of 50 ns (4-PhO), 70 ns (4-MeOC₆H₄O), 370 ns (4-MeO), 550 ns (4-EtO), 1.25 μs (4-PrⁱO), 1.56 μs (4-Bu^tO) and 1.35 μs (4-methoxynaphthyl). A nitrenium transient is not observed with the 4-halophenyl azides, probably because the lifetime is too short for detection with ns laser flash photolysis (LFP). The alkoxyphenylnitrenium ions are argued to be better represented as oxocarbocations derived from O-alkylation of the quinone imine. The 4-ethoxyphenylnitrenium ion is not quenched by 0.01 mol dm^-3 2′-deoxyguanosine, so that k₂(dG) is less than 2 × 10⁷ mol ^-1 dm³ s^-1. This contrasts with the 4-biphenylylnitrenium ion, which has a similar solvent reactivity, but reacts with k₂(dG) = 2 × 10⁹ mol^-1 dm³ s^-1. The localization of the positive charge in the alkoxy system is a possible explanation behind this difference.

Full text of DOI:10.1039/a807567k

View Article Online / Journal Homepage / Table of Contents for this issue
Photochemical generation and lifetimes in water of p-aryloxy- and
p-alkoxyphenylnitrenium ions
Pratima Ramlall and Robert A. McClelland*
Department of Chemistry, University of Toronto, Toronto, Ontario M5S 3H6
Received (in Cambridge) 29th September 1998, Accepted 27th November 1998
This paper describes product and flash photolysis studies following irradiation in aqueous solution of 4X–C6H4N3
[X = MeO (12a), EtO (12b), PriO (12c), ButO (12d), C6H5O (12e), 4-MeOC6H4O (12f), F, Cl] and 4-methoxy-1-
naphthyl azide (15). p-Benzoquinone (or 1,4-naphthoquinone) is observed as a product, in yields of 70–90% with
12a–d, 15, 40% with 12e, 26% with 4-F and 15% with 4-Cl. The quinone arises by a pathway whereby the initially-
formed singlet arylnitrene is quenched by protonation by a solvent water molecule to form a nitrenium ion.
Hydration of this cation at the para position leads through a hemiacetal (or halohydrin) to the quinone imine,
whose hydrolysis results in the final quinone product. Three kinetic processes are observed, the nitrenium hydration
on the µs time scale, the hemiacetal breakdown on the ms time scale, and the imine hydrolysis on the minutes time
scale. The nitrenium ions have lifetimes in aqueous solution of 50 ns (4-PhO), 70 ns (4-MeOC6H4O), 370 ns (4-MeO),
550 ns (4-EtO), 1.25 µs (4-PriO), 1.56 µs (4-ButO) and 1.35 µs (4-methoxynaphthyl). A nitrenium transient is not
observed with the 4-halophenyl azides, probably because the lifetime is too short for detection with ns laser flash
photolysis (LFP). The alkoxyphenylnitrenium ions are argued to be better represented as oxocarbocations derived
from O-alkylation of the quinone imine. The 4-ethoxyphenylnitrenium ion is not quenched by 0.01 mol dmϪ3 2Ј-
deoxyguanosine, so that k2(dG) is less than 2 × 107 mol Ϫ1 dm3 sϪ1. This contrasts with the 4-biphenylylnitrenium ion,
which has a similar solvent reactivity, but reacts with k2(dG) = 2 × 109 molϪ1 dm3 sϪ1. The localization of the positive
charge in the alkoxy system is a possible explanation behind this difference.
Arylnitrenium ions are a class of reactive intermediates, first
extensively investigated by Gassman and co-workers,1 that have
seen increased study over the past twenty years.2–7 This has been
in part due to the implication of these species in the metabolism
of carcinogenic arylamines.8 The majority of studies of this
class has been in the context of solvolysis reactions and the like,
where the presence of the nitrenium intermediate is inferred on
the basis of standard mechanistic criteria. Unlike the carbe-
nium class of electrophile however, there have been virtually no
successful studies of nitrenium ions in super-acid solutions,9–11
although highly stabilized examples have been reported upon
electrochemical oxidation of arylamines.12–14 Within the past
few years reports have begun to appear of arylnitrenium ions
being observed with laser flash photolysis (LFP).15–18 This
approach not only shows that these species can exist, but also it
provides rate constants for the various decay reactions that they
can undergo.19
They are however also the conjugate base of a singlet aryl-
nitrenium ion 5, the ground electronic state of most members
of the latter class.17,21–23 Product studies by the groups of
Takeuchi24 and Abramovitch25 have suggested the possibility
that arylnitrenes can be converted to arylnitrenium ions. Within
the past few years such a reaction has been successfully applied
in flash photolysis experiments for the direct observation of
such cations. Examples are the biphenyl-4-ylnitrenium ion 6
(X = H),26,27 the fluoren-2-ylnitrenium ion 7,26,27 a series of
biphenyl-4-ylnitrenium ions 6 bearing substituents in the distal
phenyl ring ranging from 4-methoxy to 4-trifluoromethyl,28 the
imidazol-2-ylnitrenium ion 8,29 a series of 2,3,5,6-tetrafluoro-
phenylnitrenium ions 9,30 and the 2,4,6-tribromophenyl-
nitrenium ion 10.31 In the case of 9 and 10, the addition of a
strong acid was necessary to trap the nitrene, but with 6–8 effi-
cient protonation by solvent water molecules occurred. For the
parent phenylnitrenium ion 11, there is evidence from product
studies,27 and from quenching of singlet phenylnitrene,31,32 that
the cation does form from the nitrene in the presence of acid,
although the cation has yet to be spectrally characterized.
Arylnitrenes 2 are obtained photochemically upon irradi-
ation of aryl azides 1. They are initially formed in a singlet
electronic state, and rapidly undergo reactions such as ring
expansion to a didehydroazepine 4 or intersystem crossing to
the triplet nitrene 3.20
N
X
+
NH
+
NH
+
NH
1
N
+
NH
N3
N
Me
6
7
8
hν
HA
F
F
F
Br
Br
X
X
X
+
NH
+
NH
+
NH
X
Br
1
2
5
F
9
10
11
3
N
N
We have reported in preliminary form that the 4-methoxy-
and 4-ethoxyphenylnitrenium ions 14a and 14b are formed in
aqueous solution upon irradiation of the corresponding azides
12a and 12b.33 In this paper we provide full details of this
work, extending the study to the 4-isopropoxy-, 4-tert-butoxy-,
X
X
3
4
Scheme 1
J. Chem. Soc., Perkin Trans. 2, 1999, 225–232
225

View Article Online
Table 1 Yields of benzoquinone upon irradiation of 4-alkoxy-, 4-aryloxy- and 4-halophenyl azides and rate constants for reactions of 4-alkoxy-
phenylnitrenium ions
4-Subst.
Yield (%)a
kw(µ = 0)b/sϪ1
kw(µ = 0.5)c,d/105 sϪ1
18
kaz(µ = 0.5)d,e/109 molϪ1 dm3 sϪ1
MeO
EtO
90
79
68
87
40
5
4
4
4
3
(2.7 0.2) × 106
(1.8 0.1) × 106
(8.0 0.1) × 105
(6.4 0.1) × 105
(1.9 0.2) × 107
1
5.4 0.2
5.3 0.5
4.0 0.3
4.5 0.3
10.9 0.4
5.3 0.3
4.6 0.1
PriO
ButO
C6H5O
f
f
—
—
65 5g
ArOh
F
—
(1.4 0.1) × 107
—
—
—
—
—
—
f
f
f
i
i
i
26
15
3
2
—
—
i
i
i
Cl
MeONPj
81 4k,l
(7.4 0.1) × 105
5.2 0.1
3.6 0.2
a Percent yield of 1,4-benzoquinone following 254 nm irradiation in pH 4.6 1:1 acetic acid:sodium acetate buffer, 0.004 mol dmϪ3 b First-order rate
.
constant for decay of cation in water alone. c First-order rate constant for decay of cation in solvent without azide. d Solution was pH, 7 1:1
NaH2PO4–Na2HPO4, 0.002 mol dmϪ3, containing 0.5 mol dmϪ3 NaClO4. e Second-order rate constant for quenching of cation by azide ion. f Not
measured. g Yield of phenol. h 4-MeOC6H4O. i No transient observed. j 4-Methoxynaphthyl. k Refers to 1,4-naphthaquinone. l Solvent contained
10% acetonitrile.
4-phenoxy- and 4-(4-methoxyphenoxy)phenyl azides 12c–f, as
well as to a 4-methoxy-1-naphthyl azide 15. In addition to
observing the decay reactions of the nitrenium ion, it has also
proved possible to see two processes that follow, including the
breakdown of a hemiacetal intermediate.
after one hour. Thus, while the quinone is the final product,
there is a long-lived intermediate that is its precursor.
The conversion of the intermediate to the quinone could also
be seen with a diode array spectrometer. With 12a–d, UV spec-
tra recorded immediately after irradiation for 30–60 seconds
had a peak with λmax at 255 nm. Repeat scans showed a change
to the spectrum for p-benzoquinone, a λmax at 245 nm with
about 80% of the intensity of the initial absorbance. This
change was identical for each of 12a–d (and is also seen with
12e, f), to the extent that it even occurred with the same rate
constant in a solution with the same pH and buffer concen-
tration.
1
+
N
NH
N3
H-OH
hν
(1)
OR
OR
OR
This spectral change is identical to one published by Corbett
in a study of the kinetics of hydrolysis of p-benzoquinone
monoimine to benzoquinone.34 Moreover the rate constants
obtained with the azide precursors are within experimental
error the same as ones calculated from equations provided by
Corbett. Thus the long-lived intermediate observed with 12a–f
can be identified as p-benzoquinone monoimine. The similar
intermediate with the naphthalene derivative 15 is presumably
1,4-naphthoquinone monoimine.
12
13
14
R = a, Me, b, Et; c, iPr; d, tBu; e, Ph; f, 4-MeOC6H4
Results
Product analysis
Product analysis was performed with HPLC following 254 nm
irradiation of pH 4–5 aqueous buffers containing ~1 × 10Ϫ4
mol dmϪ3 azide. The final product of the irradiation of 12a–d
is p-benzoquinone. The amount of this material obviously
depended on irradiation time, but the yield calculated on the
basis of azide reacted was, within experimental error, constant.
For example, irradiation of 12d for 15, 30 and 60 seconds con-
sumed respectively 27, 50 and 73% of the starting material, and
gave respectively yields of 80, 93 and 88% 1,4-benzoquinone.
The yields of benzoquinone obtained with the various pre-
cursors are given in Table 1. In the case of 12e, phenol was also
observed in the HPLC; its yield, like that of the quinone, was
independent of irradiation time. We also examined 4-fluoro-
phenyl and 4-chlorophenyl azide; benzoquinone was a product,
although in a yield lower than those with the 4-alkoxy deriv-
atives. The azide 15 gave the naphthalene analog, 1,4-naphtho-
quinone. With 12e and the two 4-halophenyl azides, there were
other peaks in the HPLC chromatogram, but these were not
characterized. With 12a–12d and 15, the only peak observed in
the HPLC other than the azide (and the quinone imine that
is its precursor—see next section) was the quinone. We are
uncertain as to the reason behind our failure to account for
100% of the azide. Irradiation in 100% acetonitrile (performed
with 12b) produced no quinone, with several unidentified peaks
in the HPLC chromatogram. These peaks did not appear in the
chromatogram following irradiation in water.
Fast kinetic phase
A short-lived species is detected with 248 nm LFP of solutions
of 12a–f and 15 in water. The intermediate forms within the 20
ns pulse width of the laser, and decays with first-order kinetics
with lifetimes (1/kw) ranging from 50 ns to 2 µs (see rate con-
stants kw in Table 1). As shown in Fig. 1A, the transient with
12b absorbs strongly near 300 nm, with a shoulder extending
out to 450 nm. This is suggested to represent a single intermedi-
ate, since the decay occurs with the same rate constant through-
out. As seen by the negative ∆OD below ~310 nm, the azide
precursor absorbs, distorting the spectrum of the transient.†
Since the product of the decay at this stage does not absorb
strongly, the true spectrum can be approximated by employing
the change in optical density in the decay, i.e. from the reading
immediately after the laser pulse to the reading at the comple-
tion of the exponential decay. Such spectra for the methoxy and
tert-butoxy systems are shown in Fig. 1C. While the two are
qualitatively similar, there is a slight difference in λmax. In fact
the trend is seen across the series 12a–d, with λmax at 305, 300,
295 and 290 nm for MeO, EtO, PriO and ButO respectively.
The transient spectrum for the 4-phenoxy azide 12e shows a
λmax in the same region, around 295 nm, but there is now weak
absorbance out into the visible (Fig. 1D). The naphthalene
Slow kinetic process
Injecting immediately after irradiation gave very little of the
quinone peak, but there was a large peak at a slightly shorter
retention time. On repeat injection this decreased in intensity,
while the peak for benzoquinone grew, becoming the only peak
† The apparatus measures the OD after the laser pulse minus the OD
at the same wavelength before the pulse. Thus, in regions where the
precursor absorbs, negative ∆ODs are possible and positive ODs do not
represent the actual OD of the intermediate.
226
J. Chem. Soc., Perkin Trans. 2, 1999, 225–232

View Article Online
Fig. 1 Transient spectra obtained on 248 nm LFP of 4-alkoxyphenyl azides in water containing 0.002 mol dmϪ3 acetic acid:0.002 mol dmϪ3 sodium
acetate. A. 4-Ethoxyphenyl azide. B. 4-Methoxy-1-naphthyl azide. C. 4-Methoxyphenyl azide and 4-tert-butoxyphenyl azide. D. 4-Phenoxyphenyl
azide and 4-ethoxyphenyl azide. For A and B, the ‘y’ axis plots the change in optical density; the spectra marked initial were obtained immediately
after the laser pulse, and the spectra marked final at a time corresponding to 10 half-lives of the first-order decay. For C and D the ‘y’ axis plots the
difference in the optical density between the ‘initial’ and ‘final’ reading. The data in these graphs have been normalized so that the value of ‘y’ at the
maximum is 0.1.
derivative 15 shows two quite strong bands decaying with the
same rate constant, a broad absorbance with λmax around 400
nm, and a sharper peak with λmax near 300 nm (Fig. 1B). There
is considerable distortion of the latter band due to the strong
absorbance of the precursor below 350 nm.
effectively plots the quantum yield for formation of the inter-
mediate relative to the quantum yield in 100% aqueous solu-
tion. As can be seen there is a regular decrease with increasing
acetonitrile, until in 100% acetonitrile the intermediate is not
observed.§ The first-order rate constants (Fig. 2B) increase with
increasing acetonitrile content until about 70% acetonitrile, and
then start to decrease.
These transients are quenched by sodium azide, with the
observed rate constants linear in the azide concentration.
Second-order rate constants for this quenching are given in
Table 1.‡ The decay is accelerated in perchloric acid solutions,
with plots of kobs versus [HClO4] linear up to acid concen-
trations of ~0.2 mol dmϪ3, at which point kobs has reached
the apparatus limit of 3–4 × 107 sϪ1. Second-order rate con-
stants kH are 2.2 × 108, 2.1 × 108, 2.0 × 108 and 1.5 × 108 dm3
molϪ1 sϪ1 for MeO, EtO, PriO and ButO respectively (at ionic
strength 1.0 mol dmϪ3 with NaClO4 and 20 ЊC). The transient
from 12b was studied in aqueous NaOH, where a first-order
dependence on hydroxide ion was observed, with a rate con-
Intermediate kinetic phase
When performed with a monitoring wavelength of 260 nm,
these LFP experiments show that there is a large bleaching
(negative ∆OD) at the completion of the decay. Thus, the
species present at this time is much less absorbing than the
precursor azide. However, recording a UV spectrum 30–60
seconds following irradiation shows a slight increase in OD at
260 nm. As discussed above this absorbance is that of the quin-
one imine. The implication though is that the imine is not the
immediate product after the decay of the 300 nm transient, but
that there is some process (or processes) that occurs at inter-
mediate time that results in its formation. This is indeed
observed with a lamp flash photolysis apparatus. A represen-
tative kinetic trace is shown in Fig. 3A. The large bleaching is
observed after the ~100 µs lamp flash; this returns to a slightly
positive ∆OD in an exponential fashion. Such an increase was
observed with all of the azides 12a–f, 15. A detailed kinetic
investigation was carried out with 12a, with rate constants
below pH 3 measured in perchloric acid solutions, and rate
constants from pH 3–5 in formate and acetate buffers. Catalysis
by the buffers was observed. This was not investigated in detail,
with data points being obtained at low buffer concentration
(0.002–0.02 mol dmϪ3) and an extrapolation carried out to
zero buffer concentration. The rate–pH profile combining the
HClO4 and buffer data is shown in Fig. 3B.
stant kOH of 8.8 × 108 dm3 molϪ1 Ϫ1. Experiments with 12b
s
were also carried out in a dilute pH 7 phosphate buffer contain-
ing varying concentrations of 2Ј-deoxyguanosine (dG). How-
ever, kobs remained unchanged at 1.8 0.1 sϪ1 throughout. The
highest concentration of dG that was employed was 0.01 mol
dmϪ3. With the assumption that a 10% increase in kobs would
have been detected, the second-order rate constant for quench-
ing by the nucleoside k2(dG) must be less than 2 × 107 molϪ1
dm3 sϪ1
.
The alkoxy derivatives 12a–d were also examined in solutions
with varying acetonitrile content (Fig. 2). Fig. 2A plots the OD
change that occurs in the exponential decay at the λmax, normal-
ized by dividing by the ∆OD measured in 100% water. This
experiment was performed with the same laser intensity and the
same concentration of azide precursor, so that the ‘y’ axis
‡ The azide rate constants for the 4-aryloxy compounds 12e,f could not
be measured accurately since the decay in the solvent was already near
the limit of the apparatus. It can also be noted that the kaz for 12a and
12b are almost two-fold higher than those of our preliminary report.33
The latter were obtained in an acetate buffer, and there was a failure to
correct for protonation of the nucleophile.
§ There is a small positive ∆OD in the region 300–400 nm in 100%
acetonitrile, but this does not decay over 100 µs. This is likely to also be
present in the other acetonitrile-rich solutions, since there is residual
absorbance at the end of the decay.
J. Chem. Soc., Perkin Trans. 2, 1999, 225–232
227

View Article Online
Fig. 2 Dependence on acetonitrile content of yield of transient (A) and first-order rate constant for decay (B). 4-MeO (closed squares), 4-EtO (open
squares), 4-PriO (closed circles), 4-ButO (open circles).
Fig. 3 (A) Absorbance change at 260 nm observed on flash lamp irradiation of 12a in 0.005 mol dmϪ3 HOAc:0.005 mol dmϪ3 NaOAc. (B) First-
order rate constants for the absorbance increase with 12a at 260 nm, at ionic strength 0.1 mol dmϪ3 (with NaClO4) and 20 1 ЊC. The data above
pH 3 are the values extrapolated to zero buffer concentration. See Discussion section for the origins of the two curves.
to the triplet. The protonation is the predominant pathway,
since the major product is the quinone that is derived from the
nitrenium ion. This protonation must occur with solvent water
Discussion
Formation of nitrenium ion
molecules, since the quinone is observed in the absence of
added acid. Grouping the ring expansion and intersystem cross-
ing together as knit, the ratio kp :knit can be estimated, since the
quinone can not form from the knit pathways. The yields of
quinone are 70–90% for 12a–d, 15, and 40% for 12e (Table 1).
With the assumption that the material unaccounted for is
associated with one of the other pathways, the ratio kp :knit
ranges from 9:1 to 7:3 for 12a–d, 15, and is 4:6 for 12e.
In every case, but particularly with the alkoxy derivatives, the
protonation pathway is very competitive. As implied by the data
in Fig. 2A, adding acetonitrile decreases its contribution.¶
Protonation still competes in 90% acetonitrile, but there is no
nitrenium ion in the absence of water.
Scheme 2 depicts our proposed mechanism, with the three kin-
1
+
NH
N3
N
NH
kp(H2O)
hν
OR
OR
OR
OR
+
knit
Fast
NH
O
O
NH
ring-expanded
didehydroazepine
or
Slow
Inter
triplet nitrene
Identification of nitrenium ion intermediate
HO OR
O
The intermediate observed with LFP at the completion of the
20 ns laser pulse is proposed to be a 4-alkoxy- (or 4-aryloxy)-
phenylnitrenium ion. Several pieces of evidence support this
interpretation.
(i) Studies with solvolytic precursors to the N-acetyl-4-
ethoxyphenylnitrenium ion show that such nitrenium ions react
with water at the 4-position, leading ultimately to p-benzo-
quinone.7,35 Not only is the quinone formed in good to excellent
Scheme 2
etic processes that are observed labelled as ‘Fast’, ‘Inter’ and
‘Slow’. There is of course an even faster kinetic process, the
decay of the singlet nitrene, which was complete well within
the pulse width of our LFP apparatus. This step was observed
using ps LFP in our studies with singlet biphenyl-4-yl- and
fluoren-2-ylnitrenes, whose lifetimes in 20% acetonitrile are 160
and 90 ps respectively.27 The rapid initial reactions of singlet
phenylnitrene have also been recently directly observed.31,32
As shown in Scheme 1, the singlet nitrene has several compet-
ing reaction channels, protonation to form the nitrenium ion,
ring expansion to a didehydroazepine and intersystem crossing
¶ Since the data in Fig. 2A measure the relative quantum yield of
nitrenium formation, this statement is true only if the quantum yield
for nitrene formation from the azide is not changed with solvent
composition. This has been shown to be approximately true in other
systems.27
228
J. Chem. Soc., Perkin Trans. 2, 1999, 225–232

View Article Online
yield from the azide precursors, but there is kinetic evidence for
both the quinone imine and the hemiacetal, the latter being the
immediate product of the nitrenium ion hydration.
CF3
CF3
C
H
H
(ii) Arylnitrenium ions, like carbocations, are quenched by
azide ion, with second-order rate constants at or very near
the diffusion limit.19,28 For biarylylnitrenium ions, this limit is
~5 × 109 dm3 molϪ1 sϪ1 at an ionic strength of 0.5.28 As shown
by the kaz data in Table 1, such behaviour is observed with the
transients in this study.
(iii) The kaz :kw ratio for the N-acetyl-4-ethoxyphenylnitren-
ium ion is 5.4 × 102 molϪ1 dmϪ3, as measured by competition
kinetics in a solvolysis reaction.6,7 Taking kaz as 5 × 109, kw
is 9.2 × 106 sϪ1, a factor of about nine faster than 14b. This
relatively small difference between NH and NAc is typical of
arylnitrenium ions. With the biphenyl-4-yl derivatives, for
example, there is a factor of 3.5 difference between NAc and
NH.18,27
(iv) As shown in Fig. 2A, the intensity of the transient
decreases with increased acetonitrile content. As discussed
above, this behaviour reflects the competition between water
protonation to form the nitrenium ion, and the pathways that
lead to other intermediates.
(v) The effect of substituent on kw takes the expected form.
The two aryloxy derivatives are more reactive than the alkoxy
ones, with 4-methoxyphenoxy slightly more stabilized than
phenoxy. In the alkoxy cations the order is 4-MeO > 4-EtO >
4-PriO > 4-ButO, the same order as seen previously with tri-
alkoxy- and dialkoxy carbocations.36,37
OMe
OMe
+
+
16
17
would be longer-lived. They have about the same lifetime as
a biphenyl-4-ylnitrenium ion (as initially noted by the Novak
group6), and yet 4-phenyl on the σϩ scale is much less cation-
stabilizing than 4-alkoxy. The 4Ј-methoxybiphenyl-4-ylnitren-
ium ion, where there is an additional phenyl ring interposed
between the methoxy group and the formal Nϩ center, has a
lifetime of almost a millisecond,28 three orders of magnitude
longer. A possible explanation is that the positive charge in the
4-alkoxyphenylnitrenium ion is largely localized on the oxy-
gen adjacent to the carbon center where the reaction with water
occurs. The biphenylyl systems on the other hand are much
more delocalized, as seen for example in the effects of substitu-
ents in the phenyl ring that is distant from the nitrogen.28 An
alternative way of putting this is to say that the 4-alkoxyphenyl-
nitrenium ions are reacting as simple oxocarbocations. The
Richard group has suggested that reactions of such cations
have lower intrinsic barriers than ones associated with more
delocalized systems.42
It is also interesting to note that the α,α-bistrifluoromethyl-4-
methoxybenzyl cation has a lifetime in water of 100 ns,43 only
about a factor of four shorter than the 4-methoxyphenyl-
nitrenium ion. Because of the electron withdrawing trifluoro-
methyl groups, the former cation is better written as the struc-
ture 17 with the positive charge largely localized at the oxygen.44
This cation also shows the feature, highly unusual for a benzylic
cation, of reacting at the ring carbon next to the oxygen leading
to a quinone methide by way of a hemiacetal.45 This sequence
is obviously very similar to that seen with the alkoxyphenyl-
nitrenium ions.
Alkyl substituent effect
The quantitative alkyl order observed with the oxocarbocations
was interpreted in terms of a polar effect rather than a steric
one.36,37 The same conclusion is reached here. The polar
constants σ* are 0.00 (Me), Ϫ0.10 (Et), Ϫ0.19 (Pri) and Ϫ0.3
(But). This predicts that there should be an incremental stabil-
ization across the series, and this is roughly what is observed. In
fact, a plot of log kw versus σ* is reasonably linear with a slope
of 2.2. The steric parameters Es are 0.00 (Me), Ϫ0.07 (Et),
Ϫ0.49 (Pri) and Ϫ1.54 (But). These predict the same order, but
they require that there be a bigger difference between OPri and
OEt than between OEt and OMe, and an even bigger difference
between OBut and OPri. This is not observed.||
The reactivity difference that is of relevance to carcino-
genicity is that with 2Ј-deoxyguanosine. The biphenyl-4-yl-
nitrenium ion reacts with this nucleoside with k2(dG) = 2 × 109
dm3 molϪ1 sϪ1, a number suggested to be close to the encounter
limit.46 Even the longer-lived 4Ј-methoxybiphenyl-4-ylnitrenium
ion has k2(dG) = 4 × 107 dm3 molϪ1 sϪ1 46
With each of these
.
Structure and reactivity of 4-alkoxyphenylnitrenium ions
cations there is a significant selectivity towards dG even in the
nucleophilic solvent water. For example, 0.01 mol dmϪ3 dG
would trap almost 90% of the 4-biphenylylnitrenium ion, and
accelerate the rate of decay by almost a factor of 10. This
selectivity is not observed with a 4-alkoxyphenylnitrenium ion.
In fact 0.01 mol dmϪ3 dG fails to show any acceleration of the
Both theory38 and experiment19,29,39,40 point to the idea that
arylnitrenium ions are better represented as iminobenzenium
ions, i.e. carbocations where the C–N bond is essentially a
double bond. In the case of a 4-alkoxyphenylnitrenium ion,
there is the additional resonance contributor shown in Scheme
2, an oxocarbocation or O-alkylated quinone imine structure.
As we commented in our preliminary report,33 the UV spectra
of these cations are consistent with this being a major contri-
butor. In particular the λmax of 290–305 nm is quite different
from those of other nitrenium ions.15–19,27,28 However, it does
approach the 285 nm λmax of C4-protonated anisole 16.41 An
additional feature seen in this study is the small decrease in λmax
from OMe to OBut. This may be associated with an even
greater importance of the oxocarbocation structure as the alkyl
group becomes more electron donating.
decay, and the upper limit on k2(dG) is 2 × 107 molϪ1 dm3 sϪ1
.
This difference may be associated with the localization of
charge in these cations. Arylnitrenium ions react with dG at the
nitrenium nitrogen forming a 8-(arylamino)guanine adduct.46,47
We speculate that in the alkoxyphenyl systems there is just too
little positive charge at this site for the reaction to be initiated.
Halophenyl azides
The 4-fluoro- and 4-chlorophenyl azides give modest yields of
p-benzoquinone on irradiation in water. This product would
arise by the same mechanism as Scheme 1, with the halide as the
leaving group after water addition at C4. The observation of the
quinone thus demonstrates that nitrene protonation is still
occurring. However, LFP with these two revealed only weak
absorbances in the region 300–400 nm, with no decaying transi-
ent characteristic of a nitrenium ion. This is not surprising. As
the above yields show, the actual amount of nitrenium ion that
would be produced is less than that from the others. More
importantly, the lifetimes of these nitrenium ions are likely too
short to permit them to be detected with our apparatus. In fact,
While the alkoxyphenylnitrenium ions do have significant
lifetimes in water, it might have been expected that they
|| The possibility that the tert-butyl system reacts by unimolecular cleav-
age of the But–O bond can be dismissed. This nitrenium ion has exactly
the same dependence on acetonitrile content as that shown by the other
three derivatives (Fig. 2B). Moreover, the intermediate kinetic phase,
the conversion of hemiacetal to the quinone imine, is still observed,
with about the same relative ∆OD as the others. The hemiacetal is not
an intermediate in the unimolecular pathway.
J. Chem. Soc., Perkin Trans. 2, 1999, 225–232
229

View Article Online
the 4-chlorophenylnitrenium ion has a lifetime estimated by the
azide clock method of only 5 ns.7
to addition of water to the nitrenium ion, the conversion of the
hemiacetal to the quinone imine is left as the only reasonable
reaction to account for the intermediate kinetic step. Hemi-
acetal breakdown is characterized by catalysis by Hϩ, OHϪ
and water (i.e. a pH independent reaction), as well as by buffer
acids and bases.49 The buffer catalysis was not investigated in
this work, but a rate–pH profile for one system was obtained
(Fig. 3B). This shows a pronounced hydroxide reaction, a pH
independent region from about pH 3 to pH 2, and the begin-
nings of the Hϩ-catalyzed region from pH 2 to pH 1.
There is a qualitative correlation between the fraction of
an arylnitrene that converts to the arylnitrenium ion by water
protonation and the stability of the ion so produced, at least as
measured by its lifetime. For the series 4-XC6H4- with X = H,
4-Cl, 4-PhO, Ph, 4-alkoxy, the lifetimes of the nitrenium ions
in water are ~0.5 ns,48 ~5 ns,7 50 ns, 300 ns,27 500–2000 ns, and
the percentages of nitrenium ion formed from the nitrene are
5,27 15, 40, ~9027 and 70–90. It is tempting to ascribe the latter
trend to an increase in the rate constant kp of Scheme 1 with
increased nitrenium ion stability. However the constant knit will
also depend on substituent, and this effect is unknown.
Initially we employed the simple equation kH[Hϩ] ϩ ko ϩ
k
OH[OHϪ]. As shown by the more poorly fitting curve in Fig.
3B, the correlation is not perfect. A better fit is obtained recog-
nizing that the imine nitrogen is protonated at the more acidic
pHs of Fig. 3B, so that there are actually two forms of the
hemiacetal, both of which in principle can react with Hϩ, water
and OHϪ [eqn. (3)].
Alkoxynitrenium ions in acid
Quenching of a nitrenium ion by hydroxide ion, as observed
with the one cation examined, is not surprising, since this sim-
ply represents the reaction with a stronger nucleophile. A faster
decay is also observed in acid, where it seems unexpected that a
cation would be quenched. The explanation invokes proton
activation, the conversion of the nitrenium ion to an aniline
dication which reacts at a greater rate [eqn. (2)]. With the
k+H[H+]
+
kH[H+]
ko
kOH[OH-]
NH2
NH
k+
o
k+OH[OH-]
K'a
(3)
HO
OR
HO
OR
+
NH2
NH
This gives rise to the more complicated kinetic expression
[eqn. (4)]. The better fitting curve in Fig. 3B shows the result of
applying this equation. The fitting provides five parameters
2+
+
kw
kw
Ka
(2)
KЈa = 7.0 × 10Ϫ5 (pKЈa = 4.2), kOH = 7.0 × 109 dm3 molϪ1 sϪ1
,
kϩH = 4.1 dm3 molϪ1 sϪ1, (kϩo ϩ KЈakH) = 2.2 sϪ1 and (kϩOHKw ϩ
KЈako) = 5.11 × 10Ϫ4 dm3 mol sϪ1. The last two parameters arise
from two kinetically equivalent mechanisms, and the individual
constants cannot be independently determined.
OR
OR
+
+
assumption that equilibrium is established between the cation
and dication, the observed rate constants are given by the
expression (kwϩ2[Hϩ] ϩ kwϩKa)/([Hϩ] ϩ Ka).
Eqn. (4) does contain a number of parameters, and thus it is
not surprising that it provides a better fit to the data. There are
however features that lend confidence in this model. The rate–
pH profile for this hemiacetal is somewhat unusual in that the
Hϩ-catalyzed limb is just barely apparent even in 0.1 mol dmϪ3
acid. This is typical of hemiacetals that are cationic, or like the
present one, become protonated in acids, since the Hϩ-catalyzed
breakdown of hemiacetals is slowed by electron withdraw-
ing substituents.50,51 The rate constant kOH represents the
hydroxide-catalyzed breakdown of the neutral hemiacetal. The
magnitude of the number suggests that this process may be
occurring by rate-limiting deprotonation of the OH group of
the hemiacetal, as has been observed with other ketone hemi-
acetals.52 Finally the pKЈa of 4.2 is quite consistent with the
structure. The hemiacetal is 0.5 log units more basic than its
quinone imine product.34
ϩ
The constant kw is the rate constant in weakly acidic and
neutral solutions, and is obviously just kw, as given in Table 1.
Moving into stronger acids there is an initial linear dependence
on [Hϩ]. This corresponds to the situation where the reaction
begins to proceed by way of the dication, but the equilibrium
has not yet begun to shift significantly to this species. The
second-order rate constant kH, the slope of a plot of kobs versus
[Hϩ], is equal to kwϩ2/Ka. At even higher acidities, where the
equilibrium begins to shift towards the dication, kobs is expected
to begin to level. With the alkoxyphenylnitrenium ions however,
there is no indication of such levelling up to HClO4 concen-
trations of 0.2 mol dmϪ3, the point where kobs becomes too
fast to be accurately measured with our apparatus. With the
assumption that a 10 percent deviation from linearity would
be detected, Ka must be greater than 2 mol dmϪ3. With kH
around 2 × 108 mol dmϪ3 sϪ1 for all four alkoxy nitrenium ions,
kw2ϩ is greater than 4 × 108 sϪ1
Some levelling was observed with the biphenyl-4-ylnitrenium
.
Conclusion
This study shows that irradiation in water of 4-alkoxyphenyl
azides and 4-aryloxyphenyl azides leads to arylnitrenium ions
through the solvent protonation of the initially formed singlet
arylnitrene. These nitrenium ions have lifetimes in water of
400–1600 ns (alkoxy) and 50–70 ns (aryloxy). Several pieces of
evidence are consistent with their structure being that of oxo-
carbocations—O-alkylated (or arylated) benzoquinone imines.
This localization of charge may be responsible for the reactivity
differences that are observed between 4-alkoxyphenylnitrenium
ions and 4-biphenylylnitrenium ions, cations that are substan-
tially more delocalized. Two kinetic processes are observed
following the hydration of the 4-alkoxyphenylnitrenium ions.
These represent the conversion of a hemiacetal to p-benzo-
quinone imine, followed by the hydrolysis of this species to
p-benzoquinone.
ion, and curve fitting then provided Ka = 0.8 mol dmϪ3, kw
=
2ϩ
6.0 × 107 sϪ1 and kwϩ = 1.1 × 106 sϪ1 (at ionic strength = 1.0).27
ϩ
Thus while the 4-alkoxyphenylnitrenium ions have a similar kw
they are less easily protonated, but once protonated, more
reactive. This difference may also be associated with charge
delocalization. The 4-aminobiphenyl dication has the second
phenyl ring where it can disperse one of the positive charges.
The 4-alkoxyaniline dication will likely localize the positive
charges on its oxygen and nitrogen atoms.
Hemiacetal kinetics
As discussed in the Results section, there can be no doubt
that the slow kinetic process represents p-benzoquinone imine
hydrolyzing to p-benzoquinone. With the fast process being due
kϩH[Hϩ]2 ϩ (kϩo ϩ KЈakH)[Hϩ] ϩ (kϩOHKw ϩ KЈako) ϩ (kOHKЈaKw/[Hϩ])
kobs
=
(4)
[Hϩ] ϩ KЈa
230
J. Chem. Soc., Perkin Trans. 2, 1999, 225–232

View Article Online
azide was a pale yellow liquid with δH (200 MHz, CDCl3,
Experimental
Me4Si) 7.1 (2H, m), 7.4 (7H, m); νmax/cmϪ1 2118; m/z (EI, 70 eV)
211 (Mϩ, 10%), 183 (Mϩ Ϫ N2, 100), 77 (Phϩ, 22). (This com-
pound has been previously reported, but no details of its prep-
aration or characterization were given (see ref. 59).)
The aryl azides were synthesized by diazotization of the aryl
amine, followed by treatment with a large excess of sodium
azide.53,54 Where the anilines were not commercially available,
they were synthesized by hydrogenation of the corresponding
nitrobenzene. In cases where this was not available, it was pre-
pared by reacting p-fluoronitrobenzene with the corresponding
alkoxide or aryloxide.55,56
4-Methoxyphenyl azide, 4-ethoxyphenyl azide, 4-fluorophenyl
azide and 4-chlorophenyl azide were known compounds and
were prepared from the commercially available anilines using
the diazotization procedure described for the isopropoxy
derivative below.53,57,58
4-(4-Methoxyphenoxy)phenyl azide was prepared by the
three step sequence. 4-Methoxyphenol (0.6 g) was deprotonated
with sodium hydride (0.3 g) in 20 cm3 of dry acetonitrile, and
4-fluoronitrobenzene (0.5 cm3) added. After stirring at room
temperature overnight, water was added, and the aqueous solu-
tion washed four times with diethyl ether. The combined ether
layers were washed with aqueous HCl, aqueous bicarbonate,
then dried over MgSO4. After filtration and drying, 4-(4-
methoxyphenyl)nitrobenzene was obtained as an orange solid
by removal of the solvent: δH (200 MHz, d6-DMSO) 3.80 (3H,
s), 6.95 (6H, m), 8.15 (2H, d). This nitrobenzene was converted
to the azide following the standard procedure. 4-(4-Methoxy-
phenyl)phenyl azide was obtained as a light brown solid: δH (200
MHz, CDCl3) 3.80 (3H, s), 6.90 (8H, m); m/z (EI, 70 eV) 241
(Mϩ, 21%), 213 (Mϩ Ϫ N2, 100), 198 (Mϩ Ϫ N2 Ϫ CH3, 18),
182 (Mϩ Ϫ N2 Ϫ OCH3, 21).
4-Methoxy-1-azidonaphthalene was synthesized in the two
step sequence from the commercially available nitro compound.
4-Methoxy-1-aminonaphthalene hydrochloride was isolated as
described above and had δH (200 MHz, d6-DMSO) 4.0 (3H, s),
7.0 (1H, d), 7.7 (3H, m), 7.95 (1H, d), 8.25 (1H, d), 10.6 (3H,
broad s). 4-Methoxy-1-azidonaphthalene was a yellow brown
oil: δH (200 MHz, CDCl3) 4.0 (3H, s), 6.95 (2H, distorted doub-
4-Isopropoxyphenyl azide was prepared by the three step
sequence. p-Fluoronitrobenzene (0.035 mol) was added to a
solution of 4 cm3 of previously dried isopropyl alcohol in
potassium tert-butoxide in tert-butyl alcohol (40 cm3 of solu-
tion of concentration 1 mol dmϪ3). After stirring overnight, the
solvent was removed, water added, and the solution extracted
three times with diethyl ether. The combined ether fractions
were washed with aqueous potassium hydroxide, and water,
then dried over MgSO4, and after filtration, the solvent
removed to give a 76% yield of crude 4-isopropoxynitrobenzene
as a red-brown liquid: δH (200 MHz, CDCl3, Me4Si) 1.35 (6H,
d), 4.65 (1H, sept), 6.90 (2H, d), 8.15 (2H, d).
This material was carried to the next stage without purifi-
cation. The liquid (0.93 g) was dissolved in methanol (25 cm3)
and hydrogenated over 10% palladium on carbon (0.03 g), until
the reaction was complete as judged by TLC. The catalyst was
removed by filtration through superflow celite, anhydrous
hydrochloric acid in diethyl ether was then added, and the sol-
vents removed to give a 96% yield of 4-isopropoxyaniline hydro-
chloride as a light brown solid: δH (200 MHz, D2O) 1.2 (6H, d),
4.60 (1H, sept), 6.98 (2H, d), 7.24 (2H, d), 12.0 (3H, broad s).
The aniline hydrochloride (0.9 g) was dissolved in 2 mol dmϪ3
aqueous HCl, and the solution cooled to 0–5 ЊC. Sodium nitrite
(1.23 g) in 10 cm3 of water was slowly added, maintaining the
temperature below 10 ЊC. After the addition was complete, the
solution was stirred for a further 10 minutes, and then a satur-
ated solution of sodium azide (10 cm3) was added dropwise.
After stirring for a further 10 minutes, the mixture was neutral-
ized with aqueous sodium bicarbonate, followed by extraction
three times with diethyl ether. After drying the ether over
MgSO4, filtration to remove the drying agent, the solvent was
removed to give a red-brown liquid of 4-isopropoxyphenyl
azide (76%). This was purified by column chromatography: δH
(200 MHz, CDCl3, Me4Si) 1.30 (6H, d, J = 8 Hz), 4.45 (1H,
let of doublets), 7.55 (2H, m), 8.05 (1H, m), 8.25 (1H, m); νmax
/
cmϪ1 2118; m/z (EI, 70 eV) 199 (Mϩ, 27), 171 (Mϩ Ϫ N2, 100),
156 (Mϩ Ϫ N2 Ϫ CH3, 55), 140 (Mϩ Ϫ N2 Ϫ OCH3, 29).
Laser flash photolysis experiments involved ca. 20 ns pulses
at 248 nm (60–120 mJ per pulse) from a Lumonics excimer laser
(KrF emission). A pulsed Xenon lamp providing monitoring
light. The sample was placed in a 4 × 1 × 1 cm cuvette, irradi-
ated with the laser on the 4 × 1 face, and monitored perpendicu-
lar so that the path length was 4 cm. The cuvette was replaced
with a fresh solution after each irradiation. After passing
through a monochromator, the signal from the photomultiplier
tube was digitized and sent to a computer for analysis.
Conventional flash photolysis experiments were performed
using an apparatus previously described,60 with the sample
being irradiated with a broad band flash lamp of ca. 100 µs
duration.
Products analyses were performed with a Waters HPLC sys-
tem using a C18 column with 2 cm3 minϪ1 flow rate with a UV–
visible detector set at 254 nm. Elution was carried out in the
programmed mode—50:50 acetonitrile–water isocratic for
4 minutes, a linear gradient to 80:20 acetonitrile–water for the
next 6 minutes, and then isocratic at the latter composition for
4 minutes. The quinone products were identified by comparison
of retention times with those of authentic samples. Quanti-
tative analyses were performed by determining the response
factor of the quinone product and azide precursor at 254 nm,
and correcting the peak areas of the unknown solutions. The
majority of the product studies were performed with air-
saturated solutions, i.e. without specifically removing oxygen.
An early experiment with the 4-ethoxyphenyl azide compared
the products under air, and in an argon saturated solution, and
showed no difference.
sept, J = 8 Hz), 6.85 (4H, distorted doublet of doublets); νmax
/
cmϪ1 2105; m/z (EI, 70 eV) 177 (Mϩ, 14%), 149 (Mϩ Ϫ N2, 26),
107 (Mϩ Ϫ N2 Ϫ C3H7, 100).
4-(tert-Butoxy)phenyl azide was prepared by the same
sequence, omitting the propan-2-ol in the first step. 4-(tert-
Butoxy)nitrobenzene had δH (200 MHz, CDCl3, Me4Si) 1.45
(9H, s), 7.03 (2H, d), 8.14 (2H, d). 4-(tert-Butoxy)aniline was
isolated as the neutral amine and had δH (200 MHz, CDCl3,
Me4Si) 1.25 (9H, s), 3.45 (2H, broad s), 7.55 (2H, d), 6.77 (2H,
d). 4-(tert-Butoxy)phenyl azide was obtained as pale yellow oil
after chromatography and had δH (200 MHz, CDCl3, Me4Si)
1.30 (9H, s), 6.95 (4H, distorted doublet of doublets); νmax/cmϪ1
2100; m/z (EI, 70 eV) 191 (Mϩ, 10%), 135 (Mϩ Ϫ C4H9, 28), 107
(Mϩ Ϫ N2 Ϫ C4H9, 100), 57 (C4H9ϩ, 91).
4-Phenoxyphenyl azide was prepared in the two step sequence
starting with commercially available 4-nitrophenyl phenyl ether.
Hydrogenation as described above gave 4-phenoxyaniline,
isolated as the free amine: δH (200 MHz, d6-DMSO) 4.95 (2H,
broad s), 6.6 (2H, d), 6.8 (4H, m), 7.0 (1H, t), 7.3 (2H, t).
4-Phenoxyaniline hydrochloride was obtained by precipitation
from diethyl ether by addition of HCl in diethyl ether, and this
salt converted to the azide as described above. 4-Phenoxyphenyl
Acknowledgements
The continued financial support of the Natural Sciences and
Engineering Research Council of Canada is gratefully
acknowledged.
References
1 P. G. Gassman, Acc. Chem. Res., 1970, 3, 26.
2 R. A. Abramovitch and R. Jeyaraman, in Azides and Nitrenes;
J. Chem. Soc., Perkin Trans. 2, 1999, 225–232
231

View Article Online
32 N. P. Gritsan, T. Yuzawa and M. S. Platz, J. Am. Chem. Soc., 1997,
Reactivity and Utility, ed. E. F. V. Scriven, Academic Press, 1984,
pp. 297–354.
119, 5059.
3 M. Novak, M. Pelecanou, A. K. Roy, A. F. Andronico, F. M.
Plourde, T. M. Olefirowicz and T. J. Curtin, J. Am. Chem. Soc., 1984,
106, 5623.
4 J. C. Fishbein and R. A. McClelland, J. Am. Chem. Soc., 1987, 109,
2824.
33 P. Sukhai and R. A. McClelland, J. Chem. Soc., Perkin Trans. 2,
1996, 1529.
34 J. F. Corbett, J. Chem. Soc. (B), 1969, 213.
35 M. Novak, M. Pelecanou and L. Pollack, J. Am. Chem. Soc., 1986,
108, 112.
5 M. Novak, M. J. Kahley, E. Eigen, J. D. Helmick and H. Peters,
J. Am. Chem. Soc., 1993, 115, 9453.
6 M. Novak, M. J. Kahley, J. Lin, S. A. Kennedy and L. A. Swanegan,
J. Am. Chem. Soc., 1994, 116, 11626.
36 S. Steenken, J. Buschek and R. A. McClelland, J. Am. Chem. Soc.,
1986, 108, 2808.
37 R. A. McClelland and S. Steenken, J. Am. Chem. Soc., 1988, 110,
5860.
7 M. Novak, M. J. Kahley, J. Lin, S. A. Kennedy and T. G. James,
J. Org. Chem., 1995, 60, 8294.
8 See S. A. Kennedy, M. Novak and B. A. Kolb, J. Am. Chem. Soc.,
1997, 119, 7654.
9 G. A. Olah and D. J. Donovan, J. Org. Chem., 1978, 43, 1743.
10 G. A. Olah, G. K. S. Prakash and M. Arvanaghi, J. Am. Chem. Soc.,
1980, 102, 6640.
38 See C. J. Cramer and D. E. Falvey, Tetrahedron Lett., 1997, 38, 1515.
39 G. Kohnstam, W. A. Petch and D. L. H. Williams, J. Chem. Soc.,
Perkin Trans. 2, 1984, 423.
40 S. Srivastava, J. P. Toscano, R. J. Moran and D. E. Falvey, J. Am.
Chem. Soc., 1997, 119, 11552.
41 A. J. Kresge, Y. Chiang and L. E. Hakka, J. Am. Chem. Soc., 1971,
93, 6167.
11 G. A. Olah, M. Arvanaghi and G. K. S. Prakash, J. Am. Chem. Soc.,
1982, 104, 1628.
12 U. Svanholm and V. D. Parker, J. Am. Chem. Soc., 1974, 96, 1234.
13 D. Serve, J. Am. Chem. Soc., 1975, 97, 432.
14 A. Riecker and B. Reiser, Tetrahedron Lett., 1990, 31, 5013.
15 G. B. Anderson and D. E. Falvey, J. Am. Chem. Soc., 1993, 115,
9870.
42 J. P. Richard, T. L. Amyes, V. Jagannadham, Y.-G. Lee and D. J.
Rice, J. Am. Chem. Soc., 1995, 117, 5198.
43 R. A. McClelland, F. L. Cozens, S. Steenken, T. L. Amyes and J. P.
Richard, J. Chem. Soc., Perkin Trans. 2, 1993, 1717.
44 J. P. Richard, Tetrahedron, 1995, 51, 1535.
45 J. P. Richard, T. L. Amyes, L. Bei and V. Stubblefield, J. Am. Chem.
Soc., 1990, 112, 9513.
16 R. J. Robbins, L. L.-N. Yang, G. B. Anderson and D. E. Falvey,
J. Am. Chem. Soc., 1995, 117, 6544.
17 S. Srivasta and D. E. Falvey, J. Am. Chem. Soc., 1995, 117, 10186.
18 P. A. Davidse, M. J. Kahley, R. A. McClelland and M. Novak,
J. Am. Chem. Soc., 1994, 116, 4513.
19 R. A. McClelland, Tetrahedron, 1996, 52, 6823.
20 G. B. Schuster and M. Platz, Adv. Photochem., 1992, 17, 143.
21 S.-J. Kim, T. P. Hamilton and H. F. Schaefer, J. Am. Chem. Soc.,
1992, 114, 5349.
46 R. A. McClelland, T. A. Gadosy and D. Ren, Can. J. Chem., in the
press.
47 S. A. Kennedy, M. Novak and B. A. Kolb, J. Am. Chem. Soc., 1997,
119, 7654.
48 J. C. Fishbein and R. A. McClelland, Can. J. Chem., 1996, 74, 1321.
49 P. E. Sorensen and W. P. Jencks, J. Am. Chem. Soc., 1987, 109, 4675
and references therein.
50 R. D. Gandour, R. A. McClelland and P. E. Sorensen, Acta Chem.
Scand., 1991, 45, 558.
22 D. A. Hrovat, E. E. Waali and W. T. Borden. J. Am. Chem. Soc.,
1992, 114, 8698.
23 C. J. Cramer, F. J. Dulles and D. E. Falvey, J. Am. Chem. Soc., 1994,
116, 9787.
24 See H. Takeuchi, S. Hirayama, M. Mitani and K. Koyama, J. Chem.
Soc., Perkin Trans. 2, 1986, 611 and references therein.
25 R. A. Abramovitch, A. Hawi, J. A. R. Rodrigues and T. R.
Trombetta, J. Chem. Soc., Chem. Commun., 1986, 283 and
references therein.
26 R. A. McClelland, P. A. Davidse and G. Hadzialic, J. Am. Chem.
Soc., 1995, 117, 4173.
27 R. A. McClelland, M. J. Kahley, P. A. Davidse and G. Hadzialic,
J. Am. Chem. Soc., 1996, 118, 4794.
28 D. Ren and R. A. McClelland, Can. J. Chem., 1998, 76, 78.
29 T. A. Gadosy and R. A. McClelland, J. Am. Chem. Soc., accepted
for publication.
51 R. A. McClelland, K. M. Engell, P. Sukhai and P. E. Sorensen, Can.
J. Chem., 1994, 72, 2333.
52 P. E. Sorensen, K. J. Pedersen, P. R. Pedersen, V. M. Kanaga-
sabapathy and R. A. McClelland, J. Am. Chem. Soc., 1988, 110,
5118.
53 H. Rupe and K. v. Majewski, Chem. Ber., 1900, 33, 3405.
54 P. A. S. Smith and B. B. Brown, J. Am. Chem. Soc., 1951, 73, 2438.
55 K. Bowden and P. N. Green, J. Chem. Soc., 1954, 1795.
56 G. Baddeley, N. H. P. Smith and M. A. Vickars, J. Chem. Soc., 1956,
2455.
57 O. Dimroth, M. Eble and W. Gruhl, Chem. Ber., 1907, 40, 2397,
2399.
58 M. Tanno, S. Sueyoshi and S. Kamiya, Chem. Pharm. Bull., 1982,
3125.
59 N. Almirante, M. Ballabio, G. Bianchetti, A. Cambiaghi and
D. Pocar, J. Chem. Res. (M), 1986, 1230.
30 J. Michalak, H. B. Zhai and M. Platz, J. Phys. Chem., 1996, 100,
60 A. A. Allen, A. J. Kresge, N. P. Schepp and T. T. Tidwell, Can. J.
Chem., 1987, 65, 1719.
14028.
31 R. Born, C. Burda, P. Senn and J. Wirz, J. Am. Chem. Soc., 1997,
119, 5061.
Paper 8/07567K
232
J. Chem. Soc., Perkin Trans. 2, 1999, 225–232