Reactivity of the 3-Nitroanisole Triplet: Methanol Inhibition of Photohydroxylation

Article abstract of DOI:10.1021/jo971188g

The reactivity of the triplet state of 3-nitroanisole (1) toward a series of quenchers has been investigated in water and water alcohol mixtures. The triplet is a relatively efficient hydrogen abstractor, a very efficient electron acceptor, but a poor electron donor. Hydroxide quenches the triplet state via nucleophillic aromatic substitution. Methanol inhibition of the S_N2 Ar* reaction of hydroxide with triplet 1 in water at pH 12 was examined by steady-state and time-resolved methods HPLC measurements show that a product in addition to 3-nitrophenolate is formed in trace amounts (≤1% of total material) when methanol is present. 2-Propanol, tert-butyl alcohol and trifluoroethanol do not quench the reaction nor does their presence lead to formation of additional products. Methanol quenching of the reaction follows Stern-Volmer behavior with K_sv = 0 36 ± 0.03 M^-1 leading to a rate constant of (2.3 ± 0.2) × 10⁶ M^-1 s^-1 for methanol quenching of product formation. Transient absorption studies show that methanol quenches triplet 1 at pH 11.5 with a rate constant of (1.8 ± 0.1) × 10⁶ M^-1 s^-1. In the absence of hydroxide, methanol does not quench the triplet state. The results are explained in terms of formation of low concentrations of methoxide in alkaline solution.

Full text of DOI:10.1021/jo971188g

J . Org. Chem. 1997, 62, 8777-8783
8777
Rea ctivity of th e 3-Nitr oa n isole Tr ip let: Meth a n ol In h ibition of
P h otoh yd r oxyla tion
Christopher H. Evans,*,†,‡ Gudru´n AÄ rnado´ttir,† and J . C. Scaiano§
Science Institute, University of Iceland, Dunhaga 3, IS-107 Reykjavik, Iceland, and Department of
Chemistry, University of Ottawa, 10 Marie Curie, Ottawa, Canada K1N-6N5
Received J uly 1, 1997X
The reactivity of the triplet state of 3-nitroanisole (1) toward a series of quenchers has been
investigated in water and water alcohol mixtures. The triplet is a relatively efficient hydrogen
abstractor, a very efficient electron acceptor, but a poor electron donor. Hydroxide quenches the
triplet state via nucleophillic aromatic substitution. Methanol inhibition of the SN2 Ar* reaction
of hydroxide with triplet 1 in water at pH 12 was examined by steady-state and time-resolved
methods. HPLC measurements show that a product in addition to 3-nitrophenolate is formed in
trace amounts (e1% of total material) when methanol is present. 2-Propanol, tert-butyl alcohol,
and trifluoroethanol do not quench the reaction nor does their presence lead to formation of
additional products. Methanol quenching of the reaction follows Stern-Volmer behavior with KSV
) 0.36 ( 0.03 M-1 leading to a rate constant of (2.3 ( 0.2) × 106 M-1 s-1 for methanol quenching
of product formation. Transient absorption studies show that methanol quenches triplet 1 at pH
11.5 with a rate constant of (1.8 ( 0.1) × 106 M-1 s-1. In the absence of hydroxide, methanol does
not quench the triplet state. The results are explained in terms of formation of low concentrations
of methoxide in alkaline solution.
In tr od u ction
interferes with the photosubstitution has not been in-
vestigated. In this report we examine the influence of
methanol on the photohydroxylation making use of both
time-resolved and steady-state methods. In this context
we have also carried out a detailed examination of the
triplet’s reactivity in hydrogen bonding solvents.
Photosubstitution of 3-nitroanisole (1) by aqueous
hydroxide is believed to proceed via the so-called SN2 Ar*
mechanism.1,2 The nucleophile attacks the nitroanisole
triplet state at the methoxy-substituted carbon yielding
an anionic σ-complex which thermally decays to yield the
photosubstitution product, 3-nitrophenolate (2), and
methanol.1,2
Exp er im en ta l Section
Ma ter ia ls. All materials were purchased from Aldrich
unless otherwise noted. 3-Nitroanisole (1) and 3-nitrophenol
were recrystallized from methanol-water mixtures. Phenol
(Fisher) was vacuum sublimed. Cyclohexadienes were distilled
before use. Cumyl alcohol, trifluoroethanol, methylviologen,
and sorbic acid (2,6-hexadienonic acid, Sigma) were used as
received. Solvents for spectroscopy (Aldrich, BDH, or Rath-
burn) were of the highest quality commercially available and
were used as received. High performance liquid chromatog-
raphy (HPLC) mobile phase solvents (Rathburn, HPLC grade)
were filtered and degassed before use. Conductivity water was
prepared by passing distilled water through a Lab-Ion L2
system. Oxygen-free nitrogen was used for sample degassing.
Inorganic salts (Merck, BDH, Fisher) were used as received.
Steady-State P h otolysis: Rate Deter m in ation . Samples
typically contained ca. 2 × 10-4 M 1 in aqueous pH 12 KCl/
NaOH buffer7 or mixtures of buffer and methanol. pH values
were monitored with a Radiometer PHM80 pH meter. Ap-
proximately 3 mL samples were placed in Pyrex test tubes.
The tubes were sealed with rubber septa and degassed by
bubbling with oxygen-free nitrogen. Photolyses were per-
formed in a Rayonet photochemical chamber reactor (RPR-
100, Southern New England Ultraviolet Co.) equipped with a
Rayonet merry-go-round (RMA-500) and operating with RPR-
3500 Å, 21 W lamps.
The quantum yield is dependent on the hydroxide
concentration but reaches a maximum value of 0.22 at
[OH-] 1 mM.1 tert-Butyl alcohol has no effect on the yield
of photosubstitution product at less than 90% tert-butyl
alcohol by volume.1,3 This has been used as evidence to
support the SN2 Ar* mechanism.1,3 By contrast, aqueous
methanol markedly reduces the product yield at concen-
trations less than ca. 20% methanol by volume.4-6
Although this behavior of methanol has been noted on
4-6
several occasions
the process by which methanol
* Corresponding Author: Christopher H. Evans Science Institute,
University of Iceland, Dunhaga 3, IS-107 Iceland. E-mail chris@
raunvis.hi.is.
† University of Iceland.
‡ Presently on leave at University of Ottawa.
§ University of Ottawa.
X Abstract published in Advance ACS Abstracts, November 15, 1997.
(1) Cornelisse, J .; Lodder, G.; Havinga, E. Rev. Chem. Intermed.
1979, 2, 231.
Formation of 3-nitrophenolate (2) as a function of irradiation
time was followed by monitoring the phenolate absorption at
420 nm, a wavelength where 1 does not absorb. Absorption
values were converted to concentrations using an absorption
coefficient of 900 M-1 cm-1 for 2.6 From these data the initial
rate of phenolate formation (d[3-nitrophenolate]/dt) could be
determined. Absorption measurements were made on a Per-
(2) Rossi, R. A.; de Rossi, R. H. Aromatic Substitution by the SRN1
Mechanism; American Chemical Society: Washington, D. C., 1983.
(3) van Riel, H. C. H. A.; Lodder, G.; Havinga, E. J . Am. Chem. Soc.
1981, 103, 7257.
(4) Nijhof, D. F. Doctoral Thesis, University of Leiden, Leiden, 1967.
(5) Bonhila, J . B. S.; Chaimovich, H.; Toscano, V. G.; Quina, F. J .
Phys. Chem. 1979, 83, 2463.
(6) Evans, C. H.; Gunnlaugsson, T. J . Photochem. Photobiol. A:
Chem. 1994, 78, 57.
(7) Handbook of Chemistry and Physics; 63rd ed.; CRC Press: Boca
Raton, 1982.
S0022-3263(97)01188-2 CCC: $14.00 © 1997 American Chemical Society

8778 J . Org. Chem., Vol. 62, No. 25, 1997
Evans et al.
kin-Elmer Lambda 3B instrument or a Hewlett-Packard 8451
diode array spectrometer.
Product formation was also followed by analytical HPLC.
HPLC was carried out on an LDC Analytical HPLC instrument
fitted with a ConstaMetric 3200 pump and a SpectroMonitor
3200 detector operating at 270 nm. Separations were achieved
on an 11 cm silica column fitted with a 10 mm silica guard
column. The mobile phase used was 15 vol % ethyl acetate in
hexane and a typical flow rate was 1.0 mL/min.
GC-MS measurements were made with a Fisons MD800
system. The GC column was a 30 m J &W DB5 with a 0.32
mm i.d. and 25 µm films. The MS used simple ion detection
and TIC integration.
After photolysis, samples for HPLC and GC-MS analysis
were neutralized with dilute HCl. The aqueous phase was
extracted several times into ether, the ether evaporated, and
the residue redissolved in a small volume of ether. This was
then injected onto the HPLC or GC column and eluted. HPLC
and GC-MS peaks for photolyzed samples were compared to
authentic samples for identification.
La ser F la sh P h otolysis Mea su r em en ts. For triplet
spectra, lifetime measurements, and quenching studies a
concentration of ca. 3 × 10-4 M 1 was used. The samples were
irradiated in 7 × 7 mm2 Suprasil quartz cuvettes with pulses
(355 nm, ≈10 ns, e30 mJ /pulse) from a frequency tripled
Surelite Nd:YAG laser. Transient decays were averages of
5-10 laser shots with monitoring at 400 nm or other suitable
wavelengths. Transient absorption spectra were constructed
from decay traces monitored at different wavelengths. In all
cases transient absorption was captured with a Tektronix 2440
transient digitizer and transferred to a Power Macintosh
computer. The computer controlled the experiments and
provided suitable processing facilities through use of home-
developed programs written in the LabVIEW-3.1 environment
from National Instruments. Further details on similar laser
systems have been provided elsewhere.8,9
F igu r e 1. Concentration of 3-nitrophenolate as a function of
steady-state irradiation time. 0.2 mM 3-Nitroanisole in pH 12
buffer containing (a) 0.5% methanol (triangles), (b) 10%
methanol (circles). Detection wavelength ) 420 nm.
In the absence of hydroxide the samples were found to be
rather stable as judged by comparison of pre- and post-laser
absorption spectra. In the presence of hydroxide (pH 11.5
phosphate buffer7) significant photoconversion was observed
so that it was necessary to flow the samples using homemade
7 × 7 mm2 Suprasil flow cells. In either case degassing was
achieved by bubbling with oxygen-free nitrogen. In the flow
systems the nitrogen was passed through a reservoir attached
to the photolysis cell with Teflon tubing. pH 11.5 rather than
12 was used in the transient studies so that the 1 triplet
lifetime would not be so short as to be difficult to measure.
Resu lts
Stea d y-Sta te Beh a vior of th e 3-Nitr oa n isole Sys-
tem . 3-Nitroanisole exhibits two absorption bands at
about 270 and 326 nm in polar solvents.10 Upon pho-
tolysis of 3-nitroanisole in pH 12 buffer (excitation
centered at 350 nm) there is a reduction in intensity of
the 1 bands, and three new bands are observed at 254,
280, and 410 nm. These are due to formation of 3-nitro-
phenolate.1,11 The initial rate of 2 formation (d[3-nitro-
phenolate]/dt) could be determined from plots of the
concentration of 2 (based on its absorbance at 420 nm)
formed as a function of irradiation time. Two such plots
are shown in Figure 1 for 0.5% and 10% methanol by
volume. The plots curve at long irradiation times as a
result of depletion of starting material and absorption
of the exciting radiation by the phenolate. The rates are
calculated from the linear portion of the curves. Clearly,
the rate of phenolate formation is surpressed by metha-
nol, in agreement with earlier reports.4-6
F igu r e 2. Stern-Volmer plot showing quenching of 3-nitro-
phenolate formation by methanol in pH 12 buffer. Based on
steady-state measurements. R refers to rate of appearance of
3-nitrophenolate.
Figure 2 is a Stern-Volmer plot showing the influence
of methanol on the formation rate of phenolate. The plot
is linear with a slope of 0.36 ( 0.03 M-1. Over this range
of added methanol the measured pH (pH meter) of the
buffer-methanol mixtures showed only slight variations
(<0.1 pH units).
No reaction occurs without hydroxide. Under these
conditions absorption spectra and HPLC indicate that
only starting material is present after 10 min photolysis.
This is true both in water and in water containing 10%
methanol by volume. When 1 is photolyzed at pH 12,
HPLC shows the presence of 1 (eluting at 1.8 min) and
2 (eluting at 3.5 min), as expected. However, when the
photolysis is carried out in alkaline solution with 7%
(8) Scaiano, J . C. J . Am. Chem. Soc. 1980, 102, 7747.
(9) Scaiano, J . C.; Tanner, M.; Weir, D. J . Am. Chem. Soc. 1985,
107, 4396.
(10) De J ongh, R. O.; Havinga, E. Rec. Trav. Chim. Pays Bas 1966,
85, 275.

3-Nitroanisole Photohydroxylation
J . Org. Chem., Vol. 62, No. 25, 1997 8779
Ta ble 1. Tr ip let 1 Lifetim e in Va r iou s Solven ts (λm on
)
400 n m )
solvent
τT, ns
CF3CH2OH
H2O
H2O, pH 12
CH3OH
2-propanol
tert-butyl alcohol
CH3CNa
8000
3500
156
760
70
50
30
a
CH3CN was not dried.
could be eliminated by repetitive recrystallization of the
nitroanisole.
The lifetime of the 3-nitroanisole triplet state depends
strongly on the solvent. It varies from about 8 µs in
trifluoroethanol (TFE) down to about 30 ns in acetonitrile
(Table 1). Attempts to observe the triplet in cyclohexane
proved futile.
This variation in lifetime is due to the influence of
hydrogen bonding on the triplet state.13 In hydrogen
bonding solvents the lowest energy triplet of 1 is believed
to be a ππ* charge transfer state with the π* electron
strongly localized on the nitro group.13 It has been
reported that strong hydrogen bonding of the nitro group
inhibits certain bending vibrations which facilitate rapid
triplet decay in non-hydrogen bonding solvents. The
triplet lifetime of 1 in cyclohexane is thus expected to be
in the subnanosecond regime.13
The large difference between the triplet 1 lifetimes in
methanol and 2-propanol is surprising. One possible
explanation is that 2-propanol, a good hydrogen-donor,14
photoreduces triplet 1. This seems unlikely as we
detected no permanent change in sample absorbance
when 1 was photolyzed in neat 2-propanol. In addition,
the transient spectrum of 1 observed in 2-propanol was
that of triplet 1 at all times after the laser pulse. The
fact that triplet 1 is also short-lived in tert-butyl alcohol,
a much poorer hydrogen donor than 2-propanol,14 also
argues against photoreduction as the source of the short
triplet lifetime in these solvents.
F igu r e 3. Triplet absorption spectrum of 1 recorded in water
1 µs after excitation of a 0.3 mM solution by 355 nm laser
pulses. The insert shows a typical decay trace recorded at 400
nm.
methanol by volume an additional, very small peak is
observed eluting at about 1.4 min. When an unphoto-
lyzed sample of 1 in pH 12 buffer containing 7% methanol
is treated by the extraction procedure outlined above and
then injected into the HPLC, no such peak is observed.
This is also the case when methanol alone is injected into
the HPLC. Prolonged irradiation (30 min) of 3-nitrophe-
nol in pH 12 buffer containing 10% methanol did not
generate this additional peak in the HPLC. Evidently
the peak at 1.4 min is due to some interaction of excited
3-nitroanisole and methanol in basic aqueous solution.
It should be emphasized that this peak corresponds to
a trace product representing no more than 1% of the total
material. Attempts to further identify this trace sub-
stance via GC-MS were unsuccessful.
The triplet lifetime of 1 is determined by the strength
of the hydrogen bond to it.13 The hydroxyl group of tert-
butyl alcohol is sterically hindered enough that it cannot
form a particularly strong hydrogen bond with triplet 1.
This results in a short triplet lifetime. A similar situation
likely pertains for 2-propanol as solvent.
Similar irradiation experiments were carried out using
15% 2-propanol or 5% tert-butyl alcohol instead of
methanol. In these systems no reduction in the rate of
product formation was observed, and no species other
than 1 and 2 were detected in HPLC.
For our purposes it is sufficient to note that the triplet
lifetime changes little over the range of alcohol concen-
trations used in the steady-state experiments (0 to 6 M
methanol, 0 to 2 M 2-propanol, 0 to 0.6 M tert-butyl
alcohol). This is illustrated in Figure 4 which shows how
the triplet decay rate varies as a function of water
concentration in water:methanol mixtures. A 6 M aque-
ous solution of methanol corresponds to a water concen-
tration of about 44 M. Clearly there is no significant
difference in triplet decay rate between pure water and
44 M water. Similar results were obtained with mixtures
of water and 2-propanol or tert-butyl alcohol.
Whatever its cause, the short lifetime of triplet 1 in
neat propanol and tert-butyl alcohol has no consequences
for triplet behavior in dilute alcohol solutions at pH 12.
Under these conditions the triplet lifetime is determined
by the hydroxide concentration.
Tr a n sien t Sp ectr u m a n d Lifetim e of th e 3-Ni-
tr oa n isole Tr ip let. Laser excitation of dilute solutions
of 1 in water with 355 nm pulses results in formation of
a transient species with the spectrum shown in Figure
3. This transient, which forms within the laser pulse
duration, exhibits two absorption bands, an intense band
at 307 nm and a weaker band at 400 nm. The spectrum
exactly matches that assigned previously to the triplet
state of 3-nitroanisole in water.12
In water the triplet decays via pseudo-first-order
kinetics, with a 3 to 4 µs lifetime; the signal returns back
to the preexcitation baseline level following triplet decay
(Figure 3, insert). A residual signal following triplet
decay had been reported previously.12 The spectrum of
this residual showed a weak absorbance near 420 nm and
a strongly rising absorbance below 340 nm. This residual
was also observed in some of our early experiments but
(11) Cornelisse, J .; Havinga, E. Chem. Rev. (Washington, D.C.) 1975,
75, 353.
(12) van Zeijl, P. H. M.; van Eijk, L. M. J .; Varma, C. A. G. O. J .
Photochem. 1985, 29, 415.
(13) Varma, C. A. G. O.; Plantenga, F. L.; Huizer, A. H.; Zwart, J .
P.; Bergwerf, P.; van der Ploeg, J . P. M. J . Photochem. 1984, 24, 133.
(14) Lissi, E. A.; Encinas, M. V. In Handbook of Organic Photo-
chemistry; Scainao, J . C., Ed.; CRC Press: Boca Raton, 1989; Vol. 2.

8780 J . Org. Chem., Vol. 62, No. 25, 1997
Evans et al.
F igu r e 4. Dependence of 3-nitroanisole triplet decay rate, kobs
,
F igu r e 5. Stern-Volmer plot of sorbic acid quenching of
triplet 1 in methanol. The triplet decay was monitored at 400
nm. [1] ) 0.3 mM.
on water concentration in mixtures of water and methanol.
The anisole concentration was 0.3 mM and the triplet decay
monitored at 400 nm.
cations of methylviologen and of the sensitizer. Based
on the monitoring of the triplet decay at 460 nm, MV2+
reacts only slowly with 1. Nonetheless, the absorption
spectrum of the MV+• radical cation, with bands at 390
and 600 nm, was readily observed following 1 triplet
decay. Further, the MV+• radical cation grew in as the
triplet decayed, monitoring at 600 nm. The rate of
growth of the MV+• signal matched the rate of triplet
decay. Thus we conclude that MV2+ quenching is due to
electron transfer and that the 1 triplet is at best a poor
electron donor. We were unable to observe evidence for
the radical cation of 1, but this is probably due to spectral
overlap. The 1 radical cation is expected to have a strong
band below 400 nm and only a weak band near 500 nm.18
1,4-Cyclohexadiene and phenol are expected to quench
triplet 1 via hydrogen abstraction.19-21 In the latter case
this should lead to formation of the readily detectable
phenoxy radical.19 In the former case the quenching
process is expected to yield the 1,4-cyclohexadienyl
radical. Both of these radicals have distinct spectra.19,22,23
Quenching by 1,4-cyclohexadiene occurs in methanol
and methanol:water with a rate constant close to 2 × 108
M-1 s-1 (Table 2). The quenching process leads to
formation of a new transient which grows in as the triplet
decays (monitoring near 300 nm), is long-lived (half-life
ca. 40 µs), and decays via second-order kinetics. The
spectrum of this transient is shown in Figure 6 and is
consistent with the known spectrum of the 1,4-cyclohexa-
dienyl radical.21,23 The insert shows the initial decay of
the triplet followed by the growth of the radical. The
Ta ble 2. Ra te Con sta n ts for Qu en ch in g of Tr ip let 1 by
Va r iou s Su bstr a tes
solvent
water
quencher
kq, ×108 M-1 s-1
O2
30
water
sorbic acid
7.5
methanol
1:1 MeOH:H2O
methanol
water
water, pH 11.5
1:1 MeOH:H2O
methanol
water
water
water
water
water, pH 11.5
water, pH 11.5
sorbic acid
15
16.3
50
36.9
49.7
1.5
1.8
0.2
4.1
1.8
1,3-cyclohexadiene
1,3-cyclohexadiene
phenol
phenolate
1,4-cyclohexadiene
1,4-cyclohexadiene
MV2+
OH-
CN-
-
NO2
51.3
0.02a
0.03b
methanol
methanol
a
b
By laser flash photolysis. By Stern-Volmer analysis of
product yield (3-nitrophenolate formation).
Qu en ch in g of th e 3-Nitr oa n isole Tr ip let. Quench-
ing data are summarized in Table 2. Unless otherwise
noted the rate constants were determined by monitoring
the triplet decay rate constant, kobs, at 400 nm as a
function of quencher concentration and fitting the decay
rates to the expression
kobs ) ko + kq[Q]
ko is the triplet decay rate constant in the absence of
added quencher, Q, and kq is the bimolecular rate
constant for the quenching process. Figure 5 shows a
typical quenching plot.
(17) Patterson, L. K.; Small J r., R. D.; Scaiano, J . C. Radiat. Res.
1977, 72, 218.
Oxygen, sorbic acid, and 1,3-cyclohexadiene can be
considered to be energy transfer quenchers,15,16 and in
each case they quench 1 without product formation.
Methylviologen (MV2+) is a well-known electron accep-
tor.17 It reacts rapidly with triplets to yield the radical
(18) Tamminga, J . J .; van den Ende, C. A. M.; Warman, J . M.;
Hummel, A. Recl. Trav. Chim. Pays-Bas 1979, 98, 305.
(19) Das, P. K.; Encinas, M. V.; Scaiano, J . C. J . Am. Chem. Soc.
1981, 103, 4154.
(20) Encinas, M. V.; Scaiano, J . C. J . Am. Chem. Soc. 1981, 103,
6393.
(21) Effio, A.; Griller, D.; Ingold, K. U.; Scaiano, J . C.; Sheng, S. J .
J . Am. Chem. Soc. 1980, 102, 6063.
(15) Carmichael, I.; Hug, G. L. In Handbook of Organic Photochem-
istry; Scaiano, J . C., Ed.; CRC Press: Boca Raton, 1989; Vol. 1.
(16) Turro, N. J . Modern Molecular Photochemistry; University
Science Books: Sausalito, 1991.
(22) Schuler, R. H.; Buzzard, G. K. Int. J . Radiat. Phys. Chem. 1976,
8, 563.
(23) Bunce, N. J .; Ingold, K. U.; Landers, J . P.; Lusztyk, J .; Scaiano,
J . C. J . Am. Chem. Soc. 1985, 107, 5464.

3-Nitroanisole Photohydroxylation
J . Org. Chem., Vol. 62, No. 25, 1997 8781
F igu r e 7. Transient spectrum of the 1 radical anion recorded
56 µs after excitation of an aqueous solution containing 0.3
mM 1 and 20 mM NaNO2.
F igu r e 6. Transient spectrum recorded following 355 nm
excitation of 0.3 mM 1 in methanol in the presence of 53 mM
1,4-cyclohexadiene. The spectrum was recorded 3 µs after the
laser pulse. The inset shows the initial decay of triplet 1
followed by a growth monitored at 300 nm. The spectrum is
probably a mixture of the cyclohexadienyl radical and the
neutral 1H• radical.
It appears that no value for the standard redox
potential of 1 in water has been reported due to the
irreversibility of single electron reduction of nitroben-
zenes in this solvent.25 Values reported for m-methyl,
m-chloro, and m-amino nitrobenzene in aqueous surfac-
tant solutions cover a narrow range between -0.4 and
-0.47 V vs NHE26 (based on the relationship VNHE ) VSCE
+ 0.24V27). If we assume that the redox potential for 1
will be roughly similar and take the redox potential for
negative signal near 340 nm occurs because the absor-
bance of ground state 1 is larger in this wavelength range
than that of the transients; as a result bleaching is
observed.
There is no conclusive evidence for the formation of
the neutral radical, 1H•, expected as a product in the
photoreduction of 1 by 1,4-cyclohexadiene. The corre-
sponding neutral radical formed in the photoreduction
of 3,5-dinitroanisole has a strong absorption band below
400 nm and a very weak band in the 400 to 500 nm
range.18 Corresponding bands of 1H• may overlap with
the spectrum of the 1,4-cyclohexadienyl radical since the
latter has only weak absorptions in the visible region.21,23
This may also account for the unusually broad UV band
in the spectrum of Figure 6.
the NO2-/NO2 couple in water as 1.03 V vs NHE28 and
•
the 1 triplet energy as 59.8 kcal/mol,29 we can apply the
Rehm-Weller equation30 to this system. Assuming the
Coulombic term is zero, we find that the electron transfer
from nitrite to triplet 1 is strongly exergonic with ∆G° <
-26 kcal/mol. Such a large negative value is consistent
with a diffusion controlled electron-transfer process. We
thus suggest that nitrite quenching of triplet 1 proceeds
via an electron-transfer mechanism leading to formation
of the 1 radical anion. The fact that the nitro group is
so electron deficient may explain why oxygen did not
quench the radical anion signal in these experiments.
When hydroxide is present, aqueous solutions of 1
undergo irreversible photoreaction. Transient absorption
measured under these conditions shows a residual ab-
sorption following decay of the triplet. The spectrum of
the residual is shown in Figure 8. A similar spectrum is
observed when methanol is present. The shape of this
residual spectrum suggests that the long-lived species in
basic medium is the photoproduct, 3-nitrophenolate, and
is in agreement with previous work.12
Phenol quenches triplet 1 at close to the diffusion
controlled rate in unbuffered water (pH ca. 6.5 in the
presence of 1.5 mM phenol). In this case the rate
constant was obtained by monitoring triplet decay at 460
nm, a wavelength where the phenoxy radical does not
absorb.19,22 After triplet decay a typical phenoxy radical
19,22
spectrum
was observed, exhibiting sharp peaks at
380 and 400 nm and a stronger absorbance rising to the
UV. Phenol was also used as a quencher under basic (pH
11.5) conditions. At this pH the phenol should be
completely in the phenolate form (pKa ) 9.9 for phenol.24)
The diffusion-controlled quenching process (monitored at
460 nm) lead to observation of the phenoxy radical
spectrum as well as to a weak absorbance extending out
to near 500 nm.
The rapid quenching of triplet 1 by NaNO2 lead to
formation of a new transient which decayed via second-
order kinetics (half-life ca. 100 µs). The spectrum of the
new transient is shown in Figure 7. It closely matches
that reported for the radical anion of 3,5-dinitroanisole.18
The location of the short wavelength band of the
nitrophenolate spectrum (Figure 8) is slightly different
than is observed in a UV-vis spectrophotometer (vide
(25) Kemula, M. D.; Kok, G. L. In Encyclopedia of Electrochemistry
of the Elements; Bard, A. J ., Lund, H., Eds.; Marcel Dekker: New York,
1979; Vol. XII.
(26) Holleck, L.; Kastening, B. Rev. Polarog. 1963, 11, 129.
(27) Kavarnos, G. J . Fundamentals of Photoinduced Electron Trans-
fer; VCH Publishers: New York, 1993.
(28) Wilmarth, W. K.; Stanbury, D. M.; Byrd, J . E.; Po, H. N.; Chua,
C. P. Coord. Chem. Rev. 1983, 51, 1983.
(29) Gunst, G. P.; Havinga, E. Tetrahedron 1973, 29, 2167.
(30) Wayner, D. D. M. In Handbook of Organic Photochemistry;
Scaiano, J . C., Ed.; CRC Press: Roca Baton, 1989; Vol. 2.
(24) Cram, D. J . Fundementals of Carbanion Chemistry; Academic
Press: New York, 1965.

8782 J . Org. Chem., Vol. 62, No. 25, 1997
Evans et al.
Under basic conditions, methanol inhibits the forma-
tion of 3-nitrophenolate and also quenches the triplet of
3-nitroanisole. The reduction in phenolate formation rate
as a function of methanol concentration yields a Stern-
Volmer constant KSV ) 0.36 ( 0.03 M-1. In the absence
of methanol the lifetime measured for the triplet anisole
at pH 12 is 156 ns. The quotient of KSV and the triplet
lifetime yields, the rate constant for methanol quenching
of the excited-state responsible for the formation of the
phenolate; in this case (2.6 ( 0.2) × 106 M-1 s-1. This is
in reasonable agreement with the directly measured
value for methanol quenching of triplet 3-nitroanisole,
kMeOH ) (1.8 ( 0.1) × 106 M-1 s-1. We conclude that
methanol inhibits phenolate formation because it quenches
triplet 3-nitroanisole at pH 12.
Discu ssion
The rate constant for oxygen quenching of triplet 1 is
near diffusion controlled as expected.14 By contrast,
sorbic acid quenching in water is rather slow. The triplet
energies of 2,4-hexadienes are near 60 kcal/mol15,16 while
the reported triplet energy for 1 is 59.8 kcal/mol.29 The
energy transfer from triplet 1 to sorbic acid will be
essentially thermoneutral and is thus expected to be
significantly below the diffusion limit.16 The rate con-
stant for this quenching process increases in methanol.
This might reflect a destabilization of triplet 1 in this
less hydrogen bonding solvent13 and/or it may simply
reflect the fact that diffusion is faster in methanol than
water. The results with 1,3-cyclohexadiene (ET ) 52.3
kcal/mol15) suggest that the increase in rate constant for
quenching in methanol is due to an increased rate of
diffusion in that solvent, as energy transfer from 1 to 1,3-
cyclohexadiene should be rather exothermic in both
methanol and water rich solvents.
F igu r e 8. Transient spectra recorded 360 ns (circles) and 7
µs (squares) after 355 nm excitation of 0.4 mM 1 aqueous
solution containing 10% methanol and 2.5 mM NaOH.
The transient spectrum shown in Figure 6 supports
hydrogen abstraction as the mechanism for quenching
of triplet 1 by 1,4-cyclohexadiene. The similarity of the
rate constants for the reaction in methanol and in water:
methanol indicate that the extent of hydrogen bonding
does not strongly influence the ability of 1 to abstract
hydrogen.
At first glance the phenol results seem straightforward
evidence for direct hydrogen abstraction from phenol by
triplet 1. However, it is interesting to compare the rate
constant observed here with that for phenol quenching
of ketone triplets in hydrogen bonding solvents.19 For
example, phenol quenches benzophenone triplet with a
rate constant of 1.3 × 109 M-1 s-1 in benzene. This is
reduced to only 8.0 × 107 in wet acetonitrile.19 This
reduction has been attributed to hydrogen bonding of the
phenolic hydroxyl group. Clearly this is not a factor in
the interaction of triplet 1 and phenol. Most likely, the
initial interaction here involves electron or charge trans-
fer.
F igu r e 9. Transient decay traces recorded at 400 nm follow-
ing 355 nm excitation of 0.3 mM 1 in pH 11.5 phosphate buffer
with 0 M (circles) and 5 M (squares) methanol added.
supra). This simply reflects that the spectra of Figure 8
are difference spectra. There is significant spectral
overlap between the nitrophenolate and the ground-state
absorption spectrum of its precursor, 3-nitroanisole, at
and below 300 nm which results in slight distortion of
the difference spectrum.
Quenching of aromatic ketone triplets by phenolate is
a well-known process31 and proceeds via an electron-
transfer mechanism to yield the phenoxy radical and the
ketone radical anion. In the case of triplet 1 the phenoxy
radical is observed. It decays via complicated kinetics
which include a major second-order component and a
residual signal at long delay times. In addition to the
sharp bands near 380 and 400 nm a weak absorption
extending to about 500 nm is observed. We attribute this
The cyanide anion is also known to act as a nucleophile
in the photosubstitution of triplet 1.1 The rate constant
for triplet quenching by cyanide is included in Table 2
simply by way of comparison with that for hydroxide
quenching.
At pH 12 methanol quenches the triplet state of
3-nitroanisole as shown in Figure 9. The rate constant
for quenching is kMeOH ) 1.8 ( 0.1 × 106 M-1 s-1. In the
absence of hydroxide the alcohol does not quench triplet
anisole at methanol concentrations up to ca. 6 M (see
Figure 4).
(31) Das, P. K.; Bhattacharyya, S. N. J . Phys. Chem. 1981, 85, 1391.

3-Nitroanisole Photohydroxylation
J . Org. Chem., Vol. 62, No. 25, 1997 8783
to the 1 radical anion. The spectrum of the 3,5-dini-
troanisole radical anion shows a band centered near 400-
410 nm and extending to slightly greater than 500 nm.18
[methoxide]
pH ) pKa + log
[methanol]
we can estimate the concentration of methoxide in
equilibrium with, for example, 4.94 M (the second largest
value used in the steady-state experiments) methanol at
pH 12. Taking 15.5 as the pKa of methanol34 we find that
[methoxide] is 1.6 mM. This means that the [OH-] will
remain essentially constant at 0.01 M. That is the main
reaction channel will be photohydroxylation. Even so
some methoxide does form and it will quench the ni-
troanisole triplet. The rate constant for hydroxide quench-
ing of triplet 3-nitroanisole is (4.1 ( 0.3) × 108 M-1 s-1
(Table 2) but for methoxide quenching in methanol it is
Nitrite quenching seems to lead to formation of the 1
radical anion (Figure 7). Clearly the band in the 400 to
500 nm region is consistent with that observed in the
phenolate quenching of triplet 1.
The question arises as to the nature of the quenching
process between methanol and the triplet anisole. One
possibility is that methanol simply lowers the hydroxide
concentration by dilution. This would tend to lengthen
the lifetime, not shorten it. In addition, no reduction in
rate of product formation was observed when tert-butyl
alcohol and 2-propanol were used. Also, this would not
explain the formation of the material with the 1.4 min
elution time observed in HPLC.
kmethoxide ) 3.4 × 109 M-1 s-1 12
.
Assuming the latter to
be solvent independent, we can predict the value of kobs
,
the rate of triplet decay, at pH 12 and 4.94 M methanol
compared to its value at pH 12 without methanol. That
is we can estimate the contribution of methoxide to the
observed rate of triplet decay. We use the Stern-Volmer
expression
Another possibility is that addition of methanol in-
creases the solution concentration of dissolved oxygen.
Great pains were taken to ensure degassing was as
complete as possible in all experiments. Further, this
effect should also be active if 2-propanol or tert-butyl
alcohol are added. As indicated previously, neither of
these had any effect on the rate of triplet decay at pH
12.
kobs ) ko + kmethoxide[methoxide]
where ko is the rate of triplet decay at pH 12 in the
absence of methanol. Its value is 1/156 ns ) 6.4 × 106
s-1. So the predicted value of kobs at 4.94 M methanol,
which corresponds to 1.6 mM methoxide, is 1.18 × 107
Additional possible interactions of methanol with
triplet 1 in basic solution which might lead to an increase
in the triplet decay rate are hydrogen abstraction by the
triplet from the alcohol or changes in the triplet lifetime
due to changes in the hydrogen bonding power of the
solvent as methanol is added. Neither of these effects
will play a role here as indicated by the results presented
in Figure 4.
s-1
. The experimentally obtained value for the same
solution is 1.20 × 107 s-1. Thus it seems reasonable to
conclude that the observed triplet quenching in strongly
basic aqueous solution and the reduction in rate of
formation of the nitrophenolate both arise because small
amounts of methoxide form, providing an alternate decay
channel for the anisole triplet. Methoxide quenching is
a reactive process which ultimately leads to stable
products as indicated in the literature32,33 and as observed
in our HPLC study. This view also explains why we see
no effect of tert-butyl alcohol or 2-propanol on the rate of
product formation and why no additional products are
detected during HPLC of the 2-propanol system. These
alcohols are simply too weak acids to form significant
yields of their corresponding alkoxides in aqueous solu-
tion, even at pH 12. In support of this we note that pKa
for tert-butyl alcohol is 19.24
A likely interaction of methanol with triplet anisole is
photoreduction leading to nitroso formation. Photore-
duction of nitrobenzenes occurs in 50% aqueous methanol
with rather high chemical yields when hydroxide is
present but is very inefficient in its absence.32,33 Pho-
toreduction is also rather efficient in neat alcohols when
alkoxide ions are present.32,33 Furthermore, the following
process, where the star refers to isotopically labeled
carbon, has been reported to occur with a rate constant
of 3.4 × 109 M-1 s-1. The products were determined at
0.25 M methoxide.3,12
The case of TFE is special in this context. Clearly this
alcohol is more acidic than methanol, and yet it has no
influence on the decay of triplet 1 in basic solution. We
explain this by noting that the same feature that makes
TFE quite acidic, i.e., the trifluromethyl group, also make
its alkoxide a poor nucleophile. Simply stated, the
negative charge on the TFE anion is stabilized by the
trifluromethyl group and is thus not particularly reactive.
In conclusion we suggest that the influence of methanol
on the rate of the alkaline photolysis of 3-nitroanisole
arises from this alcohol’s nature as a weak, but not too
weak, acid.
A process like this would be compatible with the data
reported here. Small amounts of methoxide formed in
the very basic buffer solution would be available as triplet
state quenchers leading ultimately to photoreduction
products. Yields would be relatively small because of the
low methanol concentration (e6 M) and therefore low
(potential) methoxide concentration. For example, on the
basis of the Henderson-Hasslebach equation
Ack n ow led gm en t. This work was supported in part
by the Icelandic Science Foundation and the Research
Fund of the University of Iceland. J .C.S. thanks
NSERC (Canada) for support under its research grants
program; this support has been used to assist C.H.E.’s
visit to the University of Ottawa.
(32) Frolov, A. N.; Kuznetsova, N. A.; El’tsov, A. V.; Rtischev, N. I.
J . Org. Chem. (USSR) Eng. Transl. 1973, 9, 988.
(33) The Chemsitry of Amino, Nitroso and Nitro Compounds and
their Derivatives, Chow, Y. L., Ed.; J . S. Wiley: New York, 1982; Part
1.
J O971188G
(34) Ballinger, P.; Long, F. J . Am. Chem. Soc. 1960, 82, 795.