Carbocation-like reactivity patterns in X′-substituted-4-biphenylylnitrenium ions

Article abstract of DOI:10.1139/v97-204

4-Azido-X′-substituted biphenyls (X′ = 4′-MeO, 4′-Me, 4′-F, 3′-Me, 4′-Cl, H, 3′-MeO, 3′-Cl, 4′-CF₃) have been prepared and subjected to 248 nm flash photolysis irradiation in 20:80 acetonitrile:water. Transient X′-substituted 4-biphenylylnitrenium ions 10 (Ar-C₆H₄-N⁺H) are observed, with lifetimes ranging from 0.6 ms (4′-MeO) to 26 ns (4′-CF₃). These cations are quenched by azide ion, with values of k_az ranging from 6 to 10×10⁹ M^-1 s^-1, with the majority in the range (9-10)×10⁹. This near constant k_az provides further evidence that arylnitrenium ions are quenched by azide ion at the diffusion limit. The solvent reactivities, plotted in a single-parameter Hammett plot against σ⁺(X), exhibit a poor correlation, with the points for the para π-electron donors deviating from the correlation line based on the other substituents in the direction of requiring a more negative substituent parameter. The data are more satisfactorily fit to the two-parameter Yukawa-Tsuno equation; the parameter r⁺ obtained in this fit is 2.8. Thus the resonance interaction of the para π-donor X′-substituents with the positive charge of the cation is underestimated by σ⁺, a situation that has previously been observed with benzylic-type carbenium ions. The conclusion is made that, in their reaction with water, 4-biphenylylnitrenium ions behave like benzyl cations bearing two additional stabilizing vinyl groups, i.e., as if they had the structure Ar-C⁺(C-to-C)₂ double bond. The inherent reactivity and the pattern of the aryl substituent effects are in fact similar to those in the carbocation series Ar-C⁺(Ph)₂.

Full text of DOI:10.1139/v97-204

78
Carbocation-like reactivity patterns in
X′-substituted-4-biphenylylnitrenium ions
Daniel Ren and Robert A. McClelland
Abstract: 4-Azido-X′-substituted biphenyls (X′ = 4′-MeO, 4′-Me, 4′-F, 3′-Me, 4′-Cl, H, 3′-MeO, 3′-Cl, 4′-CF3) have been
prepared and subjected to 248 nm flash photolysis irradiation in 20:80 acetonitrile:water. Transient X′-substituted
4-biphenylylnitrenium ions 10 (Ar-C6H4-N+H) are observed, with lifetimes ranging from 0.6 ms (4′-MeO) to 26 ns (4′-CF3).
These cations are quenched by azide ion, with values of kaz ranging from 6 to 10 × 109 M–1 s–1, with the majority in the range
(9–10) × 109. This near constant kaz provides further evidence that arylnitrenium ions are quenched by azide ion at the
diffusion limit. The solvent reactivities, plotted in a single-parameter Hammett plot against σ+(X), exhibit a poor correlation,
with the points for the para π-electron donors deviating from the correlation line based on the other substituents in the
direction of requiring a more negative substituent parameter. The data are more satisfactorily fit to the two-parameter
Yukawa–Tsuno equation; the parameter r+ obtained in this fit is 2.8. Thus the resonance interaction of the para π-donor
X′-substituents with the positive charge of the cation is underestimated by σ+, a situation that has previously been observed
with benzylic-type carbenium ions. The conclusion is made that, in their reaction with water, 4-biphenylylnitrenium ions
behave like benzyl cations bearing two additional stabilizing vinyl groups, i.e., as if they had the structure Ar-C+(CTC)2. The
inherent reactivity and the pattern of the aryl substituent effects are in fact similar to those in the carbocation series Ar-C+(Ph)2.
Key words: nitrenium, nitrene, aryl azide, photochemistry, lifetime.
Résumé : On a préparé des 4-azidobiphényles substitués par des groupes X′ (X′ = 4′-MeO, 4′-Me, 4′-F, 3′-M3, 4′-Cl, H,
3′-MeO, 3′-Cl, 4′CF3) et on les a soumis à une irradiation de photolyse éclair à 248 nm, en solution dans un mélange 20 : 80
d’acétonitrile : eau. On a observé la présence des ions 4-biphénylylnitrénium substitués par les groupes X′ (10)
(Ar-C6H4-N+H) dont les temps de vie varient de 0,6 ms (4′-MeO) à 26 ns (4′-CF3). On a piégé ces cations par l’ion azoture,
avec des valeurs de kaz allant de 6 à 10 × 109 M–1 s–1 alors que la majorité se trouve entre 9 et 10 × 109 M–1 s–1. Cette valeur
pratiquement constante de kaz suggère une fois de plus que les ions arylnitrénium sont piégés par l’ion azoture à la limite de la
diffusion. La courbe des réactivités de solvant avec un seul paramètre de Hammett σ+(X) donne une mauvaise corrélation
alors que les points pour les substituants para donneurs d’électrons π dévient de la droite de corrélation établie à l’aide des
autres substituants; le sens de cette déviation suggère l’utilisation d’un paramètre de substituant plus négatif. Les données
conduisent à une meilleure corrélation avec l’équation à deux paramètres de Yukawa–Tsuno; le paramètre r+ ainsi obtenu est
de 2,8. L’interaction de résonance entre les substituants X′ para donneurs d’électrons π et la charge positive du cation est
donc sous-évaluée par σ+; cette situation avait été observée antérieurement avec des ions carbénium de type benzylique. On en
conclut donc que, lors de leur réaction avec l’eau, les ions 4-biphénylnitrénium se comportent comme des cations benzyliques
portant deux groupes vinyliques additionnels, comme s’ils avaient une structure du type Ar-C+(CTC)2. L’activité inhérente et
le patron des effets des substituants aryles sont en fait semblables à ceux observés dans la série des carbocations Ar-C+(Ph)2.
Mots clés : nitrénium, nitrène, azidobenzènes, photochimie, temps de vie.
[Traduit par la rédaction]
We recently reported the direct observation of the 4-
Excellent evidence exists pointing to these nitrenium
ions as the reactive electrophiles responsible for DNA bind-
ing of the carcinogens 4-aminobiphenyl and 4-(acetylamino)
biphenyl (4, 5). The flash photolysis experiments have pro-
vided direct information regarding their lifetimes in water, and
their reactivities with added nucleophiles (6).
Like benzyl cations, additional resonance contributors for
1 can be written with the positive charge at the ortho and para
positions, the latter as illustrated in Scheme 1 by structure 1a.
There is in fact experimental evidence that these carbenium
ion contributors do indeed make a significant contribution.
(i) The cations 1 react with water at the ring carbon para to
the nitrogen (1, 2, 7) to form as the initial product the cyclo-
hexadienone imine 5. In the case of the parent nitrenium ion
(R = H) the imine is simply hydrolyzed to give the ketone 5a
(1, 7), while with the N-acyl derivative there are also
biphenylylnitrenium ion (1, R = H) (1) and its N-acetyl analog
(1, R = Ac) (2) in water using the technique of laser flash
photolysis. The N-acetyl derivative was obtained upon pho-
tolysis of N-chloro and N-sulfatoxy precursors 2, and the N-H
derivative upon irradiation of the azide 3, the cation forming
via rapid solvent protonation of the singlet nitrene 4 formed
by nitrogen extrusion (3).
Received June 13, 1997.
D. Ren and R.A. McClelland.1 Department of Chemistry,
University of Toronto, Toronto, ON M5A 3H6, Canada.
1
Author to whom correspondence may be addressed.
Telephone: (416) 978-3592. Fax: (416) 978-3592. E-mail:
rmcclell@alchemy.chem.utoronto.ca
Can. J. Chem. 76: 78–84 (1998)
© 1998 NRC Canada

79
Ren and McClelland
Scheme 1.
rearrangements to restore aromaticity (7). Also indicative of
the delocalization of the positive charge to the ring is the very
small difference in reactivity between the NH and N-acetyl
cations, a factor of only three ((1, 7, 8). In other words, sub-
stituting the nitrogen with an electronegative group has little
effect on reactivity, as if there was little positive charge local-
ized on this nitrogen.
(8, R1 = Et, R2 = H). Substitution at the para carbon (8,
R1 = H, R2 = Me) resulted in a pronounced rate acceleration.
We present in this paper an examination of the lifetimes of
4-biphenylylnitrenium ions 10 bearing substituents in the
phenyl ring distant from the formal nitrenium ion center. These
cations can be generated from the appropriate azide precursor,
in the same manner as the parent (1). Pronounced substituent
effects are observed. Our analysis suggests that the substi-
tuents in this ring interact with the positive charge much as
they do in benzylic carbocations, in other words, as if the posi-
tive charge was placed at the para carbon as in structure 10a.
(ii) Novak et al. obtained aqueous lifetimes of a series of
p-substituted phenylnitrenium ions 6 using the “azide-clock”
approach (9). These lifetimes were found to correlate very
poorly with the σ+ constant of the para substituent. This lack
of correlation is seen also with carbenium ions (see later).
However, the plots for the nitrenium ions are highly scattered,
indicating that the para substituents interact with the positive
charge in a way that is very different from the interaction in
an arylcarbenium ion.
(iii) The kinetic data for the decay of 1, R = H, in acid
solutions show that the cation equilibrates with a conjugate
acid, presumably the dication 7. The acidity constant Ka for 7
calculated from the kinetics is 0.8 (2). This is not inconsistent
with a structure for the cation such as 1a, an imine whose
basicity is reduced because of a nearby positive charge.
Results and discussion
Preparation of precursors
(iv) Williams and co-workers (10) reached similar conclu-
sions about the importance of carbenium resonance contribu-
tors through the comparison of the rate constants for formation
of nitrenium ions from hydroxylamines 8 under acidic condi-
tions. They found little difference between the parent system
(8, R1 = H = R2) and the system with an N-alkyl substituent
The substituents X chosen for this study are given in Table 1.
The precursor azides 9 were prepared as shown in Scheme 2.
The key step, the formation of the biphenyl skeleton, was
achieved by palladium-catalyzed cross-coupling of p-bro-
moaniline and an appropriately substituted phenylboronic acid
(11). The latter was prepared by the reaction of the appropriate
© 1998 NRC Canada

80
Can. J. Chem. Vol. 76, 1998
Scheme 2.
Table 1. Rate constants at 20°C and absorption maxima for
X′-substituted 4-biphenylylnitrenium ions in 20% acetonitrile:water.
Substituent
kw (s–1)a
kaz(M–1 s–1)b
λ
max(nm)
4′-Methoxy
4′-Methyl
4′-Fluoro
3′-Methyl
3′-Methoxy
4′-Chloro
1.58 × 103
2.74 × 105
1.32 × 106
1.50 × 106
2.46 × 106
2.53 × 106
2.70 × 106 c
1.30 × 107
3.8 × 107
6.1 × 109
9.2 × 109
9.4 × 109
9.2 × 109
7.9 × 109
10.0 × 109
9.6 × 109
10.2 × 109
500
490
470
470
475
480
465
475
None
3′-Chloro
4′-Trifluoromethyl
Not measuredd Not measuredd
a First-order rate constant for decay in the solvent alone.
b Second-order rate constant for quenching by sodium azide.
c 2.84 × 106 s–1 (ref. 2).
d Because of very rapid decay in solvent alone.
400 nm represents some species forming in the competing re-
action (or reactions), possibly a triplet nitrene or a ring-ex-
panded 1,2-didehydroazepine 13 (15).
Absorption maxima
As seen in Table 1, the substituents do have a small effect on
the λmax of the nitrenium ions 10, λmax in general increasing
with greater electron donation. This trend is similar to that
observed in benzylic-type cations. Thus, for example, the λmax
(in acetonitrile) for diarylmethyl cations 4-X-C6H4C+HPh are
425 nm (X = CF3), 435 (H), 436 (F), 456 (Cl), 450 (Me), and
455 (MeO) (16). Similarly, cumyl cations (4-X-
C6H4C+(CH3)2) in (CF3)2CHOH have λmax at 325 nm (X = H),
340 (4-Me), and 360 (4-MeO) (17). While it is risky to base
structural conclusions on absorption maxima, the similar be-
haviour with 10 is at least consistent with the idea that there is
a significant benzylic-type interaction between the ring bear-
ing the substituent and the positive charge.
Grignard reagent with trimethyl borate, followed by hydrolysis
(12). The 4-aminobiaryls were converted to the azides 9 by
diazotization, followed by the addition of sodium azide (13).
Flash photolysis experiments
Flash photolysis experiments were carried out with laser irra-
diation at 248 nm in an aqueous solution containing 20% ace-
tonitrile. As illustrated by the four representative spectra in
Fig. 1, the substituted biphenyl azides show behaviour qualita-
tively similar to that previously seen with the parent 4-azido-
biphenyl (1). This consists of a relatively intense transient with
Also worth noting is the λmax at 502 nm of the persistent
biphenylylnitrenium ion 11 obtained by anodic oxidation of
the corresponding aniline derivative in acetonitrile (18). This
cation is obviously analogous to 10 except for the additional
tert-butyl groups ortho to the formal nitrenium ion center.
While the effect of these groups is unknown, the λmax for 11
clearly falls within the same pattern that we observe with 10.
One interesting feature with 11 is that the tert-butyl groups
λ
max in the region from 460 to 500 nm. This decays with first-
order kinetics, with rate constants given in Table 1. At the end
of the decay there is little residual absorbance above 400 nm.
Also present is a second peak in the region 300–350 nm that
does not decay over 100 µs.
Based upon the similarity with the parent, the transient
above 400 nm can be attributed to the appropriate arylni-
trenium ion 10. This assignment is corroborated by the dem-
onstration that these transients are effectively quenched by
azide ion, indeed with second-order rate constants similar to
the value obtained for the parent (Table 1). Quenching by az-
ide ion is one criterion that points to a cationic intermediate
(14). In fact, in the case of the parent 4-biphenylylnitrenium
ion, there is excellent agreement between the kaz:kw ratio ob-
tained with flash photolysis and the value determined by prod-
uct analysis for a ground state solvolysis reaction (7).
1
exhibit nonequivalent signals in the H NMR spectrum (18).
The identity of the non-decaying transient below 400 nm is
uncertain. This absorbance is also seen with the parent system,
where product analysis reveals that there is not a 100% yield
of nitrenium ion on irradiation in 20% acetonitrile (1). The
latter observation indicates that the protonation forming the
nitrenium ion (as shown in Scheme 1) is not the only fate of
the singlet nitrene. We assume that the absorbance below
© 1998 NRC Canada

81
Ren and McClelland
Fig. 1. Transient spectra following 248 nm irradiation of 50 µM solutions of 3′- or 4′-substituted 4-azidobiphenyls 10 in 20% acetonitrile.
Spectra were obtained from the optical density change immediately after completion of the laser pulse.
This clearly points to a nonlinear -N+-H moiety, whose geome-
try is fixed on the NMR time scale by extensive charge delo-
calization, such as shown in structure 11a.
Azide has also been employed to “clock” reactivities of
arylnitrenium ions (6–9, 23). On the basis of the absolute kaz
values for 4-biphenylyl- and 2-fluorenylnitrenium ions, the reac-
tion with azide has indeed been suggested to be diffusion-limited
(1, 2). The data in Table 1 now show that this is the case.
Despite a change of over four orders of magnitude in the sol-
vent rate constants, the values of kaz all lie within the range
6–10 × 109 M–1 s–1, with the majority in the range 9–10 × 109
M–1 s–1. The limit, ~1 × 1010 M–1 s–1, is slightly higher than
those of the di- and triarylmethyl cations. The rate constant
calculated from the Debye–Smoluchowski equation for a dif-
fusion-controlled cation–anion encounter in this solvent is
~1.5 × 1010 M–1 s–1 (22). The smaller observed limits can be
explained through a model where there are nonproductive en-
counters in which the anion and cation cannot actually com-
bine (22). The slightly higher limit with the arylnitrenium ion
would suggest that there are fewer of these nonproductive en-
counters with this system.
Rate constants for azide trapping
Azide ion has seen extensive use as a “clock” for obtaining
absolute rate constants from selectivities measured by
competition kinetics. The assumption is made that this nucleo-
phile reacts with short-lived cations at a constant, diffusion-
limited, rate constant, with assignment of a value of 5 × 109
M–1 s–1 for that limit. This approach was originally applied to
reactivities of carbenium ions (19, 20), and still sees consider-
able use in this regard (21). Laser flash photolysis has provided
direct evidence that the assumption is valid, through the obser-
vation that rate constants kaz for series of structurally related
carbocations become independent of substituent (22). The lim-
its are typical of encounter-controlled reactions, although in a
given solvent there is a slight dependence on structural type.
Thus, diarylmethyl cations in 20% acetonitrile have a limiting
kaz of ~7 × 109 M–1 s–1, while with triarylmethyl cations the
limit would be ~5 × 109 M–1 s–1 (22).
It can be noted that the limit of 1 × 1010 M–1 s–1 obtained
from the data in Table 1 refers to solutions that have effec-
tively zero ionic strength. Addition of salts can have significant
© 1998 NRC Canada

82
Can. J. Chem. Vol. 76, 1998
Fig. 2. Linear free energy correlations. (a) A single-parameter
correlation of log kw versus the Hammett σ+ substituent constant.
The line drawn in this figure (slope = 2.0) is the least-squares line
calculated for the solid squares, representing the 4′-CF3, 3′-Cl,
3′-MeO, 3′-Me, and unsubstituted cations. (b) Plots log kw versus
σ + r+(σ+ – σ), where the value of r+ = 2.8 has been obtained from
application of the two-parameter Yukawa–Tsuno equation (25).
The line drawn in this figure (slope = 1.8) is the least-squares line
for all the points. Values of σ and σ+ are taken from ref. 26.
Table 2. Parameters from two-parameter analysis of rate constants
for solvent addition to benzylic-type cations 12–14a and to
biphenylylnitrenium ions 10.
ρ
+2.7
+6.1
2.3
+1.3
+4.7
3.6
+1.3
+2.8
2.2
+1.8
+5.0
2.8
ρ+ = r+ ρ
r+
ks(X = H) (s–1) ~1 × 1011 b
1.5 × 105
1.5 × 107
2.7 × 106
aData for 12 refer to 50:50 2,2,2-trifluoroethanol:water and are from ref.
20, using a value for X = 4-Me2N from ref. 28. Data for 13 refer to 1:2
acetonitrile:water and are from ref. 27. Data for 14 refer to 1:4
acetonitrile:water and are from ref. 28.
bEstimate (ref. 20).
effects on the azide reactivity (24). In 20% acetonitrile con-
taining 0.5 M NaClO4, the limiting rate for the biphenylylni-
trenium ions is ~5 × 109 M–1 s–1 (1).
effectively a measure of the resonance effect of the substi-
tuents on the reaction (20). This comparison shows very
clearly that substituents in 10 interact with the positive charge
very much as if this cation were a benzylic cation, i.e., as if
the structure were 10a. This conclusion is particularly evident
when the “inherent” reactivities, the rate constant for the de-
rivative with X = H, are considered. The least stable and most
reactive cation is phenethyl, which has only one other stabiliz-
ing substituent, a methyl group, on the C+ center. This system
has the largest values of ρ and ρ+ , indicating the greatest
interaction with the aryl ring. The most stable and least reac-
tive cation is triphenylmethyl, where there are two additional
stabilizing phenyl groups on the C+ center. As expected, the
series ArC+Ph2 has smaller values of ρ and ρ+ since there is
less demand for stabilization by substituents in the one ring.
The biphenylylnitrenium ions, written as 10a, can be viewed
as benzylic cations with two stabilizing vinyl substituents. The
reactivity of this system lies intermediate between phenethyl
and triphenylmethyl, albeit closer to the latter. The substituent
effects are also intermediate, again closer to triphenylmethyl.
The fluorenyl cations 13 are not considered in these com-
parisons, since the aryl ring is significantly twisted from co-
planarity (28), attenuating the substituent effects. In this
respect it is interesting that the biphenylylnitrenium ions do
show such a strong resonance interaction with the substituents
in the distal ring. Biphenyls would normally be regarded as
being significantly twisted but this clearly is not the case for
the nitrenium ions. For a substituent such as 4′-methoxy to
have such a strong stabilizing effect, the two rings must be
fairly close to being coplanar. This is obviously part of the
reason that 4-biphenylylnitrenium ions can be formed so read-
ily both thermally and photochemically.
Solvent reactivities
Despite the fact that the change is taking place in the phenyl
ring remote from the formal N+ center, substituents have a
pronounced effect on the lifetime of 10. From the least reac-
tive, 4′-methoxy, to the most reactive, 4′-trifluoromethyl, there
is a change of over four orders of magnitude. In attempting a
quantitative correlation, we initially constructed a single-pa-
rameter Hammett plot, employing σ+ as the substituent con-
stant. As shown in Fig. 2a, this correlation is poor, with the
points for the para π-electron donors, especially 4′-methoxy,
deviating significantly from the correlation line based on the
other substituents. Moreover, the deviation of these points is in
the unusual direction of requiring a more negative substituent
parameter in order for the point to fit on the line. In other
words, the σ+ constant is underestimating the effect of the
substituent.
This type of behaviour has previously been observed (20,
27, 28) for benzylic-type cations 12–14 (see Table 2). The data
are more satisfactorily fit to the two-parameter Yuk-
awa–Tsuno equation (eq. [4]),
N
log(kw ⁄ kw) = ρ [ σ + r+ ( σ+ − σ)]
[4]
N
where kw is the rate constant for the unsubstituted compound, ρ has
its normal meaning, and r+ is a scaling parameter that effec-
tively provides substituent parameters other than σ+ and σ. The
r+ parameter was intended to lie between zero and one, repre-
senting substituent interactions intermediate between those de-
scribed by σ and σ+. As shown in Table 2, values significantly
greater than one are required for the cation water reactivities.
This is also true for 10, where r+ = 2.8 provides the best fit for
the experimental data. This fit is illustrated in Fig. 2b; as
shown in this figure, the methoxy substituent in 10 behaves as
if it has a “σ” substituent constant of –1.7.
Conclusions
Table 2 provides a comparison of the ρ and r+ parameters
obtained with 10 and those previously found with 12–14. Also
listed is the ρ+ parameter, the product of ρ and r+, which is
The qualitative observation of very significant substituent ef-
fects on the lifetime and the quantitative comparisons just dis-
cussed provide further experimental support for the idea that
© 1998 NRC Canada

83
Ren and McClelland
arylnitrenium ions are better regarded as carbocations, with a
significant portion of the positive charge at the ring carbons.
The 4-biphenylylnitrenium ions, in their reaction with water,
behave very much like benzyl cations bearing two additional
stabilizing vinyl groups, i.e., Ar-C+(CTC)2. Their inherent
reactivity and the pattern of their aryl substituent effects are
very similar to those in the carbocation series Ar-C+(Ph)2.
d, J = 8.5 Hz), 6.95 (2H, d, J = 8.5 Hz), 3.83 (3H, s);
HRMS m/z, calcd. for C13H11N3O: 225.0902; found:
225.0891; calcd. for C13H11NO (M –N2): 197.0841; found:
197.0837.
4-Azido-4-methylbiphenyl: 1H NMR (500 MHz, CDCl3), δ:
7.54 (2H, d, J = 8.5 Hz), 7.44 (2H, d, J = 8.5 Hz), 7.23 (2H, d,
J = 8.5 Hz), 7.06 (2H, d, J = 8.5 Hz), 2.37 (3H, s); HRMS m/z,
calcd. for C13H11N3: 209.0953; found: 209.0951.
Experimental section
4-Azido-4′-fluorobiphenyl: 1H NMR (500 MHz, CDCl3), δ:
7.51–7.47 (4H, m), 7.12–7.06 (4H, m); HRMS m/z, calcd. for
C12H8N3F: 213.0702; found: 213.0694.
Flash photolysis experiments were carried out in the normal
manner with ca. 20 ns pulses at 248 nm (ca. 60 mJ per pulse)
from a Lumonics excimer laser (KrF emission). Solutions of
the 4-azidobiphenyl precursors in 20% acetonitrile were
20–50 µM, and the experiments were carried out under air.
The rate constants kaz were calculated as the slope of the plot
of kdecay versus [NaN3], for five or six concentrations of sodium
azide over the range 0–1 mM. Rate constants for the 4′-
methoxy compound were determined with both the laser flash
photolysis apparatus where the rate is at about the longest time
limit possible, and with a conventional flash photolysis appa-
ratus (29). Excellent agreement was observed between the two.
4-Azido-4′-chlorobiphenyl: 1H NMR (500 MHz, CDCl3), δ:
7.52 (2H, d, J = 8.5 Hz), 7.46 (2H, d, J = 8.5 Hz), 7.38 (2H, d,
J = 8.5 Hz), 7.08 (2H, d, J = 8.5 Hz); HRMS m/z, calcd. for
C12H8N3Cl: 229.0407; found: 229.0409.
4-Azido-4′-trifluoromethylbiphenyl: 1H NMR (500 MHz,
CDCl3), δ: 7.67 (2H, d, J = 8.5 Hz), 7.64 (2H, d, J = 8.5 Hz),
7.57 (2H, d, J = 8.5 Hz), 7.11 (2H, d, J = 8.5 Hz); HRMS m/z,
calcd. for C13H8N3F3: 263.0670; found: 263.0668.
Synthesis. General coupling procedure
4-Azido-3′-chlorobiphenyl: 1H NMR (500 MHz, CDCl3), δ:
7.54–7.52 (3H, m), 7.42–7.4 (1H, m), 7.334 (1H, t, J =
8.0 Hz), 7.31–7.29 (1H, m), 7.08 (2H, d, J = 8.5 Hz);
HRMS m/z, calcd. for C12H8N3Cl: 229.0407; found:
229.0411.
A 50 mL flask was charged under argon with 0.35 g of
Pd(PPh3)4 (0.3 mmol), 20 mL of benzene, 1.72 g of p-bro-
moaniline (10 mmol), and 10 mL of 2 M aqueous Na2CO3
solution, and then 10 mmol of arylboronic acid in a minimum
amount of 95% ethanol was added. The mixture was refluxed
overnight under vigorous stirring. After the reaction was com-
pleted, the product was extracted with ether, washed with a
saturated NaCl solution, dried over NaSO4, and the solvent
was removed. The crude product was further purified by col-
umn chromatography on silica gel, eluting with ethyl acetate –
hexanes, and was recrystallized in aqueous ethanol.
4-Azido-3′-methylbiphenyl: 1H NMR (500 MHz, CDCl3), δ:
7.55 (2H, d, J = 8.5 Hz), 7.36–7.3 (3H, m), 7.16–7.14 (1H, m),
7.07 (2H, d, J = 8.5 Hz); HRMS m/z, calcd. for
C12H8N3Cl: 209.0953; found: 209.0957.
4-Azido-3′-methoxybiphenyl: 1H NMR (500 MHz, CDCl3),
δ: 7.59 (2H, d, J = 8.5 Hz), 7.37 (1H, t, J = 8.0 Hz), 7.18–7.08
(3H, m), 6.94–6.90 (1H, m), 3.88 (3H, s); HRMS m/z, calcd.
for C13H11N3O: 225.0902; found: 225.0891; calcd. for
C13H11NO (M–N2): 197.0841; found:197.0897.
General procedure for the syntheses of biaryl azides
The 4-aminobiphenyl (10 mmol) was dissolved in 40 mL of
glacial acetic acid containing 10 mL of concentrated sulfuric
acid. The solution was cooled to below 5°C in an ice bath and
diazotized with a solution of 0.76 g of NaNO2 (11 mmol) in
7 mL of distilled water. After 1 h of stirring, 100 mL of ice-
water was added. Enough urea was added to destroy the excess
nitrous acid, and 0.5 g of Norite was then added. The cold
suspension was stirred for 15 min and was then rapidly filtered
into a flask immersed in an ice bath. The clear, yellow filtrate
was treated with 1.34 g of NaN3 (20 mmol) in 10 mL of water
at 0°C. Nitrogen was immediately evolved, the solution be-
came turbid, and a light tan precipitate formed in a few min-
utes. The mixture was kept in the ice-bath and stirred for 1 h
after the addition of NaN3 was completed, and was then al-
lowed to warm up to room temperature and stand overnight.
The precipitate was filtered off, and washed first with 10%
Na2CO3 and then with water. The crude product was purified
by column chromatography on silica gel eluted with ethyl ace-
tate – hexanes and was recrystallized in hexanes.
Acknowledgement
Financial support from the Natural Sciences and Engineering
Research Council of Canada is gratefully acknowledged.
References
1. R.A. McClelland, M.J. Kahley, P.A. Davidse, and G. Hadzialic.
J. Am. Chem. Soc. 118, 4794 (1996).
2. P.A. Davidse, M.J. Kahley, R.A. McClelland, and M. Novak. J.
Am. Chem. Soc. 116, 4513 (1994).
3. G.B. Schuster and M.S. Platz. Adv. Photochem. 17, 143 (1992).
4. A. Dipple. Carcinogenesis, 16, 437 (1995).
5. M. Novak and S. Kennedy. J. Am. Chem. Soc. 117, 574 (1995).
6. R.A. McClelland, M.J. Kahley, and P.A. Davidse. J. Phys. Org.
Chem. 9, 355 (1996).
7. M. Novak, M.J. Kahley, E. Eiger, J.S. Helmick, and H.E. Peters.
J. Am. Chem. Soc. 115, 9453 (1993).
8. M. Novak, M.J. Kahley, J. Lin, S.A. Kennedy, and T.G. James.
J. Org. Chem. 60, 8294 (1995).
9. M. Novak, M.J. Kahley, J. Lin, S.A. Kennedy, and L.A.
The following azides were prepared by the procedure de-
scribed above.
4-Azido-4′-methoxybiphenyl: 1H NMR (500 MHz, CDCl3),
δ: 7.51 (2H, d, J = 8.5 Hz), 7.47 (2H, d, J = 8.5 Hz), 7.05 (2H,
Swanegan. J. Am. Chem. Soc. 116, 11626 (1994).
© 1998 NRC Canada

84
Can. J. Chem. Vol. 76, 1998
10. G. Kohnstam, W.A. Petch, and D.L.H. Williams. J. Chem. Soc.
Perkin Trans 2, 423 (1984).
21. J.P. Richard, S.-S. Lin, J.M. Buccigros, and T.L. Amyes. J. Am.
Chem. Soc. 51, 12603 (1996).
11. R.B. Millar and S. Dugar. Organometallics, 3, 1261 (1984).
12. R.M. Washburn, E. Levens, C.F. Albright, F.A. Billing, and E.S.
Cernak. Adv. Chem. Ser. 23, 102 (1959).
13. P.A.S. Smith and B.B. Brown. J. Am. Chem. Soc. 73, 2438
(1951).
14. R.A. McClelland. Tetrahedron, 52, 6823 (1996).
15. G.B. Schuster and M.S. Platz. Adv. Photochem. 17, 143 (1992).
16. J. Bartl, S. Steenken, H. Mayr, and R.A. McClelland. J. Am.
Chem. Soc. 112, 6918 (1990).
17. R.A. McClelland, C. Chan, F. Cozens, A. Modro, and S.
Steenken. Angew. Chem. Int. Ed. Engl. 30, 1337 (1991).
18. A. Riecker and B. Speiser. Tetrahedron Lett. 31, 5013 (1990).
19. J.P. Richard and W.P. Jencks. J. Am. Chem. Soc. 104, 4689
(1982).
22. R.A. McClelland, V.M. Kanagasabapathy, N. Banait, and S.
Steenken. J. Am. Chem. Soc. 113, 1009 (1991).
23. J.C. Fishbein and R.A. McClelland. J. Am. Chem. Soc. 109,
2824 (1987).
24. E.O. Alonso, L.J. Johnston, J.S. Scaiano, and V.G. Toscano. J.
Am. Chem. Soc. 112, 1270 (1990).
25. Y. Yukawa, Y. Tsuno, and M. Sawada. Bull. Chem. Soc. Jpn. 39,
2274 (1966).
26. H.C. Brown, D.P. Kelley, and M. Periasamy. Proc. Natl. Acad.
Sci. U.S.A. 77, 6956 (1980).
27. R.A. McClelland, V.M. Kanagasabapathy, N. Banait, and S.
Steenken. J. Am. Chem. Soc. 111, 3966 (1989).
28. F.L. Cozens, N. Mathivanan, R.A. McClelland, and S. Steenken.
J. Chem. Soc. Perkin Trans. 2, 2083 (1992).
20. J.P. Richard, M.E. Rothenberg, and W.P. Jencks. J. Am. Chem.
Soc. 106, 1361 (1984).
29. A.A. Allen, A.J. Kresge, N.P. Schepp, and T.T. Tidwell. Can. J.
Chem. 65, 1719 (1987).
© 1998 NRC Canada