Amine Photoreduction of 3-Phenylquinoxalin-2-ones
droanthraquinone and iminium ion, occurring via a
J . Org. Chem., Vol. 65, No. 23, 2000 7957
Ra d ia n t F lu x Deter m in a tion . A solution of Aberchrome
40, ca. 1.0 mM in toluene, was used to estimate the radiant
flux from the Hg lamp at 366 nm. Absorbance at 494 nm was
measured every 30 s of irradiation, and the photon flux was
5
sequence of electron and hydrogen atom transfer, has an
activation free energy of 25.9 kcal/mol.20 This value is in
agreement with the activation free energy obtained for
the MeNQ/TEA system and supports a sequence of
electron and hydrogen atom transfer for the thermal
calculated from the slope of the plot A494 vs time, with ꢀ494
200 M cm and φ ) 0.20.
)
-1
-1
34,35
8
Qu a n tu m Yield . The quinoxalin-2-one photobleaching
-
recovery of XNQH to the parent quinoxalin-2-ones.
quantum yield was estimated by using the time-integrated
absorbed light at 366 nm and the absorbance at 360 nm, with
Φ
XNQH- ) [(∆A360/ꢀ360)/(I° × S)](v/1000), where ∆A360 is the
Con clu sion
absorbance difference at 360 nm, ꢀ360 is the molar extinction
coefficient of quinoxalin-2-one at 360 nm, I° is the photon flux
The preceding results support the mechanism shown
in Scheme 2. Photoexcitation of XNQ leads to its first
-
1
in Einstein s , S is the time-integrated absorbed light, and v
is the volume of irradiated solution.
Gen er a l P r oced u r e for P h otoblea ch in g Stu d ies. Solu-
tions (3 mL) of quinoxalin-2-one with absorbances between
singlet state, which is converted into the triplet via an
efficient intersystem crossing, step 1, with kISC ) 1010
s
-1
for MeNQ. The triplet state is quenched by amine via
electron transfer, step 2, giving the triplet radical-ion
pair. The ion-radical pair can then decay to the starting
compounds, step 5, but back electron transfer seems to
0
2
.20 and 1.40 at 360 nm were purged with N for 20 min in a
10 mm fluorescence quartz cell sealed with a septum. Im-
mediately after purging an aliquot of pure or diluted amine
was added through the septum. The solutions were photolyzed
with a Black Ray UV lamp with a 366 nm filter, and the
absorbance was measured every 30 s at 360 nm on ATI Unicam
UV2 or UV4 spectrophotometers with Vision 2.11 software.
In experiments at high amine concentrations, neutral filters
were used to reduce the photon flux from the lamp to keep
total bleaching times within 5-10 min. For each experiment,
the radiant flux was determined as described previously.
Da r k Recover y. This determination was carried out with
solutions prepared as described for bleaching. A thermostated
cell holder was used to control temperature before and after
bleaching. In these experiments the lamp was at a very short
distance in order to obtain photobleaching times less than 60
s, and the temperature was varied between 20 and 50 °C.
Bleaching was stopped when the blue fluorescence of the
quinoxalin-2-one disappeared to the eye. In all experiments
the final absorbance at 360 nm (A360) was about 0.020 at the
end of bleaching. Absorbance at 360 nm was measured at time
intervals of 60 s up to 5 min and then each 5 min for 1 h. To
minimize further photobleaching by the spectrophotometer
beam, a black shutter interrupted the beam between measure-
ments.
be inefficient, probably as a result of a spin forbidden
transition. A rapid proton transfer, step 3, to yield XNQH•
and the N-methylene radical pair follows formation of
the ion-radical pair in step 2. The radical pair can give
by a second electron transfer, step 4, the ion pair in a
more efficient reaction than step 6 involving homolytic
bond cleavage to transfer a hydrogen atom. Reactions 4
and 6 require intersystem crossing from the presumed
triplet radical pair that precedes them. The quantum
-
-
yield of XNQH , ΦXNQH ) 1.0, shows that the proton-
transfer step is faster than competitive disproportion-
ation of a radical pair in step 6. Also step 6 fits formation
2
of dihydro-quinoxalin-2-ones, XNQH . The ion pair formed
in step 4 can diffuse out of the solvent cage, and in the
presence of water, the iminium ion will be hydrolyzed to
+
H
2
N R
2
R
3
and R
1
CHO in step 7, as supported by detec-
tion of DEA and acetaldehyde in the photolysis of MeNQ/
TEA and HNQ/TEA.
The reverse dark reaction could follow a stepwise
pathway, step 8 of Scheme 2. An electron transfer from
Meta sta ble P r od u ct Detection . Experiments were per-
formed in a Bruker Avance DRX-300 (300 MHz) spectrometer.
-
XNQH to the iminium ion could be induced thermally
Reactions were carried out by direct photobleaching of N
purged solutions containing a weighed amount of the respec-
tive quinoxalin-2-one and excess TEA in CD CN. Solutions
2
-
to regenerate the singlet radical pair, followed by back
hydrogen atom transfer, leading to the starting quinoxa-
lin-2-ones. The activation free energies of 17.4 and 25.9
kcal/mol in acetonitrile and benzene seem reasonable for
the proposed sequence of electron and atom transfers for
3
were prepared directly in a NMR tube and sealed with a
1
septum. During photobleaching some H NMR spectra were
taken in order to test for the maximum concentration of the
metastable species. COSY spectra were taken when the
lifetime of the metaestable photoproduct was long enough. It
2
0
this thermal back reaction.
1
3
The high yield of dihydro products XNQH
2
on continu-
was not possible to obtain a C NMR spectrum of the
irradiated samples of XNQ/TEA because of the low concentra-
ous irradiation reported by Nishio17 can be explained by
-
tion of XNQH generated and the necessary number of scans.
-
considering the high stationary concentration of XNQH
that can react thermally to give XNQH
portionation of the radical pair in equilibrium with the
ion pair.
2
or from dispro-
F lu or escen ce Qu en ch in g a n d En h a n cem en t of Em is-
sion . Measurements were made in a Spex Fluorolog Tau 2
system with DM 3000 software, all at 20 °C, by using the
corrected fluorescence method. Quenching experiments were
2
performed at absorbance 0.2 at 360 nm, in aerated or N -
Exp er im en ta l Section
purged solutions. The entire spectra were scanned to test for
possible bathochromic shifts. Enhancement of emission was
also measured by using the corrected fluorescence method in
Acetonitrile, hexane, chloroform, ethanol, methanol, dichlo-
romethane, and N,N′-dimethylformamide were Merck HPLC
2
N -purged solutions. A 10 mm square quartz cell sealed with
grade. Benzene-d
9.5% were Merck spectroscopic grade. These solvents were
used as received.
6
, acetonitrile-d
3
99%, and chloroform-d
1
a septum was used in both experiments, and pure or diluted
amine was added through the septum. The absorbance of the
solutions was checked between scans to show that quinoxalin-
2-one was not consumed.
9
Triethylamine, diethylamine, isopropylamine, n-butylamine,
tertbutylamine, and aniline (Fluka) were stored over potas-
sium hydroxide pellets and vacuum distilled trap-to-trap,
F lu or escen ce Yield s. Yields were determined from the
-
4
sealed into glass tubes at 10 mmHg, and stored at -18 °C.
Before each experiment a new tube was opened to ensure
freshness of the amine. DABCO from Aldrich was used as
received. DABCO solutions were prepared immediately before
use.
(
34) Heller, H. G.; Langan, J . R. J . Chem. Soc., Perkin Trans. 2 1981,
41-343.
35) The Use of Aberchrome 540 in Chemical Actinometry; Aber-
3
(
chromic Ltd., School of Chemistry and Applied Chemistry, University
of Wales, College of Cardiff, PO Box 912, Cardiff CF1 3TP, U.K.