The photodecomposition mechanism of tert-butyl-9-methylfluorene-9- percarboxylate: New insight from femtosecond IR spectroscopy

Article abstract of DOI:10.1039/b806359a

The ultrafast photodissociation of tert-butyl-9-methylfluorene-9- percarboxylate (TBFC) is studied by mid-infrared transient absorption spectroscopy after UV excitation at 266 nm. By means of ¹³C-labeled TBFC and additional DFT calculations transient IR bands in the fingerprint region were unambiguously assigned to the methylfluorenyl radical. The experiments show that the fragmentation is controlled by the S ₁-lifetime of TBFC and, dependent on the solvent, within 0.8-2.1 ps leads to tert-butyloxy and methylfluorenyl radicals plus CO₂via concerted bond breakage of the O-O and the fluorenyl-C(carbonyl) bond. In accordance, the CO₂ quantum yield is determined to be unity.

Full text of DOI:10.1039/b806359a

View Article Online / Journal Homepage / Table of Contents for this issue
PAPER
www.rsc.org/pccp | Physical Chemistry Chemical Physics
The photodecomposition mechanism of tert-butyl-9-methylfluorene-
9-percarboxylate: new insight from femtosecond IR spectroscopy
a
ab
Christian Reichardt, Jorg Schroeder and Dirk Schwarzer*a
¨
Received 15th April 2008, Accepted 22nd May 2008
First published as an Advance Article on the web 27th June 2008
DOI: 10.1039/b806359a
The ultrafast photodissociation of tert-butyl-9-methylfluorene-9-percarboxylate (TBFC) is studied
by mid-infrared transient absorption spectroscopy after UV excitation at 266 nm. By means of
13C-labeled TBFC and additional DFT calculations transient IR bands in the fingerprint region
were unambiguously assigned to the methylfluorenyl radical. The experiments show that the
fragmentation is controlled by the S1-lifetime of TBFC and, dependent on the solvent, within
0.8–2.1 ps leads to tert-butyloxy and methylfluorenyl radicals plus CO2 via concerted bond
breakage of the O–O and the fluorenyl–C(carbonyl) bond. In accordance, the CO2 quantum yield
is determined to be unity.
rule, with decarboxylation rates becoming very rapid for
Introduction
highly stable product radicals being formed, but probably
not exceeding 2 ꢁ 1010
s
ꢂ1. This conclusion by Falvey and
Technical polymerization processes on a large scale make use
of organic peroxides as initiators, which as such have been the
subject of numerous studies providing much of the experi-
mental basis of free radical chemistry.1,2 Both a stepwise and a
concerted mechanism have been suggested for peroxide
thermal fragmentation, i.e., initial O–O-bond cleavage in the
peroxy compound RC(O)O–OR0 followed by subsequent dec-
Schuster found support in later photolysis studies showing
that aroyloxyl radicals generally decarboxylate on time scales
ranging from about 50 ns to a few microseconds.6–11 Also a
recent study of the photolysis of organic peroxides containing
coumarin or bis(phenylethynyl)benzene chromophores
showed that CO2 formation occurs on a microsecond time-
scale indicating sequential fragmentation via intermediate
carbonyloxyl radicals.12,13
ꢀ
arboxylation of the RCO2 radical, and a concertꢀed two bond
cleavage yielding instantaneously Rꢀ, CO2, and OR0. While
trapping experiments established the existence of the benzoyl-
oxyl radical C6H5CO2 as an intermediate,3 Bartlett and co-
However, 308 nm photolysis of TBFC as studied by pico-
second UV-pump/IR-probe spectroscopy was shown to gen-
erate CO2 within a few picoseconds.14 Equally rapid CO2
formation was also observed in the UV-photolysis of several
diaroylperoxides, tert-butyl-benzoylperoxide, and tert-butyl-
phenylperoxycarbonate (TBPC).15,16 These results would be
consistent with a concerted fragmentation mechanism via the
electronically excited singlet state without formation of the
ꢀ
workers could show that the thermal decomposition rate in a
series of peroxyesters increases with increasing stability of the
alkyl radical formed in this process, which led them to suggest
that concerted two bond cleavage occurs whenever a suffi-
ciently stable radical is formed, and that the intermediate
RCO2 is not involved.4
ꢀ
In an attempt to look more closely into this problem, Falvey
and Schuster noted that peroxyester photolysis ‘‘leads to
chemical reactions that appear to mimic closely those of
their. . . thermolyses’’,5 and in a landmark experiment studied
the laser flash photolysis of tert-butyl-9-methylfluorene-9-
percarboxylate (TBFC) at 266 nm excitation wavelength with
B20 ps time resolution. They reported a first order transient
absorption rise with a time constant of 55 ps, which they
attributed to the 9-methylfluorenyl radical MFꢀ formed in the
decarboxylation of the intermediate aroyloxyl radical
ꢀ
RCO2 intermediate. On the other hand, a slower build-up of
CO2 with a time constant of 30 ps was found in the photolysis
of tert-butyl-2-naphthylperoxycarbonate15 (TBNC), which
one might consider as being due to fast sequential bond
breakage on a time scale suggested by Falvey and Schuster
for TBFC.
Of course, the ability to distinguish between concerted and
sequential mechanisms heavily depends on the time scale
accessible to observation. Looking at peroxide decomposition
with higher time resolution by femtosecond pump–probe
spectroscopy in the visible to near IR spectral range17,18 led
to the suggestion that, in fact, the reaction mechanism involves
an ultrafast sequence of elementary steps.19 A quantitative
model was developed in which instantaneous O–O-bond scis-
ꢀ
MFCO2 . They also pointed out that because of the very
high stability of the MFꢀ radical, according to Bartlett’ꢀs
proposition, the observed lifetime of the MFCO2
radical should represent
a lower limit for aroyloxyl
radical lifetimes in general. As a consequence, a two step
sequential mechanism as depicted in Scheme 1 would be the
ꢀ
sion generates vibrationally highly excited RCO2 that may
rapidly decarboxylate while simultaneously undergoing vibra-
tional energy relaxation (VER). The rates of simultaneous
a Abteilung Spektroskopie und Photochemische Kinetik, Max-Planck-
ꢀ
CO2 formation and RCO2 decay depend on vibrational
ꢀ
Institut fur Biophysikalische Chemie, 37070 Gottingen, Germany
b Institut fur Physikalische Chemie, Georg-August-Universitat
¨
¨
excess energy, VER rates and barrier heights to RCO2
¨
Gottingen, Tammannstraße 6, 37077 Gottingen, Germany
¨
decarboxylation. This model successfully described the
¨
¨
ꢃc
This journal is the Owner Societies 2008
5218 | Phys. Chem. Chem. Phys., 2008, 10, 5218–5224

View Article Online
Scheme 1 Suggested reaction pathways of TBFC photodecomposition.
photodecomposition of TBFC as monitored by femtosecond
mid-infrared pulses were generated by difference frequency
time-resolved spectroscopy in the visible to near IR spectral
region.20 One should note that also in these experiments TBFC
decomposition was found to occur on a timescale of a few
picoseconds.
mixing of idler and signal from an optical parametric ampli-
fier23,24 pumped by 250 mJ pulse energy of the regenerative
amplifier output. The IR light was split in a probe and a
reference beam and focused into the sample cell. At the sample
position the probe pulse was superimposed on the pump beam.
To avoid signal contributions arising from rotational relaxation
of molecules the relative plane of polarization of both pulses was
adjusted to 54.71. After passing the cell both IR beams were
spectrally dispersed in a grating polychromator and indepen-
dently imaged on a HgCdTe detector (2 ꢁ 32 elements). The
whole pump–probe setup was purged with dry nitrogen to avoid
spectral and temporal distortion of the IR pulses by CO2 and
water absorptions in air. All experiments were performed in a
stainless steel flow cell equipped with 1 mm thick CaF2 windows
(the path length inside the cell was 0.6 mm).
TBFC and carbonyl-13C-labeled TBFC were prepared from
a reaction of the corresponding acid chlorides with sodium
tert-butyl peroxide according to ref. 25. The synthesis of these
acid chlorides started from methylation of fluorene to
9-methyl-9H-fluorene26 and subsequent carboxylation to 9-
methyl-9H-fluorene-9-carboxylic acid by treatment of
9-methyl-9H-fluorene with buthyllithium in THF at ꢂ78 1C
followed by discharging gaseous CO2 into the solution.27 For
the synthesis of 13C-TBFC the latter reaction step was carried
out with 13CO2. The acids were transformed into acid chlor-
ides by means of PCl5.
The validity of interpreting this ultrafast decomposition rate
in terms of a sequential mechanism, however, rests upon the
correct assignment of transient spectra in the visible and near
IR. Radical absorption spectra and excited state absorption
are known for a very few cases only, and spectral overlap as
well as time-dependent spectral shifts and band shapes com-
plicate the analysis. As we have shown recently, such compli-
cations can be largely avoided by using broadband IR probe
pulses, in particular if one focuses on the spectral region in the
vicinity of the CO-group stretch band.21,22 There we studied
the photodissociation mechanism of peroxycarbonates TBNC
and TBPC by monitoring the temporal evolution of vibra-
tional marker bands. In both peroxides UV excitation popu-
lates the S1 state of the parent molecule state which
subsequently decays at the same rate as the final products
CO2, ROꢀ and tBOꢀ appear. A sequential dissociation
mechanism could be ruled out unequivocally. Also the some-
what slower fragmentation of TBNC (25–50 ps) is a conse-
quence of the correspondingly longer S1 lifetime.21 A similar
study on dibenzoylperoxide revealed the same initial pattern of
the photofragmentation mechanism.22
In the present paper we report results of applying the same
technique to investigate the photodecomposition of TBFC
with the goal of finally settling the issue about its photofrag-
mentation pathway. For that purpose we also study the
appropriately 13C-labelled compound to verify the mechanism
and the assignment of transient bands. This seems of parti-
cular relevance, as this reaction is often quoted as a kind of
benchmark when referring to organic peroxide decarboxyl-
ation mechanisms and rates.
The measured IR spectra were compared with quantum
chemical calculations of TBFC and the corresponding radi-
cals. We applied density functional theory (DFT) using
the ORCA program package28,29 developed by Neese and
coworkers. To enhance the efficiency of the calculations the
resolution-of-identity (RI) approximation was invoked.30
Since this procedure requires a non-hybrid functional, we have
chosen the Becke–Perdew functional, BP86.31,32 The Ahlrichs’
triple-z valence basis set,33 TZVPP, with three sets of polar-
ization functions on all atoms was applied, to which diffuse
functions from Dunning’s correlation-consistent polarized
triple-z basis set could be added (hence, aug-TZVPP).34 In
addition, the RI-approximation requires an auxiliary
Coulomb fitting basis (TZV/J) for the triple-z valence basis
set, which was taken from the TURBOMOLE library.35
Following the recommendations by Neugebauer and Hess36
for calculations at the RI-DFT/TZVPP level of theory, a
correction factor of 1.004 for conversion from harmonic to
fundamental vibrational frequencies was used.
Experimental technique
The femtosecond infrared experiments were carried out with a
laser system based on a 1 kHz Ti : sapphire oscillator/regenera-
tive amplifier system (wavelength 800 nm, pulse width 100 fs,
pulse energy 700 mJ). Pump pulses at 267 nm were produced by
frequency tripling part of the 800 nm light. The UV light was
attenuated to energies of 0.5–2 mJ and used to excite 0.5–2 mM
solutions of TBFC dissolved in CD3CN (Deutero GmbH,
99.6% deuteration grade), CCl4, or n-heptane. Tunable
ꢃc
This journal is the Owner Societies 2008
Phys. Chem. Chem. Phys., 2008, 10, 5218–5224 | 5219

View Article Online
Results and discussion
Fig. 1a shows linear absorption spectra of TBFC and 13C-
TBFC in CD3CN. The spectrum of TBFC is characterized by
a strong carbonyl stretching band at 1766 cmꢂ1 which upon
labelling red-shifts to 1723 cmꢂ1. Absorption bands in
the fingerprint region correspond to vibrations of the
fluorenyl and tert-butyl rest and are hence almost unaffected
by 13C-labeling of the carbonyl group.
In Fig. 1b transient difference spectra recorded 1.3 ns after
UV excitation of TBFC and 13C-TBFC in CD3CN, respec-
tively, are presented for the frequency range 1300–1820 cmꢂ1
.
Because of the limited bandwidth of the IR probe pulse they
were constructed from several overlapping spectra of
B100 cmꢂ1 width. The estimated error of the relative ampli-
tudes of spectral features arising from this procedure is
estimated to be o20%. At 2220–2400 cmꢂ1 where the solvent
CD3CN is opaque separate transient spectra of TBFC and
13C-TBFC in n-heptane are shown. This part of the spectrum
is not to scale. The transient spectra show negative peaks
caused by bleaching of the peroxide and positive absorptions
due to the formation of photoproducts. Most prominent is the
bleaching of the CO stretching bands but also vibrations at
1369 and 1453 cmꢂ1 clearly show the disappearance of TBFC
and 13C-TBFC, respectively, as indicated by full circles. The
strongest increase of absorption arises from the asymmetric
stretching vibration of newly formed CO2 at 2337 cmꢂ1 and
13CO2 at 2273 cmꢂ1, respectively (open circle in Fig. 1b). In the
Fig. 2 Calculated IR spectra of (a) TBFC, (b) MFꢀ, (c) tBOꢀ, and (d)
MFCO2ꢀ; computed frequencies and intensities (bars, right scale) were
convoluted with a uniform Lorentzian line shape of 8 cmꢂ1 width (FWHM).
Fig. 1 (a) Linear absorption spectra of TBFC (black) and 13C-TBFC (red) in CD3CN; (b) transient difference spectra 1.3 ns after 267 nm
excitation of TBFC (black) and 13C-TBFC (red) in CD3CN (at 2220–2380 cmꢂ1 spectra were recorded in n-heptane, because CD3CN is opaque in
this range); symbols indicate bleaching of TBFC (filled circles) and formation of the products CO2 (open circles) and methylfluorenyl radical MFꢀ
(stars); the insert shows the calculated difference spectrum of e(MFꢀ) þ e(tBOꢀ) ꢂ e(TBFC) (see Fig. 2).
ꢃc
This journal is the Owner Societies 2008
5220 | Phys. Chem. Chem. Phys., 2008, 10, 5218–5224

View Article Online
Fig. 3 CO2 quantum yield of TBFC photodissociation in n-heptane at 1.3 ns. Solid lines represent extinction coefficients (left scale) of TBFC CO-
stretch (black) and of CO2 asymmetric stretch (red). The transient difference spectrum at 1.3 ns pump–probe delay (right scale, filled circles) is
normalized to the CO-stretch extinction coefficient. The relative amplitude of the absorption bands suggests a CO2 quantum yield of
FCO ¼ 0:95 ꢄ 0:10.
2
fingerprint region bands at 1434 cmꢂ1 and, additionally,
weaker absorptions at 1544 and 1559 cmꢂ1 for both, TBFC
and 13C-TBFC appear as marked by asterisks. As no isotope
shift can be detected we attribute these bands to the MFꢀ
radical.
residue of MFCO2 . These findings preclude a sequential
ꢀ
ꢀ
dissociation mechanism leading to vibrationally hot MFCO2
radicals from which an appreciable amount cools down to
thermal equilibrium and lives for many nanoseconds.20
The photofragmentation mechanism of TBFC may be
deduced from the time evolution of transient spectra in the
picosecond time range. The contour diagram of Fig. 4 shows
that immediately upon UV excitation, CO stretch bands in the
electronic ground and first excited state of TBFC appear as a
superposition of bleach (blue, centered at 1767 cmꢂ1) and a
red-shifted absorption (red), respectively. With increasing
pump–probe delay the excited state population decays and
only the ground state bleach remains. An exponential fit to the
integrated band intensity (upper panel in Fig. 4) gives a S1
lifetime of tS1 ¼ (0.8 ꢄ 0.2) ps. The S1 spectrum was
reconstructed by subtracting the ground state bleach (i.e. the
spectrum at 8 ps) from the t ¼ 0.38 ps spectrum and then
extrapolating the amplitude to t ¼ 0. The result is presented in
the right panel of Fig. 4 and demonstrates that the CO-stretch
band in the S1 state peaks at 1760 cmꢂ1 and is slightly broader
than the bleach component at 8 ps resembling the ground state
spectrum.
This assignment is supported by our DFT calculations
presented in Fig. 2. Here the IR intensities (stick spectra, right
ordinate) of TBFC (a) and all the radicals to be considered
(b–d) are plotted. For a better comparison with the experiment
smooth spectra were generated (full lines, left ordinate) assum-
ing a uniform Lorentzian line shape of 8 cmꢂ1 width
(FWHM). The calculated spectrum of TBFC appears to be
in fairly good agreement with the experimental one of Fig. 1a.
Note that in the range 1280–1640 cmꢂ1 the absorptions oꢀf
TBFC, MFꢀ, and tBOꢀ are relatively weak, whereas MFCO2
shows two strong bands at 1483 and 1592 cmꢂ1. These bands
do not appear in the transient spectrum (Fig. 1b). Instead, the
absorptions marked by asterisks in Fig. 1b agree well (both, in
frequency and intensity) with the calculated spectrum of MFꢀ.
Assuming that the photodecomposition of TBFC is complete
after 1.3 ns the difference spectrum can be calculated from the
sum of spectra of MFꢀ þ tBOꢀ ꢂ TBFC from Fig. 2 (neglect-
ing CO2 which does not absorb in this spectral range). The
result is shown in the insert of Fig. 1b and is in very good
agreement with the measured difference spectrum.
In CCl4 and n-heptane the S1 lifetime of TBFC is found to
be (1.5 ꢄ 0.3) and (2.1 ꢄ 0.3) ps, respectively. Such a
pronounced solvent effect on tS1 was also observed for the
photofragmentation of peroxycarbonates and discussed in
terms of a possible polarity effect on the barrier height of
the dissociation coordinate.21 As for the peroxycarbonates,
however, it is not known for TBFC either whether the
fragmentation is actually a barrier-controlled process. Tem-
perature dependent measurements may help to clarify this
point. On the basis of Fig. 3 we have concluded that each
photo-bleached TBFC molecule generates one CO2. The short
S1 lifetime of TBFC suggests that other deactivation processes
repopulating the ground state (fluorescence, internal
The conversion efficiency of TBFC to the final products can
also be determined by measuring the relative CO2 quantum
yield. To this end in Fig. 3 the absorption coefficients of TBFC
(at the CO-stretching band, black line) and CO2 (n3 band, red
line) in n-heptane are compared with the scaled transient
difference spectrum (filled circles) at 1.3 ns pump–probe delay.
Fig. 3 clearly shows that the ratio of transient TBFC bleach
and n3 absorption matches the ratio of the corresponding
extinction coefficients, suggesting that every decomposed
TBFC generates just one CO2 molecule and no significant
ꢃc
This journal is the Owner Societies 2008
Phys. Chem. Chem. Phys., 2008, 10, 5218–5224 | 5221

View Article Online
Fig. 4 Spectral evolution of the CO stretch absorption band of TBFC in CD3CN. At early times the spectrum consists of a superposition of
ground state bleach (blue) and red shifted S1 absorption (red). The reconstructed spectra are shown in the right panel. Upper panel: integrated
band intensity (red curve: exponential fit to the data with time constant of 0.8 ꢄ 0.2 ps).).
conversion) cannot compete with fragmentation. Conse-
quently, the quantum yield of direct CO2 formation from
the S1 state has to be close to unity.
demonstrates a very fast production of the radical within the S1
lifetime of TBFC. Subsequently no further increase of the MFꢀ
concentration can be observed up to 1.3 ns after excitation.
A similar behaviour is found for the production of CO2 not
shown here. Again the fragment is generated within the S1-lifetime
of the parent molecule. In accordance with earlier IR studies on
the decarboxylation of TBFC14 and other peroxides15,16,21,22 the
CO2 molecule is generated with high vibrational excess energy. A
quantitative analysis of the transient IR spectra on the basis of a
model of anharmonically coupled oscillators described pre-
viously16,21,37 reveals an initial CO2 vibrational temperature of
T0 ¼ (2500 ꢄ 200) K.
Fig. 5 presents formation of the photoproduct MFꢀ at a
probe frequency of around 1550 cmꢂ1 after UV excitation of
TBFC in CD3CN. The contour diagram shows that until a
pump–probe delay of 20 ps the absorption is broad and red-
shifted with respect to the final spectrum which is characterized
by a double peak. This behaviour is a clear signature of a hot
species relaxing by vibrational energy transfer to the surround-
ing solvent. The rate of MFꢀ formation is deduced from the
band integral as plotted in the upper panel of Fig. 5 and
Fig. 5 Spectral evolution in the 1515–1585 cmꢂ1 range after photodissociation of TBFC in CD3CN; right panel: transient spectrum at 100 ps
pump–probe delay; upper panel: integrated band intensity.
ꢃc
This journal is the Owner Societies 2008
5222 | Phys. Chem. Chem. Phys., 2008, 10, 5218–5224

View Article Online
here show unequivocally that it cannot be considered as evidence
ꢀ
for an intermediate MFCO2 radical.
Acknowledgements
The authors are grateful to Jens Schimpfhauser and Jurgen
¨
Bienert for the preparation and purification of TBFC and its
13C-labeled analogue. We also thank Jurgen Troe for contin-
¨
uous support of this work.
References
1 K. Fujimori, in Organic Peroxides, ed. W. Ando, Wiley,
New York, 1992, p. 319.
2 Y. Sawaki, in Organic Peroxides, ed. W. Ando, Wiley, New York,
1992, p. 425.
3 G. G. Hammond and L. M. Soffer, J. Am. Chem. Soc., 1950, 72.
4 P. D. Bartlett and R. R. Hiatt, J. Am. Chem. Soc., 1958, 80,
1398–1405.
5 D. E. Falvey and G. B. Schuster, J. Am. Chem. Soc., 1986, 108,
7419–7420.
6 H. Misawa, K. Sawabe, S. Takahara, H. Sakuragi and K.
Tokumaru, Chem. Lett., 1988, 357–360.
7 J. Chateauneuf, J. Lusztyk and K. U. Ingold, J. Am. Chem. Soc.,
1988, 110, 2877–2885.
8 J. Chateauneuf, J. Lusztyk and K. U. Ingold, J. Am. Chem. Soc.,
1988, 110, 2886–2893.
9 J. Wang, T. Tateno, H. Sakuragi and K. Tokumaru, J. Photochem.
Photobiol., A, 1995, 92, 53–59.
10 J. Wang, H. Itoh, M. Tsuchiya, K. Tokumaru and H. Sakuragi,
Tetrahedron, 1995, 51, 11967–11978.
11 T. Najiwara, J. Hashimoto, K. Segawa and H. Sakuragi, Bull.
Chem. Soc. Jpn., 2003, 76, 575–585.
Fig. 6 Spectral evolution of the main transient absorption band in the
visible after UV-excitation of TBFC in propylene carbonate solution.20
The bottom graph shows the decay of the band integral over the
wavelength range shown in the contour plot. tcc denotes the pump–p-
robe cross correlation width fitted to the observed absorption rise.
12 D. E. Polyansky, E. O. Danilov, S. V. Voskresensky, M. A. J.
Rodgers and D. C. Neckers, J. Phys. Chem. A, 2006, 110,
4969–4978.
13 D. E. Polyansky and D. C. Neckers, J. Phys. Chem. A, 2005, 109,
2793–2800.
Summary and conclusion
Our results for the TBFC photo-fragmentation can be sum-
marized as follows:
14 J. Aschenbrucker, U. Steegmuller, M. Buback, N. P. Ernsting and
¨
¨
J. Schroeder, J. Phys. Chem. B, 1998, 102, 5552–5555.
15 J. Aschenbrucker, U. Steegmuller, M. Buback, N. P. Ernsting and
J. Schroeder, Ber. Bunsen-Ges. Phys. Chem., 1998, 102, 965.
¨
¨
(1) The S1-lifetime of the photo-excited TBFC molecule is
solvent dependent and amounts to tS1 ¼ (0.8 ꢄ 0.2), (1.5 ꢄ
0.3), and (2.1 ꢄ 0.3) ps in CD3CN, CCl4, and n-heptane,
respectively.
16 M. Buback, M. Kling, M. T. Seidel, F. D. Schott, J. Schroeder and
U. Steegmuller, Z. Phys. Chem., 2001, 215, 717–736.
¨
17 B. Abel, J. Assmann, P. Botschwina, M. Buback, M. Kling, R.
Oswald, S. Schmatz, J. Schroeder and T. Witte, J. Phys. Chem. A,
2003, 107, 5157–5167.
18 B. Abel, J. Assmann, M. Buback, C. Grimm, S. Schmatz, J.
Schroeder and T. Witte, J. Phys. Chem. A, 2003, 107, 9499–9510.
19 M. Buback, M. Kling, S. Schmatz and J. Schroeder, Phys. Chem.
Chem. Phys., 2004, 6, 5441–5455.
20 B. Abel, M. Buback, M. Kling, S. Schmatz and J. Schroeder, J.
Am. Chem. Soc., 2003, 125, 13274–13278.
21 C. Reichardt, J. Schroeder and D. Schwarzer, J. Phys. Chem. A,
2007, 111, 10111–10118.
(2) CO2 and the radical MFꢀ are formed with the same rate
as the electronically excited TBFC molecule decays. Both
fragments are generated with excess vibrational energy. For
CO2 this energy corresponds to an initial vibrational tempera-
ture of T0 ¼ (2500 ꢄ 200) K.
(3) The quantum yield of CO2 is FCO ¼ 0:95 ꢄ 0:10.
2
In contrast to the original implication by Falvey and Schuster,
a sequential bond breakage mecꢀhanism with a spectroscopically
accessible intermediate MFCO2 fragment is clearly not opera-
tive in this reaction. Therefore, in terms of photochemical
kinetics the mechanism may be considered as concerted. The
question whether both bonds break synchronously cannot be
answered on the basis of our experiments. The product rise time
constant of about 25 ps reported by Falvey and Schuster, which
at that time was at the limit of the experimental time resolution,
could not be found, as already noted in previous studies.14,20 The
main part of the transient visible absorption band centred
around 650 nm decays on a time scale of about 0.75 ps in
propylene carbonate solution as illustrated in Fig. 6. Therefore,
in our view the band has to be reassigned to excited state
absorption from the S1 state of TBFC. The results reported
22 C. Reichardt, J. Schroeder, P. Vohringer and D. Schwarzer, Phys.
¨
Chem. Chem. Phys., 2008, 10, 1662–1668.
23 R. A. Kaindl, M. Wurm, K. Reimann, P. Hamm, A. M. Weiner
and M. Woerner, J. Opt. Soc. Am. B, 2000, 17, 2086–2094.
24 P. Hamm, R. A. Kaindl and J. Stenger, Opt. Lett., 2000, 25,
1798–1800.
25 J. P. Lorand and P. D. Bartlett, J. Am. Chem. Soc., 1966, 88,
3294–3302.
26 D. Zhang, L. Chen, J. Chen, Y. Liang and H. Zhou, Synth.
Commun., 2005, 35, 2609–2613.
27 C. Garms and W. Francke, in Treatment of contaminated soils,
ed. R. Stegmann, G. Brunner, W. Calmano and G. Matz, Springer,
Berlin, 2001, pp. 95–131.
28 F. Neese, University of Bonn, 2007.
29 F. Neese, J. Am. Chem. Soc., 2006, 128, 10213.
30 F. Neese, J. Comput. Chem., 2003, 24, 1740.
31 A. D. Becke, Phys. Rev. A: At., Mol., Opt. Phys., 1988, 38, 3098.
ꢃc
This journal is the Owner Societies 2008
Phys. Chem. Chem. Phys., 2008, 10, 5218–5224 | 5223

View Article Online
32 J. P. Perdew, Phys. Rev. B: Condens. Matter Mater. Phys., 1986,
33, 8822.
35 K. Eichkorn, O. Treutler, H. Ohm, M. Haser and R. Ahlrichs,
Chem. Phys. Lett., 1995, 240, 283.
33 A. Schaefer, H. Horn and R. Ahlrichs, J. Chem. Phys., 1992, 97,
2571.
36 J. Neugebauer and B. A. Hess, J. Chem. Phys., 2003, 118
7215.
34 R. A. Kendall, T. H. Dunning, Jr and R. J. Harrison, J. Chem.
Phys., 1992, 96, 6769.
37 P. Hamm, S. M. Ohline and W. Zinth, J. Chem. Phys., 1997, 106,
519–529.
ꢃc
This journal is the Owner Societies 2008
5224 | Phys. Chem. Chem. Phys., 2008, 10, 5218–5224