Photochemical generation of cyclopentadienyliron dicarbonyl anion by a nicotinamide adenine dinucleotide, dimer analogue

Article abstract of DOI:10.1021/ic0009627

Irradiation of the absorption band of an NAD (nicotinamide adenine dinucleotide) dimer analogue, 1-benzyl-1,4-dihydronicotinamide dimer, (BNA)₂, in acetonitrile containing a cyclopentadienyliron dicarbonyl dimer, [CpFe-(CO)₂]₂, results in generation of 2 equiv of the cyclopentadienyliron dicarbonyl anion, [CpFe(CO)₂]^-, accompanied by the oxidation of (BNA)₂ to yield 2 equiv of BNA⁺. The studies on the quantum yields, the electrochemistry, and the transient absorption spectra have revealed that the photochemical generation of [CpFe(CO)₂]^- by (BNA)₂ proceeds via photoinduced electron transfer from the triplet excited state of (BNA)₂ to [CpFe(CO)₂]₂.

Full text of DOI:10.1021/ic0009627

Inorg. Chem. 2001, 40, 1213-1219
1213
Photochemical Generation of Cyclopentadienyliron Dicarbonyl Anion by a Nicotinamide
Adenine Dinucleotide Dimer Analogue
Shunichi Fukuzumi,*,† Kei Ohkubo,† Mamoru Fujitsuka,‡ Osamu Ito,‡
Marcus C. Teichmann,§ Emmanuel Maisonhaute,§ and Christian Amatore*,§
Department of Materials and Life Science, Graduate School of Engineering, Osaka University, CREST,
Japan Science and Technology Corporation, Suita, Osaka 565-0871, Japan, Institute for Chemical
Reaction Science, Tohoku University, CREST, Japan Science and Technology Corporation,
Sendai, Miyagi 980-8577, Japan, and De´partement de Chimie, Ecole Normale Supe´rieure, UMR CNRS
8640 “Pasteur”, 24 rue Lhomond, F-75231 Paris Cedex 05, France
ReceiVed August 22, 2000
Irradiation of the absorption band of an NAD (nicotinamide adenine dinucleotide) dimer analogue, 1-benzyl-
1,4-dihydronicotinamide dimer, (BNA)2, in acetonitrile containing a cyclopentadienyliron dicarbonyl dimer, [CpFe-
(CO)2]2, results in generation of 2 equiv of the cyclopentadienyliron dicarbonyl anion, [CpFe(CO)2]-, accompanied
by the oxidation of (BNA)2 to yield 2 equiv of BNA+. The studies on the quantum yields, the electrochemistry,
and the transient absorption spectra have revealed that the photochemical generation of [CpFe(CO)2]- by (BNA)2
proceeds via photoinduced electron transfer from the triplet excited state of (BNA)2 to [CpFe(CO)2]2.
Introduction
the other hand, photochemistry of metal carbonyl compounds
has provided a valuable synthesis technique in organometallic
The cyclopentadienyliron dicarbonyl anion, [CpFe(CO)2]-
(Cp ) η5-C5H5), has played an important role in the develop-
ment of inorganic and organometallic chemistry1-3 because it
is readily alkylated, acylated, or metalated by reaction with an
appropriate electrophile.1-3 However, strong reducing reagents
such as Na/Hg amalgam,4 Na/K alloy,5 Na dispersion,6 and
trialkylborohydrides7 have so far been required to produce
[CpFe(CO)2]- by the chemical reductive cleavage of the
cyclopentadienyliron dicarbonyl dimer, [CpFe(CO)2]2. Alter-
natively, [CpFe(CO)2]- can be produced by the electrochemical
reduction of [CpFe(CO)2]2 at highly negative potentials.8,9 On
chemistry.10-17 In particular, photochemistry of [CpFe(CO)2]2
has received much attention because there are a diversity of
reaction pathways leading to mononuclear, dinuclear, and
even ionic products.12,15-17 Loss of CO and homolysis of the
Fe-Fe bond are primary photochemical processes known for
[CpFe(CO)2]2.12-17 However, there has so far been no report
on the photochemical reduction of [CpFe(CO)2]2 by an electron
donor to produce [CpFe(CO)2]-.
We report herein a convenient method for generation of
[CpFe(CO)2]- by the photochemical reduction of [CpFe(CO)2]2
by a unique organic two-electron donor, that is, an NAD
* To whom correspondence should be addressed. E-mail for S.F.:
fukuzumi@ap.chem.eng.osaka-u.ac.jp.
christian.amatore@ens.fr.
† Osaka University.
E-mail
for
C.A.:
(9) (a) Dessy, R. E.; Weissman, P. M.; Pohl, R. L. J. Am. Chem. Soc.
1966, 88, 5117. (b) Legzdins, P.; Wassink, B. Organometallics 1984,
3, 1811. (c) Miholova´, D.; Vlcek, A. A. Inorg. Chim. Acta 1980, 41,
119. (d) Mora´n, M.; Cuadrado, I.; Losada, J. J. Organomet. Chem.
1987, 320, 317.
‡ Tohoku University.
§ Ecole Normale Supe´rieure.
(1) (a) Ellis, J. E. J. Organomet. Chem. 1975, 86, 1. (b) Ellis, J. E.
ComprehensiVe Organic Synthesis; Trost, B. M., Ed.; Pergamon
Press: Oxford, 1991; Vol. 4.
(10) (a) Meyer, T. J. Prog. Inorg. Chem. 1975, 19, 1. (b) Bullock, J. P.;
Palazotto, M. C.; Mann, K. R. Inorg. Chem. 1991, 30, 1284.
(11) Pugh, J. R.; Meyer, T. J. J. Am. Chem. Soc. 1992, 114, 3784.
(12) Wrighton, M. Chem. ReV. 1974, 74, 401.
(13) (a) Meyer, T. J.; Caspar, J. V. Chem. ReV. 1985, 85, 187. (b) Caspar,
J. V.; Meyer, T. J. J. Am. Chem. Soc. 1980, 102, 7794.
(14) Geoffroy, G. L.; Wrighton, M. S. Organometallic Photochemistry;
Academic Press: New York, 1979.
(15) (a) Abrahamson, H. B.; Palazzotto, M. C.; Reichel, C. L.; Wrighton,
M. S. J. Am. Chem. Soc. 1979, 101, 4123. (b) Blaha, J. P.; Bursten,
B. E.; Dewan, J. C.; Frankel, R. B.; Randolph, C. L.; Wilson, B. A.;
Wrighton, M. S. J. Am. Chem. Soc. 1985, 107, 4561. (c) Hooker, R.
H.; Mahmoud, K. A.; Rest, A. J. J. Chem. Soc., Chem. Commun. 1983,
1022.
(16) (a) Bursten, B. E.; McKee, S. D.; Platz, M. J. Am. Chem. Soc. 1989,
111, 3428. (b) Zhang, S.; Brown, T. L. J. Am. Chem. Soc. 1993, 115,
1779. (c) Kvietok, F. A.; Bursten, B. E. J. Am. Chem. Soc. 1994, 116,
9807. (d) Tyler, D. R.; Schmidt, M. A.; Gray, H. B. J. Am. Chem.
Soc. 1979, 101, 2753.
(2) (a) King, R. B. AdV. Organomet. Chem. 1964, 2, 157. (b) King, R.
B. Acc. Chem. Res. 1970, 3, 417.
(3) (a) Nitay, M.; Priester, W.; Rosenblum, M. J. Am. Chem. Soc. 1978,
100, 3620. (b) Rosenblum, M. Acc. Chem. Res. 1974, 7, 122. (c)
Rosenblum, M. J. Organomet. Chem. 1986, 300, 191. (d) Davis, R.;
Khazaal, N. M. S.; Maistry, V. J. Chem. Soc., Chem. Commun. 1986,
1387.
(4) King, R. B. Organometallic Synthesis; Academic Press: New York,
1965; Vol. 1.
(5) Ellis, J. E.; Flom, E. A. J. Organomet. Chem. 1975, 99, 263.
(6) (a) Piper, T. S.; Wilkinson, G. J. Inorg. Nucl. Chem. 1956, 3, 104.
(b) Fischer, E. O.; Bo¨ttcher, R. Z. Naturforsch., B 1955, 10, 600. (c)
Reger, D. L.; Fauth, D. J.; Dukes, M. D. Synth. React. Inorg. Met.-
Org. Chem. 1977, 7, 151.
(7) Gladysz, J. A.; Williams, G. M.; Tam, W.; Johnson, D. L.; Parker, D.
W.; Selover, J. C. Inorg. Chem. 1979, 18, 553.
(8) (a) Connelly, N. G.; Geiger, W. E. AdV. Organomet. Chem. 1984,
23, 1. (b) Ferguson, J. A.; Meyer, T. J. Inorg. Chem. 1971, 10, 1025.
(c) Johnson, E. C.; Meyer, T. J.; Winterton, N. Inorg. Chem. 1971,
10, 1673. (d) Pugh, J. R.; Meyer, T. J. J. Am. Chem. Soc. 1988, 110,
8245.
(17) (a) George, M. W.; Dougherty, T. P.; Heilweil, E. J. J. Phys. Chem.
1996, 100, 201. (b) Dixon, A. J.; George, M. W.; Hughes, C.;
Poliakoff, M.; Turner, J. J. J. Am. Chem. Soc. 1992, 114, 1719. (c)
Dixon, A. J.; Healy, M. A.; Poliakoff, M.; Turner, J. J. J. Chem. Soc.,
Chem. Commun. 1986, 994.
10.1021/ic0009627 CCC: $20.00 © 2001 American Chemical Society
Published on Web 02/13/2001

1214 Inorganic Chemistry, Vol. 40, No. 6, 2001
Fukuzumi et al.
spherical joints also fitted with Kalretz O-rings. The working electrode
was a homemade platinum disk microelectrode of 10 µm radius.24 The
counter electrode consisted of a platinum spiral, and the quasi-reference
electrode was a silver spiral. The quasi-reference electrode drift was
negligible for the time required by a single experiment. Both the counter
and the reference electrodes were separated from the working electrode
by ∼0.5 cm. Potentials were measured with ferrocene standard and
are always referenced to SCE. E1/2 values correspond to (Epc + Epa)/2
from CV curves. Ferrocene was also used as an internal standard for
checking the electrochemical reversibility of a redox couple.
Fluorescence Quenching. Fluorescence measurements were carried
out on a Shimadzu RF-5000 spectrofluorophotometer. The solution was
deoxygenated by argon purging for 10 min prior to the measurements.
The fluorescence decay of (BNA)2 was measured using a Horiba NAES-
1100 time-resolved spectrofluorophotometer.
(nicotinamide adenine dinucleotide) dimer analogue 1-benzyl-
1,4-dihydronicotinamide dimer [(BNA)2].18,19 Combination of
the photochemical and electrochemical results obtained in this
study provides confirmative bases to elucidate the reaction
mechanism of the photochemical reduction of [CpFe(CO)2]2 by
(BNA)2.
Experimental Section
Materials. Cyclopentadienyliron dicarbonyl dimer, [CpFe(CO)2]2
(Cp ) η5-C5H5), was purchased from Tokyo Kasei Organic Chemicals,
Japan, and from Aldrich, France, for the experiments performed in
Osaka and in Paris, respectively. In both cases it was used as received.
Pentamethylcyclopentadienyliron dicarbonyl dimer, [Cp*Fe(CO)2]2
(Cp* ) η5-C5Me5), was obtained from Strem Chemicals, Inc., U.S.
[CpFe(CO)2]-NBu4 used in the voltammetric experiments was pre-
+
Laser Flash Photolysis. The measurements of transient absorption
spectra of 3[(BNA)2]* in the presence of [CpFe(CO)2]2 were performed
according to the following procedures. The (BNA)2 solution (1.0 ×
10-3 M) was excited by a Nd:YAG laser (Quanta-Ray, GCR-130, 6 ns
fwhm) at 350 nm with a power of 7 mJ. A pulsed Xenon flash lamp
(Tokyo Instruments, XF80-60, 15 J, 60 ms fwhm) was used for the
probe beam, which was detected with a Si-PIN photodiode (Hamamat-
su G5125-10) after passing through the photochemical quartz vessel
(10 mm × 10 mm) and a monochromator. The output from Si-PIN
photodiode was recorded with a digitizing oscilloscope (HP 54510B,
300 MHz). Since the solution of (BNA)2 in acetonitrile disappeared
by each laser shot (350 nm; 7 mJ) in the presence of [CpFe(CO)2]2,
the transient spectra were recorded using fresh solutions for each laser
excitation. All experiments were performed at 298 K. The solution was
deoxygenated by argon purging for 10 min prior to the measurements.
ESR Measurements. ESR spectra of the photolyzed MeCN solution
of [CpFe(CO)2]2 (5.0 × 10-4 M) and (BNA)2 (5.0 × 10-4 M) were
taken on a JEOL JES-RE1XE and were recorded under nonsaturating
microwave power conditions. A sample solution was irradiated by using
a high-pressure mercury lamp (USH-1005D) focusing at the sample
cell in the ESR cavity. The magnitude of modulation was chosen to
optimize the resolution and signal-to-noise (S/N) ratio of the observed
spectra. The g values were calibrated using a Mn2+ marker.
Theoretical Calculations. Density functional calculations were
performed on a Compaq DS20E computer using the Amsterdam density
functional (ADF) program, version 1999.02, developed by Baerends
et al.25 The electronic configurations of the molecular systems were
described by an uncontracted triple-ú Slater-type orbital basis set (ADF
basis set IV) with a single polarization function used for each atom.
Core orbitals were frozen through 1s (C, O) and 3p (Fe). The
calculations were performed using the local exchange-correlation
potential by Vosko et al.26 and the nonlocal gradient corrections by
Becke27 and Perdew28 during the geometry optimizations. First-order
scalar relativistic correlations were added to the total energy. Final
geometries and energetics were optimized by using the algorithm of
Versluis and Ziegler29 provided in the ADF package and were
considered converged when the changes in bond lengths between
subsequent iterations fell below 0.01 Å.
pared by pre-electrolysis in a two-compartment cell of a solution of
[CpFe(CO)2]2 at -1.6 V vs SCE in deaerated acetonitrile in the presence
of 0.3 M NBu4BF4. An amount of 10 mL of the electrolyzed solution
was transferred with a syringe to the voltammetric cell. 1-Benzyl-1,4-
dihydronicotinamide dimer [(BNA)2] was prepared according to the
literature.20 Acetonitrile (MeCN) used as a solvent was purified and
dried according to the standard procedure.21
Reaction Procedure. Typically, [CpFe(CO)2]2 (1.0 × 10-3 M) and
(BNA)2 (1.0 × 10-3 M) were added to an NMR tube that contained
deaerated CD3CN solution (0.60 cm3) under an atmospheric pressure
of argon and the solution was irradiated with a Xe lamp (Ushio model
V1-501C) through a UV cut-off filter (Toshiba UV-31) transmitting λ
> 300 nm at 298 K for 30 min. The reaction products of [CpFe(CO)2]2
and [Cp*Fe(CO)2]2 with (BNA)2 were identified by the 1H NMR spectra
by comparing them with those of authentic samples. The 1H NMR
measurements were performed using a JNM-GSX-400 (400 MHz) NMR
spectrometer. 1H NMR (CD3CN): [CpFe(CO)2]- δ (Me4Si, ppm), 4.16
(s, 5H); [Cp*Fe(CO)2]-, δ 1.79 (s, 15H); BNA+, δ 5.76 (s, 2H), 7.49
(m, 5H), 8.11 (t, 1H, J ) 6.8 Hz), 8.80 (m 2H,), 9.16 (s, 1H).
Quantum Yield Determinations. A standard actinometer (potassium
ferrioxalate)22 was used for the quantum yield determination of the
photoreduction of [CpFe(CO)2]2 by (BNA)2. Square quartz cuvettes (10
mm i.d.), which contained a deaerated MeCN solution (3.0 cm3) of
(BNA)2 (5.6 × 10-4 M) together with [CpFe(CO)2]2 at various
concentrations, were irradiated with monochromatized light of λ ) 350
nm from a Shimadzu RF-5000 fluorescence spectrophotometer. Under
the conditions of actinometry experiments, the actinometer and (BNA)2
absorbed essentially all the incident light of λ ) 350 nm. The light
intensity of monochromatized light of λ ) 350 nm was determined as
2.32 × 10-9 einstein s-1 with a slit width of 20 nm. The photochemical
reaction was monitored using a Hewlett-Packard 8452A diode-array
spectrophotometer. The quantum yields were determined from an
increase in absorbance due to the cyclopentadienyliron dicarbonyl anion,
[CpFe(CO)2]-. To avoid the contribution of light absorption of the
products, only the initial rates were determined for determination of
the quantum yields.
Cyclic Voltammetry. Cyclic voltammetry measurements were
performed with a homemade potentiostat23 and a waveform generator,
PAR model 175. The fast-scan cyclic voltammograms were recorded
with a Nicolet 3091 digital oscilloscope. The one-compartment
electrochemical cell was of airtight design with high-vacuum glass
stopcock fitted with either Teflon or Kalretz (Du Pont) O-rings in order
to prevent contamination by grease. The connections to the high-vacuum
line and to the Schlenck containing the solvent were obtained by
Results and Discussion
Photoreduction of [CpFe(CO)2]2. Irradiation of an aceto-
nitrile (MeCN) solution containing (BNA)2 (7.0 × 10-5 M, λmax
) 350 nm) and [CpFe(CO)2]2 (7.0 × 10-5 M) with UV-visible
light (λ ) 350 nm) results in the disappearance of absorbance
due to [CpFe(CO)2]2 and (BNA)2 accompanied by the appear-
ance of a new broad absorption band at ca. λ ) 450 nm with
(18) Patz, M.; Kuwahara, Y.; Suenobu, T.; Fukuzumi, S. Chem. Lett. 1997,
567.
(19) Fukuzumi, S.; Suenobu, T.; Patz, M.; Hirasaka, T.; Itoh, S.; Fujitsuka,
M.; Ito, O. J. Am. Chem. Soc. 1998, 120, 8060.
(20) Wallenfels, K.; Gellrich, M. Chem. Ber. 1959, 92, 1406.
(21) Perrin, D. D.; Armarego, W. L. F. Purification of Laboratory
Chemicals; Butterworth-Heinemann: Oxford, 1988.
(22) Hatchard, C. G.; Parker, C. A. Proc. R. Soc. London, Ser. A 1956,
235, 518.
(24) Amatore, C.; Thouin, L.; Bento, M. F. J. Electroanal. Chem. 1999,
463, 45.
(25) (a) Baerends, E. J.; Ellis, D. E.; Ros, P. Chem. Phys. 1973, 2, 41. (b)
te Velde, G.; Baerends, E. J. J. Comput. Phys. 1992, 99, 84.
(26) Vosko, S. H.; Wilk, L.; Nusair, M. Can. J. Phys. 1980, 58, 1200.
(27) Becke, A. Phys. ReV. A 1988, 38, 3098.
(28) (a) Perdew, J. P. Phys. ReV. B 1986, 33, 8822. (b) Perdew, J. P. Phys.
ReV. B 1986, 34, 7406.
(29) Versluis, L.; Ziegler, T. J. Chem. Phys. 1988, 88, 322.
(23) Amatore, C.; Lefrou, C.; Pflu¨ger, F. J. Electroanal. Chem. 1989, 270,
43.

Cyclopentadienyliron Dicarbonyl Anion
Inorganic Chemistry, Vol. 40, No. 6, 2001 1215
Figure 1. Electronic absorption spectra observed in the photochemical
reaction of [CpFe(CO)2]2 (7.0 × 10-5 M) with (BNA)2 (7.0 × 10-5
M) in deaerated MeCN at 298 K.
clean isosbestic points as shown in Figure 1. The oxidation and
reduction products of (BNA)2 and [CpFe(CO)2]2 were identified
1
as BNA+ and [CpFe(CO)2]- by the H NMR spectra of the
product solution, respectively (see Experimental Section). Thus,
the stoichiometry of the photochemical reaction is given by eq
1, where [CpFe(CO)2]2 is known to exist predominantly as a
Figure 2. (a) Dependence of the quantum yield (Φ) on [Fp2] for the
photoreduction of [CpFe(CO)2]2 (Fp2) by (BNA)2 (5.6 × 10-5 M) in
deaerated MeCN at 298 K. (b) Plot of Φ-1 vs [Fp2]-1
.
Φ-1 ) Φ∞-1[1 + (Kobs[Fp2])-1]
(3)
and the linear plot of Φ-1 vs [Fp2]-1 is shown in Figure 2b.
From the slope and intercept are obtained the Φ∞ and Kobs values
cis isomer in the two-carbonyl bridged form.30-32 When [CpFe-
(CO)2]2 was replaced by the pentamethylcyclopentadienyliron
dicarbonyl dimer, [Cp*Fe(CO)2]2 (Cp* ) η5-C5Me5), the
photoreduction of [Cp*Fe(CO)2]2 by (BNA)2 hardly occurred.
as 4.8 × 10-3 and 3.3 × 104 M-1
.
Irradiation of the absorption band (λmax ) 350 nm) of (BNA)2
results in fluorescence (λmax ) 390 nm). From these values is
obtained the excitation energy as 3.36 eV. Since the one-electron
oxidation potential (E0ox) of (BNA)2 is 0.26 V (vs SCE),18 the
E0ox value of the singlet excited state [1(BNA)2*] is determined
The irradiation with light of λmax of (BNA)2 is essential for
the selective formation of [CpFe(CO)2]- without loss of CO.
The quantum yield of the photogeneration of [CpFe(CO)2]- was
determined from an increase in absorbance due to [CpFe(CO)2]-
under irradiation with light at λ ) 350 nm. The quantum yield
(Φ) increased with an increase in the [CpFe(CO)2]2 concentra-
tion to approach a limiting value (Φ∞), as shown in Figure 2a.
Such a saturated dependence of Φ on the [CpFe(CO)2]2
concentration is expressed by
as -3.1 V by subtracting the excitation energy from the E0
ox
value of the ground state. On the other hand, a fast-scan cyclic
voltammetry was used to determine the one-electron reduction
potential (E0red) of [CpFe(CO)2]2 in MeCN at 298 K. Slow-
scan voltammograms of [CpFe(CO)2]2 at temperatures above
248 K have been reported to show an irreversible reduction wave
due to the instability of an initially produced dimer radical anion
[CpFe(CO)2]2•-, which dissociates into a mononuclear anion,
[CpFe(CO)2]-, and a further reducible radical, [CpFe(CO)2]•,
Φ∞Kobs[Fp2]
Φ )
(2)
1 + Kobs[Fp2]
[CpFe(CO)2]2•- f [CpFe(CO)2]• + [CpFe(CO)2]- (4)
where Fp2 denotes [CpFe(CO)2]2 and Kobs is the quenching
constant that can be converted to the corresponding rate con-
stant (kobs) provided that the lifetime of the excited state involved
in the reaction (τ) is known: kobs ) Kobsτ-1. Equation 2 is
rewritten as
resulting in a net two-electron reduction.9,33,34 As the temperature
is lowered, the reduction of [CpFe(CO)2]2 becomes reversible.34
Electrochemical Oxidation of [CpFe(CO)2]-. At moderate
scan rates (V < 102 V s-1) the steady-state limit of the
voltammetric pattern obtained upon continuous cycling between
-0.42 and -2.1 V vs SCE on a bulk solution of [CpFe(CO)2]-
(30) Hepp, A. F.; Blaha, J. P.; Lewis, C.; Wrighton, M. S. Organometallics
1984, 3, 174.
(31) Bullitt, J. G.; Cotton, F. A.; Marks, T. J. J. Am. Chem. Soc. 1970, 92,
2155.
(33) Davies, S. G.; Simpson, S. J.; Parker, V. D. J. Chem. Soc., Chem.
Commun. 1984, 352.
(32) Adams, R. D.; Cotton, F. A. J. Am. Chem. Soc. 1973, 95, 6589.
(34) Dalton, E. F.; Ching, S.; Murray, R. W. Inorg. Chem. 1991, 30, 2642.

1216 Inorganic Chemistry, Vol. 40, No. 6, 2001
Fukuzumi et al.
Scheme 1
Figure 3. Cyclic voltammogram for the oxidation of the monomer
anion [CpFe(CO)2]- (3.1 × 10-3 M) in deaerated MeCN containing
to dimerize with a rate constant of 3.2 × 109 M-1 s-1 13
,
so
Bu4NBF4 with Pt electrode at 298 K. Sweep rate is 3280 V s-1
.
intercepting its fast dimerization under the millimolar conditions
used here would require scan rates ca. 100 times larger than
that used in Figure 3.35 Moreover, if the [CpFe(CO)2]• radical
were detected under the conditions of Figure 3, its ESR spectrum
would be observable under irradiation of a mixture of (BNA)2
and [CpFe(CO)2]2 (vide infra) with an intensity being ap-
proximately 20% that of [CpFe(CO)2]2•- based on their relative
lifetimes. Such an ESR signal is not observed (vide infra), which
points to a much larger reaction rate for [CpFe(CO)2]•, in
agreement with a previous report.13 Henceforth, under the
conditions used in Figure 3, the [CpFe(CO)2]• radical must
undergo a quantitative dimerization and none can be reduced
at wave Ic. This indicates that the species detected at wave Ic is
an intermediate on the way to [CpFe(CO)2]2 in which the Fe-
Fe bond is already made.
exhibits two major waves as reported previously.34,35 One, wave
Ia, is anodic and features the one-electron oxidation of
[CpFe(CO)2]- at ca. -0.75 V vs SCE.33,34 The other, wave IIc
of approximately equal current intensity, is observed around
ca. -1.75 V vs SCE and features the reduction of the [CpFe-
(CO)2]2 dimer as evidenced by comparison with an authentic
sample. This pattern confirms that the radical [CpFe(CO)2]•
formed upon the one-electron oxidation of [CpFe(CO)2]- at
wave Ia undergoes a fast dimerization to afford [CpFe(CO)2]2
that is then reduced through an overall two-electron reduction
at wave IIc to regenerate the parent anion, thus giving rise to
the impression of the occurrence of a reversible couple with a
large peak-to-peak separation (ca. 1 V).
Besides the facts that both peak potentials experience
displacements with the scan rate (ca. -20 mV per log V for
wave Ia, and ca. -30 mV per log V for wave IIc) and that the
peak of wave Ia depends on the concentration of [CpFe(CO)2]•
(ca. -20 mV per log C), further evidence that the system Ia/IIc
is not a canonical reversible couple is given by the growth of
a reversible anodic wave, IIa, coupled to wave IIc, as soon as
the scan rate is increased above a few 102 V s-1. Above a few
kV s-1, wave IIa has a peak current intensity comparable to
that of wave IIc, showing that a full reversible behavior is
achieved for the redox couple [CpFe(CO)2]2/[CpFe(CO)2]2•- as
shown in Figure 3. This allows the determination of the rate
constant kc ≈ 5 × 102 s-1 of the Fe-Fe bond cleavage in
[CpFe(CO)2]2•-, as well as of the standard reduction potential
of [CpFe(CO)2]2 at E1/2 ) -1.55 V vs SCE.
In fact, [CpFe(CO)2]2 is known to predominantly exist not
only in the stable two-carbonyl bridged form but also in a less
stable unbridged form.30-32 With the unbridged form having a
significantly weaker Fe-Fe bond than the bridged form, it is
reasonable for such species to be reducible at a much lesser
negative potential than its more stable [CpFe(CO)2]2 isomer
(note that the peak potential separation between waves Ic
and IIc is ca. 0.8 V). Furthermore, its one-electron reduction at
wave Ic should produce a more unstable anion radical than
•-
[CpFe(CO)2]2 that is detected at wave IIa, so its reduction
should regenerate its parent anion [CpFe(CO)2]- more quickly
than that of the more stable [CpFe(CO)2]2 isomer. In other
words, although it is impossible for us to further characterize
this species, it seems highly probable that this species detected
by its reduction at wave Ic is the unbridged dimer and that
the 140 µs half-lifetime corresponds to the bridging rate con-
stant (5 × 103 s-1) of the two carbonyl ligands as shown in
Scheme 1.
When the scan rate approaches the kV s-1 range, a second
cathodic wave (Ic) develops, being apparently the anodic
counterpart of wave Ia. The growth of this wave upon increasing
scan rate is concomitant with a comparable decay of the system
of waves IIc/IIa, showing that the species reduced at wave Ic is
an intermediate on the way to [CpFe(CO)2]2. At V ≈ 3 kV s-1
(Figure 3), the peak current intensity of wave Ic is approximately
equal to those of waves IIc and IIa. This establishes that the
intermediate reduced at wave Ic has a half-lifetime of ca. 140
µs. Such voltammetric behavior would suggest that this
intermediate is the [CpFe(CO)2]• radical formed upon the one-
electron oxidation of [CpFe(CO)2]- at wave Ia. However, this
interpretation must be ruled out because this radical is reported
When an authentic sample of the stable [CpFe(CO)2]2 isomer
is reduced voltammetrically, the existence of the equilibrium
between the unbridged and bridged isomers in Scheme 130-32
should lead to a CE sequence as soon as the electrode potential
reaches wave Ic. Therefore, [CpFe(CO)2]2 could be reduced via
the continuous displacement of the equilibrium (Scheme 1) to
the side of the unbridged isomer owing to the consumption of
the unbridged isomer at wave Ic, unless the backward rate
constant is too small.36 We did not observe any evidence of
such a behavior even for the smallest scan rates used in this
(35) Amatore, C.; Jutand, A.; Pflu¨ger, F. J. Electroanal. Chem. 1987, 218,
(36) Bard, A. J.; Faulkner, L. R. Electrochemical Methods; J. Wiley &
361.
Sons: New York, 1980.

Cyclopentadienyliron Dicarbonyl Anion
Inorganic Chemistry, Vol. 40, No. 6, 2001 1217
Figure 5. Fluorescence decay of (BNA)2 (1.0 × 10-4 M) at λ ) 390
nm in deaerated MeCN at 298 K.
The fluorescence decay of 1(BNA)2* obeys first-order kinetics
as shown in Figure 5. The fluorescence lifetime was determined
1
as 7.4 ns. If the singlet excited state (BNA)2* is responsible
for the photoreduction of [CpFe(CO)2]2 by (BNA)2, the Kobs
value (3.3 × 104 M-1) obtained from saturated dependence of
Φ on [Fp2] (Figure 2) would be converted to the corresponding
rate constant (kobs ) 4.5 × 1012 M-1 s-1) using the relation,
kobs ) Kobsτ-1 and τ ) 7.4 ns. The estimated kobs value is much
larger than the diffusion-limited value (2.0 × 1010 M-1 s-1).37
This indicates that the excited state involved in the photore-
duction of [CpFe(CO)2]2 by (BNA)2 should have a much longer
lifetime than the singlet excited state, and therefore, it may be
the triplet excited state of (BNA)2.
The existence of the triplet excited state of (BNA)2 is shown
by the laser flash photolysis of an MeCN solution of (BNA)2
with 355 nm laser light. The result is shown in Figure 6a where
a new absorption band at 430 nm, which is attributed to the
triplet-triplet (T-T) absorption of the triplet excited state
[3(BNA)2*], is observed upon laser excitation. The T-T
absorption decays obeying first-order kinetics as shown in Figure
6b. The triplet lifetime is determined as τT ) 20 µs.
Figure 4. Ball and stick diagrams of the optimized structures of (a)
the bridged and (b) unbridged isomers of [CpFe(CO)2]2 and the LUMO
orbitals.
study (50 mV s-1). Therefore, the backward rate constant of
equilibrium (Scheme 1) has to be much less than 1 s-1
.
Henceforth, the equilibrium constant of CO bridging has to be
much larger than 5 × 103 (viz., ∆G° < -5.0 kcal mol-1). Within
this guideline, the E1/2 ) -0.88 V vs SCE deduced from the
half-sum of the peak potentials of waves Ia and Ic does not
correspond at all to the standard oxidation potential of [CpFe-
(CO)2]• but to a kinetic potential corresponding to the pseu-
doreversible sequence in Scheme 1.
In the presence of [CpFe(CO)2]2, the T-T absorption
observed at 0.25 µs after laser excitation (λmax ) 430 nm) is
changed to a new transient absorption band at 2.5 µs (λmax
)
470 nm), which may be attributed to [CpFe(CO)2]2•- produced
3
by photoinduced electron transfer from (BNA)2* to [CpFe-
(CO)2]2 as shown in Figure 7a. On the other hand, the oxidized
product of (BNA)2 [(BNA)2•+] is known to be converted to
BNA• and BNA+ via facile C-C bond cleavage.18,19 Since the
transient absorption spectrum of NAD• was reported to appear
at λmax ) 500 nm,38 the absorption spectrum of BNA• may be
overlapped with that of [CpFe(CO)2]2•- at 470 nm. An increase
in absorbance due to [CpFe(CO)2]2•- at 470 nm obeys pseudo-
first-order kinetics as shown in Figure 6b and the pseudo-first-
order rate constant increases linearly with increasing the
[CpFe(CO)2]2 concentration. From the slope of a linear cor-
relation of the pseudo-first-order rate constant vs the [CpFe-
(CO)2]2 concentration is obtained the rate constant (ket) of
photoinduced electron transfer from 3(BNA)2* to [CpFe(CO)2]2
as 1.8 × 109 M-1 s-1. On the other hand, the Kobs value (3.3 ×
104 M-1) obtained from the saturated dependence of Φ on [Fp2]
(Figure 2) is converted to the corresponding rate constant (kobs
) 1.6 × 109 M-1 s-1) provided that the triplet excited state
3(BNA)2* is responsible for the photoreduction of [CpFe(CO)2]2
The thermodynamic stabilities of the bridged and unbridged
isomers of [CpFe(CO)2]2 were evaluated by the Amsterdam
density functional (ADF) calculations (see Experimental
Section).25-29 The final geometries and energetics were obtained
by optimizing the total molecular energy with respect to all
structural variables as shown in Figure 4. Consistent with the
above estimation of the ∆G° value (< -5.0 kcal mol-1), we
find that the bridged isomer is more stable than the unbridged
trans isomer by 12.3 kcal mol-1. The unbridged cis isomer is
less stable than the unbridged trans isomer by 13.7 kcal mol-1
.
The LUMO (lowest unoccupied molecular orbital) of the
unbridged isomer consists of the Fe-Fe σ antibonding orbital,
whereas the LUMO of the bridged isomer involves the Fe-
CO d-π* bonding orbital. This may be the reason for the facile
cleavage of the Fe-Fe bond of the unbridged isomer upon the
one-electron reduction (Scheme 1).
Mechanism of Photoreduction of [CpFe(CO)2]2 by (BNA)2.
1
Judging from the one-electron redox potentials of (BNA)2*
(E0 ) -3.10 V) and [CpFe(CO)2]2 (E0 ) -1.55 V), the
ox
red
(37) Rehm, A.; Weller, A. Isr. J. Chem. 1970, 8, 259.
(38) (a) Czochralska, B.; Lindqvist, L. Chem. Phys. Lett. 1983, 101, 297.
(b) Lindqvist, L.; Czochralska, B.; Grigorov, I. Chem. Phys. Lett. 1985,
119, 494.
photoinduced electron transfer from 1(BNA)2* to [CpFe(CO)2]2
is highly exergonic because the free energy change of electron
transfer (∆G0et) is -1.55 eV.

1218 Inorganic Chemistry, Vol. 40, No. 6, 2001
Fukuzumi et al.
Figure 6. (a) T-T absorption spectrum of (BNA)2 (1.0 × 10-3 M)
obtained by the laser flash photolysis in deaerated MeCN at 298 K.
(b) Kinetic trace for the T-T absorption at 430 nm in deaerated MeCN.
Inset: first-order plot.
Figure 7. (a) Transient absorption spectra observed in the photore-
duction of [CpFe(CO)2]2 (3.0 × 10-4 M) by (BNA)2 (1.0 × 10-3 M)
after laser excitation in deaerated MeCN at 298 K. (b) Kinetic trace
for formation of [CpFe(CO)2]2•-. Inset: first-order plot.
-1
by (BNA)2, using the relation kobs ) KobsτT and τT ) 20 µs.
This kobs value agrees with the ket value (1.8 × 109 M-1 s-1
)
determined directly from the appearance of the transient
absorption at 470 nm due to [CpFe(CO)2]2•- (Figure 7a). Such
an agreement strongly indicates that the photoreduction of
[CpFe(CO)2]2 by (BNA)2 proceeds via photoinduced electron
3
transfer from (BNA)2* to [CpFe(CO)2]2.
•-
The Fe-Fe bond of [CpFe(CO)2]2 generated in photoin-
duced electron transfer from 3(BNA)2* to [CpFe(CO)2]2 is
cleaved to give [CpFe(CO)2]• and [CpFe(CO)2]- (Scheme 1).33,34
However, the cleavage rate is relatively slow as estimated by
the electrochemical measurements (Scheme 1).33,34 In such a
•-
case [CpFe(CO)2]2 may be stable enough to be detected by
ESR at a low temperature. In fact, a broad isotropic ESR signal
(g ) 2.0073) is detected under photoirradiation of an MeCN
solution of (BNA)2 (5.0 × 10-4 M) and [CpFe(CO)2]2 (5.0 ×
10-4 M) with a high-pressure mercury lamp at 243 K as shown
in Figure 8. When the temperature is lowered to 173 K, the
isotropic signal is changed to the anisotropic signal in a frozen
medium with g1 ) 2.0554, g2 ) 2.0031, and g3 ) 1.9635. The
averaged value (2.0073) agrees with the isotropic value in
solution. The ESR spectrum with a g value of 2.0073, which is
much larger than the free spin value, may be attributed to
[CpFe(CO)2]2•-. No organic radicals such as BNA• was detected
probably because of the fast dimerization.19,39 In such a case it
is unlikely that [CpFe(CO)2]• is detected by ESR because
dimerization of [CpFe(CO)2]• is known to be as rapid as that
of BNA•.13,19
•-
Figure 8. ESR spectra of [CpFe(CO)2]2 observed under irradiation
of a deaerated MeCN solution containing (BNA)2 (5.0 × 10-4 M) and
[CpFe(CO)2]2 (5.0 × 10-4 M) (a) at 243 K and (b) at 173 K (frozen).
as shown in Scheme 2. The triplet excited state 3(BNA)2*
generated by intersystem crossing (ISC) upon photoexcitation
of (BNA)2 is quenched by electron transfer to [CpFe(CO)2]2 to
give the radical ion pair in competition with the decay to the
On the basis of the above results, the reaction mechanism of
the photoreduction of [CpFe(CO)2]2 by (BNA)2 is summarized
ground state.40 The C-C bond of (BNA)2 is readily cleaved
•+
to produce BNA• and BNA+. In the case of the photoinduced
(39) Fukuzumi, S.; Koumitsu, S.; Hironaka, K.; Tanaka, T. J. Am. Chem.
Soc. 1987, 109, 305.
electron transfer from (BNA)2 to C70, the C-C bond cleavage

Cyclopentadienyliron Dicarbonyl Anion
Inorganic Chemistry, Vol. 40, No. 6, 2001 1219
Scheme 2
By application of the steady-state approximation to the
reactive intermediates in Scheme 2, the dependence of quantum
yields on the substrate concentration [Fp2] is derived as shown
in
[kc/(kc + kb)]ketτT[Fp2]
Φ )
(5)
1 + ketτT[Fp2]
which agrees with the experimental observation in eq 2. The
efficiency for the photochemical reaction is thereby determined
by the Fe-Fe bond cleavage rate in competition with the back
•-
electron transfer from [CpFe(CO)2]2 to BNA+, which may
be diffusion-limited judging from the highly negative ∆G0
et
value of the electron transfer (vide supra).43 In the case of
[Cp*Fe(CO)2]2, the Fe-Fe bond cleavage rate in the radical
anion is known to be 2 orders of magnitude slower compared
to the cleavage rate of [CpFe(CO)2]2.34 This may be the reason
that the photoreduction of [Cp*Fe(CO)2]2 by (BNA)2 hardly
occurred (vide supra).
of (BNA)2•+ is reported to be much faster than the back electron
In conclusion, this study provides a new method for photo-
generation of [CpFe(CO)2]- by a unique organic two-electron
donor [(BNA)2] via photoinduced electron transfer from
•-
transfer from C70 to (BNA)2•+, leading to formation of the
products with 100% quantum efficiency.41 In the present case
3
as well, the photoinduced electron transfer from (BNA)2* to
3(BNA)2* to [CpFe(CO)2]2 followed by Fe-Fe bond cleavage
[CpFe(CO)2]2 results in formation of BNA•, BNA+, and
•-
in [CpFe(CO)2]2
.
[CpFe(CO)2]2•-. Since the E0 value of BNA+ (-1.08 V vs
red
SCE)39 is less negative than the E0 value of [CpFe(CO)2]2
•-
ox
Acknowledgment. This work was partially supported by a
Grant-in-Aid for Scientific Research Priority Area (No. 11228205)
from Ministry of Education, Science, Sports and Culture, Japan.
This work was also supported in part by CNRS (UMR 8640
“PASTEUR”), Ecole Normale Supe´rieure, and the French
Ministry of Research. M.C.T. welcomes the award of a Marie
Curie doctoral grant by the European Community.
(-1.55 V), back electron transfer from [CpFe(CO)2]2•- to BNA+
(∆G0 ) 0.47 eV) may occur to regenerate [CpFe(CO)2]2
et
accompanied by formation of BNA•, which dimerized to
reproduce (BNA)2. In competition with such a back electron
transfer process (kb), the Fe-Fe bond cleavage (kc) occurs to
give [CpFe(CO)2]- and [CpFe(CO)2]•.42 The electron transfer
from BNA• to [CpFe(CO)2]• gives BNA+ and [CpFe(CO)2]-.
Thus, the net reaction is two-electron reduction of [CpFe(CO)2]2
by (BNA)2 to yield 2 equiv of [CpFe(CO)2]- and BNA+
(Scheme 2).
IC0009627
(43) The photoreduction of [CpFe(CO)2]2 by (BNA)2 may also proceed
via photoinduced electron transfer from 3(BNA)2* to the unbridged
trans isomer, which exists as a much less stable form, leading to facile
Fe-Fe bond cleavage in the dimer radical anion. If this is the major
pathway, the rate constant (kobs ) KobsτT-1) derived from the
dependence of Φ on [Fp2] (eq 2) would be much smaller than the
rate constant of the photoinduced electron transfer determined directly
from the appearance of the transient absorption at 470 nm due to
(40) Under our experimental conditions, irradiation at 350 nm results in
mostly formation of 3(BNA)2*, and the photocleavage of the Fe-CO
bond of [CpFe(CO)2]2 is negligible on the basis of the stoichiometry
of the photochemical reaction (see Experimental Section).
(41) Fukuzumi, S.; Suenobu, T.; Hirasaka, T.; Sakurada, N.; Arakawa, R.;
Fujitsuka, M.; Ito, O. J. Phys. Chem. A 1999, 103, 5935.
•-
[CpFe(CO)2]2 (Figure 7b). The agreement between the kobs and ket
(42) Acetonitrile used as the solvent for the photochemical reaction via
photoinduced electron transfer may be coordinated to Fe in coordi-
natively unsaturated intermediates in Scheme 2.
values obtained in this study indicates that a reaction pathway via a
cleavage precursor such as an unbridged isomer of [CpFe(CO)2]2 is
not a major pathway in the photoreduction of [CpFe(CO)2]2 by (BNA)2.