TR-FTIR absorption spectroscopy of transition metal carbonyl radicals generated by photodissociation of metal-metal bonds, by halogen abstraction or by radical ligand substitution

Article abstract of DOI:10.1016/j.jorganchem.2005.06.015

Three methods of obtaining time-resolved Fourier-Transform infrared (TR-FTIR) absorption spectra of transition metal carbonyl radicals in hexane are reported here. For the first method, CpM(CO)₂L and Cp*M(CO)₂L (M = Mo, W; L = CO, PR

Full text of DOI:10.1016/j.jorganchem.2005.06.015

Journal of Organometallic Chemistry 690 (2005) 4132–4138
www.elsevier.com/locate/jorganchem
TR-FTIR absorption spectroscopy of transition metal
carbonyl radicals generated by photodissociation of metal–metal
bonds, by halogen abstraction or by radical ligand substitution
*
Thiam Seong Chong, Peng Li, Weng Kee Leong, Wai Yip Fan
Department of Chemistry, National University of Singapore, 3 Science Drive 3, Singapore 117543, Singapore
Received 30 April 2005; received in revised form 6 June 2005; accepted 16 June 2005
Abstract
Three methods of obtaining time-resolved Fourier-Transform infrared (TR-FTIR) absorption spectra of transition metal car-
bonyl radicals in hexane are reported here. For the first method, CpM(CO)2L and Cp*M(CO)2L (M = Mo, W; L = CO, PR3) rad-
icals have been generated by photodissociation of the corresponding metal–metal bonded dimers. Radicals of formula M(CO)4L
(M = Mn, Re; L = CO, PR3, AsPh3, SbPh3) and CpM(CO)n (M = Fe, Mo; n = 2, 3) have been produced via the second method
which is halogen abstraction of the transition metal carbonyl halides using CpMo(CO)3 radical. For the third method, fast radical
ligand substitution kinetics has been exploited to generate CpMo(CO)2PR3 radicals from CpMo(CO)3 in the presence of free phos-
phines. An assessment of the three methods with respect to TR-FTIR spectroscopic detection of radicals was also discussed.
Ó 2005 Elsevier B.V. All rights reserved.
Keywords: TR-FTIR radical spectra; Metal–metal bond dissociation; Halide abstraction; Ligand substitution kinetics
1. Introduction
tion rates of key radical intermediates directly impli-
cated in transition metal carbonyl-catalysed reactions
[15–17].
The direct observation of 17-electron transition metal
carbonyl radicals and other reactive intermediates using
time-resolved infrared spectroscopy (TRIR) has contin-
ued to receive much attention [1–6]. This technique has
often been applied to the detection of CpM(CO)nL
[M = Fe, Mo, W; n = 2, 3; L = CO, PR3 (for Fe only)]
and M(CO)5 (M = Mn, Re) radicals produced via pho-
tolysis of metal–metal bonded dimers [7–13]. The advent
of time-resolved Fourier-Transform infrared (TR-
FTIR) absorption technique has further widened the
scope for simultaneous detection of radical spectra
across the mid-IR region [14]. TR-FTIR techniques
could be invaluable for determining lifetimes and reac-
In this work, we focus on obtaining TR-FTIR spec-
troscopic data of transition metal carbonyl radicals by
utilising three different methods of radical production.
Firstly, photodissociation of metal-bonded dimers has
continued to be a convenient way of acquiring spectra
of radicals such as CpM(CO)2PR3 or Cp*M(CO)3
[M = Mo, W]. Secondly, radicals generated via halogen
abstraction of transition metal carbonyl halides by
CpMo(CO)3 have turned out to be suitable candidates
for TR-FTIR studies. This method is especially useful
for studying M(CO)4(EPh3) (M = Mn, Re; E = P, As,
Sb) radicals without the concomitant presence of
decarbonylation products. Thirdly, we have made use
of fast radical ligand substitution kinetics to generate
phosphine derivatives of CpMo(CO)3 itself. For the sec-
ond and third methods, CpMo(CO)3 has to be generated
*
Corresponding author. Tel.: +65 68746823/96385110; fax. +65
67791691.
E-mail address: chmfanwy@nus.edu.sg (W.Y. Fan).
0022-328X/$ - see front matter Ó 2005 Elsevier B.V. All rights reserved.
doi:10.1016/j.jorganchem.2005.06.015

T.S. Chong et al. / Journal of Organometallic Chemistry 690 (2005) 4132–4138
4133
in situ first by visible laser photolysis. The suitability of
all three methods towards detection of certain classes of
transition metal carbonyl radicals will be discussed.
Although a wealth of kinetic data could be extracted
from TR-FTIR studies, we have left the detail analysis
of the reaction rates and lifetimes of radicals to a forth-
coming paper. Only brief observations pertaining to
radical kinetics will be mentioned here.
each side of the cell allowed for N2 bubbling through
the solution if required.
The sample preparations for all three methods were
carried out in the dark under nitrogen atmosphere. The
precursors
Cp2Mo2(CO)5PR3,
[CpM(CO)2PR3]2,
[Cp*M(CO)2PR3]2 (M = Mo or W; R = Ph, OEt, Me)
and M(CO)4EPPh3Br (M = Mn, Re; E = P, As, Sb)
compounds were prepared according to literature meth-
ods [18–22]. All other organometallic compounds ob-
tained from Strem and Aldrich were used directly
without purification. For method 1, the metal–metal
bonded precursor was dissolved in hexane and trans-
ferred into the air-tight IR cell with the optical path-
length set at 2 mm. Usually about 10–20 mg of the
precursor in 30 ml solvent (10ꢀ2–10ꢀ3 M) was sufficient
to generate free radical spectra. For the second method,
a quantity of [CpMo(CO)3]2 was added into the hexane
solution already containing the transition metal carbonyl
halides. Again the same concentrations of precursors of
10ꢀ2–10ꢀ3 M have been used for this method. However
since the reactions are irreversible, the precursor concen-
tration could be depleted very quickly depending on the
laser pulse energy. Hence if the concentrations of radicals
had to be maintained for longer observations, saturated
2. Experiment
A brief description of the cell used for performing
time-resolved and linear FTIR spectroscopy on the reac-
tion mixture is given here (see Fig. 1). It is a stainless
steel, static, liquid cell with CaF2 windows for passage
of the IR probe beam. Absorbance pathlengths of 0.2–
5 mm could be accommodated by manual adjustment
of the window holders. The cell also allowed for place-
ment of quartz or glass windows to permit photodisso-
ciation by a laser beam propagated at right angle to
the IR probe beam. The cell could hold a volume of
about 40 ml of solvent, together with a magnetic bar
to provide constant stirring. Inlet and outlet ports on
Fig. 1. (a) Schematic setup of the TRS experiment and (b) extended schematic diagram of our home-made IR cell.

4134
T.S. Chong et al. / Journal of Organometallic Chemistry 690 (2005) 4132–4138
solutions with additional solid precursor remaining in
the reaction cell could be prepared. This would ensure
continuous observation of radical signals until all the sol-
ids have dissolved and reacted. The experiments could
also be performed using different mol ratio of [CpMo-
(CO)3]2 to halide precursors. Usually, a ratio of 1:1 was
used for spectroscopic measurements although compara-
ble radical signals have also been produced using ratios
from 1:3 to 3:1. On the other hand, free phosphines
(PPh3, P(OEt)3, P(C6F5)3) were added into a solution
containing [CpMo(CO)3]2 for carrying out experiments
according to method 3. About 10–25 mg of the metal
precursors and phosphines in 30 ml of hexane solvent
were used for method 3, similar to method 2.
copy of the previously detected CpMo(CO)3 radical
was carried out so that the sensitivity of the experimen-
tal setup could be tested while its chemical lifetime could
also be employed as references for comparison with new
radical species [10]. Unless stated otherwise, the lifetime
regardless of the kinetic order of the decay, is simply de-
fined as the time taken for the absorption intensity of the
radical to decrease to half of the initial value. Having
successfully identified the vibrational bands of CpMo-
(CO)3 (1914, 2010 cmꢀ1), the TR-FTIR spectra of other
radicals were then recorded. The second-order decay
lifetime of CpMo(CO)3 radical of 4.6 ls is also in very
good agreement with previous studies in which radi-
cal–radical recombination was postulated to be its main
loss process in n-heptane [10].
A frequency-doubled Q-switched Nd-YAG pulsed-
laser (Continuum Surelite III-10, 532 nm, 10 Hz,
25–30 mJ/pulse, 10 ns) was focused elliptically by a
cylindrical lens into the region between the CaF2 win-
dows in the cell for maximum overlap with the IR beam.
The generation of radical signals could be optimized at
higher laser energies (>30 mJ/pulse) although at such
energies, the quartz windows through which the laser
passed, would be damaged more quickly. For the collec-
tion of TR-FTIR spectra, an AC-coupled detector
(TRS-20 MHz MCT detector) was used to capture
changes in the IR absorption down to 1 ls while the syn-
chronization of the spectrometer with the firing of the
pulsed laser was controlled by a Digital Delay Genera-
tor (Stanford SRS DG 535). Radical signals could be de-
tected most easily at short time intervals below 1 ls, i.e.,
before their decay fully set in after the nanosecond laser
pulse has stopped firing. Although the spectrometer has
the capability of recording spectra at 20 ns time interval,
it was not a necessity to set the time interval to below
1 ls for spectroscopic measurements since most radicals
observed here have lifetimes of a few microseconds.
However for more precise kinetic studies, this feature
would be essential. The detection range and resolution
of a step-scan Nicolet Nexus 870 FTIR spectrometer
were set at 1000–4000 and 4 cmꢀ1 respectively. We have
found that only one scan of the spectrum which lasted
about 30 min was sufficient to produce detectable signals
of the radicals [14]. We have found that averaging radi-
cal signals over a few scans only improved them slightly
but at the expense of time. Linear scan FTIR measure-
ments over the same wavelength and resolution were
also carried out to record the evolution of stable species
in the solution throughout the photolysis.
Fig. 2(a) shows an example of a TR-FTIR spectrum
recorded about 10 ls after 532 nm laser photolysis of
the mono-phosphine complex, Cp2Mo2(CO)5PPh3.
Apart from the bands belonging to CpMo(CO)3, two
other bands at 1834 and 1924 cmꢀ1 have been detected.
These bands have been assigned to CpMo(CO)2PPh3
based on several observations. Metal–metal bond pho-
tolysis of Cp2Mo2 (CO)5PPh3 should result in the simul-
taneous generation of these two radicals. The
vibrational band positions are consistent with a shift
to lower wavenumbers upon phosphine substitution of
CpMo(CO)3. The measured lifetime of the phosphine-
containing radical (5.6 ls) was close to that of CpMo-
(CO)3, indicating similarity in reactivity. The mCO
frequencies for structurally similar stable species such
as CpMo(CO)2PPh3C2H5 (1843, 1930 cmꢀ1) showed
very good agreement with those belonging to the radical
[23–26]. In fact, these analyses have been applied suc-
cessfully to the identification of all the TR-FTIR radi-
cals in Table 1 regardless of the method of generation.
A linear FTIR scan has been recorded of the reaction
mixture in order to observe the formation of long-lived
Table 1
TR-FTIR spectral data of transition metal carbonyl radicals detected
using one of the three methods
Method
Radicals
IR bands (cmꢀ1
)
1–3
1, 3
1, 3
1, 3
1, 2
1
1
1
1
1
CpMo(CO)3
2010, 1912
1924, 1834
1936, 1853
1921, 1838
2004, 1938
2000, 1900
1915, 1827
1990, 1898
1961, 1840
1984, 1887
1996, 1977
1986, 1970
1975, 1914
1977, 1924
1971, 1913
CpMo(CO)2PPh3
CpMo(CO)2P(OEt)3
CpMo(CO)2PMe3
CpFe(CO)2
CpW(CO)3
CpW(CO)2PPh3
Cp*Mo(CO)3
Cp*Mo(CO)2PPh3
Cp*W(CO)3
Re(CO)5
Mn(CO)5
Mn(CO)4PPh3
Mn(CO)4AsPh3
Mn(CO)4SbPh3
3. Results and discussion
2
2
2
2
3.1. Photodissociation of metal–metal bonded dimers
The mCO infrared bands of the detected radicals are
presented in Table 1. Initially, the TR-FTIR spectros-
2

T.S. Chong et al. / Journal of Organometallic Chemistry 690 (2005) 4132–4138
4135
Fig. 2. TR-FTIR spectra of CpMo(CO)3 (2010, 1914 cmꢀ1) and CpMo(CO)2PPh3 (1924, 1834 cmꢀ1) upon (a) Cp2Mo2(CO)5PPh3 metal–metal bond
photolysis or (b) PPh3 ligand substitution in the presence of [CpMo(CO)3]2. The band pointing up at 1962 cmꢀ1 is due to [CpMo(CO)3]2.
metal carbonyl species during photolysis. For example,
with Cp2Mo2(CO)5PR3 as the precursor, the linear scan
revealed increased concentrations of [CpMo(CO)3]2 and
[CpMo(CO)2 PR3]2 with concomitant decrease of precur-
sor concentration. The formation of these dimers was due
to radical–radical self-recombination processes of
CpMo(CO)3 and CpMo(CO)2PPh3. However it also im-
plied that contributions to the CpMo(CO)3 and CpMo-
(CO)2PR3 radical signals came from both precursor
and the dimer products especially during the later stages
of photolysis. However, when a symmetrical precursor
such as [CpMo(CO)2(PR3)]2 was used, the IR spectra ap-
peared unchanged even after long hours of photolysis,
indicating once again the dominance of radical–radical
recombination pathway to regenerate the precursor.
The same trend could be applied to the phosphine-deriv-
atives of CpW(CO)3 and to the Cp*-containing radicals.
Preliminary studies have shown that slightly longer
lifetimes of the phosphine containing radicals (5 to
7 ls) were determined compared to that of CpMo(CO)3.
Steric and electronic effects due to the coordinated PR3
on the recombination rate of these radicals appear not
to be very significant. These results suggest that the ori-
entation of the phosphine groups in the radicals relative
to the radical recombination trajectory is most likely
similar to the alkylated counterparts or the metal dimers
themselves, i.e., these PR3 groups do not restrict or
block the approach of another radical species for recom-
bination. Similar observations could also be ascribed to
the Cp*-containing radicals. Since a small activation
barrier for the recombination process is implied by these
measurements, diffusion-limited kinetics could also have
played a major role in determining the lifetimes of the
radicals.
3.2. Halogen abstraction by CpMo(CO)3 radicals
In this method the CpMo(CO)3 radical generated
upon 532 nm photolysis of [CpMo(CO)3]2 has been used
to abstract the halogen atom from the transition metal
carbonyl halide precursors to form MLn radicals.
CpMoðCOÞ3 þ MLnX ! MLn þ CpMoðCOÞ3X
ðX ¼ halogen; L ¼ other ligandsÞ
The infrared mCO bands and the lifetimes of the orga-
nometallic radicals detected via halogen abstraction are
also shown in Table 1. For example, the TR-FTIR spec-
trum recorded about 1 ls after 532 nm laser photolysis
of a mixture containing Mn(CO)5Br and [CpMo(CO)3]2
showed vibrational bands due to Mn(CO)5, CpMo(CO)3

4136
T.S. Chong et al. / Journal of Organometallic Chemistry 690 (2005) 4132–4138
and gauche-[CpMo(CO)3]2 (Fig. 3). Bands belonging to
the CpMo(CO)3 dimer and the mixed-metal complex
CpMo(CO)3–Mn(CO)5 were also observed but pointed
to the opposite direction of the spectrum which indi-
cated their formation upon radical–radical recombina-
tion processes.
ter to a first-order rather than a second-order decay. In
the absence of other species, the main loss process for
Mn(CO)5 is its self-recombination to form Mn2(CO)10.
The presence of CpMo(CO)3 species obviously renders
the kinetic data to be more difficult to interpret since
it could also combine efficiently with Mn(CO)5. Indeed
the mixed metal complex, CpMo(CO)3–Mn(CO)5 ob-
served in the system most likely came from the cross-
radical recombination reaction.
We have found that reactions between CpMo(CO)3
and many transition metal carbonyl halides were not
reversible although this condition is essential for acquir-
ing TR-FTIR spectra in a static cell. Choosing the
CpMo(CO)3/Mn(CO)5Br (1:1 mole ratio) system as an
example, the linear scan FTIR spectroscopy recorded
of the reaction mixture two hours after photoinitiation
only indicated the presence of Mn2(CO)10, CpMo(CO)3
Br and a lesser amount of the mixed-metal complex,
CpMo(CO)3–Mn(CO)5; all of which are not susceptible
to 532 nm photolysis (see Fig. 4(b)) [33–35]. For such
cases, TR-FTIR signals of Mn(CO)5 and CpMo(CO)3
species should also gradually decrease as the photolysis
progresses towards completion. Indeed TR-FTIR spec-
tra taken at later stages of photolysis showed much
lower signals of both radicals, consistent with the linear
scan measurements.
The identity of the newly formed radicals could be
made by examining their chemical lifetimes, effect of li-
gand substitution on the vibrational frequencies and
comparisons to structurally similar 18-electron alkyl
counterparts [27–32]. In the CpMo(CO)3/Mn(CO)5Br
system for example, a band at 1986 cmꢀ1 has been ob-
served close to the literature values for Mn(CO)5
(m = 1988 cmꢀ1
,
cyclohexane; 1990 cmꢀ1
,
heptane)
[27,28]. The band did not show up if either Mn(CO)5Br
or CpMo(CO)3 dimer was removed. Its short lifetime of
about 8.1 ls also suggested a radical species as the carrier
and thus this band was assigned to Mn(CO)5. In addition,
a weaker second band at 1970 cmꢀ1 could also be as-
signed to Mn(CO)5 based on the same kinetic behavior.
The halogen abstraction method of producing
Mn(CO)5 radical can be compared to the case where me-
tal–metal bond photodissociation of Mn2(CO)10 was used
instead. Because of the absence of strong electronic tran-
sitions in the visible region (>420 nm), the dimer is nor-
mally photolysed at uv wavelengths where the
competitive decarbonylation pathway produces signifi-
cant amount of Mn2(CO)9 [27–30]. In our case, a search
for Mn2(CO)9 based on the previously obtained IR values
turned out to be unsuccessful since halogen abstraction by
CpMo(CO)3 is not expected to promote decarbonylation.
A close examination of the decay curve of CpMo-
(CO)3 and Mn(CO)5 in Fig. 4(a) shows that they fit bet-
Similar results were obtained for the CpMo(CO)3/
CpFe(CO)2I system. Although 532 nm photodissocia-
tion of the [CpFe(CO)2]2 product is possible, reversibil-
ity could not be achieved because of inefficient iodine
atom abstraction by CpFe(CO)2 from the CpMo(CO)3I
product. We conclude that at least for the Mn/Mo
and Mo/Fe systems, the reverse abstraction process is
0.40
0.20
(e)
(e)
(d)
(d)
Abs
(c)
(a)
(c)
-0.20
-0.40
-0.60
(c)
(a)
(b)
(b)
2200
2100
2000
Wavenumbers (cm-1)
1800
1900
Fig. 3. TRS data (1 ls after laser pulse) shows the production of (a) Mn(CO)5 at 1988 and 1973 cmꢀ1, (b) CpMo(CO)3 at 2010 and 1914 cmꢀ1
,
(c) gauche-[CpMo(CO)3]2 at 2020, 1932 and 1880 cmꢀ1, (d) CpMo(CO)3–Mn(CO)5 at 2085 and 1994 cmꢀ1 from 532 nm photolysis of (e) trans-
[CpMo(CO)3]2 and excess of Mn(CO)5Br in hexane solution via halide abstraction.

T.S. Chong et al. / Journal of Organometallic Chemistry 690 (2005) 4132–4138
4137
was photolysed at 532 nm. However disproportionation
reactions induced by uv photolysis of the dimer have
been shown to lead to ionic phosphine-containing com-
plexes in polar solvents [36]. The formation of the
PPh3- containing products in hexane is more likely due
to fast phosphine ligand substitution into CpMo(CO)3 it-
self. Trogler [37] has reviewed similar substitution pro-
cesses for V(CO)6 and Mn(CO)5 radicals in which
facile formation of a two-center three-electron bond
where the odd electron interacts with an electron pair
on an entering nucleophile, can lead to large substitution
rates. In our case, it appears that CpMo(CO)3 radical
kinetics conform to this trend although further phos-
phine substitution into the CpMo(CO)2PPh3 radical is
either very slow or does not take place.
The decay lifetimes measured for both CpMo(CO)3
(4.8 ls) and CpMo(CO)2PPh3 (5.5 ls) in a 1:1 [CpMo-
(CO)3]2/PPh3 system are essentially the same as those
obtained via method 1. This is probably not surprising
because the radicals were produced under rather similar
chemical environment especially during the later stages
of photolysis. Since the latter radical must have origi-
nated from CpMo(CO)3, there should be a difference
in the appearance or risetime of both radicals. We have
been unable to observe the differences presumably be-
cause the ligand substitution reaction proceeded too
quickly. However the subsequent generation of CpMo-
(CO)2PPh3 during the later stages of the TR-FTIR scan
came from photodissociation of the mono and diphos-
phine-substituted molybdenum complex. The linear
FTIR kinetic scan shows that it takes only a short while
(<10 min, 25 mJ laser energy) for a significant amount
of the dimeric phosphine derivatives to build up at the
expense of the CpMo(CO)3 dimer. However, depending
on the relative donor strength, not all phosphines can be
substituted into the CpMo(CO)3 radical. For example,
both CpMo(CO)3 and CpMO(CO)2(POEt3) radical sig-
nals could be observed while the presence of only
CpMo(CO)3 signals were detected when P(C6F5)3 was
used instead.
Fig. 4. (a) The TR-FTIR decay curves of CpMo(CO)3 (b) and
Mn(CO)5 (d) radicals and (b) the spectral changes of stable organo-
metallic components during the halide abstraction process in the
[CpMo(CO)3]2/Mn(CO)5Br system.
endothermic enough to prevent any overall reaction
reversibility. Nevertheless, we have shown here that
the reversibility conditions need not be maintained or
even required in order to detect TR-FTIR radical sig-
nals. Optimal signals are best acquired during the initial
stages of photolysis while the reactants are still present
in large amount.
3.3. Radical ligand substitution
Detection of TR-FTIR spectra of transition metal
carbonyl radicals via fast ligand substitution is also pos-
sible. Fig. 2(b) shows the TR-FTIR spectrum of CpMo-
(CO)2PPh3 generated together with CpMo(CO)3 upon
[CpMo(CO)3]2 photolysis in the presence of excess
PPh3 (>3:1) in hexane. As expected the same radical spe-
cies were expected to be detected via this method when
compared to those obtained by method 1, i.e., metal–me-
tal bond dissociation of Cp2Mo2(CO)5PPh3 (Fig. 2(a)).
A linear FTIR scan showed fast depletion of CpMo-
(CO)3 dimer signals upon photolysis accompanied by
facile formation of its mono and diphosphine deriva-
tives. Initially we thought that the substitution proceeded
via the decarbonylation product, Cp2 Mo2(CO)5 but no
evidence of it was seen even when pure [CpMo(CO)3]2
3.4. Comparison
The three methods used for TR-FTIR radical detec-
tion are briefly assessed. Photodissociation of metal-
bonded dimers (method 1) is straightforward with
interpretable radical–radical recombination kinetics. It
is the only method described here where reversible kinet-
ics could be achieved. However, unless the metal-
bonded dimers are commercially available, syntheses
of the dimers have to be carried out but interpretation
of the kinetics data could be brought into question if
there are significant synthetic side products. Low quan-
tum yields and decarbonylation pathways upon metal–
metal bond dissociation are also among some factors
causing low radical concentrations.

4138
T.S. Chong et al. / Journal of Organometallic Chemistry 690 (2005) 4132–4138
Halogen abstraction by CpMo(CO)3 radical is the
References
most versatile of the three methods presented here.
Many transition metal carbonyl halides are commer-
cially available, relatively air stable and easily deriva-
tized with phosphines and other two-electron donors.
The long wavelength required to produce CpMo(CO)3
also prevents extensive halide photodecomposition.
While the abstraction process works well for first row
halides, it could be difficult to apply to later row species
such as CpRu(CO)2Cl and (Indenyl)Ru(CO)2I. We have
failed to carry out halogen abstraction of these ruthe-
nium species by CpMo(CO)3 presumably due to the
stronger ruthenium–halide bond. Compared to the first
method, the interpretation of the irreversible halogen
abstraction process could be more difficult as well.
Radical ligand substitution (method 3) is excellent for
generating spectroscopic data of phosphine-containing
transition metal carbonyl radicals. Unlike method 1,
the syntheses of high purity phosphine-containing me-
tal-bonded precursors are not required. Ligand substitu-
tion rate studies have so far been restricted to a very few
classes of radicals. We hope to provide some impetus for
such studies since both radicals before and after substi-
tution could be directly detected and characterized.
Obviously the shortcomings are reflected in a more com-
plicated kinetic scheme in which nanosecond TR-FTIR
timescale may not be sufficiently short to detect differ-
ences in the risetime between the two radicals. A pico-
second TR-FTIR system will be of more general use
for such measurements.
The three methods used here have widened the scope
for spectroscopic detection of transition metal carbonyl
radicals. All three methods should be exploited in a
complementary manner so that a thorough understand-
ing of the spectroscopy and kinetics of transition metal
carbonyl radicals could be achieved. The requirement
of complete reversibility in the reactions is also not nec-
essary if methods 2 and 3 were employed, hence reduc-
ing considerable amount of time in devising such
reaction schemes for TR-FTIR studies. Work is already
underway to understand in detail the kinetics of the
reactions carried out according to all three methods.
We also plan to record spectra of radical species directly
implicated in homogeneous catalytic process as well as
radicals that contain two-electron donors such as isocy-
anides and Arduengo carbenes.
[1] M.S. Wrighton, Chem. Rev. 74 (1974) 401.
[2] M.C. Baird, Chem. Rev. 88 (1988) 1217.
[3] M. Poliakoff, E. Weitz, Adv. Organomet. Chem. 25 (1986) 277.
[4] M.W. George, M. Poliakoff, J. Turner, J. Anal. 119 (1994) 551.
[5] H. Hermann, F.W. Grevels, A. Henne, K. Schaffner, J. Phys.
Chem. 86 (1982) 5151.
[6] Y. Ishikawa, P.A. Hackett, D.M. Rayner, J. Mol. Struct. 174
(1988) 113.
[7] A.J. Dixon, M.A. Healy, P.M. Hodges, B.D. Moore, M.
Poliakoff, M.B. Simpson, J.J. Turner, M.A. West, J. Chem.
Soc., Faraday Trans. 2 82 (1986) 2083.
[8] A.J. Dixon, M.W. George, C. Hughes, M. Poliakoff, J.J. Turner,
J. Am. Chem. Soc. 114 (1992) 1719.
[9] B.D. Moore, M.B. Simpson, M. Poliakoff, J.J. Turner, J. Chem.
Soc. Chem. Commun. 15 (1984) 972.
[10] J. Peters, M.W. George, J. Turner, J. Organomet. 14 (1995)
1503.
[11] I.G. Virrels, M.W. George, F.P.A. Johnson, J.J. Turner, J.R.
Westwell, Organometall. 14 (1995) 5203.
[12] X.Z. Sun, S.M. Nikiforov, A. Dedieu, M.W. George, Organome-
tall. 20 (2001) 1515.
[13] B.D. Moore, M. Poliakoff, M.B. Simpson, J.J. Turner, J. Phys.
Chem. 89 (1985) 850.
[14] X.Z. Sun, S.M. Nikiforov, J. Yang, C.S. Colley, M.W. George,
Appl. Spectrosc. 56 (2002) 31.
[15] M. Jezequel, V. Dufaud, M.J. Ruiz-Garcia, F. Carrillo-Hermo-
silla, U. Neugebauer, G.P. Niccolai, F. Lefebvre, F. Bayard, J.
Corker, S. Fiddy, J. Evans, J.P. Broyer, J. Malinge, J.M. Basset,
J. Am. Chem. Soc. 123 (2001) 3520.
[16] S.L. Buchwald, E.V. Anslyn, R.H. Grubbs, J. Am. Chem. Soc 107
(1985) 1766.
[17] S.J. Sturla, N.M. Kablaoui, S.L. Buchwald, J. Am. Chem. Soc.
121 (1999) 1976.
[18] R.J. Haines, C.R. Nolte, J. Organomet. Chem. 24 (1970) 725.
[19] D.S. Ginley, C.R. Bock, M.S. Wrighton, Inorg. Chim. Acta 23
(1977) 85.
[20] J. Wachter, J.G. Riess, A. Mitschler, Organometallics 3 (1984)
714.
[21] W. Hieber, G. Faulhaber, F. Theubert, Z. Anorg. Allg. Chem.
314 (1962) 125.
[22] R.J. Angelici, F. Basolo, J. Am. Chem. Soc. 84 (1962) 2495.
[23] H.G. Alt, M.E. Eichner, J. Organomet. Chem. 212 (1981) 397.
[24] R.H. Hill, J.D. Debad, Polyhedron 10 (1991) 1705.
[25] O.G. Adeyemi, N. Coville, J. Organomet. 22 (2003) 2284.
[26] A.S. Goldman, D.R. Tyler, J. Am. Chem. Soc. 108 (1986) 89.
[27] S.P. Church, H. Hermann, F.W. Grevels, K. Schaffner, J. Chem.
Soc. Chem. Commun. 12 (1984) 785.
[28] S.P. Church, M. Poliakoff, J.A. Timney, J.J. Turner, J. Am.
Chem. Soc. 103 (1981) 7515.
[29] S. Firth, P.M. Hodges, M. Poliakoff, J. Turner, J. Inorg. Chem.
25 (1986) 4608.
[30] H. Yesaka, T. Kobayashi, K. Yasufuku, S. Nagakura, J. Am.
Chem. Soc. 105 (1983) 6249.
[31] C.S. Kraihanzel, P.K. Maples, J. Am. Chem. Soc. 87 (1965)
5267.
[32] C.S. Kraihanzel, P.K. Maples, Inorg. Chem. 7 (1968) 1806.
[33] T. Madach, H. Vahrenkamp, Chem. Ber. 113 (1980) 2675.
[34] R.H. Hill, A. Becalska, N. Chiem, Organometallics 10 (1991)
2104.
[35] W. Xia, L.B. Goetting, J.D. Debad, B.J. Palmer, R.H. Hill, J.
Photochem. Photobiol. A 71 (1993) 221.
[36] C.E. Philbin, A.S. Goldman, D.R. Tyler, Inorg.Chem. 25 (1986)
4434.
[37] W.C. Trogler, Int. J. Chem. Kinet. 19 (1987) 1025.
Acknowledgements
This work was supported by the Agency of Science,
Technology and Research (ASTAR) under Metallocene
Catalysis Grant No. 012-101-0035. T.S. Chong thanked
the Institute of Chemical and Engineering Science
(ICES) of Singapore for a research scholarship.