Structural characterization of selectively prepared cationic iron complexes bearing monodentate and bidentate ether ligands using infrared photodissociation spectroscopy

Article abstract of DOI:10.1016/j.cplett.2003.12.080

The infrared spectroscopy of gaseous ions Fe(CH₃OCH ₃)₂⁺ and Fe(CH₃OCH ₂CH₂OCH₃)_n⁺ (n=1-2) has been studied in the 800-2000 cm^-1 energy range using the coupling of the free electron laser CLIO and an FTICR mass spectrometer, and compared to spectra calculated by ab initio quantum chemistry. The match between experimental and theoretical infrared spectra appears to be very good, the experimental spectra corresponding to the most stable structure predicted by the calculations. Characteristic absorption bands of functional groups have been evidenced, allowing this coupling to be a very powerful tool for structure elucidation in the gas phase.

Full text of DOI:10.1016/j.cplett.2003.12.080

Chemical Physics Letters 385 (2004) 273–279
www.elsevier.com/locate/cplett
Structural characterization of selectively prepared cationic
iron complexes bearing monodentate and bidentate ether
ligands using infrared photodissociation spectroscopy
S. Le Ca €e r, M. Heninger, J. Lemaire, P. Boissel, P. Ma ^ı tre, H. Mestdagh *
Laboratoire de Chimie Physique (UMR 8000, associated to CNRS), Universit ꢀe Paris-Sud, B ^a t. 350, 91405 Orsay Cedex, France
Received 14 November 2003; in final form 18 December 2003
Published online: 20 January 2004
Abstract
þ
þ
The infrared spectroscopy of gaseous ions Fe(CH3OCH3)2 and Fe(CH3OCH2CH2OCH3)n (n ¼ 1–2) has been studied in the
ꢀ
8
00–2000 cm 1 energy range using the coupling of the free electron laser CLIO and an FTICR mass spectrometer, and compared to
spectra calculated by ab initio quantum chemistry. The match between experimental and theoretical infrared spectra appears to be
very good, the experimental spectra corresponding to the most stable structure predicted by the calculations. Characteristic ab-
sorption bands of functional groups have been evidenced, allowing this coupling to be a very powerful tool for structure elucidation
in the gas phase.
ꢀ
2003 Elsevier B.V. All rights reserved.
1
. Introduction
method can be broadened by using the Ômessenger atomÕ
technique [2–5], the complex of interest being bound to a
rare gas atom which is ejected upon absorption of one
photon in the mid-IR. The latter way, known as infrared
multiphoton dissociation (IRMPD), requires a high
Infrared spectroscopy is a powerful tool to charac-
terize the structure of chemical species. This is particu-
larly true in the gas phase, since the experimental
infrared spectra can be directly compared with those
calculated by ab initio quantum chemistry. However, in
the case of ionic species, the low density available in the
gas phase prevents usual absorption measurements. In
recent years much effort has been put into the develop-
ment of experimental methods to study the infrared
spectroscopy of gas phase ions, most of them being
based on ion photodissociation coupled with mass
spectrometric detection. Dissociation induced by light
absorption requires either the presence of a low energy
dissociation pathway which can be reached upon ab-
sorption of one IR photon, or a high irradiation power
so that absorption of several photons may occur and
induce fragmentation. The former way has been used to
get structural information upon complexes presenting
weak electrostatic interactions [1]. The scope of this
fluence, initially available only with CO lasers. These
2
lasers provide knowledge of the photofragmentation
processes [6–8], but their narrow wavelength range
prevents spectroscopic studies. In the recent years, the
application of free electron lasers (FEL) which provide
high peak power and wide tunability has enabled to
record infrared spectra in a large energy range [9–13].
Under these conditions, infrared absorption can be
probed by monitoring IRMPD by mass spectrometry.
In this respect, Fourier transform ion cyclotron reso-
nance (FT-ICR) mass spectrometers are ideal tools, al-
lowing to selectively prepare ions and to distinguish
fragmentation pathways yielding ions of close masses
such as H or H2 loss, because of their high resolution.
Coupling of an FT-ICR mass spectrometer with an in-
frared FEL, recently performed in our group [14–16],
opened the way to the direct characterization of selec-
tively prepared ions, including reactive intermediates
[17].
*
Corresponding author. Fax: +33-16-9153053.
E-mail address: helene.mestdagh@lcp.u-psud.fr (H. Mestdagh).
0
009-2614/$ - see front matter ꢀ 2003 Elsevier B.V. All rights reserved.
doi:10.1016/j.cplett.2003.12.080

274
S. Le Ca €e r et al. / Chemical Physics Letters 385 (2004) 273–279
Transition metal complexes bearing bidentate ligands
such as dimethoxyethane CH3OCH2CH2OCH3 (DXE)
center of the ICR cell with a 1 m focal length mirror.
The beam alignment and mirror position have been set
so as to optimize the photofragmentation rates of or-
raise interesting structural questions since the dihapto
2
þ
(
g ) coordination mode does not allow a trans geometry
which is generally the most stable form of a ML2
ganometallic species, such as Fe(CO) .
5
þ
complex. This might induce either ligand activation with
1
2.3. Experimental sequence
C–O bond cleavage, or stabilization of the g form, the
low enthalpic loss being counterbalanced by the entropic
The conditions used for ion preparation are specified
below:
gain.
In order to directly determine the DXE coordination
mode, we present here the infrared spectra of the cat-
þ
• Ions Fe(CO) (n ¼ 0–5) are prepared by electron ion-
n
ization of Fe(CO)5.
þ
þ
þ
þ
ionic complexes Fe(DXE) and Fe(DXE) , along with
the corresponding ab initio calculations. The IR spec-
• The Fe(DME)2 (Fe(DXE) ) ions are mass-selected
after reaction for 1 s of mass-selected Fe(CO)2 ions
2
þ
þ
trum of the analogous complex Fe(DME) , bearing the
monodentate ligand dimethylether CH OCH (DME),
with DME (resp. DXE) pulsed for 100 ms. The
þ
2
Fe(DXE)2 ions are mass selected after reaction
þ
3
3
is also described for comparison. Ion–molecule reactions
of DME [18] and DXE [19] with iron cationic complexes
have been studied in our group, showing that C–O bond
activation occurs in these systems. In the case of the
of Fe(CO)n (n ¼ 0–5) ions with DXE pulsed for
800 ms. The flow of pulsed reactant gases corresponds
ꢀ
6
to 10 Torr static pressure.
• To ensure ion relaxation, a delay of 500–1000 ms is
allotted before irradiation.
þ
Fe(DME) /DME reaction, two pathways are observed,
ꢁ
corresponding either to the loss of CH or CH4 by the
þ
InfraRed MultiPhoton Dissociation (IRMPD) is ev-
idenced by the appearance of photofragment ions fol-
lowing IR excitation. The fragmentation efficiency,
defined as )ln(Iparent) where Iparent is the relative intensity
of the parent ion, is recorded as a function of the
wavelength to obtain the infrared spectrum of the parent
ion.
3
Fe(DME) ]* intermediate complex. Therefore it is in-
[
teresting to determine whether soft energy deposition in
2
þ
Fe(DME) by IRMPD leads to the same products.
2
2
. Experimental and theoretical methods
.1. FT-ICR mass spectrometer
2
2.4. Theoretical calculations
A compact and easily transportable FT-ICR mass
Theoretical calculations are performed with the
GAUSSIAN 98 suite of programs [22] using the B3LYP
spectrometer has been developed in our laboratory [20].
It is well suited for temporary coupling to a FEL ma-
chine. It is based on a structured permanent magnet
with a nominal field of 1.25 T. The mass resolution
m=Dm is better than 70,000 at mass 131. The trap differs
from the usual cubic FT-ICR cell by its open structure,
allowing wide optical access to the cell center. Primary
ions are generated in the cell by electron impact on
neutral species admitted through pulsed valves.
hybrid density functional. For iron, we used the
[8s6p4d1f] contraction of the (14s11p6d3f) primitive
basis set as recommended by Bauschlicher [23]. A po-
larized double-zeta basis set was used for the ligands
[24]. All the structures calculated were characterized as
1
1
þ
ꢀ
minima except for (g ; g )Fe(DXE) , which presents a
2
1
small imaginary frequency (i179 cm ), corresponding
to the rotation of the methyl group directly attached to
the oxygen atom bound to the metallic center.
2
.2. FEL
The harmonic vibrational frequencies calculated are
presented with a scaling factor of 0.98, determined by
comparing the calculated and experimental [25] infrared
lines of the free dimethylether ligand in the 800–2000
Tunable infrared radiation is produced by the FEL
CLIO [21]. The tuning range is 110–3300 cm 1, although
ꢀ
only the region from 800 to 2000 cm 1 is used in the
present experiments. Continuous tunability is obtained
over a spectral range Dk=k ꢂ 2:5. The FEL relative
bandwidth is typically 0.3% when the optical cavity is
tuned in order to minimize it (at the expense of the peak
power). The FEL temporal structure consists of mac-
ropulses of 8 ls duration, at a repetition rate of 25 Hz.
Each macropulse contains a series of 1–2 ps micropulses
at a repetition rate of 62.5 MHz. A macropulse contains
up to 40 mJ of energy, with a peak power of 40 MW in
each micropulse. The infrared light is focused in the
ꢀ
cm range. Each line is convoluted by a lorentzian
ꢀ
ꢀ1
1
profile of 50 cm fwhm, in order to facilitate compar-
ison with IRMPD spectra.
3. Results and discussion
The three ions studied lead to several photofrag-
ments, resulting from successive and/or parallel frag-
mentations. The primary processes are the only one
reported here. They are identified in different ways, for

S. Le Ca €e r et al. / Chemical Physics Letters 385 (2004) 273–279
275
þ
ꢀ1
instance by decreasing the laser power so that the pri-
mary photoproducts do not reabsorb light and pre-
dominate over the others. Discussion of the sequential
fragmentation mechanisms will be published separately.
(DME)Fe(CH3)(OCH3) is predicted to lie 21 kJ mol
higher in energy. The strongest band in its calculated
infrared spectrum, appearing at 1110 cm 1, is absent
ꢀ
from the experimental spectrum, which confirms the
þ
Fe(DME) structure predicted by theory.
2
þ
3
.1. Fe(CH OCH )
3
The remarkable consistency obtained between exper-
imental and predicted spectra was not obviously ex-
pected concerning band intensities: while the calculations
correspond to one-photon absorption, the IRMPD
mechanism represents a much more complex process.
The conditions required for this agreement correspond
to a relative fragmentation amount comprised be-
tween 3% and 20% per macropulse in the absorp-
tion peaks. In this range, the fragmentation efficiency
was observed to vary almost linearly with the laser
power.
3
2
þ
The Fe(DME)2 ions were irradiated for 1 s with
photons of wavelength ranging from 800 to 1600 cm 1.
The IRMPD spectrum, obtained with an average laser
power of 600–800 mW is shown in Fig. 1, and the cor-
responding predicted and experimentally observed IR
band positions are listed in Table 1.
ꢀ
The most stable calculated structure for this complex
þ
is Fe(DME)2 in which two intact DME ligands lie in
trans position around the metallic center (D2d symmetry
þ
and quartet spin state, as expected from the calculations
þ
The primary photofragments are Fe(OCH2)(DME)
þ
on Fe(H2O)2 [26]). There is an excellent agreement be-
(major) and Fe(OCH3)(DME) , resulting from loss of
ꢁ
tween the predictions of theory and the measurements,
as reproduced in Fig. 1 and Table 1, both in terms of
positions and relative intensities of the infrared transi-
tions. The isomer with the inserted structure
CH4 and CH , respectively. The corresponding
3
branching ratio remains comprised between 75% and
95% (generally close to 85%) methane loss, for all
wavelengths leading to fragmentation. Interestingly, no
DME loss was observed. The constancy of the branch-
ing ratio throughout the spectrum, i.e. whatever the
nature of the vibration initially excited by IR radiation,
is consistent with the hypothesis of fast internal vibra-
tional redistribution (IVR) resulting in energy redistri-
bution over the ion vibration modes prior to the
fragmentation process.
þ
3.2. Fe(CH OCH CH OCH )
3 2 2 3
þ
Mass-selected Fe(DXE) ions were irradiated for
00 ms with photons of wavelength ranging from 800 to
5
1
800 cm 1. Primary fragmentation consists in a unique
ꢀ
process, ethylene loss:
þ
þ
FeðCH OCH CH OCH Þ ! ½Fe; C ; H ; O ꢃ þ C H
3
2
2
3
2
6
2
2
4
The experimental spectrum, recorded for an aver-
age laser power of 680 mW, is depicted in Fig. 3, along
with the predicted spectra of the lowest energy lying
structures.
Fig. 1. Experimental (fine line with markers) and theoretical (thick
þ
line) infrared spectrum of Fe(DME) . The calculated structure of
þ
2
Fe(DME)2 is also given.
Table 1
Band positions and relative intensities of the experimental and theoretical spectra of Fe(DME) , along with an approximate description of each
mode
þ
2
ꢀ1
Description of the vibrations
Symmetry
Position (cm
)
Relative intensity
Calculated
Calculated
Experimental
Experimental
C–O sym. str.
C–O asym. str.
b
e
2
2
2
871
1017
1137
1248
1454
1453
1473
852
1015
1155
1258
1458
1.00
0.94
0.19
0.12
0.48
1.00
0.93
0.24
0.15
0.31
CH
CH
CH
CH
3
3
3
3
rock.
rock.
d-def.
d-def.
e
b
e
a
e
b
a
Type: CH2 scissoring.

276
S. Le Ca €e r et al. / Chemical Physics Letters 385 (2004) 273–279
1
.46
1.47
1
.41
1
.47
1
.45
1
.42
2
.00
2
.02
η1) Fe(DXE)+ Sextet
1
+
(
(
η ) Fe(DXE) Quartet
1
.45
.46
1
.45
.44
η2) Fe(DXE)+ Quartet
7
6.0˚
8
2.8˚
1
1
2
.11
2
.04
(η ) Fe(DXE)+ Sextet
2
(
1
2
þ
Fig. 2. Calculated structures of the most stable DXE conformer as well as (g )- and (g )Fe(DXE) ions. Distances are given in angstr o€ ms and angles
in degrees.
þ
1
000
compared to the one calculated for Cu(DXE) (297
ꢀ1
1
1
0
0
.5
.5
kJ mol ) by Koizumi et al. [27] at similar level of the-
þ
ory, also predicting a bidentate coordination. The Fe -
1
(
g )DXE binding energy at 0 K is calculated to be
8
00
00
00
00
0
ꢀ
1
195 kJ mol . This value is larger but of the same order
of magnitude than the value found by Ricca et al. in the
ꢀ1
case of a water ligand (140 kJ mol ) [26].
As can be seen in Fig. 2, the experimental spectrum
is in very good agreement with the spectrum calcu-
lated for (g )Fe(DXE) in both its sextet and quartet
spin state. The corresponding predicted and experi-
mentally observed IR band positions are given in Table
2. The strong band appearing at 1019 cm corresponds
to C–O asymmetric stretching, whereas the weaker
band at 1460 cm can be assigned to a combination
of bands corresponding to dissymmetric CH deforma-
6
4
2
2
þ
ꢀ
1
ꢀ
1
3
tion and CH scissoring. There is a better agreement
2
8
00
1000
1200
1400
1600
1800
between the experimental and the calculated spec-
-
1
2
þ
Energy (cm
)
trum of the sextet (g )Fe(DXE) isomer, but the
predicted difference in the positions of the C–O asym-
metric stretching, for example, are too small to con-
clude. On the other hand, the energetically disfavored
Fig. 3. Experimental (fine line with markers) and theoretical (thick
þ
lines) infrared spectrum of Fe(DXE) . The black/red, grey/blue and
2
dashed/green theoretical spectra correspond to sextet-(g )-, quartet-
2
1
þ
1
þ
(
g )- and sextet-(g )Fe(DXE) , respectively.
(g )Fe(DXE) presents an IR spectrum clearly different
from the experimental spectrum. In particular, there
is a strong band at 1150 cm 1, which is not observed
ꢀ
þ
Two non-inserted forms of Fe(DXE) were consid-
experimentally.
ered, the DXE molecule being bound by either one or
both oxygen atoms. The corresponding calculated
structures are depicted in Fig. 2, as well as the most
stable DXE conformer. They are termed as
(g )Fe(DXE) and (g )Fe(DXE) , where g refers to
the actual DXE hapticity in the complex. The
þ
3.3. Fe(CH
3
OCH
2
CH
2
OCH
3
)
2
þ
Mass-selected Fe(DXE)2 ions were irradiated for
500 ms with photons of wavelength ranging from 800 to
1
þ
2
þ
2000 cm 1. The experimental spectrum, recorded for an
average laser power of 510 mW, is depicted in Fig. 5,
along with the calculated spectra of three potential iso-
mers (see below). The corresponding predicted and ex-
perimentally observed IR band positions are shown in
Table 3.
ꢀ
2
þ
(
g )Fe(DXE) isomer in a sextet spin state has been
found to be lowest energy isomer, but the lowest quartet
spin state only lies +8 kJ mol 1. Fe(DXE) isomers
presenting a g coordination are much higher in energy
(+94 kJ mol ), and the lowest spin state is also a sextet,
although a quartet spin state isomer has been found only
+4.7 kJ mol
binding energy at 0 K is 289 kJ mol . This value can be
ꢀ
þ
1
ꢀ
1
þ
Three isomers of Fe(DXE) corresponding to the
2
2
ꢀ
1
þ
2
1
1
1
2
2
higher in energy. The Fe -(g )DXE
three (g ; g ) (g ; g ) and (g ; g ) combinations of the
coordination mode of the two DXE ligands have been
ꢀ
1

S. Le Ca €e r et al. / Chemical Physics Letters 385 (2004) 273–279
277
Table 2
Band positions and relative intensities of the experimental and theoretical spectra of Fe(DXE) , along with an approximate description of each mode
þ
1
þ
2
þ
2
þ
Approximate description
Sextet-(g )Fe(DXE)
Sextet-( g )Fe(DXE)
Quartet-(g )Fe(DXE)
Experimental
Position
)
Relative
intensity
Position
ꢀ1
Relative
intensity
Position
ꢀ1
Relative
intensity
Position
ꢀ1
Relative
intensity
ꢀ
1
(
cm
(cm
)
(cm
)
(cm
)
C
C
ter–O sym. str.; CH
2
rock
870
979
076
0.86
1.00
0.32
0.78
0.43
841
1024
1073
0.26
1.00
844
1039
1074
0.19
1.00
n.d.
n.d.
1.00
ter–O asym. str.; C–C str.
1019
1074
b
1
0.19
0.08
CH
CH
2
twist; CH
d-def.; CH
3
rock.
scissoring
1150
1459
<0.03
0.28
<0.03
0.25
a
3
2
1457
1456
1461
0.20
a
Several close absorption bands of the same nature in the calculated spectra, leading to a unique peak after lorentzian convolution.
Appears as a shoulder in the main peak.
b
Table 3
Band positions and relative intensities of the experimental and theoretical spectra of Fe(DXE)2
þ
1
1
þ
1
2
þ
2
2
þ
(g , g )Fe(DXE)
2
Approximated description
(g , g )Fe(DXE)
(g , g )Fe(DXE)
Experimental
2
2
Position
(
Relative
intensity
Position
ꢀ1
Relative
intensity
Position
ꢀ1
Relative
intensity
Position
ꢀ1
Relative
intensity
ꢀ
1
cm
)
(cm
)
(cm
)
(cm
)
a
C
ter–O sym. str.; CH
2
rock
858
915
0.26
0.35
0.33
866
0.20
0.06
n.d.
n.d.
1.00
9
1000
07
0.88
1.00
a
b
C
C
ter–O asym. str.; C–C str.
ter–O asym. str.
1019
1025
1067
999
a
a
1088
0.52
1059
1.00
1.00
0.14
1056
1093
b
1
102
a
CH
CH
CH
3
2
3
rock; CH
twist; CH
d-def.; CH
2
twist
1147
1184
1461
0.98
0.48
0.42
1146
1184
1461
0.53
0.21
0.34
1176
1233
1462
0.06
0.05
0.22
1188
1244
1473
0.20
0.20
0.1
a
3
rock
a
2
scissoring
a
Several close absorption bands of the same nature in the calculated spectra, leading to a unique peak after lorentzian convolution.
Appears as a shoulder in the main peak.
b
1
considered, and the optimized structured are given in
Fig. 4. In these three isomers, oxygens of two different
ligands bind in trans from each other and a quartet spin
ground state is found. The lowest energy structure cor-
noted however that the binding energies of g -DXE and
2
2
þ
ꢀ1
g -DXE to (g -DXE)Fe (144 and 191 kJ mol ) are
significantly smaller than the corresponding binding
energies to Fe (195 and 289 kJ mol ). Both values are
slightly higher than those calculated for the corre-
sponding copper complexes (125 and 168 kJ mol , re-
spectively [27]).
þ
ꢀ1
2
þ
2
responds to the (g -DXE)Fe -(g -DXE) isomer. The
2
þ
1
1
þ
1
ꢀ1
(
isomers lie respectively 47 and 111 kJ mol higher in
g -DXE)Fe -(g -DXE) and (g -DXE)Fe -(g -DXE)
ꢀ
1
1
energy. As expected, the binding energy of the g -DXE
in (g -DXE)Fe -(g -DXE) (170 kJ mol ) is of the
same order of magnitude as Fe -(g -DXE). It should be
The experimental spectrum nicely matches with the
2
1
þ
1
ꢀ1
2
IR spectrum calculated for the bis-bidentate (g ; g )
þ
1
þ
2
þ
Fe(DXE)2 structure. As for the (g )Fe(DXE) , the
1
.46
1
.45
1
.44
.44
1.46
7
6.0˚
1.45
.12
1
.96
1
.15
2.02
.10
2
2
1
.96
2
1
.46
OFeO=76.3˚
1
.46
2
2
+
1
2
+
1
1
+
(
η η ) Fe(DXE)
2
Quartet
(η η ) Fe(DXE) Quartet
2
(η η ) Fe(DXE) Quartet
2
1
1
1
2
2
2
þ
2
Fig. 4. Calculated structures of (g ; g )-, (g ; g )- and (g ; g )Fe(DXE) ions. Distances are given in angstr o€ ms and angles in degrees.

278
S. Le Ca €e r et al. / Chemical Physics Letters 385 (2004) 273–279
þ
Primary fragmentation of Fe(DXE)2 consists exclu-
sively in ethylene loss, as for Fe(DXE) :
þ
1
200
000
2
1
0
þ
þ
FeðDXEÞ ! FeðOCH Þ ðDXEÞ þ C H
2
3
2
2
4
1
For comparison, IRMPD of crown ether complexes
þ
ꢀ1
of alkali metals or H3O at 943 cm , reported by Peiris
et al. [28], follows very different schemes: dissociation of
crown ether alkali metal complexes leads to the loss of
the ligand at low laser fluences, the dissociation of the
crown ether being not observed. Dissociation of crown
800
600
400
200
0
þ
ether H3O complexes proceeds via loss of water to
generate protonated crown ether, which is fragmented at
high fluence through loss of C2H4O units. The particular
fragmentation behaviour of our DXE complexes may
þ
be explained by the possibility for Fe insertion into the
C–O bond.
8
00
1000
1200
1400
Energy (cm
1600
1
)
1800
2000
4. Conclusion
-
Infrared spectra of three cationic iron complexes
bearing ether and diether ligands have been obtained in
the gas phase. They appear in very good agreement with
the theoretical spectra corresponding to the most stable
structure predicted by ab initio calculations. This re-
mains true when the fragmentation observed involves
prior isomerization of the complex (as in the case of
Fig. 5. Experimental (fine line with markers) and theoretical (thick
þ
lines) infrared spectrum of Fe(DXE) . The black/red, grey/blue and
2
2
2
1
2
dashed/green theoretical spectra correspond to (g ; g )-, (g ; g )- and
1
1
þ
(
cm help for comparison.
g ; g )Fe(DXE) , respectively. The vertical lines at 915 and 1146
2
ꢀ
1
2
2
þ
2
(
g ; g )Fe(DXE) IR spectrum is characterized by two
þ
CH loss from Fe(DME) , implying insertion of the
4
2
main features. First, a broad band corresponding to
metal in the C–O bond followed by H transfer). The
spectrum is characteristic of the structure since an es-
sential condition for photofragmentation is absorption
of the first photons by the starting stable ion in its
ground state, which occurs exclusively at the absorption
wavelengths specific of its structure. Upon further
multiphoton absorption, isomerisation may and does
occur, allowing fragment formation through the lowest
energy pathway.
deformations of CH and CH scissoring at 1462 cm 1,
ꢀ
3
2
corresponding to the band at 1473 cm 1 in the experi-
mental spectrum. Second, the most intense band ob-
served in the experimental IRMPD spectrum at 1056
cm could be attributed to the C–O asymmetric
stretching modes, the most intense being predicted at
ꢀ
ꢀ
1
ꢀ
1
1067 cm . The two other (weaker) experimental fea-
ꢀ
tures at 1188 and 1244 cm could also be assigned from
1
2
2
þ
the calculated IR spectrum of (g ; g )Fe(DXE) , the
former corresponding to CH3 rock (1176 cm ) and the
2
The three infrared spectra exhibit similarities: a
ꢀ1
ꢀ
1
strong band between 1000–1100 cm corresponding to
ꢀ1
latter to CH2 twist (1233 cm 1).
ꢀ
a C–O stretching mode, and a band around 1460 cm
corresponding to CH dissymmetric deformation and
1
2
þ
The (g ; g )Fe(DXE) structure can be excluded
2
3
since it should lead to photodissociation at 1146 cm 1,
where a strong IR absorption band is predicted,
whereas this energy corresponds to a minimum in the
IRMPD absorption efficiency (see Fig. 5). Similarly,
ꢀ
CH2 scissoring modes. Comparison with the infrared
spectra of the free ligands shows that upon metal co-
ꢀ1
ordination the former band is shifted by ꢂ100 cm
towards lower energies, as expected from calculations.
Such absorption bands appear characteristic of func-
tional groups or structural patterns, similarly to con-
densed phase infrared spectroscopy. This indicates that
even if no calculated spectrum is available, infrared
spectroscopy of gas phase ions is a useful tool for
structural elucidation.
1
1
þ
the (g ; g )Fe(DXE) isomer can also be excluded
since the IRMPD photofragmentation signal vanishes
2
ꢀ
1
at 900 cm , whereas it is a maximum in the IR ab-
þ
1
1
sorption spectrum of the (g ; g )Fe(DXE) isomer.
As in the case of Fe(DME)2 and Fe(DXE) , the
2
þ
þ
þ
IRMPD spectrum of Fe(DXE)2 is in good agreement
with the IR absorption spectrum of the lowest energy
isomer predicted by ab initio calculations. It should be
noted that a similar (g ; g ) structure has been ex-
perimentally assigned to Cu(DXE)2 based on the
References
2
2
þ
[1] M. Okumura, L.I. Yeh, Y.T. Lee, J. Chem. Phys. 83 (1985) 3705.
[2] O. Dopfer, R.V. Olkhov, R.V. Roth, J.P. Maier, Chem. Phys.
Lett. 296 (1998) 585.
comparison of collision-induced dissociation results
and ab initio calculations [27].

S. Le Ca €e r et al. / Chemical Physics Letters 385 (2004) 273–279
279
[
[
[
[
[
3] R.V. Olkhov, S.A. Nizkorodov, O. Dopfer, Chem. Phys. 239
1998) 393.
4] G. Gregoire, J. Velasquez, M.A. Duncan, Chem. Phys. Lett. 349
2001) 451.
[16] W. Jones, P. Boissel, B. Chiavarino, M.E. Crestoni, S. Fornarini,
J. Lemaire, P. Ma ^ı tre, Angew. Chem., Int. Ed. Engl. 42 (2003)
2057.
(
(
[17] L. MacAleese, A. Simon, J. Lemaire, P. Boissel, P. Ma ^ı tre, in
preparation.
5] W.H. Robertson, E.A. Price, J.M. Weber, J.-W. Shin, G.H.
Weddle, M.A. Johnson, J. Phys. Chem. A 107 (2003) 6527.
6] R.L. Woodin, D.S. Bomse, J.L. Beauchamp, J. Am. Chem. Soc.
[18] S. Le Ca €e r, M. Heninger, P. Pernot, H. Mestdagh, Phys. Chem.
Chem. Phys. 4 (2002) 1855.
100 (1978) 3248.
[19] S. Le Ca €e r, Thesis, Universit ꢀe Paris-Sud (France), 2003.
[20] G. Mauclaire, J. Lemaire, P. Boissel, G. Bellec, M. Heninger, Eur.
J. Mass Spectrom., DOI:10.1255/ejms.620, in press.
[21] R. Prazeres, F. Glotin, C. Insa, D. Jaroszynski, J.M. Ortega, Eur.
Phys. J. D 3 (1998) 87.
7] M.A. Hanratty, E.A. Carter, J.L. Beauchamp, I. Goddard, A.J.
Illies, M.T. Bowers, Chem. Phys. Lett. 123 (1986) 239.
8] Y. Huang, B.S. Freiser, J. Am. Chem. Soc. 115 (1993) 737.
9] K.R. Asmis, M. Br u€ mmer, C. Kaposta, G. Santambrogio, G. von
Helden, G. Meijer, K. Rademann, L. W o€ ste, Phys. Chem. Chem.
Phys. 4 (2002) 1101.
[
[
[22] M.J. Frisch et al., GAUSSIAN 98 (Revision A.6), Gaussian Inc,
Pittsburgh PA, 1998.
[
[
[
[
[
10] J. Oomens, A.J.A. van Roij, G. Meijer, G. von Helden, Astro-
phys. J. 542 (2000) 404.
11] J. Oomens, G. Meijer, G. von Helden, J. Phys. Chem. A 105
[23] C.W. Bauschlicher Jr., Theor. Chim. Acta 92 (1995) 183.
[24] P.J. Hay, J. Chem. Phys. 66 (1977) 4377.
[25] NIST Mass Spec Data Center, S.E. Stein, Infrared Spectra in
NIST Chemistry WebBook, NIST Standard Reference Database
Number 69, P.J. Linstrom, W.G. Mallard (Eds.), March 2003,
National Institute of Standards and Technology, Gaithersburg
MD, 20899. Available from <http://webbook.nist.gov>.
[26] A. Ricca, C.W. Bauschlicher Jr., J. Phys. Chem. 99 (1995)
9003.
(2001) 8302.
12] D. van Heijnsbergen, M.A. Duncan, G. Meijer, G. von Helden,
Chem. Phys. Lett. 349 (2001) 220.
13] D. van Heijnsbergen, G. von Helden, G. Meijer, P. Ma ^ı tre, M.A.
Duncan, J. Am. Chem. Soc. 124 (2002) 1562.
14] J. Lemaire, P. Boissel, M. Heninger, G. Mauclaire, G. Bellec, H.
Mestdagh, A. Simon, S. Le Ca €e r, F. Glotin, J.M. Ortega, P.
Maitre, Phys. Rev. Lett. 89 (2002) 273002.
[27] H. Koizumi, P.B. Armentrout, J. Am. Soc. Mass Spectrom. 12
(2001) 480.
[
15] P. Ma ^ı tre, S.L. Ca €e r, A. Simon, J. Lemaire, H. Mestdagh, M.
Heninger, G. Mauclaire, P. Boissel, J.-M. Ortega, R. Prazeres, F.
Glotin, Nucl. Instrum. Meth. Phys. Res. A 507 (2003) 541.
[28] D.M. Peiris, Y. Yang, R. Ramanathan, K.R. Williams, C.H.
Watson, J.R. Eyler, Int. J. Mass Spectrom. Ion Proc. 157–158
(1996) 365.