Photochemical investigation of Roussin's red salt esters: Fe2(μ-SR)2(NO)4

Article abstract of DOI:10.1021/ic020309e

The photochemistry of various Roussin's red ester compounds of the general formula Fe₂(SR)₂(NO)₄, where R = CH₃, CH₂CH₃, CH₂C₆H₅, CH₂CH₂OH, and CH₂CH₂SO₃^-, were investigated. Continuous photolyses of these ester compounds in aerated solutions led to the release of NO with moderate quantum yields for the photodecomposition of the ester (Φ_RSE = 0.02-0.13). Electrochemical studies using an NO electrode demonstrated that 4 mol of NO are generated for each mole of ester undergoing photodecomposition. Nanosecond flash photolysis studies of Fe₂(SR)₂(NO)₄ (where R = CH₂CH₂OH and CH₂CH₂SO₃^-) indicate that the initial photoreaction is the reversible dissociation of NO. In the absence of oxygen, the presumed intermediate, Fe₂(SR)₂(NO)₃, undergoes second-order reaction with NO to regenerate the parent cluster with a rate constant of k_NO = 1.1 x 10⁹ M^-1 s^-1 for R = CH₂CH₂OH. Under aerated conditions the intermediate reacts with oxygen to give permanent photochemistry.

Full text of DOI:10.1021/ic020309e

Inorg. Chem. 2003, 42, 2288−2293
Photochemical Investigation of Roussin’s Red Salt Esters:
Fe2(µ-SR)2(NO)4
Christa L. Conrado, James L. Bourassa, Christian Egler, Stephen Wecksler, and Peter C. Ford*
Department of Chemistry and Biochemistry, UniVersity of California,
Santa Barbara, California 93106-9510
Received May 2, 2002
The photochemistry of various Roussin’s red ester compounds of the general formula Fe2(SR)2(NO)4, where R )
CH , CH CH , CH C H , CH CH OH, and CH CH SO -, were investigated. Continuous photolyses of these ester
3
2
3
2
6
5
2
2
2
2
3
compounds in aerated solutions led to the release of NO with moderate quantum yields for the photodecomposition
of the ester (ΦRSE ) 0.02−0.13). Electrochemical studies using an NO electrode demonstrated that 4 mol of NO
are generated for each mole of ester undergoing photodecomposition. Nanosecond flash photolysis studies of
Fe2(SR)2(NO)4 (where R ) CH CH OH and CH CH SO -) indicate that the initial photoreaction is the reversible
2
2
2
2
3
dissociation of NO. In the absence of oxygen, the presumed intermediate, Fe2(SR)2(NO)3, undergoes second-order
reaction with NO to regenerate the parent cluster with a rate constant of kNO ) 1.1 × 109 M-1 s-1 for R )
CH CH OH. Under aerated conditions the intermediate reacts with oxygen to give permanent photochemistry.
2
2
-
2-
Introduction
Fe4S3(NO)7 (RBS) and “red salt” dianion Fe2S2(NO)4
(RRS). Both RBS and RRS are photoactive toward NO
release;3 however, cell culture experiments show RBS to be
fairly toxic.3a The red salt is less toxic and more photoactive
(ΦRRS values ranging from 0.09 to 0.40 depending upon
solvent and other conditions), and visible light photolysis
of hypoxic V79 Chinese hamster cell cultures that had been
treated with a submillimolar solution of RRS indicated
marked sensitization of γ-radiation induced cell death. This
effect was attributed to sensitization of radiation damage by
NO released during the simultaneous photolysis.3a However,
RRS is relatively unstable and, as a consequence, not likely
to have practical application in more complex systems.
Related to RRS are a series of Roussin’s “red esters”
(RSEs), which can be prepared via the alkylation of the anion
by reaction with an alkyl halide (eq 1).6
Nitric oxide (NO, also known as nitrogen monoxide) has
been shown to sensitize cells to ionizing radiation;2 thus,
nitrosyl containing compounds are of interest as potential
NO sources for physiological targets. One ongoing research
theme of this laboratory has been the development of
photochemical strategies to deliver nitric oxide to such targets
on demand,3,4 and we and others have probed a number of
different compounds as potential photochemical precursors
of NO.3-5 Among these were the Roussin’s “black salt” anion
* To whom correspondence should be addressed. E-mail: ford@
chem.ucsb.edu.
(1) Undergraduate student from Westfaelische Wilhelms-Universitaet,
Muenster, Germany.
(2) (a) Howard-Flanders, P. Nature (London) 1957, 180, 1191. (b)
Mitchell, J. B.; Wink, D. A.; DeGraff, W.; Gamson, J.; Keefer, L. K.;
Krishna, M. C. Cancer Res. 1993, 53, 5845.
(3) (a) Bourassa, J.; DeGraff, W.; Kudo, S.; Wink, D. A.; Mitchell, J. B.;
Ford, P. C. J. Am. Chem. Soc. 1997, 119, 2853-2860. (b) Kudo, S.;
Bourassa, J. L.; Boggs, S. E.; Sato, Y.; Ford, P. C. Anal. Biochem.
1997, 247, 193-202. (c) Bourassa, J.; Lee, B.; Bernard, S.; Schoonover,
J.; Ford, P. C. Inorg. Chem. 1999, 38, 2947-2952. (d) Bourassa, J.
L.; Ford, P. C. Coord. Chem. ReV. 2000, 200-202, 887-900. (e)
Bourassa, J. L. Ph.D. Dissertation, University of California, Santa
Barbara, 1998.
(4) (a) Ford, P. C.; Bourassa, J.; Miranda, K.; Lee, B.; Lorkovic, I.; Boggs,
S.; Kudo, S.; Laverman, L. Coord. Chem. ReV. 1998, 171, 185-202.
(b) Lorkovic, I. M.; Miranda, K. M.; Lee, B.; Bernhard, S.;
Schoonover, J. R.; Ford, P. C. J. Am. Chem. Soc. 1998, 120, 11674-
11683. (c) De Leo, M.; Ford P. C. J. Am. Chem. Soc. 1999, 121, 1980-
1981. (d) Works, C. F.; Ford, P. C. J. Am. Chem. Soc. 2000, 122,
7592-7593.
Preliminary experiments indicated that the esters were
more stable than RRS while remaining photoactive toward
NO release. Herein are described the results of photochemical
studies of the following red ester compounds: the dimethyl
2288 Inorganic Chemistry, Vol. 42, No. 7, 2003
10.1021/ic020309e CCC: $25.00 © 2003 American Chemical Society
Published on Web 03/06/2003

Photochemistry of Roussin’s Red Salt Esters
photolyzed did not lead to significant 510 nm absorbance changes
over the time frame of the analytical procedure.
ester, Fe2(µ-SCH3)2(NO)4 (1), the diethyl ester, Fe2(µ-SCH2-
CH3)2(NO)4 (2), the dibenzyl ester Fe2(µ-CH2C6H5)2(NO)4
(3), the 2-hydroxyethyl ester, Fe2(SCH2CH2OH)2(NO)4 (4),
and the 2-sulfonato-ethyl ester, Na2[Fe2(SCH2CH2SO3)2-
(NO)4] (5).
Flash photolysis experiments were carried out on a time-resolved
optical (TRO) apparatus described previously.4b The pump source
was a Continuum NY60-20 nanosecond pulse Nd:YAG laser
operating in the third harmonic (355 nm) and attenuated to 10-30
mJ/pulse. The probe source was a high pressure, short arc xenon
lamp. Kinetic traces were obtained using a SPEX Doublemate
monochromator and RCA model 8852 PMT for detection. The
eventual photolysis product is an orange solid that would build up
on the laser cell after repetitive laser flashes, and formation of such
solids could potentially compromise the data collection for a given
experiment. For this reason, fresh 3 mL aliquots of the ester
solutions were flashed a maximum of 10 times each, and the
solutions were shaken between each flash experiment. Such
procedures reduced the opportunity for improving signal/noise ratio
by signal averaging.
Experimental Section
Materials. Water for spectroscopic studies was purified by a
Millipore system. Methanol was distilled under nitrogen from I2
and Mg. Acetonitrile was distilled under dinitrogen from CaH2 and
used immediately. Chromatographic grade argon was purchased
from Air Liquide and used as received. All inert atmosphere
manipulations were performed on a vacuum manifold. Deaeration
of the solutions was achieved by 3 freeze-pump-thaw (f-p-t)
cycles. Nitric oxide (99.0%) was purchased from Matheson and
passed through a stainless steel column that contained Ascarite II
(NaOH on a silicate carrier) to remove higher nitrogen oxides.
Stainless steel tubing and fittings were used to transfer gases.
Manifold connections were accomplished by using stainless steel-
to-glass fittings.
Syntheses. The Roussin’s red ester compounds were prepared
by literature methods.6b These esters are moderately air and light
sensitive and require long-term storage in the dark and under inert
atmosphere.
Instruments. UV-vis absorption spectra were recorded for
solutions in 1 cm path length quartz cells using a HP8572 diode
array spectrophotometer. Infrared spectra were obtained using a
BioRad FTS 60 SPC 3200 FTIR spectrometer and IR cells equipped
with CaF2 windows.
Photochemical Experiments. Solutions for continuous photoly-
sis experiments were generally about 10-4 M in RSE concentration
while those for flash experiments were in the range 35-40 mM.
All continuous and flash experiments were carried out at 295 ((1)
K. Quartz photolysis cells (1.0 cm) were used in both types of
experiments. When a controlled atmosphere (Ar, NO, or O2) was
necessary, the photolysis cell consisted of a triple O-ring Chemglass
Teflon stopcock, a four-sided quartz cuvette, and an O-ring adapter
for connection to a gas/vacuum manifold. Concentrations were
calculated from published solubility data.7
Continuous photolyses (CW) were performed using an optical
train with an Oriel 200 W Hg arc lamp in an Oriel lamp housing
model 66033 as the light source. Interference filters were used to
isolate the appropriate excitation wavelengths. Chemical actinometry
was performed with ferric oxalate solutions.8 All solutions in the
CW experiments were stirred continuously and kept from exposure
to extraneous light.
The quantitative determination for the photochemical production
of Fe2+ from the ester complexes was accomplished by adding a
1,10-phenanthroline solution and recording the intensity of the
characteristic 510 nm absorbance of the Fe(phen)32+ species formed.
Addition of 1,10-phen to RSE solutions which had not been
Fe2(µ-SCH3)2(NO)4 (1), Fe2(µ-CH2CH3)2(NO)4 (2), Fe2-
(µ-CH2C6H5)2(NO)4 (3), and Fe2(µ-SCH2CH2OH)2(NO)4 (4).6
The methyl ester (1) was prepared by the procedure described by
Seyferth and co-workers.6b Similar procedures were followed to
prepare Fe2(µ-SCH2CH3)2(NO)4 (2) and Fe2(µ-SCH2C6H5)2(NO)4
(3) with the exception that iodoethane, benzyl iodide, or 2-iodo-
ethanol was used, respectively, instead of iodomethane. These
compounds were characterized by both IR and UV-vis spectros-
copy. (1). UV-vis (MeOH) (λmax in nm (ꢀ in M-1 cm-1)): 312
(8.3 × 103), 364 (7.8 × 103). IR (CH3CN) (νmax in cm-1 (ꢀ in M-1
cm-1)): 1778 (5.8 × 103), 1752 (5.3 × 103), 1800 (w). (2). UV-
vis (CH3CN): 310 (9.1 × 103), 364 (8.7 × 103). IR (CH3CN): 1776
(5.9 × 103), 1750 (5.2 × 103), 1800 (w). (3). UV-vis (CHCl3):
312 (9.3 × 103), 364 (8.5 × 103). IR (CHCl3): 1778 (5.5 × 103),
1751 (5.0 × 103), 1800 (w). FAB (NBA matrix): m/z 448 (Fe2-
(SCH2C6H5)2(NO)3)+, 418 (Fe2(µ-SCH2C6H5)2(NO)2+. (4). UV-
vis (MeOH): 312 (9.3 × 103), 362 (8.5 × 103). IR (MeOH): 1778
(5.8 × 103), 1752 (5.6 × 103), 1800 cm-1 (w). ESI (methanol/
water): m/z 386 (Fe2(SCH2CH2OH)2(NO)4]-).
Na2[Fe2(µ-SCH2CH2SO3)2(NO)4] (5). A solution of Na2[Fe2S2-
(NO)4]‚8H2O (RRS) (1.0 g; 2.1 mmol) plus the sodium salt of
bromoethanesulfonic acid (1.0 g; 4.7 mmol) in 75 mL of deoxy-
genated water was prepared. The mixture was stirred under argon,
and then the solution was heated at reflux under argon for 1 h. The
volume was reduced by rotary evaporation to approximately 10
mL and was then cooled to 5 °C. The result was a crop of shiny,
brownish-red crystals. The product was recrystallized from a
minimum volume of hot water and stored under argon in a
refrigerator. The product was characterized by UV-vis and IR
spectroscopy and ESI MS. The UV-vis spectrum was similar to
those seen for the other ester compounds.6 UV-vis (H2O) (λmax in
nm (ꢀ in M-1 cm-1)): 310 (9.2 × 103), 364 (8.9 × 103). IR (CH3-
CN) (νmax in cm-1 (ꢀ in M-1 cm-1)): 1782 (5.6 × 103), 1757 (5.8
× 103), 1821 (w). Aqueous solution negative ion ESI-MS m/z
(5) (a) Flitney, F. W.; Megson, I. L.; Thomson, J. L. M.; Kennovin, G.
D.; Butler, A. R. Br. J. Pharmacol. 1996, 117, 1549. (b) Carter, T.;
Bettache, N.; Corrie, J. E. T.; Ogden, D.; Trentham, D. R. Methods
Enzymol. 1996, 268, 266-281. (c) Namiki, S.; Kaneda, F.; Ikegami,
M.; Arai, T.; Fujimori, K.; Asada, S.; Hama, H.; Kasuyua, Y.; Goto,
K. Bioorg. Med. Chem. 1999, 7, 1695-1702. (d) Fukuhar, K.;
Kurihara, M.; Miyata, N. J. Am. Chem. Soc. 2001, 123, 8662-8666.
(6) (a) Glidewell, C.; Lambert, R. J.; Harman, M. E.; Hursthouse, M. B.
J. Chem. Soc., Dalton Trans. 1990, 2685-2690. (b) Seyferth, D.;
Gallagher, M. K.; Cowie, M. Organometallics 1986, 5, 539-548. (c)
Sung, S. S.; Glidewell, C.; Butler, A. R.; Hoffman, R. Inorg. Chem.
1985, 24, 3856. (d) Butler, A. R.; Glidewell, C.; Hyde, A. R.;
McGinnis, J. Inorg. Chem. 1985, 24, 2931-2934. (e) Thomas, J. T.;
Robertson, J. H.; Cox, E. G. Acta Crystallogr. 1958, 11, 599.
(7) Battino, R.; Clever, H. L.; Young, C. L. IUPAC Solubility Data Series;
Pergamon Press: New York, 1985; Vol. 8, p 261.
(relative intensity): 535 (18) [(Fe2(SCH2CH2SO3)2(NO)4)Na]1-
,
505 (21) [(Fe2(SCH2CH2SO3)2(NO)3)Na]1-, 483 (12) [(Fe2-
(SCH2CH2SO3)2(NO)3),H]1-, 372 (5) [(Fe2(SCH2CH2SO3)(NO)4)]1-
,
342 (8) (Fe2(SCH2-CH2SO3)1(NO)3]1-), 256 (48)([Fe-
(SCH2CH2SO3)(NO)2]1-), 226 (100) ([Fe1(SCH2-CH2SO3)1(NO)1]1-).
CHN elemental analysis Calcd for C4H8N4O10S4Fe2Na2‚(3 H2O):
C, 7.84; H, 2.29; N, 9.15. Found (UCSB Marine Biology Analytical
Labs): C, 7.87; H, 2.05; N, 8.67. Iron analysis by flame atomic
(8) Calvert, J. G.; Pitts, J. N. Photochemistry; Wiley: New York, 1967;
pp 783-786.
Inorganic Chemistry, Vol. 42, No. 7, 2003 2289

Conrado et al.
absorption (Varian Techtron AA6 spectrophotometer) Calcd: Fe,
18.3. Found: Fe, 19.1 (( 1.0). An X-ray crystal structure has been
obtained for the Ph4As+ salt of the Fe2(µ-SCH2CH2SO3)2(NO)4
2-
anion with an R-factor about 6.3% confirming that this species has
a structure consistent with the other esters as illustrated qualitatively
in eq 1.9
Results
Spectroscopic Properties of RSE Compounds. The
infrared NO stretching bands νNO of the RSE compounds
Fe2(µ-SR)2(NO)4 are largely insensitive to the nature of the
R group5,6 but are markedly shifted to higher frequency
compared to the bands seen for Na2[RRS] (∼1665 cm-1,
1690 cm-1). For example, the FTIR spectra of Fe2(µ-SCH3)2-
(NO)4 (1), Fe2(µ-SCH2CH3)2(NO)4 (2), Fe2(µ-CH2C6H5)2-
(NO)4 (3), and Fe2(SCH2CH2OH)2(NO)4 (4) each display two
strong nitrosyl stretching bands of similar intensity at
∼1750-1754 cm-1 (ꢀ ∼ 5.0-5.6 × 103 M-1 cm-1) and
∼1776-1778 cm-1 (∼5.5-5.9 × 103 M-1 cm-1) as well as
a weak band at 1800 cm-1 (200 M-1 cm-1).6 The nitrosyl
stretches of the sulfonated species, Na2[Fe2(SCH2CH2SO3)2-
(NO)4] (5), are shifted slightly to 1757, 1782, and 1815 cm-1
in acetonitrile. The νNO bands show typical broadening in
polar solvents; for instance, the bandwidths at half-maximum
intensity for Fe2(µ-SCH3)2(NO)4 range from 4 cm-1 in
cyclohexane to 12 cm-1 in acetonitrile, but the band strengths
and positions are not affected significantly by changing
solvent. Notably, the much lower νNO band frequencies for
the red salt anion [Fe2(µ-S)2(NO)4]2- are consistent with the
greater back-donation of electron density from the dianionic
Fe2(µ-S)2 core into NO π* orbitals.
Figure 1. Spectral changes recorded during the continuous photolysis (λirr
) 365 nm) of Fe2(SCH2CH2OH)2(NO)4 (85 µM, initial concentration) in
aerated methanol ([O2] ) 1.8 mM, T ) 294 K). The second, third, and
fourth spectra were recorded after successive 2 min photolysis intervals.
Table 1. Quantum Yields for Ester Photodecomposition under Various
Conditionsa
medium
λ
irr (nm)
Ia (Ein/L s)
Φd (mol/Ein)
Fe2(µ-SCH3)2(NO)4] (1)
3.0 × 10-6
MeOH
CH3CN
CH2Cl2
THF
365
365
365
365
0.14 ( 0.02
0.23 ( 0.05
0.07( 0.03
0.28 ( 0.02
3.0 × 10-6
3.0 × 10-6
3.0 × 10-6
Fe2(µ-SCH2CH3)2(NO)4] (2)
CH3CN
CH3CNb
CH3CNc
CHCl3
365
1.1 × 10-6
1.1 × 10-6
1.1 × 10-6
1.1 × 10-6
0.076 ( 0.004
0.12 ( 0.01
365
365
546
<0.003
(1.7 ( 0.2) × 10-4
Fe2(µ-SCH2C6H5)2(NO)4] (3)
CHCl3
CHCl3
436
546
1.1 × 10-6
(3.2 ( 0.1) × 10-4
The electronic spectra of the esters are also essentially
independent of the R-group in the near-UV-vis region.
Solutions of 1, 2, 3, and 4 each show absorption bands with
1.1 × 10-6
(1.9 ( 0.2) × 10-4
Fe2(µ-SCH2CH2OH)2(NO)4] (4)
MeOH
MeOH
MeOH
MeOHb
MeOH
MeOH
MeOH
365
365
365
365
405
405
405
1.1 × 10-6
3.0 × 10-6
8.2 × 10-6
1.1 × 10-6
6.0 × 10-7
1.7 × 10-6
1.0 × 10-6
0.079 ( 0.005
0.045 ( 0.003
0.020 ( 0.003
0.103 ( 0.009
0.132 ( 0.01
0.056 ( 0.003
0.052 ( 0.003
λ
max at ∼312 nm (∼9 × 103 M-1 cm-1) and ∼362 nm (∼8.5
× 103 M-1 cm-1) with a shoulder at ∼430 nm (∼4 × 103
M-1 cm-1) that tails out into the green (∼5 × 102 M-1 cm-1
at 550 nm). These spectra are analogous to the spectrum of
aqueous RRS, which shows bands at 314 nm (8.5 × 103
M-1 cm-1) and 374 nm (10.4 × 103 M-1 cm-1) as well as
a stronger band at 242 nm (2.8 × 104 M-1 cm-1). For the
Roussin’s salts, EHMO calculations suggested that the
LUMO for the iron-sulfur cluster has Fe-NO antibonding
character consistent with the photolability seen.6c We assume
that the electronic transitions for these RSE compounds are
similar to those seen for RRS.
Na2[Fe2(µ-SCH2CH2SO3)2(NO)4] (5)
H2O
H2Oc
H2O
365
365
365
1.1 × 10-6
1.1 × 10-6
3.0 × 10-6
0.024 ( 0.001
0.052 ( 0.005
0.048 ( 0.004
a Quantum yields for RSE decomposition determined from changes in
UV-vis spectra. Experiments in aerated solutions at 294 K, except where
noted. Values reported are from 3 or more independent determinations. Initial
RSE concentrations for these continuous photolysis experiments were ∼10-4
M. b Solutions equilibrated with O2 (1 atm). c Deaerated solution.
Photoreactions of Roussin’s Red Esters: CW Photoly-
ses. In the dark, the RSE compounds are thermally stable
over a period of 30 h in deoxygenated solutions and show
only ∼10% thermal decomposition over 3 h in aerated
solutions. As noted already, the extinction coefficients for
all the ester compounds are nearly 104 M-1 cm-1 for each
(Φd) were determined from the quantitative UV-Vis absorp-
tion changes upon photolysis for solutions equilibrated with
oxygen or with air and with RSE concentrations of 76-100
µM. These are summarized in Table 1 for various esters
studied under a variety of conditions. In general, the IR
spectra of photolyzed RSE solutions under aerated conditions
showed the loss of the parent νNO bands, and no new bands
in the NO stretching frequency region appeared suggesting
the absence of a long-lived metal-nitrosyl intermediate or
product. (Free NO has a low extinction coefficient and would
not have been detected under these conditions.)
λ
max. However, these compounds were reactive under con-
tinuous photolysis with near-UV or visible light when
dissolved in aerated solvents. This is illustrated by Figure
1, which shows the typical spectral changes across the near-
UV-vis spectrum when one of the esters, in this case
methanolic Fe2(SCH2CH2OH)2(NO)4, is irradiated in aerated
solution. The quantum yields for RSE photodecomposition
The Φd values for the various RSEs under continuous
2290 Inorganic Chemistry, Vol. 42, No. 7, 2003

Photochemistry of Roussin’s Red Salt Esters
2+
addition of 1,10-phenanthroline (phen) to give Fe(phen)3
with a characteristic strong absorbance at 510 nm. Aerated
aqueous solutions of 5 (∼70 mM) were monitored spectro-
photometrically during 365 nm photolysis until the RSE
photodecomposition was approximately 20%. Then, a 1.0
mL aliquot of this solution was diluted with a 1.0 mL volume
of water, and the UV-vis absorption spectrum was taken
(for a baseline). Another 1.0 mL aliquot was added to a 1.0
mL volume of phen solution (5.5 mM), and the spectrum
was recorded after allowing the solutions to develop. The
result was the characteristic, strong absorptions at the λmax
(510 nm, ꢀ510 ) 1.11 × 104 M-1 cm-1) of Fe(phen)32+. RSE
solutions that had not been irradiated gave negligible positive
tests for Fe(II) formation after correcting for the baseline
solution spectrum. The absorbance changes in the photolysis
solution of 5 indicated the amount of RSE decomposed,
while the phenanthroline analyses determined the concentra-
tion of Fe2+ generated and from these a product yield ratio
of 2.4 ( 0.3 mol of Fe(II) generated per mole of 5
photodecomposed was calculated. The final photoproduct
solution is thus presumed to include various ferrous-thiol
complexes. Consistent with this view was the observation
that the product spectrum can be duplicated by mixing ferrous
solutions with mercaptoethanol and nitrite.
Figure 2. Quantitative determination of NO production upon irradiation
of Roussin’s salt ester Na2[Fe2(µSCH2CH2SO3)2(NO)4]. UV-vis absorbance
changes and NO sensor amperogram upon white light photolysis of an
aerated pH 7 phosphate buffered solution at 294 K. ∆[NO]/∆[ester] ) 3.7.
photolysis proved to be systematically dependent upon
several key factors (Table 1) including excitation wavelength,
λirr, and intensity, Ia. For example, Φd tended to fall off at
longer excitation wavelength λirr and at higher Ia. Another
factor was the oxygen concentration in the solution. For
example, there is little net photodecomposition (Φd < 0.003)
in deaerated acetonitrile solutions of 2 with Ia ) 6.1 × 10-6
Ein/L‚s at 365 nm, but Φd rose to 0.076 in aerated solution
and to 0.12 in solution equilibrated with O2 at 1 atm under
otherwise analogous conditions. This behavior parallels the
quantitative photochemistry of Roussin’s red and black salts3
and can be attributed to the primary photoreaction step being
the reversible labilization of one NO to give transient species
that must be trapped by O2 in order to give permanent
photoreaction (see later). The oxygen dependence also
explains the quantum yield results in different solvents. One
can see from Table 1 that solvents in which O2 is more
soluble reveal the larger Φd values in air equilibrated systems.
Products from the CW Photolysis of 5. Continuous
photolysis of aqueous Na2[Fe2(µ-SCH2CH2SO3)2(NO)4] dem-
onstrated that the ester complexes release 4 NOs for every
molecule of ester decomposed (Figure 2). The quantitative
analyses were carried out for aerated aqueous phosphate
buffer solutions (pH ) 7) using the combination electrode
NO sensor. The Pyrex and IR filtered output of a mercury
arc lamp with a 365 nm interference filter was used for the
photolysis. The change in [5] upon photolysis was evaluated
by changes in the solution absorption spectra, and the [NO]
produced was evaluated from the amperogram. These
procedures gave the reaction stoichiometry ratio ∆[NO]/
∆[RSE] ) 4.1 ( 0.2 for 5 in aqueous solution. In the absence
of other NO sinks, it is presumed that aqueous autoxidation
leads primarily to nitrite as the eventual product. A similar
study with Fe2(µ-SCH2CH2OH)2(NO)4 in 5% aq methanol
gave the ratio 3.8 ( 0.3.
Flash Photolysis Experiments. Flash photolysis studies
were carried out for solutions of Fe2(µ-SCH3)2(NO)4 (1) in
acetonitrile, Fe2(µ-SCH2CH2OH)2(NO)4 (4) in methanol, and
Na2[Fe2(µ-SCH2CH2SO3)2(NO)4] (5) in water. The behavior
for all three systems was qualitatively similar. Namely, 355
nm flash photolysis led to bleaching of the visible range
absorbance monitored at 450 nm, which decayed quickly
back to the baseline absorbance when excess NO was present,
and much more slowly under an inert atmosphere. The latter
observation is consistent with the result of the continuous
photolysis experiments that indicated virtually no net pho-
toreaction in the absence of oxygen. In this context, when
the flash photolysis was carried out using aerated solutions,
only partial recovery of the initial absorbance at 450 nm was
observed, again consistent with the continuous photolysis
experiments that demonstrated oxygen to be required for net
photoreaction.
An example of the behavior in deaerated solution would
be the 355 nm flash photolysis of methanolic Fe2(µ-SCH2-
CH2OH)2(NO)4 under argon. Transient bleaching of the
absorbance at 450 nm was seen, and this decayed back to
baseline over the course of several hundred microseconds.
The decay did not fit well to an exponential function but
could be fit adequately according to second-order kinetics
(1/abs versus t is linear) as expected if the primary photo-
reaction is reversible NO labilization to give the reactive
intermediate Fe2(µ-SR)(NO)3 (I1) and this is followed by
second-order recombination of I1 and NO (eq 2).
Other products of CW photolysis of aqueous Na2[Fe2-
-
(µ-SCH2CH2SO3)2(NO)4] include NO2 , free ligand
-
-
HSCH2CH2SO3 , and minor amounts of NO3 as seen by
electrospray ionization mass spectral (ESI-MS) analysis.
Additionally, the formation of Fe2+ was verified by the
Inorganic Chemistry, Vol. 42, No. 7, 2003 2291

Conrado et al.
Figure 3. Flash photolysis of Fe2(µ-SCH2CH2OH)2(NO)4 (36.3 µM) in
aerated methanol as monitored at 420 nm ([O2] ) 1.8 mM; T ) 294 K).
The solid line is the exponential fit for the temporal absorbance change.
The residual to that fit is illustrated at the top of the figure.
Figure 4. Flash photolysis of Na2[Fe2(µ-SCH2CH2SO3)2(NO)4] (36.3 µM)
in aqueous solution under NO (1.07 mM) as monitored at 420 nm. The
solid line is the exponential fit for the temporal absorbance change. The
residual to that fit is illustrated at the top of the figure.
The 355 nm flash photolysis of 4 in aerated methanol
showed analogous initial transient bleaching at 420 nm but
with much faster absorbance changes. While the absorbance
due to the ester recovers partially, it does not return to the
initial value within the monitoring time (500 µs). The decay
Scheme 1
of this transient bleach fits a simple exponential with kobs
)
2.3 × 104 s-1 (295 K) (Figure 3). If one assumes that this
decay is largely the result of O2 trapping of I1, an estimate
of kO2 ) 1.3 × 107 M-1 s-1 can be made for the rate constant
for this reaction (see later). For comparison, a kO2 value of
5.6 × 107 M-1 s-1 was determined previously for the O2
trapping of analogous intermediates formed in the aqueous
solution flash photolysis of RRS.3
Figure 4 illustrates the situation when the flash experiment
was carried out under added NO. The example shown in
this case is the 355 nm flash photolysis of aqueous
Na2[Fe2(µ-SCH2CH2SO3)2(NO)4] (5) equilibrated with a NO
partial pressure PNO of 613 Torr ([NO] ) 1.07 mM). Again,
immediate transient bleaching was observed (as monitored
in this case at 420 nm), followed by decay back to the initial
absorbance to regenerate the parent. However, in contrast
to Figure 3, the reaction is completed within a few
microseconds. The temporal absorbance follows exponential
kinetics, and the fit shown in Figure 4 gives a kobs of 1.7 ×
106 s-1 (T ) 295 K). When analogous experiments were
carried out with aqueous 5 at different PNO values, the kobs
values obtained were markedly dependent on [NO]. A plot
of kobs versus [NO] for five values of [NO] ranging from
0.1 to 2 mM proved to be linear, consistent with the
relationship kobs ) kNO[NO] that is predicted by eq 2. The
slope of this line gave kNO ) (1.1 ( 0.1) × 109 M-1 s-1 (R2
) 0.998), only about an order of magnitude smaller than
the diffusion limit under these conditions.
O2. The flash data can be interpreted according to Scheme
1, which indicates the competition between NO and O2 for
the intermediate I1. It is assumed in this model that O2 traps
this intermediate to give one or more new species X that
decompose irreversibly to the final products. NO is also
known to react with O2 via third-order kinetics to give (in
aqueous solution) the nitrite ion.10 However, while NO
autoxidation is the likely fate of the NO generated in an
aerobic environment, this process must be orders of mag-
nitude slower than the O2 trapping of I1 inferred by the
continuous and flash photolysis experiments described here.
Given that kNO is about 109 M-1 s-1, the fast reaction of I1
with the generated NO competes with trapping of the
intermediate I1 by the dissolved O2 (k ∼ 1.1 × 107 M-1 s-1).
According to the scheme, the overall quantum yield for
RSE photodecomposition will depend on the quantum yield
for the primary NO labilization (φ1), the concentration of
O2 present, and the steady state concentration of NO formed
The flash data recorded under aerated conditions dem-
onstrated that, in contrast to the situation under a NO or
argon, permanent photochemistry occurs in the presence of
(9) Wecksler, S.; Bu, X.; Ford, P. C. To be published
(10) Ford, P. C.; Wink, D. A.; Stanbury, D. M. FEBS Lett. 1993, 326,
1-3.
2292 Inorganic Chemistry, Vol. 42, No. 7, 2003

Photochemistry of Roussin’s Red Salt Esters
under continuous photolysis ([NO]ss) as indicated by eq 3:
If the intensity of absorbed light (Ia) is constant, then
increasing [O2] increases Φd as was observed (Table 1).
However, the steady state concentrations of I1 and of NO
([NO]ss)11 are both dependent on Ia, so at higher intensity,
the competitive back reaction between these is enhanced
relative to the pseudo-first-order trapping by oxygen. Quali-
tatively, at constant [O2] (e.g., an aerated solution), Φd should
decrease with increasing intensity, and this was indeed
observed (Table 1). In the limiting condition of very low Ia,
-d[RSE]
kox[O2]
dt
Ia
Φd )
)
φ1
(3)
[
]
kox[O2] + kNO[NO]ss
RSE decomposition range from ∼0.02 to 0.13, depending
on the ester, the intensity of light absorbed, and the solvent
medium. Production of 4 mol of NO for each 1 mol of ester
undergoing photoconversion was demonstrated by an elec-
trochemical NO sensor. Flash photolysis experiments indi-
cated that the primary photoreaction is the labilization of a
single NO to give intermediate I1, presumably Fe2(µ-SR)2-
(NO)3, which is trapped by oxygen in competition with the
very fast second-order back reaction to generate the initial
RSE. Ongoing studies in this laboratory are concerned with
the preparation and photochemical properties of new esters
tailored to have pendent R-groups with defined biological
or photooptical properties. For example, such pendant groups
may act as an antenna to increase the efficiency of longer
wavelength irradiation for the photochemical generation of
NO or display some aspect of biological specificity.
the kO [O2] term would be much greater than kNO[NO], and
2
Φd ∼ φ1. A more quantitative test of this model was
previously carried out with aqueous solutions of RRS.3a
In summary, Roussin’s red salt esters have the potential
to be efficient photochemical NO generators, and the pattern
of photochemical behavior appears qualitatively the same
among the various RSEs described here. Quantum yields for
(11) A referee has correctly noted that, in the absence of another sink for
NO, the trapping of I1 by O2 would lead to a build up of [NO], rather
than a steady state concentration, [NO]ss. However, the electrochemical
measurement shown in Figure 2 illustrates that other processes, mostly
autoxidation9 in competition with diffusion out of the solution, deplete
NO on a time scale (minutes) comparable to the continuous photolysis
experiment, so [NO] has the potential to achieve a steady state
dependent on the intensity of absorbed light, Ia. Although the function
describing the temporal concentration of NO must be complex owing
to the competing processes, the qualitative behavior of Φd in response
to changes in Ia or [O2] is consistent with the description offered by
eq 3.
Acknowledgment. These studies were supported by the
National Science Foundation (Grants CHE 9726889 and
CHE 0095144). We thank Dr. X. Bu of this department for
confirming the identity of the anion of compound 5 by
determining the X-ray crystal structure.
IC020309E
Inorganic Chemistry, Vol. 42, No. 7, 2003 2293