Substituent effects on the thermal cis-to-trans isomerization of 1,3-diphenyltriazenes in aqueous solution

Article abstract of DOI:10.1021/jo011094d

The thermal cis-to-trans isomerization of some symmetrically p,p′-disubstituted 1,3-diphenyltriazenes has been studied by means of laser-flash photolysis techniques. The geometric isomerization is catalyzed by general acids and general bases as a result of acid/base-promoted 1,3-prototropic rearrangements. Acid catalysis becomes more prominent as the electron-donating character of the para substituent increases, while base catalysis becomes more important as the electron-withdrawing character of the para substituent increases. In addition, the rate ascribed to the interconversion of neutral cis rotamers through hindered rotation around the nitrogen-nitrogen single bond is found to decrease as the electron-withdrawing character of the para substituent increases. Rates of interconversion of neutral cis rotamers are also found to decrease with decreasing solvent polarity, which is indicative of the involvement of a polar transition state. On the other hand, kinetic investigations of the acid-catalyzed decomposition of target triazenes are consistent with an A1 mechanism.

Full text of DOI:10.1021/jo011094d

J . Org. Chem. 2002, 67, 2271-2277
2271
Su bstitu en t Effects on th e Th er m a l Cis-to-Tr a n s Isom er iza tion of
1,3-Dip h en yltr ia zen es in Aqu eou s Solu tion
Nan Chen, Mo´nica Barra,* Ivan Lee, and Navjot Chahal
Department of Chemistry, University of Waterloo, Waterloo, Ontario, Canada N2L 3G1
mbarra@sciborg.uwaterloo.ca
Received November 21, 2001
The thermal cis-to-trans isomerization of some symmetrically p,p′-disubstituted 1,3-diphenyltria-
zenes has been studied by means of laser-flash photolysis techniques. The geometric isomerization
is catalyzed by general acids and general bases as a result of acid/base-promoted 1,3-prototropic
rearrangements. Acid catalysis becomes more prominent as the electron-donating character of the
para substituent increases, while base catalysis becomes more important as the electron-
withdrawing character of the para substituent increases. In addition, the rate ascribed to the
interconversion of neutral cis rotamers through hindered rotation around the nitrogen-nitrogen
single bond is found to decrease as the electron-withdrawing character of the para substituent
increases. Rates of interconversion of neutral cis rotamers are also found to decrease with decreasing
solvent polarity, which is indicative of the involvement of a polar transition state. On the other
hand, kinetic investigations of the acid-catalyzed decomposition of target triazenes are consistent
with an A1 mechanism.
Sch em e 1
In tr od u ction
Triazenes, compounds characterized by having a dia-
zoamino group (-NdN-N<), commonly adopt the trans
configuration in the ground state. However, these com-
pounds are also known to undergo reversible changes in
double-bond configuration as a result of photoinduced and
thermally induced trans-cis-trans isomerization.1-4
Photochromic materials of this type are of interest for
potential applications, among others, in molecular elec-
tronic devices.5
Previous work from our laboratory has shown that cis-
1,3-diphenyltriazene (cis-HDP T), generated upon pho-
toisomerization of trans-HDP T in 2% MeOH aqueous
solutions at 6 < pH < 14, undergoes cis-to-trans isomer-
ization catalyzed by general acids and general bases.4
Acid catalysis is attributed to rate-limiting proton trans-
fer to the diazo group, whereas base catalysis is at-
tributed to rate-limiting base-promoted ionization of the
amino nitrogen leading to geometric isomerization (Scheme
1). In addition, a process ascribed to the interconversion
of cis rotamers through hindered rotation around the
nitrogen-nitrogen single bond becomes rate-limiting at
stronger basic conditions.4
Ch a r t 1
The work described here focuses on the isomerization
of some symmetrically p,p′-disubstituted 1,3-diphenyl-
triazenes (Chart 1) to further probe the isomerization
mechanism reported and thus gain a better understand-
ing of the rate-controlling factors. Substrate susceptibility
to acid decomposition is also presented.
Resu lts a n d Discu ssion
Acid -Ca ta lyzed Decom p osition . Due to the limited
solubility of target triazenes in pure water, geometric
isomerization studies were carried out in 30:70 (v/v) THF/
water solutions, typically at pHs between 6 and 14.
However, since it is well documented that triazenes are
sensitive to the presence of acids,6 to minimize substrate
depletion while handling triazene aqueous solutions, a
quantitative investigation of the stability of target sub-
(1) Le Fevre, R. J . W.; Liddicoet, T. H. J . Chem. Soc. 1951, 2743.
(2) Baro, J .; Dudek, D.; Luther, K.; Troe, J . Ber. Bunsen-Ges. Phys.
Chem. 1983, 87, 1155.
(3) Scaiano, J . C.; Chen, C.; McGarry, P. F. J . Photochem. Photobiol.
A: Chem. 1991, 62, 75.
(4) Barra, M.; Chen, N. J . Org. Chem. 2000, 65, 5739.
(5) Martin, P. J . In An Introduction to Molecular Electronics; Petty,
M. C., Bryce, M. R., Bloor, D., Eds.; Oxford University Press: New
York, 1995; Chapter 6.
(6) Benson, F. R. The High Nitrogen Compound; J ohn Wiley &
Sons: New York, 1984; Chapter 6.
10.1021/jo011094d CCC: $22.00 © 2002 American Chemical Society
Published on Web 03/13/2002

2272 J . Org. Chem., Vol. 67, No. 7, 2002
Chen et al.
strates in the solvent employed in this study was also
undertaken.
When the intensity corresponding to the longest-
wavelength absorption band of target triazenes in acidic
media is monitored as a function of time, a decrease in
signal intensity is observed due to triazene decomposi-
tion. In all cases, the rate of the reactions follows first-
order kinetics, and the corresponding observed rate
constants (kobs) are found to increase with proton con-
centration (Table S1).7 The acid catalysis is interpreted
in terms of an A1 mechanism (eq 1), in accordance with
reports in the literature.8-10 The corresponding expres-
sion for the observed rate constant is given by eq 2, where
Ka, kcleavage, and γH+ represent, respectively, the acid
dissociation equilibrium constant for the diazoammonium
ion, the first-order rate constant for cleavage of the
nitrogen-nitrogen single bond, and the activity co-
efficient for a proton.
F igu r e 1. Hammett dependence of the second-order rate
coefficient for acid-catalyzed decomposition (b) and of the
observed rate constant for cis-to-trans isomerization in NaOH
solutions (O) of symmetrically p,p′-disubstituted 1,3-diphenyl-
triazenes.
k
cleavage[H+]
k
cleavage10-pH
γ
Η+Ka + 10-pH
kobs
)
)
(2)
Ka + [H+]
The linear dependence observed between kobs and 10-pH
values (plots not shown) would indicate that, under the
experimental conditions of this study, the factor γH+Ka
is at least 10 times larger than the highest 10-pH value
for each series studied; thus, the slope of the linear plots
corresponds to the second-order rate coefficient kcleavage
/
(γH+Ka). Values for the latter (Table S2)7 are displayed
in Figure 1 (filled circles) vs the corresponding Hammett
substituent constants.11 Not surprisingly, triazene stabil-
ity increases with increasing electron-withdrawing char-
acter of the para substituent; i.e., the basicity of the
amino nitrogen and the stability of the diazonium ion
intermediate both decrease with increasing electron-
withdrawing character of the para substituent. It should
be pointed out here that the slope value corresponding
to Figure 1 (i.e., F ) -(3.1 ( 0.1)) is in excellent
agreement with reaction constants determined for acid-
catalyzed decomposition of similar systems, e.g., disub-
stituted 1,3-diphenyltriazenes -2.9712 (20 °C, 20% aque-
ous ethanol),9 disubstituted 3-(N-methylcarbamoyl)-1,3-
diphenyltriazenes -2.39 (25 °C, water),13 and 1-substituted
1,3-diphenyl-3-methyltriazenes -3.70 (25 °C, 40% aque-
ous ethanol).10
F igu r e 2. Transient absorption spectra for CH 3ODP T (in
0.05 M NaOH solution) obtained 3.2 (b), 8.4 (O), 16 (2), and
71.2 µs (4) after laser pulse. Inset: kinetic traces recorded at
390 nm at (a) pH ) 6.80 (0.01 M phosphate buffer) and (b)
0.197 M NaOH solution.
30:70 (v/v) THF/water solutions leads to negative changes
in absorbance (bleaching), followed by complete recovery
of the initial absorbance of the solutions (e.g., Figure 2).
The resulting transient absorption spectra are in fact
mirror images of the corresponding ground-state absorp-
tion spectra (e.g., Figure 3). It is noticed that the shape
of these spectra is pH dependent only in the case of
XDP T having electron-withdrawing substituents, i.e., X
) Cl, CF3, CN, and NO2 (e.g., Figures 3 and 4). These
pH-dependent spectral changes are consistent with
changes in the relative concentrations of neutral trans
triazene and of its conjugate anionic form, as would be
expected under the experimental conditions of this work
(i.e., 6 < pH < 14) by taking into account the correspond-
ing pKatrans value for deprotonation of the amino nitrogen
(Table 1). In all cases, the absorption band corresponding
to the anionic trans form is red-shifted with respect to
that of its conjugate neutral form.
Th er m a l Cis-Tr a n s Isom er iza tion . Laser excita-
tion (355 nm) of any of the target triazenes in buffered
(7) Supporting Information: see paragraph at the end of this paper
regarding availability.
(8) Zverina, V.; Remes, M.; Divis, J .; Marhold, J .; Matrka, M. Collec.
Czech. Chem. Commun. 1973, 38, 251.
(9) Benes, J .; Beranek, V.; Zimprich, J .; Veternik, P. Collec. Czech.
Chem. Commun. 1977, 42, 702.
(10) Svoboda, P.; Pytela, O.; Vecera, M. Collec. Czech. Chem.
Commun. 1986, 51, 553.
(11) March, J . Advanced Organic Chemistry; J ohn Wiley & Sons:
New York, 1992; p 280.
(12) Reaction constant for Hammett relation in the Yukawa-Tsuno
modification.
(13) Pytela, O.; Vecera, M.; Vetesnik, P. Collec. Czech. Chem.
Commun. 1980, 45, 2108.

Cis-to-Trans Isomerization of 1,3-Diphenyltriazenes
J . Org. Chem., Vol. 67, No. 7, 2002 2273
Ta ble 1. Equ ilibr iu m Con sta n ts for Dep r oton a tion of
th e Am in o Nitr ogen of tr a n s-XDP T a n d Aver a ge Va lu es
for th e Obser ved Ra te Con sta n ts for In ter con ver sion of
cis-XDP T Rota m er s in Aqu eou s Solu tion a
kfaverage, 104 s-1
k
OH, 104 s-1
trans b
X
pKa
CH3O
CH3
H
Cl
CF3
CN
10.8 ( 0.8
7.7 ( 0.6
6.4 ( 0.6
2.3 ( 0.4
9.9 ( 0.4
6.9 ( 0.2
5.1 ( 0.2
1.8 ( 0.1
1.2 ( 0.1
1.0 ( 0.1c
>14
13.2 ( 0.1
11.71 ( 0.05
10.52 ( 0.02
9.46 ( 0.04
NO2
a
Solvent contains 30% THF, µ ) 0.5 M (NaCl), T ) 21 °C.
b Determined spectrophotometrically. c Determined by curve fitting
from data in Figure 7; see text.
respect to the trans form.14 On the other hand, the time-
resolved bleaching detected with the most acidic triazenes
is only observed in the spectral region corresponding to
the absorption band of the anionic trans form, and it is
ascribed to the acid/base equilibrium of the ground-state
trans isomer due to the preferential photoinduced trans-
to-cis isomerization of the neutral form (vide infra).
The complete recovery of the initial absorbance of the
solutions clearly indicates that the geometric isomeriza-
tion is reversible in all cases. Recovery traces were
collected as a function of pH and of total buffer concen-
tration. Results obtained using XDP T with X ) CH3O,
CH3, H, and Cl are overall very similar to those of our
previous study on HDP T in 2% MeOH aqueous solu-
tions:4 (a) at any given pH and buffer concentration,
kinetic traces do not vary with monitoring wavelength
(λmonitoring); (b) recovery traces are fitted to one or two
exponential terms (depending on the pH); (c) when two
first-order processes are detected (typically at pH < 10),
the observed rate constant corresponding to the faster
F igu r e 3. Ground-state absorption spectra for trans-CNDP T
at pH ) 6.30 (A) and 11.02 (B). Transient absorption spectra
for CNDP T obtained at pH ) 6.30 (0.05 M phosphate buffer)
36 (b), 140 (O), 344 (2), and 728 µs (4) after laser pulse (C)
and at pH ) 11.02 (0.05 M carbonate buffer) 3.2 (b), 24 (O),
69.6 (2), and 146 µs (4) after laser pulse (D).
f
process (kobs ) is independent of pH and buffer concentra-
tion (Tables S3, S6, S9, and S12),7 while the observed
rate constant for the slower process (kobss) varies with pH
and buffer concentration (Tables S4, S7, S10, and S13);7
(d) the preexponential factors corresponding to the two
first-order processes just mentioned are independent of
pH and buffer concentration; (e) only one first-order
process, which is independent of pH, is observed in NaOH
solutions (Tables, S5, S8, S11, and S14);7 and (f) the
average rate constant for the latter (kOH) is in very good
f
agreement with the corresponding kobs average value
obtained at lower pHs (Table 1). Furthermore, with any
s
of these substrates, plots of kobs vs buffer concentration
are fairly linear, and data are interpreted in terms of eq
3,
s
F igu r e 4. Transient absorption spectra for CNDP T at pH )
9.62 (0.05 M carbonate buffer) obtained 7.2 (b), 28 (O), 68.8
(2), and 146 µs (4) after laser pulse. Inset: kinetic traces
recorded at pH ) 9.62 (0.1 M carbonate buffer) at (a) 390 and
(b) 445 nm.
kobs ) kO + kC[buffer]
(3)
where kC and kO represent, respectively, the catalytic rate
coefficient for the buffer species and the reaction through
solvent-related species. In the case of data for phosphate
buffer (i.e., pH < 8), values of kC and kO are found to
decrease with increasing pH (Table 2), which is indicative
of general acid catalysis, as illustrated in Figures 5A and
6 (diagonal portion of slope -1 of the V-shaped profile).15
Despite the limited number of points, the clear nonlinear
dependence of kC on the molar fraction of free acid of
In the case of XDP T with X ) CH3O, CH3, H, and Cl,
bleaching is instantaneous at all pHs (e.g., Figure 2
inset). However, in the case of XDP T with X ) CF3, CN,
and NO2, in addition to instantaneous bleaching, a time-
resolved bleaching (e.g., Figure 4, inset b) can also be
trans
observed at pH ≈ pKa
( 1. In all cases, the instan-
taneous bleaching is ascribed to photoinduced trans-to-
cis isomerization, consistent with previous studies on
HDP T,3,4 the compound for which the cis isomer has been
shown to have a lower absorptivity at λ > 350 nm with
(14) Baro, J .; Dudek, D.; Luther, K.; Troe, J . Ber. Bunsen-Ges. Phys.
Chem. 1983, 87, 1161.
(15) The slowest rate constant that can be determined with our
laser-flash photolysis system is ca. 2000 s-1
.

2274 J . Org. Chem., Vol. 67, No. 7, 2002
Chen et al.
Ta ble 2. Ca ta lytic Ra te Coefficien ts (kC) a n d Ra te Con sta n ts for Rea ction th r ou gh Solven t-Rela ted Sp ecies (kO) for
Cis-Tr a n s Isom er iza tion of Sym m etr ica lly p,p′-Disu bstitu ted 1,3-Dip h en yltr ia zen es in Aqu eou s Solu tion a
pH
kC,b 104 M-1 s-1
kO,b 103 s-1
pH
kC,b 104 M-1 s-1
kO,b 103 s-1
pH
kC,b 104 M-1 s-1
kO,b 103 s-1
CH3ODP T
6.14c
HDP T
6.04c
6.27c
6.80c
7.35c
7.92c
9.62e
9.79e
10.26e
10.75e
11.08e
11.20e
ClDP T
5.74c
6.01c
6.25c
6.82c
7.50c
7.74c
8.13d
9.64e
10.39e
CF 3DP T
5.77c
15.4 ( 0.5
11.7 ( 0.6
6.9 ( 0.3
7.4 ( 0.3
6.5 ( 0.4
2.7 ( 0.2
5.8 ( 0.1
8.7 ( 0.5
65 ( 1
7 ( 1
9.4 ( 0.7
6.4 ( 0.4
2.8 ( 0.2
2.4 ( 0.1
1.9 ( 0.1f
4.5 ( 0.2
6.1 ( 0.2
19.4 ( 0.4
23.5 ( 0.9
44 ( 1
3.1 ( 0.1
3.0 ( 0.1
2.9 ( 0.1
3.6 ( 0.2
5.0 ( 0.3
11.8 ( 0.1
6.27c
3.6 ( 0.6
1.3 ( 0.3
6.27c
6.80c
7.91c
9.88e
9.50e
0.7 ( 0.5
3.4 ( 0.5
10.28e
11.33e
CH3DP T
6.14c
4.6 ( 0.9
9.94e
0.8 ( 0.3
1.8 ( 0.3
10.97e
CNDP T
5.77c
7.1 ( 0.4
5.9 ( 0.4
7.1 ( 0.2
5.5 ( 0.3
3.0 ( 0.1
4.3 ( 0.1
9.3 ( 0.3
39 ( 1
0.48 ( 0.09
0.53 ( 0.08
2.3 ( 0.2
3 ( 1
2.6 ( 0.1
2.2 ( 0.1
2.0 ( 0.1
8.9 ( 0.7
10.6 ( 0.1
11.3 ( 0.3
6.26c
6.28c
6.81c
2.0 ( 0.2
7.91c
9.89e
0.65 ( 0.20
49.7 ( 0.2
9.62e
10.35e
11.33e
10.26e
11.07e
NO2DP T
5.78c
5 ( 1
2.0 ( 0.6
8.5 ( 0.7
5.4 ( 0.4
4.4 ( 0.2
0.8 ( 0.3
1.7 ( 0.2
10 ( 1
1.9 ( 0.2
1.9 ( 0.1
2.3 ( 0.6
24 ( 3
2.50 ( 0.06
6.27c
2.16 ( 0.07f 7.91c
1.92 ( 0.05f 9.62e
1.9 ( 0.2
31 ( 5
1.0 ( 0.2
1.3 ( 0.6
4 ( 1
5.8 ( 0.4
10 ( 1
a
b
Solvent contains 30% THF, µ ) 0.5 M (NaCl), T ) 21 °C. Uncertainty corresponds to standard error. c Phosphate buffer.
Tris(hydroxymethyl)aminomethane buffer. e Carbonate buffer. f Observed value corresponds indeed to the response limit of our laser-
d
flash photolysis system (1/τ ca. 2000 s-1).
buffer component xHA is attributed to acid catalysis by
H2PO4 as well as by H3PO4, as previously reported.4
-
Interestingly, at any given pH, the catalytic rate co-
efficient kC diminishes as the electron-donating character
of the para substituent also diminishes. On the other
hand, in the case of data obtained at 8 < pH < 11 (i.e.,
tris(hydroxymethyl)aminomethane and carbonate buff-
ers), values of kC (if at all significant) and kO (diagonal
portion of slope +1 of the V-shaped profile of Figure 6)
are found to increase with increasing pH (Table 2), which
is indicative of general base catalysis. Furthermore, as
illustrated in Figure 6 for HDP T, the downward bend
observed at high pHs in the pH-rate profile is indicative
of a change in rate-controlling step. All these observations
are consistent with the reaction mechanism given in
Scheme 1; i.e., the general acid/base catalysis being
observed is attributed to acid/base-promoted 1,3-proto-
tropic rearrangements leading to geometric isomeriza-
tion, while the pH- and buffer-independent process is
ascribed to the interconversion of neutral cis isomers. The
latter process becomes rate-controlling typically at pH
> 11. As shown in Figure 6 for HDP T, and from
comparison of the values given in Table 1, kfaverage seems
to be larger than kOH. The small difference between these
two sets of values, if at all significant, might indicate
that, under the experimental conditions of this study,
both k1 and k-1 should be considered; therefore, values
of kfaverage would correspond to (k1 + k-1), whereas values
of kOH would correspond to k1.
It should be pointed out here that the preexponential
factor corresponding to the faster first-order process
observed typically at pH < 10 (ascribed to the intercon-
version of cis rotamers) diminishes as the electron-
donating character of the para substituent decreases. For
instance, in the case of CH3ODP T, it represents 25% of
the total signal, while in the case of ClDP T, it represents
only 10%. This trend would indicate that the difference
in absorptivity between the neutral cis rotamers at λ )
390 nm (the shortest λmonitoring of this study), the amount
of the less stable cis rotamer generated upon photoexci-
F igu r e 5. Plot of catalytic rate coefficients vs molar fraction
of acid for cis-to-trans isomerization of p,p′-disubstituted 1,3-
diphenyltriazenes in phosphate buffer.
tation, or the combination of these two factors diminishes
with decreasing electron-donating character of the para

Cis-to-Trans Isomerization of 1,3-Diphenyltriazenes
J . Org. Chem., Vol. 67, No. 7, 2002 2275
F igu r e 7. Plot of observed rate constants vs HO- concentra-
tion for cis-to-trans isomerization of CNDP T in aqueous
solutions at pH > 10.
F igu r e 6. Rate profile for cis-to-trans isomerization of HDP T
in aqueous solution: b and 2, rate constants for reaction
through solvent-related species (obtained from the intercept
of buffer dependence plots for phosphate and carbonate buffers,
respectively); 4, observed rate constants for the faster process
observed at pH < 10 in phosphate and carbonate buffers; O,
observed rate constants for the single process detected in
NaOH solutions. Arrow indicates direction of possible move-
ment of point, which is a maximum value.15
excitation results in instantaneous formation of the
corresponding neutral cis isomer. On the other hand,
trans
trans-XDP T with X ) Cl, CF3, CN, and NO2 (i.e., pKa
< 14) would also exist in its deprotonated form (depend-
ing on pH), and therefore, photoisomerization may lead
to the formation of the corresponding neutral and/or
anionic cis isomers. In the case of XDP T with X ) CF3,
CN, and NO2, photoisomerization of the neutral trans
form seems to occur preferentially over the anionic form.
This is inferred from the fact that when working at pH
≈ pKatrans ( 1, at the region where only the anionic trans
form absorbs, not only is an instantaneous bleaching
observed but also a time-resolved bleaching (e.g., Figure
4 inset b). The instantaneous bleaching, as already
mentioned, corresponds to the photoinduced trans-to-cis
isomerization, while the time-resolved bleaching is at-
tributed to the acid/base equilibrium of the trans isomer
as a result of the preferential removal of the neutral form
upon photoisomerization. This interpretation is further
supported by the fact that (a) the transient absorption
spectra, recorded within the time frame corresponding
to the time-resolved bleaching, are exact mirror images
of the ground state absorption spectra, which includes a
clearly defined isosbestic point (Figure 8) and (b) the rate
constants for the time-resolved bleaching (Tables S16,
S19, and S22)7 are in excellent agreement with the values
corresponding to the faster component of the recovery
substituent. Thus, it is not surprising to find that in the
case of CF 3DP T, CNDP T, and NO2DP T, recovery traces
are very well reproduced by a single-exponential term.16
The observed rate constants for cis-to-trans isomerization
of these three substrates (Tables S15, S18, and S21)7
increase with increasing buffer concentration and pH,
consistent with general base catalysis. Plots of kobs vs
buffer concentration are fairly linear, and data are
therefore interpreted in terms of eq 3 (Table 2). Interest-
ingly, the catalytic rate coefficient kC increases as the
electron-withdrawing character of the para substituent
also increases (e.g., Figure 5B). On the other hand, the
first-order rate constants observed in NaOH solutions are
pH independent in the case of CF 3DP T (Table S17),7
while values for CNDP T (Table S20)7 vary as shown in
Figure 7; no signal could be observed in the case of
NO2DP T in solutions at pH > 10.
The amino nitrogen of cis-XDP T is expected to be less
acidic than that in the corresponding trans isomeric form
due to restricted resonance delocalization as a result of
steric hindrance. This assumption is supported, for
instance, by the differences in acidity reported for azo
compounds such as 2-hydroxy-5-methylazobenzene
traces recorded under the same conditions at λmonitoring
)
(pKatrans ) 9.4, pKacis ) 10.7)17 and 4-[4′-(dimethylamino)-
390 nm (λ at which the neutral trans form is the main
absorbing species). As already mentioned, in the case of
NO2DP T no signal could be detected with solutions at
pH > 10. The lack of signal probably indicates a low
quantum yield for trans-to-cis isomerization of the an-
ionic trans form. This assumption is consistent with the
observation that under conditions where both the neutral
and anionic trans-NO2DP T forms are present in solution,
no instantaneous bleaching can be observed in the region
where the anionic trans form is the only absorbing species
and only time-resolved bleaching is recorded (Figure 8
inset). Furthermore, it is noticed that for CF 3DP T and
CNDP T, the contribution of the time-resolved bleaching
trans
phenylazo]benzenesulfonate (pKa
) 2.7018-2.87,19
pKa ) 5.020). Thus, under the experimental conditions
cis
of this study (i.e., 6 < pH < 14), trans-XDP T with X )
trans
CH3O, CH3, and H (i.e., pKa
> 14) exists in solution
essentially in its neutral form, and therefore, laser
(16) At pH ≈ pKatrans ( 1, rate constants were obtained from kinetic
traces collected at λ > 400 nm, where a time-resolved bleaching
followed by signal recovery is observed; traces were reproduced by two
well-separated exponential terms.
(17) Wettermark, G.; Langmuir, M. E.; Anderson, D. G. J . Am.
Chem. Soc. 1965, 87, 476.
(18) Tawarah, K. M.; Abu-Shamleh, H. M. Dyes and Pigments 1991,
16, 241.
(19) Reeves, R. L. J . Am. Chem. Soc. 1966, 88, 2240.
(20) Sanchez, A. M.; Barra, M.; de Rossi, R. H. J . Org. Chem. 1999,
64, 1604.
relative to that of the instantaneous bleaching at pH ≈
trans
pKa
increases in the order CF 3DP T < CNDP T.

2276 J . Org. Chem., Vol. 67, No. 7, 2002
Chen et al.
Ta ble 3. Obser ved Ra te Con sta n ts for In ter con ver sion
of Cis Rota m er s in Na OH Aqu eou s Solu tion s a s a
F u n ction of Or ga n ic Cosolven ta
k
OH, 105 s-1 b
cosolvent
triazene
CH3OH
(CH3)2SO
CH3CN
(CH3)2CHOH
CH3ODPT 4.4 ( 0.1 2.94 ( 0.08
2.8 ( 0.1
2.54 ( 0.08
1.89 ( 0.04
1.59 ( 0.06
CH3DPT
HDPT
3.7 ( 0.1
2.5 ( 0.1
2.16 ( 0.07
3.4 ( 0.2 1.94 ( 0.09 1.62 ( 0.04
a
Solvent contains 30% organic cosolvent, µ ) 0.5 M (NaCl),
b
T ) 21 °C. Values correspond to the average of 5 independent
experiments.
k1) clearly shows that the reactivity diminishes as the
electron-donating character of the para substituent di-
minishes as well. The Hammett plot represented in
Figure 1 (open circles) is characterized by a F value of
-(0.53 ( 0.04). Reported data for restricted rotation of
trans-1-aryl-3,3-dialkyltriazenes in organic solvents lead
to F ≈ -2;21-23 the restricted rotation around the nitrogen-
nitrogen single bond of triazenes is ascribed to the partial
double-bond character that results form 1,3-dipolar con-
tributions as shown in eq 5.21-27 Thus, the decrease in
the rate for interconversion of neutral cis isomers with
decreasing electron-donating character of the para sub-
stituent suggests predominant conjugation between the
diazoamino group and the aromatic ring attached to the
sp2 nitrogen. This observation is also consistent with the
increase in basicity of the amino nitrogen in the cis
isomer relative to that in the trans form, as shown in
the case of CNDP T, attributed to restricted conjugation
between the amino nitrogen and the phenyl ring attached
to it.
F igu r e 8. Transient absorption spectra for NO2DP T at pH
) 9.62 (0.05 M carbonate buffer) obtained within 0.2 and 3.4
µs (from a to b) after laser pulse. Inset: kinetic trace recorded
at 570 nm (pH ) 9.62, 0.01 M carbonate buffer).
Sch em e 2
The pH dependence illustrated in Figure 7 is inter-
preted in terms of the reaction mechanism shown in
Scheme 2; i.e., the observed rate constants in NaOH
solutions would correspond to rate-limiting interconver-
sion of conjugate rotamers of cis-CNDP T according to
the rate law given by eq 4, in which k1, k2, and Ka are
defined as shown in Scheme 2 and Kw represents the
equilibrium constant for self-ionization of water.
We noticed that the value of k1 previously reported for
HDP T in 2% MeOH, namely, (3.39 ( 0.05) × 105 s-1,4 is
significantly higher than the one presented here for
HDP T in 30% THF. Thus, rate constants ascribed to the
interconversion of neutral cis rotamers were also deter-
mined in NaOH solutions as a function of the nature and
concentration of organic cosolvent. Rate constants are
found to decrease (a) as the polarity and hydrogen
bonding ability of the corresponding organic cosolvent
decrease (Tables 3 and 4, last row) and (b) as the
percentage of THF increases (Table 4). These results
clearly indicate the involvement of a polar transition
state.
cis
k1[H+] + k2Ka
k1 + [HO-]k2Kacis/Kw
1 + [HO-]Kacis/Kw
cis
kobs
)
)
(4)
[H+] + Ka
cis
Curve fitting according to eq 4 leads to values of (1.0
( 0.1) × 104 s-1, (8 ( 2) × 104 s-1, and 28 ( 3 for k1, k2,
and Kacis/Kw, respectively; i.e., the anionic cis form of
CNDP T isomerizes more rapidly than the neutral form,
and the cis isomer is clearly less acidic than the trans
In summary, the thermal cis-to-trans isomerization of
symmetrically p,p′-disubstituted 1,3-diphenyltriazenes is
catalyzed by general acids and general bases. Catalytic
rate coefficients increase as the pKa difference between
cis
form (since pKw > 14 in 30% THF, then pKa > 12.6).
As already mentioned, in the case of CF 3DP T, the
(21) Marullo, N. P.; Mayfield, C. B.; Wagener, E. H. J . Am. Chem.
Soc. 1968, 90, 510.
(22) Akhtar, M. H.; McDaniel, R. S.; Feser, M.; Oehlschlager, A. C.
Tetrahedron 1968, 24, 3899.
(23) Lippert, Th.; Wokaun, A.; Dauth, J .; Nuyken, O. Magn. Reson.
Chem. 1992, 30, 1178.
(24) Lunazzi, L.; Cerioni, G.; Foresti, E.; Macciantelli, D. J . Chem.
Soc., Perkin Trans. 2 1978, 686.
(25) Sieh, D. H.; Wilbur, D. J .; Michejda, C. J . J . Am. Chem. Soc.
1980, 102, 3883.
(26) Golding, B. T.; Kemp, T. J .; Narayanaswamy, R.; Waters, B.
W. J . Chem. Res., Synop. 1984, 130.
(27) Panitz, J .-C.; Lippert, Th.; Stebani, J .; Nuyken, O.; Wokaun,
A. J . Phys. Chem. 1993, 97, 5246.
observed rate constants in NaOH solutions are indepen-
cis
dent of pH; therefore, one must conclude either that pKa
for CF 3DP T is significantly larger than 14 (fully consis-
tent with the decrease in Ka of, at least, 2 orders of
cis
magnitude just described for cis-CNDP T (pKa > 12.6)
trans
vs trans-CNDP T (pKa
) 10.52)) or, if indeed both
conjugate forms of cis-CF 3DP T are present in solution,
that k1 ≈ k2.
Interestingly, comparison of the rate constants for
interconversion of neutral cis rotamers (Table 1, kOH
)

Cis-to-Trans Isomerization of 1,3-Diphenyltriazenes
J . Org. Chem., Vol. 67, No. 7, 2002 2277
Ta ble 4. Obser ved Ra te Con sta n ts for In ter con ver sion
of cis Rota m er s in Na OH Aqu eou s Solu tion s a s a
F u n ction of THF Con cen tr a tion a
measured using a combination electrode that had been cali-
brated with standard aqueous buffers.
Acid dissociation equilibrium constants for deprotonation
of the amino nitrogen of target triazenes were determined by
the classical spectrophotometric method, by measuring the
changes in absorbance at the wavelength of maximum absorp-
tion of each conjugate species as a function of pH; resulting
values are in very good agreement with literature data for
similar systems.9
Acid-catalyzed decomposition reactions were carried out
under pseudo-first-order conditions and followed until at least
80-90% conversion of the starting material was observed.
Typical substrate concentrations were in the order of (1-6) ×
10-5 M. Rates of reaction were determined from the change
in absorbance at the wavelength of maximum absorption of
each target triazene (i.e., CH3ODP T ) 370 nm, CH3DP T )
362 nm, HDP T ) 365 nm, ClDP T ) 363 nm, CF 3DP T ) 357
nm, CNDP T ) 372 nm, and NO2DP T ) 406 nm) as a function
of time.
Laser experiments were carried out using a Q-switched Nd:
YAG laser (Continuum, Surelite I) operated at 355 nm (4-6
ns pulses, <30 mJ /pulse) for excitation. Further details on this
laser system have been reported elsewhere.20 Solutions were
contained in quartz cells constructed of 7 × 7 mm2 Suprasil
tubing. Transient absorption spectra were collected under flow
conditions in order to ensure the irradiation of fresh portions
of sample by each laser pulse. For each solution, kinetic traces
were recorded at different time domains (typically between
0.08 and 8 µs/point) and then combined for analysis; if at all
necessary, traces were previously normalized to account for
slight differences in signal intensity due to laser power
fluctuations. All measurements were carried out at 21 ( 1 °C.
All observed rate constants were determined by using the
general curve-fitting procedure of Kaleidagraph 3.0.5 software
from Abelbeck Software.
k
OH, 105 s-1 b
% THF
(v/v)
CH3ODP T
CH3DP T
HDP T
CF 3DP T
5
10
20
30
4.6 ( 0.1
c
c
3.3 ( 0.1d
2.7 ( 0.2
4.28 ( 0.06
0.32 ( 0.03
2.12 ( 0.07 1.52 ( 0.04 1.26 ( 0.08 0.174 ( 0.006
1.00 ( 0.03 0.66 ( 0.02 0.56 ( 0.02 0.098 ( 0.006
a
b
µ ) 0.5 M (NaCl), T ) 21 °C. Values correspond to the
average of 5 independent experiments. c Measurements could not
be performed due to the low solubility of the triazene. 2% THF.
d
the triazene and the Bro¨nsted catalyst also increases.
Thus, acid catalysis becomes more prominent as the
electron-donating character of the para substituent in-
creases (e.g., Figure 5A), whereas base catalysis becomes
more important as the electron-withdrawing character
of the para substituent increases (e.g., Figure 5B). In all
cases, interconversion of cis rotamers becomes rate
controlling at high pHs. Interestingly, the rate for inter-
conversion of neutral cis isomers decreases as the electron-
withdrawing character of the para substituent increases,
which suggests predominant conjugation between the
diazo group and the phenyl ring attached to it; this
conclusion is in agreement with the expected increase in
cis
pKa relative to pKatrans, as indeed observed in the case
of CNDP T.
Exp er im en ta l Section
1,3-Diphenyltriazene (Aldrich) was purified by treatment
with Cd(OH)2 generated in situ from Cd(NO3)2 (Allied Chemi-
cal) in basic methanol (BDH, ACS grade) aqueous solution,
as described in the literature.28 Symmetrically p,p′-disubsti-
tuted 1,3-diphenyltriazenes were synthesized by means of
aniline/diazonium salts coupling via the classical diazotization
Molar fractions, such as those displayed in Figure 5, were
calculated from the observed pH by using the ionization
equilibrium constant for the corresponding buffer determined
potentiometrically under our experimental conditions. Thus,
ionization constants were obtained by extrapolating the pH
values of solutions containing the corresponding conjugate acid
and base in a 1:1 ratio to zero buffer concentration; resulting
method with NaNO2 (X ) NO2) or using isoamyl nitrite30 (X
29
) CH3O, CH3) or sodium hexanitrocobaltate(III)31 (X ) Cl, CF3,
CN), as described in the literature. All p,p′-disubstituted 1,3-
diphenyltriazenes were purified by recrystallization (Table
S25);7 melting points and 1H NMR data (Table S25)7 are in
very good agreement with reported values.
-
pKa values are 6.76 ( 0.01 and 10.18 ( 0.01 for H2PO4 and
HCO3-, respectively.
Ackn owledgm en t. Financial support from the Natu-
ral Sciences and Engineering Research Council (NSERC)
of Canada is gratefully acknowledged.
Kinetic studies were carried out using 30:70 (v/v) THF/water
as the solvent (unless stated otherwise). Aqueous buffers were
prepared using analytical-grade salts, water purified in a
Millipore apparatus, and spectrophotometric-grade organic
solvents (BDH, Omnisolv grade). The ionic strength of the
solutions was kept constant at 0.5 M using NaCl as a
compensating electrolyte. The pH of these solutions was
Su p p or tin g In for m a tion Ava ila ble: Rate constants for
acid-catalyzed decomposition and thermal cis-trans isomer-
ization of symmetrically p,p′-disubstituted 1,3-diphenyltriaz-
enes in buffered aqueous solutions as a function of pH and/or
buffer concentration, as well as solvents for recrystallization,
melting points, and 1H NMR data of symmetrically p,p′-
disubstituted 1,3-diphenyltriazenes synthesized. This material
is available free of charge via the Internet at http://pubs.acs.org.
(28) Dwyer, F. P. J . Chem. Soc. Ind., London 1937, 56T, 70.
(29) Vogel, A. I. Practical Organic Chemistry; Longmans: New York,
1956; p 626.
(30) Vernin, G.; Siv, C.; Metzger, J . Synthesis 1977, 691.
(31) Stefane, B.; Kocevar, M.; Polanc, S. J . Org. Chem. 1997, 62,
7165.
J O011094D