The Primary Photodynamics of Aqueous Nitrate: Formation of Peroxynitrite

Article abstract of DOI:10.1021/ja030135f

We have examined the photochemical reactions occurring after irradiation at 200 nm of the aqueous nitrate ion, NO₃^-(aq). Using femtosecond transient absorption spectroscopy over the range 194-388 nm, we have characterized the formation and subsequent relaxation of the primary photoproducts of nitrate photolysis. The dominant photoproduct is the cis-isomer of peroxynitrite, which accounts for 48% of the excited state molecules initially produced. A slightly smaller fraction, 44%, of the excited molecules return to the electronic ground state of NO₃^- and relax to the vibrational ground state in 2 ps. The remaining 8% of the molecules initially excited react via the ?NO + ?O₂ ^- or the NO^- + O₂ dissociation channels. Formation of NO₂^- and ?NO₂ is not observed, suggesting that the previous observations of these species in steady-state photolysis are caused by reactions occurring on a longer time scale.

Full text of DOI:10.1021/ja030135f

Published on Web 11/18/2003
The Primary Photodynamics of Aqueous Nitrate: Formation
of Peroxynitrite
Dorte Madsen,† Jane Larsen,‡ Svend Knak Jensen, Søren R. Keiding, and
Jan Thøgersen*
Contribution from the Department of Chemistry, UniVersity of Aarhus,
Langelandsgade 140, DK-8000, Aarhus, Denmark
Received February 24, 2003; Revised Manuscript Received October 13, 2003; E-mail: thogersen@chem.au.dk
Abstract: We have examined the photochemical reactions occurring after irradiation at 200 nm of the
aqueous nitrate ion, NO3-(aq). Using femtosecond transient absorption spectroscopy over the range 194-
388 nm, we have characterized the formation and subsequent relaxation of the primary photoproducts of
nitrate photolysis. The dominant photoproduct is the cis-isomer of peroxynitrite, which accounts for 48% of
the excited state molecules initially produced. A slightly smaller fraction, 44%, of the excited molecules
return to the electronic ground state of NO3- and relax to the vibrational ground state in 2 ps. The remaining
-
8% of the molecules initially excited react via the •NO + •O2 or the NO- + O2 dissociation channels.
Formation of NO2- and •NO2 is not observed, suggesting that the previous observations of these species
in steady-state photolysis are caused by reactions occurring on a longer time scale.
to a recent comprehensive review by Mack and Bolton,3 the
Introduction
-
overall products resulting from the photolysis of aqueous NO3
The photochemistry of the nitrate ion, NO3-, has been studied
extensively since the early experiments by Thiele back in 1907.1
Later work in this area has included modeling of the Martian
surface to explain the result of the Viking Lander experiments,2
while more recent photochemical investigations of nitrate have
been spurred by its involvement in the disinfection of drinking
water.3,4 Currently, there is an intense interest in the bio-
chemistry of the nitrate isomer peroxynitrite, ONOO-, since it
has been implicated in the onset of numerous diseases.5
Physiologically, ONOO- is formed by reaction of superoxide,
•O2-, with nitric oxide, •NO. The need for additional informa-
tion about these detrimental processes warrants detailed inves-
-
are NO2 and O2. Nevertheless, fundamental issues regarding
the kinetics of primary and intermediate reactions remain
unsolved. This is, in part, due to the complexity of the nitrate
chemistry, but also is a consequence of secondary reactions
obscuring the initial steps of the nitrate photolysis in studies
with modest time resolution. In the present investigation, the
photodynamics of aqueous nitrate are studied by far-UV
transient absorption spectroscopy with femtosecond time resolu-
tion, thus facilitating the characterization of the primary
photoproducts.
The isolated nitrate ion is planar (D3h) with the excess
negative charge delocalized over the molecule. The lowest
excited states of nitrate are 3A′′, 1A′′, 3E′, and 1E′ with Franck-
tigations of reactions involving these species. The present study
1
1
-
is thus motivated by the possible formation of ONOO-, •O2
,
Condon excitation maxima of 3.6, 4.1, 4.7, and 6.0 eV, relative
and •NO from far-ultraviolet photolysis of the nitrate ion.6-9
to the X1A′ ground state.10 In polar solvents, the distribution
1
Considering that the chemistry of the nitrate ion has been
studied for a long time, it may seem surprising that uncertainty
concerning its primary photochemistry still prevails. According
of solvent molecules surrounding the nitrate ion tends to localize
the negative charge, thereby reducing the symmetry of the
ground state to an essentially planar structure with one N-O
bond different from the others (C2V) or with all N-O bonds
different (Cs).11-13 The hydrated nitrate ion has a weak n f π*
absorption band around 302 nm,14 and a much stronger π f
π* band with absorption maximum at 200 nm (corresponding
† Present address: FOM-Institute for Atomic and Molecular Physics
(AMOLF), Kruislaan 407, 1098 SJ Amsterdam, The Netherlands.
‡ Present address: Department of Chemical Physics, Chemical Centre,
Lund University, P.O. Box 124, SE-22100 Lund, Sweden.
(1) Thiele, H. Ber. Dtsch. Chem. Ges. 1907, 40, 4914-16.
(2) Plumb, R. C.; Tantayanon, R.; Libby, M.; Xu, W. W. Nature 1989, 338,
633-635.
1
1
to the X A′ f E′ transition in D3h symmetry).11,15 Excitation
1
of both transitions is accompanied by substantial weakening of
(3) Mack, J.; Bolton, J. R. J. Photochem. Photobiol. 1999, A128, 1-13.
(4) Mark, G.; Korth, H.-G.; Schuchmann, H.-P.; Sonntag, v. C. Photochem.
Photobiol. A 1996, 101, 89-103.
(10) Nevostruev, V. A. High Energy Chem. 1986, 20, 329-333.
(11) Waterland, M. R.; Kelley, A. M. J. Chem. Phys. 2000, 113, 6760-6773.
(12) Waterland, M. R.; Stockwell, D.; Kelley, A. M. J. Chem. Phys. 2001, 114,
6249-6258.
(5) Flint, B. M. Free Radical Biol. Med. 2002, 32, 797-803.
(6) Pape`e, H. M.; Petriconi, G. L. Nature 1964, 204, 142-145.
(7) Shuali, U.; Ottolenghie, M.; Rabani, J.; Yelin, Z. J. Phys. Chem. 1969, 73,
3445-3450.
(13) Shen, M.; Xie, Y.; Schaefer, H. F., III; Deakyne, C. A. J. Chem. Phys.
1990, 93, 3379-3388.
(8) Barat F.; Gilles, L.; Hickel, B.; Sutton, J. J. Chem. Soc. A 1970, 1982-
1986.
(14) Blandamer, M. J.; Fox M. F. Chem. ReV. 1970, 70, 59-93.
(15) Maria, H. J.; McDonald, J. R.; McGlynn, S. P. J. Am. Chem. Soc. 1973,
95, 1050-1056.
(9) Richard, J.; Bradley, E.; Lanzendorf, McCarthy, M. I.; Orlando, T. M.;
Hess, W. P. J. Phys. Chem. 1995, 99, 11715-11721.
9
10.1021/ja030135f CCC: $25.00 © 2003 American Chemical Society
J. AM. CHEM. SOC. 2003, 125, 15571-15576
15571

A R T I C L E S
Madsen et al.
the N-O bonds, and breaking of N-O bonds is expected to
ensue.10,11 However, the dissociation processes following the
photoexcitation of aqueous nitrate have had a history of
considerable controversy over the past 35 years. Early photolysis
studies by Daniels et al.16 suggested that excitation in the n f
π* band at 300 nm results in the two primary product channels:
NO3- + hν f NO2- + O
•NO2 + •O-
(1)
(2)
The primary formation of •NO2 was also reported by Barat et
al.8 and Shuali et al.7 upon excitation of the π f π* band,
-
whereas the same studies suggested that NO2 arises from
secondary reactions and not from the direct photodissociation
process (eq 1). These findings were later contested by Bayliss
and Bucat,17 who, on the basis of the excitation studies of both
Figure 1. Absorption spectra of the species potentially involved in the
primary photolysis of aqueous nitrate.
-
the n f π* and the π f π* transition, concluded that NO2
The femtosecond transient absorption measurements presented
in this work produce direct evidence for peroxynitrite formation
by photoisomerization of nitrate within the solvent cage. The
measurements also reveal that the formation of peroxynitrite is
(eq 1) rather than •NO2 (eq 2) is the primary photodissociation
process. Still, more recent steady-state photolysis experiments
performed at 254 nm (right at the intersection of the π f π*
and n f π* absorption bands) indicate that •NO2 + •O- is the
dominant dissociation channel with a quantum yield of 9%,
while NO2 is produced in much smaller yields of less than
0.1%.4 Hence, it appears the photodissociation channels of
aqueous nitrate are not well-established.
accompanied by a minor yield of •O2 or NO-. In contrast,
-
-
-
neither NO2 nor •NO2 is identified among the primary
photolysis products. Apart from unveiling the primary photo-
kinetics of nitrate, the transient absorption measurements offer
a rare opportunity to follow the constitutional isomerization and
vibrational relaxation of small molecular ions.
In addition to photodissociation, numerous photochemical
investigations have shown that, upon excitation of the π f π*
band, the nitrate ion can isomerize to peroxynitrite:6-8
The present investigation of the nitrate photolysis is based
on femtosecond transient absorption spectroscopy, which identi-
fies the species involved in the photodynamics by their steady-
state absorption spectra, depicted in Figure 1. The spectrum of
NO3-(aq) peaks at 200 nm with an extinction coefficient of ꢀ
) 9900 M-1 cm-1, while NO2-(aq) has a maximum extinction
coefficient of ꢀ ) 5500 M-1 cm-1 at 210 nm.3,28 •NO2(aq) and
NO3- + hν f ONOO-
(3)
This photoisomerization could occur directly or by recombina-
tion of primary photoproducts. Considering the latter possibility,
formation of peroxynitrate could in principle occur through
reaction of •NO2 and •O-. However, experiments using high
concentrations of •O- scavengers have deemed this reaction
•O-(aq) both have weak absorption bands (ꢀ < 250 M-1 cm-1
)
with maxima at 400 and 240 nm, respectively,29-31 whereas
the oxygen atom produced in reaction 1 does not absorb within
the spectral range probed. •NO(aq) absorbs only weakly in the
UV-vis region, while its anion, NO-(aq), absorbs below 300
nm and has an extinction coefficient of 1200 M-1 cm-1 at 260
nm.32 Finally, peroxynitrite potentially exists as planar cis- and
trans-isomers. The absorption band of cis-peroxinitrite is
centered at 302 nm with a maximum extinction coefficient of
channel insignificant.4 Likewise, ONOO- may be formed
through reaction of NO- + O2
or by the radical reaction
18-20
-
•NO + •O2 f ONOO-.21-25 So far, however, neither of the
reactants involved in these processes have been identified as
primary products of the aqueous nitrate photolysis. On the other
hand, steady-state photolysis studies of aqueous nitrate,4 am-
monium nitrate,26 and potassium nitrate27 films suggest that
peroxynitrite is formed by direct photoisomerization. These
findings are supported by resonant Raman spectroscopy of the
π f π* excitation, which indicates that excited-state aqueous
nitrate is pyramidal, thus bringing the oxygen atoms closer
together and facilitating the ONOO- formation.11,17
1670 M-1 cm-1 24,33,34 trans-ONOO- has not been observed in
.
aqueous solution, but guided by measurements and calcula-
tions on KOONO in solid Ar-matrices, we estimate that the
absorption spectrum of trans-ONOO-(aq) is shifted toward
longer wavelengths by ∼50 nm relative to that of the cis-
isomer.35,36 The extinction coefficient of trans-ONOO- has not
been determined.
(16) Daniels, M.; Meyers, R. V.; Belardo, E. V. J. Phys. Chem. 1968, 72, 389-
399.
(17) Bayliss, N. S.; Bucat, R. B. Aust. J. Chem. 1975, 28, 1865-1878.
(18) Donald, C. E.; Hughes, M. N.; Thompson, J. M.; Bonner, F. T. Inorg.
Chem. 1986, 25, 2676-2677.
(28) Fischer, M.; Warneck, P. J. Phys. Chem. 1996, 100, 18749-18756.
(29) Lymar, S. V.; Schwartz, H. A.; Czapski, G. J. Phys. Chem. A 2002, 106,
7245-7250.
(19) Hughes, M. N.; Nicklin, H. G. J. Chem. Soc. A 1971, 164-171.
(20) Hughes, M. N.; Biochem. Biophys. Acta 1999, 1411, 263-272.
(21) Beckman, J. S.; Beckman, T. W.; Chen, J.; Marshall, P. A.; Freeman, B.
A. Proc. Natl. Acad. Sci. U.S.A. 1990, 87, 1620-1624.
(30) Gra¨tzel, M. V.; Henglein, A.; Lilie, J.; Beck, G. Ber. Bunsen-Ges. Phys.
Chem. 1969, 73, 646-653.
(22) Goldstein, S.; Czapski, G. Free Radical Biol. Med. 1995, 19, 505-510.
(23) Huie, R. E.; Padmaje, S. Free Radical Res. Commun. 1993, 18, 195-199.
(24) Kissner, R.; Nauser, T.; Bugnon, P.; Lye, P. G.; Koppenol, W. H. Chem.
Res. Toxicol. 1997, 10, 1285-1292.
(31) Rabani, J. AdV. Chem. Ser. 1968, 81, 131-152.
(32) Seddon, W. A.; Fletcher, J. W.; Sopchyshyn, F. C. Can. J. Chem. 1973,
51, 1123-1130.
(33) Løgager, T.; Sehested, K. J. Phys. Chem. 1993, 97, 6664-6669.
(34) Hughes, M. N.; Nicklin, H. G. J. Chem. Soc. A 1968, 450-452.
(35) Lo, W.-J.; Lee, Y.-P.; Tsai, J.-H. M.; Beckman, J. S. Chem. Phys. Lett.
1995, 242, 147-152.
(25) Blough, N. V.; Zafiriou, O. C. Inorg. Chem. 1985, 24, 3504-3505.
(26) Koch, T. G.; Holmes, N. S.; Roddis, T. B.; Sodeau, J. R. J. Phys. Chem.
1996, 100, 11402-11407.
(27) Plumb, R. C.; Edwards, J. O. J. Phys. Chem. 1992, 96, 3245-3247.
(36) Krauss, M. Chem. Phys. Lett. 1994, 222, 513-516.
9
15572 J. AM. CHEM. SOC. VOL. 125, NO. 50, 2003

The Primary Photodynamics of Aqueous Nitrate
A R T I C L E S
Experimental Section
The double-beam transient absorption spectrometer utilized in this
work is improved relative to that of previous studies.37 Briefly, a 1
kHz Titanium-Sapphire laser system emitting 90 fs pulses with a pulse
energy of 0.75 mJ is frequency quadrupled to generate the 200 nm
pulse used to initiate the photolysis. The ∼5 µJ pump pulse is modulated
at 0.5 kHz by a mechanical chopper synchronized to the 1 kHz pulse
repetition rate and sent through a scanning delay line and a λ/2
waveplate before it is focused through the sample by a f ) 50 cm
parabolic mirror. The probe pulses covering the spectral range from
460 to 3000 nm are generated by a two-stage optical parametric
amplifier (OPA) pumped at 400 nm. Probe pulses ranging from 230 to
460 nm are produced by frequency doubling the pulses from the OPA
in a 0.2 mm BBO crystal, while the spectral region between 214 and
300 nm is covered by frequency mixing the 400 nm pulses with pulses
from the OPA. Likewise, mixing the residual 266 nm pulses with OPA
pulses generates probe pulses in the spectral range from 194 to 220
nm. The probe beam is then split into a signal and a reference beam.
The signal beam is focused onto the sample by an f ) 20 cm CaF2
lens and probes the sample inside the area defined by the pump beam.
Signal and reference pulses are then detected by matched photodiodes
and boxcar integrators before being processed by a digital lock-in
amplifier referenced to the 0.5 kHz modulation of the pump pulse.
Figure 2. Contour plots of the induced transient absorption of aqueous
NO3 when excited at 200 nm. The absorption transients cover the initial
-
18 ps and are measured at 0.2 eV intervals in the spectral range from 194
to 368 nm with 250 fs time resolution. Blue, green, and yellow colors denote
negative absorption induced for instance by the disappearance of absorbing
species, while orange and red areas indicate increased absorption pertaining
to the formation of new species. Thus, the negative absorption with
maximum at 206 nm pertains to the removal of ground-state nitrate ions
by the photolysis pulse. The red area spanning the first 2 ps in the spectral
range from 230 to 360 nm indicates the formation of vibrationally excited
The sample consisted of a ∼0.1 mm jet of 0.014 M aqueous KNO3
solution. The flow was adjusted to give a fresh sample for every laser
pulse. No measurable degradation of the KNO3 solution was observed
during the measurements, but to avoid potential build-up of permanent
photoproducts the solution was replaced regularly. The reproducibility
of the transient absorption data was tested among consecutive scans as
well as by repeating the measurements on different days using different
samples. From numerous measurements, we found that the data could
be measured on a common absorption scale with an uncertainty of less
than (10%. The uncertainties quoted in the remainder of this work
refer only to uncertainties pertaining to our measurements, as most
referenced spectral data are reported without error estimates.
-
NO3 molecules, while the induced absorption peaking at 310 nm
represented by the wide orange band results from the formation of
peroxynitrite.
The experimental data are dominated by a strong negative
absorption from 194 to 225 nm caused by the removal of
ground-state NO3-(aq) molecules by the photolysis pulse. The
transient absorption reaches its minimum immediately after
termination of the two-photon absorption peak and recovers to
56% of its minimum value on a 2 ps time scale. Concurrent
with the removal of ground-state nitrate molecules, strong
induced absorption emerges in the spectral range from 230 to
360 nm, as indicated by the red area in Figure 2. This absorption
feature appears within 250 fs and rapidly shifts toward shorter
wavelengths, approaching the absorption maximum of ground-
state NO3-(aq) at 200 nm as the probe delay increases. Similar
spectral behavior was observed in the photolysis of •ClO2(aq)42
and CS2(aq).43 This was assigned to vibrationally relaxing •ClO2
and CS2 molecules on the basis of excellent agreement with
calculated transient absorption dynamics of these species.
Independent resonant Raman spectroscopy studies on photo-
excited •ClO2(aq)44 further supported this assignment. The lack
of potential energy surfaces of NO3-(aq) prevents a comparison
with calculated absorption dynamics of vibrationally relaxing
NO3-(aq), but on the basis of the strong resemblance with the
spectral shifts reported in the well-established cases of •ClO2(aq)
and CS2(aq), we ascribe the induced absorption transient ranging
from 230 to 360 nm to vibrationally excited NO3-. Conse-
Results
The photoinduced transient absorption, ∆A(λ,t), of aqueous
-
NO3 produced by the 200 nm photolysis pulse and measured
at 0.2 eV (1613 cm-1) intervals in the spectral range from 194
to 368 nm is presented in the color contour plot of Figure 2.
The transient absorption is also depicted at four selected probe
wavelengths, 326, 270, 230, and 206 nm, in Figure 3 to clearly
display all absorption dynamics. Common to all curves in
Figures 2 and 3 is an initial sharp peak, which mainly originates
from the coherent absorption of one pump and one probe photon
in water.38 On the basis of the prompt response of the two-
photon absorption signal, we utilize the position and width of
this peak to determine the point of zero delay and the temporal
resolution (250 fs), respectively. Additional measurements on
-
aqueous NO3 at a probe wavelength of 700 nm (not shown)
found no sign of hydrated electrons, indicating that electron
photodetachment of NO3-(aq) is insignificant.39,40 Finally,
measurements performed on neat water showed that photo-
products pertaining to the photolysis of water contribute less
than 1 mOD across the investigated spectrum.41
-
quently, we infer that the excited NO3 molecules transfer to
the electronic ground state within 250 fs and vibrationally
equilibrate during the subsequent 2 ps. The return to the
electronic ground state may involve intermediate electronic
states, but assuming the Franck-Condon excitation energies of
(37) Thomsen, C. L.; Madsen, D.; Thøgersen, J.; Byberg, J. R.; Keiding, S. R.
J. Chem. Phys. 1999, 111, 703-710.
(38) Thomsen, C. L.; Madsen, D.; Keiding, S. R.; Thøgersen, J. J. Chem. Phys.
1998, 110, 3453-3462.
(42) Thøgersen, J.; Thomsen, C. L.; Poulsen, J. Aa.; Keiding, S. R. J. Phys.
Chem. A 1998, 102, 4186-4191.
(39) Meyerstein, D.; Treinin, A. Trans. Faraday Soc. 1961, 57, 2104-2112.
(40) Barat, F.; Hickel, B.; Sutton, J. Chem. Commun. 1969, 125-126.
(41) Thomsen, C. L.; Madsen, D.; Keiding, S. R.; Thøgersen, J. J. Chem. Phys.
1998, 110, 3453-3462.
(43) Thomsen, C. L.; Madsen, D.; Thøgersen, J.; Byberg, J. R.; Keiding, S. R.
J. Chem. Phys. 1999, 111, 703-710.
(44) Hayes, S. C.; Philpott, M. P.; Mayer, S. G.; Reid, J. P. J. Phys. Chem. A
1999, 103, 5534-5546.
9
J. AM. CHEM. SOC. VOL. 125, NO. 50, 2003 15573

A R T I C L E S
Madsen et al.
Figure 4. The absorption spectrum measured after 18 ps (squares)
fitted by the steady-state absorption spectra of NO3-(aq), cis-ONOO-(aq),
and •O2-(aq) assuming quantum yields of 44%, 48%, and 8%, respec-
tively.
After 18 ps, the transient spectrum is constant, reflecting that
all photoproducts are thermalized. Hence, the remaining 56%
of the primary products may be identified by comparing the
absorption spectrum after 18 ps shown in Figure 4 with the
spectra presented in Figure 1. Apart from the negative induced
absorption caused by photoexcitation of NO3-, the thermalized
absorption spectrum has a pronounced absorption from 250 to
388 nm with a maximum at 295 nm. This absorption profile is
nearly identical to the spectrum of cis-ONOO-(aq).24,33 The
spectrum of trans-ONOO- in Ar-matrices is shifted by about
50 nm relative to its cis-conformer. Assuming that a similar
spectral shift pertains to aqueous solution, we found that the
spectrum of trans-ONOO-(aq) is incompatible with the meas-
ured absorption. Because none of the other candidates absorb
significantly above 300 nm, the absorption is unambiguously
assigned to cis-ONOO-.
Figure 3. The first 10 ps of the induced transient absorption measured at
326, 270, 230, and 206 nm. The measurement at 326 nm shows the prompt
appearance of vibrationally excited NO3-, and the formation of peroxynitrite
lasting 2 ps. At 270 nm, the absorption resulting from vibrationally excited
NO3- reaches its maximum after 0.6 ps, while the constant absorption after
4 ps originates from peroxynitrate. The absorption caused by vibrationally
The assignment of the ONOO-(aq) absorption leaves a minor
excess absorption from 220 to 270 nm. Detailed analysis reveals
that the channels, NO2 + O (eq 1) and •NO2 + •O- (eq 2),
-
-
excited NO3 appears at longer delays as the wavelength approaches the
suggested by previous experiments to be primary reactions, are
unable to account for the extra absorption. Guided by photolysis
experiments of nitrate salts,9 we instead suggest that the
superoxide ion, •O2-, produced by the reaction
ground-state NO3- absorption maximum at 200 nm, and at 230 nm it peaks
after 1.5 ps. This absorption feature is followed by a slightly negative
absorption signifying the onset of the nitrate absorption band. The negative
absorption at 206 nm indicates the immediate removal and subsequent partial
reformation of ground-state NO3- occurring on a 2 ps time scale. Note the
different absorption scales. See text for details.
-
NO3- + hν f •NO + •O2
(4)
the isolated nitrate ion pertain to aqueous solution as well, we
assumed that the lowest excited state is about 3.6 eV above the
ground state. Hence, the final energy dissipation has to proceed
by vibrational relaxation. The incessant spectral shift of vibra-
may be responsible for the remaining absorption. We find that
a yield of Φ(ONOO-) ) 48 ( 5% cis-ONOO-(aq) and Φ(•O2
)
-
) 8 ( 3% gives the very satisfactory fit to the thermalized
absorption spectrum at 18 ps shown in Figure 4. Hence, no
further species are needed to account for the observed absorp-
tion. Even though the experimental data are accurately modeled
-
tionally relaxing NO3 and the fact that all measured features
will be accounted for by ground-state molecular species exclude
significant population of intermediate electronic states having
lifetimes in excess of 1 ps. Likewise, the (negative) transient
absorption spectrum at 0.5 ps below 230 nm is well ap-
proximated by the ground-state NO3-(aq) spectrum, indicating
that contributions to the transient absorption from excited
electronic and vibrational states of nitrate are of little or no
importance in this region. Hence, we derive a yield of excited
NO3-(aq) molecules returning to the electronic ground state of
Φ(NO3-) ) 44 ( 8% and note that contributions to the
absorption from •O2-(aq) or NO-(aq), to be determined below,
reduce this value by less than 2%.
-
by the •NO + •O2 channel, the spectral feature tentatively
assigned to •O2-(aq) has an uncertainty preventing its un-
ambiguous assignment. Thus, spectral analysis based on the
rather poorly determined absorption spectrum of NO-(aq)32 with
yields of Φ(NO- + O2) ) 11% and Φ(OONO-) ) 45% also
reproduces the experimental data. However, unlike the •NO +
•O2- channel, the formation of NO- + O2 as primary products
of the nitrate photolysis has no precedence in the literature. The
•O2- (or NO-) absorption appears after 3 ps, once the obscuring
-
absorption from vibrationally excited NO3 has passed, and
9
15574 J. AM. CHEM. SOC. VOL. 125, NO. 50, 2003

The Primary Photodynamics of Aqueous Nitrate
A R T I C L E S
remains constant (within the experimental uncertainty) during
the subsequent 15 ps. Because of the ambiguity of the
dissociation channel, we refer to it as [NO + O2]- in the
remainder of this paper.
The assignment of the absorption spectrum at 18 ps to
ONOO-(aq) and •O2 (or NO-(aq)) raises the question if
-
species that are not positively identified could be hiding
inside the experimental uncertainty. We have addressed this
issue by introducing a small variable fraction of the channels
-
NO2 + O and •NO2 + •O- into our data analysis. It appears
-
that our data analysis deteriorates if the fraction of the NO2
+
O and •NO2 + •O- channels exceeds 7% and 5%, respec-
tively. Similarly, if we assume that the unknown extinction
coefficient of trans-ONOO-(aq) is similar to that of the cis-
isomer, we find that a maximum of 5% of the excited NO3
molecules are expected to form trans-ONOO- in the primary
photolysis.
-
Figure 5. The primary photodynamics of NO3-(aq) when excited at 200
nm. The parentheses around •NO and O2 indicate that either •NO or O2 is
negatively charged. The red arrows indicate two alternative reaction paths
for the [NO + O2]- channel. The experiments cannot differentiate between
these two reaction paths.
The cis-ONOO-(aq) absorption transient is seen in Figure 2
as the absorption band centered around 300 nm. The subpico-
second dynamics of ONOO- is obscured by the overlapping
absorption from vibrationally excited nitrate ions. However,
absorption pertaining to ONOO- is clearly present after 1 ps
and attains its equilibrium spectrum without the spectral shift
characteristic of vibrational relaxation (see Figures 2 and 3 at
326 nm). The fast rise of the ONOO- concentration indicates
that the isomerization occurs within the solvent cage, thus
making peroxynitrite a primary product of the nitrate photolysis.
To further elucidate the dynamics of the ONOO- concentration,
we measured with improved statistical sampling the transient
absorption at 310 nm out to 500 ps. The resulting ONOO-
concentration increases steadily by 3 ( 1% from its value after
1 ps, indicating a slow production of ONOO- from secondary
reactions. The small increase of the ONOO- concentration in
the interval between 1 and 500 ps demonstrates that contribu-
tions to the peroxinitrite concentration from secondary geminate
recombination, like the biologically important radical reaction
-
Accordingly, NO3 requires in excess of 3 eV to isomerize to
ONOO-. The excited nitrate molecule thus arrives at the
isomerization barrier after having dissipated about one-half its
6.2 eV excitation energy to the solvent. At the barrier, the ion
may relax into the vibrational manifolds of NO3 or ONOO-.
-
The resulting ONOO- molecule is formed with 1.4 eV of
vibrational energy, which is sufficient for dissociation into [NO
+ O2]-. Strong coupling between the low-frequency ONOO-
vibrations and the aqueous environment efficiently dissipates
this excess energy to the solvent, resulting in the rapidly
thermalized ONOO- transient absorption spectrum observed
experimentally.48-50
Alternatively, early isomerization via states connecting the
excited NO3 ion with the electronic ground state of ONOO-
-
precedes the vibrational energy dissipation. According to this
picture, the excitation energy of the newly formed ONOO-
isomer is well above its dissociation energy, and dissociation
into [NO + O2]- is possible. However, in most cases, the water
cage inhibits the free dissociation, thereby forcing the majority
of the excited molecules to remain inside the cage as NO3- or
ONOO-. Only for cage configurations where the solvent barrier
is sufficiently low may the dissociation be completed and the
fragments escape. For this explanation to comply with the
absence of absorption from vibrationally excited ONOO-, the
vibrational deexcitation is required to finish within 1 ps.
Although this is remarkably fast, the short vibrational relax-
ation time of NO3-(aq) combined with the stronger solvent-
solute coupling of the low-frequency modes of ONOO- 48-50
suggest that sub-picosecond relaxation of ONOO-(aq) could
be feasible.
•NO + •O2 f ONOO-, are only of minor importance.
-
Discussion
In accordance with the assignments made in the previous
section, we suggest the primary photodynamics of NO3-(aq)
depicted in Figure 5. When aqueous nitrate is photolyzed at
200 nm (6.2 eV), 44% of the excited NO3- molecules return to
the vibrationally equilibrated electronic ground state in 2 ps.
-
Concurrently, 48% of the excited NO3 molecules isomerize
to ONOO-, while the remaining molecules dissociate to [NO
+ O2]-.
Two possible reaction pathways may be derived from the
transient absorption data with the aid of supplementary results
from studies of solid-state and gas-phase nitrate. The measure-
ments show that ONOO- appears after NO3 has reached the
-
In addition to direct photoisomerization within the solvent
cage, a small contribution to the ONOO- concentration from
diffusive recombination is observed. The secondary formation
vibrationally excited electronic ground state, suggesting that
ONOO- is formed during the vibrational relaxation of ground-
state NO3-. ONOO-(aq) has an enthalpy of 1.6 eV relative to
ground-state NO3-, while the activation energy of the isomer-
(48) The lowest vibration frequency of ONOO- is ∼50% smaller than that of
NO3-. Tsai, J.-H. M.; Harrison, J. G.; Martin, J. C.; Hamilton, T. P.; der
Woerd, M. V.; Jablonsky, M. J.; Beckman, J. S. J. Am. Chem. Soc. 1994,
116, 4115-4116.
-
ization ONOO- f NO3 has been estimated to ∼1.4 eV.45-47
(45) Manuszak, M.; Koppenol, W. H. Thermochim. Acta 1996, 273, 11-15.
(46) Ray, J. D. J. Inorg. Nucl. Chem. 1962, 24, 1159-1162.
(47) Knak Jensen, S. J.; Ma´tyus, P.; McAllister, M. A.; Csizmadia, I. G. J. Phys.
Chem. A 2001, 105, 9029-9033.
(49) Larsen, J.; Madsen, D.; Thøgersen, J.; Poulsen, T. D.; Keiding, S. R. J.
Chem. Phys. 2002, 116, 7997-8005.
(50) Harris, A. L.; Brown, J. K.; Brown, C. B. Annu. ReV. Phys. Chem. 1988,
39, 341-366.
9
J. AM. CHEM. SOC. VOL. 125, NO. 50, 2003 15575

A R T I C L E S
Madsen et al.
of ONOO- may result from the diffusion-limited radical
reaction51
ucts. Although not observed experimentally, it is possible that
-
•NO2 + •O- and NO2 + O briefly exist as transient
dissociation fragments inside the solvent cage. If so, the
experimental data clearly show that they disappear within 250
•NO + •O2- f ONOO-
or from the reaction
NO- + O2 f ONOO-
(5)
-
fs, likely through recombination. Hence, the reactions NO3
+
hν f NO2- + O (eq 1) and hν + NO3- f •NO2 + •O- (eq 2)
have only little, if any, significance for the primary photo-
dynamics in aqueous solution.
Instead, we detect the formation of •O2 (or NO-) and
(6)
-
If indeed the reactants in eq 6 are produced in the primary
photolysis, they may be formed in the triplet ground state, 3NO-,
3O2, or singlet excited state, 1NO-, 1O2. The spin-allowed
recombination of ground-state 3NO- with 3O2 is known to result
in a near-diffusion limited (k ) 2.7 × 109 M-1 s-1) production
measure a primary quantum yield of ONOO-, which is 5 times
that reported by Mark et al.4 The different ONOO- quantum
yields may in part be related to the different excitation
wavelengths used in the two studies. The present investigation
excites the π f π* transition at 200 nm, while in the work by
Mark et al.4 nitrate was photolyzed right at the overlap between
the π f π* and n f π* transitions at 254 nm. The ONOO-
quantum yield resulting from the n f π* excitation is expected
to be significantly lower than that following the excitation of
the π f π* transition10 and is thus in qualitative agreement
with the lower yield observed by Mark et al.4 Moreover, the
femtosecond time resolution of the present work eliminates
competing reactions with products from other reactions and
impurity species such as for instance O2. The smaller ONOO-
quantum yield observed by steady-state photolysis4 may there-
fore reflect peroxynitrite depleting secondary reactions.
1
of ground-state peroxynitrite, ONOO-.52 On the other hand,
the rate of the reaction 1NO- + 1O2 f ONOO- is, to the best
1
of our knowledge, unknown. NO- is a very strong base (pKa
≈ 23),52 and protonation by the surrounding water molecules
would consequently lead to the rapid formation of hydroxyl
anions according to the reaction52
1NO- + H2O f 1HNO + OH-
(7)
OH-(aq) has a relatively intense absorption band below 210
nm53 and is in principle detectable by our far-UV femtosecond
spectrometer, but the small yield of NO- together with the large
absorption background from nitrate prevents its identification.
Accordingly, the present measurements are unable to determine
the spin multiplicity of NO-. Assuming reactions 5 and 6 are
diffusion controlled, we have simulated the recombination
processes using a spherical symmetric diffusion process54 and
adopted reasonable parameters for the diffusion constants (D
) 2 × 10-5 cm2/s), reaction radius (5 Å), and initial separation
of the fragments (10 Å).55,56 We find that both reactions 5 and
6 can account quantitatively for the increase in ONOO-
concentration. The associated decrease in the concentration of
•O2- or NO- is about 15% in 500 ps. Although such a decrease
is easily accommodated by the experimental data, the relatively
large uncertainty prevents its definite confirmation.
Conclusions
The femtosecond transient absorption measurements presented
here lead to a much simplified reaction scheme for the primary
UV photolysis of aqueous NO3-. Upon excitation of NO3-(aq),
44 ( 8% of the molecules return nonadiabatically to the
electronic ground state within 250 fs. The remaining 56% of
the excited molecules proceed to form cis-ONOO- (48%) or
-
dissociate to •NO + •O2 (8%) or NO- + O2 (11%). The
peroxynitrite isomer attains its ground state in 1 ps as a primary
photoproduct of the NO3- photolysis, and the relative contribu-
tion to the peroxynitrite yield from secondary reactions during
the subsequent 500 ps amounts to only 3%. Contrary to previous
-
studies, the present investigations find no evidence of NO2
+
In contrast to earlier studies,4,7,8,16,17,57 the present measure-
ments do not find NO2- or •O- among the primary photoprod-
O or •NO2 + •O-, indicating that the NO2- formation observed
in steady-state photolysis experiments must be due to subsequent
reactions.58,59
(51) Kobayashi, K.; Miki, M.; Tagawa, S. J. Chem. Soc., Dalton Trans. 1995,
2885-2889.
Acknowledgment. The outstanding technical support of Per
Strand is highly appreciated.
(52) Shafirovich, V.; Lymar, S. V. Proc. Natl. Acad. Sci. U.S.A. 2002, 99, 7340-
7345.
(53) Nielsen, S. O.; Michael, B. D.; Hart, E. J. J. Phys. Chem. 1976, 80, 2482-
2488.
JA030135F
(54) Krissinel, E. B.; Agmon, N. J. Comput. Chem. 1996, 17, 1085-1098.
(55) Madsen, D.; Thomsen, C. L.; Poulsen, J. Aa.; Poulsen, Knak Jensen, S. J.;
Thøgersen, J.; Keiding, S. R.; Krissinel, E. B. J. Phys. Chem. 2003, 107,
in press.
(57) Edwards, J. O.; Plumb, R. C. In Progress in Inorganic Chemistry; Karlin,
K. D., Ed.; John Wiley & Sons: New York, 1994; Vol. 41, pp 599-635.
(58) Mere´nyi, G.; Lind, J.; Goldstein, S.; Czapski, G. J. Phys. Chem. A 1999,
103, 5685-5691.
(56) CRC Handbook of Chemistry and Physics, 74th ed.; CRC Press: Boca
Raton, FL, 1993; pp 5-91.
(59) Kissner, R.; Koppenol, W. H. J. Am. Chem. Soc. 2002, 124, 234-239.
9
15576 J. AM. CHEM. SOC. VOL. 125, NO. 50, 2003