Reactions of ground state Ti atoms with NO: Insertion versus complexation. An IR matrix isolation study

Article abstract of DOI:10.1016/S0301-0104(00)00058-6

The reaction of ground state Ti atoms with NO during condensation in solid argon has been reinvestigated. The NTiO molecule, already characterized in reactions of laser-ablated Ti, is the only product observed for the reaction between one Ti atom and one nitric oxide molecule. Isotopic data on ν₁, ν₂, ν₃, 2ν₁ and 2ν₂ have been measured in the mid- and far-infrared regions. This enables a complete harmonic force-field calculation based on a bent geometry, in agreement with the conclusions of the previous study. No evidence is found, however, of a metastable nitrosyl complex intermediate, as previously proposed. This study confirms that the insertion reaction proceeds directly from the ground electronic state reagents, with no or very little activation energy. (C) 2000 Published by Elsevier Science B.V.

Full text of DOI:10.1016/S0301-0104(00)00058-6

Chemical Physics 254 (2000) 267±274
www.elsevier.nl/locate/chemphys
Reactions of ground state Ti atoms with NO: insertion versus
complexation. An IR matrix isolation study
1
*
Lahouari Krim , Christophe Prot , Esma õ l M. Alikhani, Laurent Manceron
LADIR/Spectrochimie Mol eꢀ culaire, UMR 7075 CNRS-Universit eꢀ Pierre et Marie Curie, Bo ^õ te 49, 4 Place Jussieu, 75252 Paris, Cedex
0
5, France
Received 13 January 2000
Dedicated to the memory of Christophe Prot
Abstract
The reaction of ground state Ti atoms with NO during condensation in solid argon has been reinvestigated. The
NTiO molecule, already characterized in reactions of laser-ablated Ti, is the only product observed for the reaction
1 2 3 1 2
between one Ti atom and one nitric oxide molecule. Isotopic data on m , m , m , 2m and 2m have been measured in the
mid- and far-infrared regions. This enables a complete harmonic force-®eld calculation based on a bent geometry, in
agreement with the conclusions of the previous study. No evidence is found, however, of a metastable nitrosyl complex
intermediate, as previously proposed. This study con®rms that the insertion reaction proceeds directly from the ground
electronic state reagents, with no or very little activation energy. Ó 2000 Published by Elsevier Science B.V. All rights
reserved.
Ã
1. Introduction
TiO ‡ N reaction was studied in the single colli-
sion regime using pulsed, crossed molecular beams
by Naulin et al. [4]. The Ti ‡ NO, O and N O
The reactivity of transition metal atoms in the
2
2
gas phase or in isolated conditions is a challenging
object of study, due to the diculty of carrying out
structural spectroscopic studies in very dilute me-
dia, and to the presence of numerous low-lying
electronic con®gurations. For reactions involving
titanium atoms, oxidation by O , NO , N O and
reactions have also been the object of several ki-
netic studies in fast ¯ow reactors at room tem-
perature, probing the kinetics of disappearance of
Ti atoms in the ground [5,6] or excited states [7] or
the TiO product concentrations. These studies
yielded e€ective rate constants of the bimolecular
reactions and concluded the existence of low acti-
vation energy (2±4 kcal/mol) to the oxygen atom
abstraction reaction Ti ‡ XO ! TiO ‡ X. Inter-
estingly, this barrier was found to be the lowest for
X ˆ N [5,6]. In analogy with the mechanism in-
volved in the alkaline-earth-metal atom abstrac-
tion reactions with oxygen-containing molecules,
harpooning or electron-jump mechanisms were
assumed as the onset of the metal±molecule
2
2
2
O has been studied by chemiluminescence from
3
Ã
the excited TiO product, in ¯ames by Gole
and coworkers [1,2], in crossed-beam studies by
Parson et al. [3], and more recently, the Ti ‡ NO !
*
Corresponding author.
E-mail address: krim@ccr.jussieu.fr (L. Krim).
Deceased 4 October 1999.
1
0301-0104/00/$ - see front matter Ó 2000 Published by Elsevier Science B.V. All rights reserved.
PII: S 0 3 0 1 - 0 1 0 4 ( 0 0 ) 0 0 0 5 8 - 6

2
68
L. Krim et al. / Chemical Physics 254 (2000) 267±274
interaction. Another investigation explored the
temperature and pressure dependence of Ti ‡ O
O, NO, CO , SO or NO oxidation reactions in
matrix up to 25 K after Ti atom deposition, it was
concluded that this species could be formed with
very little or no activation energy from the ground
state reagents. This left, however, the existence of a
second, metastable titanium nitrosyl species, pre-
dicted some 36 kcal/mol higher in energy, more
dicult to account for. The ability of titanium
atoms to insert spontaneously into an O@X mul-
tiple bond has been documented before in matrix
2
,
N
2
2
2
2
a fast ¯ow reactor [8]. The observed pressure de-
pendences of the Ti ‡ NO and Ti ‡ CO reaction
2
rates indicate a contribution from a termolecular
association channel, in addition to the abstraction
reactions already documented. For NO, the evo-
lution of the e€ective rate constants with temper-
ature for the reaction,
isolation studies of the Ti ‡ CO
2
reaction [11,12],
and investigated theoretically [13].
Ti ‡ NO $ ‰TiNOŠ ! TiO ‡ N
In the course of parallel studies of the reactivity
of ground state, thermally produced transition
metal atoms with various diatomic molecules in
solid argon [15±18], the reaction of NO molecule
with ground state Ti has been reinvestigated. In
this article, we show that the oxonitride insertion
product is the only product observed in the reac-
tion of a ground state Ti atom with a nitric oxide
molecule in an argon matrix. Our results com-
suggested exothermicity for the ®rst step, while the
abstraction activation energy remains globally
slightly positive. Until recently, however, nothing
was known about the chemical nature or electronic
structure of TiXO (X ˆ N, O, N ) intermediates.
2
Also, as discussed in Ref. [6], the reason for the
increase in rate constants in Ti ‡ NO vs. Ti ‡ O
2
reaction remains unclear and inconsistent with
simple electron-transfer mechanisms, which sug-
gests that the nature and bonding of these inter-
mediates are important in that respect.
pletely con®rm the assignments of the m
and m
1
2
fundamentals by Kushto et al. [10], but the larger
product yield with the thermal e€usion technique
enables us to detect the lower frequency funda-
Recently, in a theoretical systematic study of
the NO interaction with ®rst row transition metal
atoms, using density functional theory, predictions
were made of the binding energies, structures and
spectroscopic properties of a TiNO, titanium ni-
trosyl species [9]. This study predicted the existence
of a linear, nitrogen-bound nitrosyl complex with
a fairly large Ti±NO binding energy (ꢀ50 kcal/
mental, the bending vibration m
2m overtones. A quadratic harmonic force ®eld is
also presented, based on the isotopic data obtained
3
1
, and the 2m and
2
46
47
48
49
50
on all fundamentals for the Ti/ Ti/ Ti/ Ti/ Ti,
18
O/ O, N/ N, isotopic species of NTiO. These
16
14
15
data (bond force constants and interaction force
constants) are interesting for the discussion of the
bonding within transition metal-containing mole-
cules [18], and give precious reference points for
testing the various theoretical modelling methods
[19].
2
mol) in the R ground state, fairly short Ti±N
ꢁ
ꢁ
(
ꢀ1.7 A) and elongated N@O (ꢀ1.228 A) dis-
tances.
Even more recently, Kushto and coworkers in-
vestigated the reaction of laser-ablated Ti atoms
with nitric oxide in rare-gas matrices, and found
two reaction products of the lowest stoichiometry,
Ti ‡ NO, along with ionic reaction products and
larger neutral species. Based on their experimental
2. Experimental
®
ndings and on further DFT calculations, they
The Ti ‡ NO samples were prepared by co-
condensing Ti vapour and NO/Ar mixtures (0.2±
2% molar ratios) onto a cryogenic metal mirror
maintained around 10 K. The experimental
methods and set-up have been previously de-
scribed [20]. Brie¯y, here, Ti was vapourized from
a 1 mm diameter TiMo alloy wire (Goodfellow,
UK) heated resistively to ꢁ1700°C. Metal depo-
identi®ed these species as a titanium nitrosyl spe-
cies, TiNO, characterized by an NO stretching
mode near 1615 cm , and a titanium oxonitride
species, NTiO, characterized by the two valence
�
1
�
1
stretching vibrations in the 900±700 cm region
10]. From atomic or molecular di€usion experi-
ments proceeding by slight warming of the argon
[

L. Krim et al. / Chemical Physics 254 (2000) 267±274
269
sition rates, monitored with a microbalance, were
3. Results
typically of the order of 0.10±0.40 lg/min.
Argon gas (LÕAir liquide, France, 99.995%)
was used without further puri®cation, NO (LÕAir
In samples where Ti atoms were codeposited
with NO in an argon matrix, both in relatively
diluted conditions, infrared spectra presented no
absorption in the metal nitrosyl N±O stretching
15
Liquide, 99.9%) and NO gas (Isotec, USA, with
an isotopical purity of 97.8%) were puri®ed, in
order to remove N , N O and NO , by using
�
1
region (1500±1900 cm ) apart from those attrib-
uted to unreacted NO, (NO) and their isotopo-
2
2
2
14
18
trap-to-trap vacuum distillations. N O was
O
2
1
8
prepared in the laboratory by addition of
16
2
mers [22,23]. The contrast with the results obtained
with a typical late ®rst-row transition metal such as
Ni is striking. Fig. 1 presents the spectra obtained
in equivalent conditions ꢂM=NO=Ar ꢀ 0:5=1=100†
depositing either Ti or Ni ground state atoms with
NO. With Ni atoms, the reaction products were
identi®ed mainly by absorptions in the 1600±1900
14
14
ꢂ16‡18†
onto N O gas to form
N O
2
which was
14
18
14
16
subsequently reduced to N O and N O by
reaction with mercury [21]. The resulting gas is a
14
16
mixture containing approximately 52% N O
18
14
and 48% N O. The purity of the samples
was con®rmed spectroscopically. For experi-
ments performed with O and N , 99.998% purity
�
1
cm region, characteristic of metal±nitrosyl com-
plexes [14,22] (Fig. 1). Under the same experi-
mental conditions, the reaction of Ti with NO does
not lead to the formation of any nitrosyl complex,
but only to the appearance of new bands in the
2
2
gases from Air Liquide, France, were used with-
out puri®cation.
In general, after 2 h deposition time, infrared
spectra of the resulting sample were recorded in
the transmission±re¯ection mode between 5000
�
1
700±920 cm region, described recently by Kushto
et al., and attributed to the primary oxonitride
insertion product, NTiO. Our experiments pre-
�
1
and 70 cm using a Bruker 120 FTIR spectrom-
eter and suitable combinations of CaF /Si, KBr/
2
�
1
Ge or 6 lm Mylar beamsplitters with either liquid
N -cooled InSb or narrow band HgCdTe photo-
2
sented no sign of a band at 1614.7 cm attributed
to a titanium mononitrosyl species, TiNO, as dis-
cussed in Ref. [10]. This species, was calculated as
stable or metastable in Refs. [9,10]. In dilu-
ted conditions (Ti/NO/Ar ˆ 0:2/0.2/100), the only
diodes or a liquid He-cooled Si±B bolometer, ®tted
with cooled bandpass ®lters. The resolution was
�
1
varied between 0.1 and 0.5 cm . Bare mirror
backgrounds, recorded at 10 K from 5000 to 70
�
1
cm prior to sample deposition, were used as
references in processing the sample spectra. Also,
absorption spectra in the near-, mid- and far-
infrared were collected on the same samples
through either CaF
mounted on a rotatable ¯ange separating the in-
2
, CsI or polyethylene windows
�
2
terferometer vacuum (10 mbar) from that of the
�
7
cryostatic cell (10 mbar). The spectra were sub-
sequently subjected to baseline correction to
compensate for infrared light scattering and in-
terference patterns.
The sample could be next irradiated using a 200
W mercury±xenon high pressure arc lamp and in-
terference or broadband ®lters as it was found that
some of the products could be photolysed or
converted by UV±visible light. Infrared spectra of
the photolysed samples or next after sample an-
nealing to 35 K in several steps were recorded
Fig. 1. Comparison of infrared absorption spectra in the ni-
�
1
� 1
trosyl (1500±2000 cm ) and oxonitride regions (700±950 cm )
for argon matrices containing NO ‡ Ni, NO ‡ Ti, and NO only
at 10 K; from bottom to top: NO/Ar ˆ 1/100, Ti/NO/Ar ˆ 0.5/1/
100 and Ni/NO/Ar ˆ 0.5/1/100.
�
1
between 5000 and 70 cm as outlined above.

2
70
L. Krim et al. / Chemical Physics 254 (2000) 267±274
signals which could be detected were due to water,
present as an impurity in the vacuum system. The
2
Experiments were run replacing NO by O to
compare the reactivity of ground state Ti with
these two molecules under the very same experi-
mental conditions. Our results for the Ti ‡ NO
�
1
band located at 1614.7 cm , assigned in Ref. [10]
to a linear TiNO complex, is detected only when
the matrix is heated up to 35 K or when the NO
molar ratio is increased. This band thus belongs to
a larger nitrosyl complex.
and Ti ‡ O reactions are contrasted in Fig. 2. In
2
the related insertion reaction Ti ‡ O ! OTiO, the
2
product yield after deposition of ground state Ti
Other species could be observed, which present
a second-order dependence in NO, such as the one
responsible for absorptions at 1488, 1208 and
2
and O is relatively low but dramatically enhanced
upon Ti atom photoexcitation. This shows that, at
the di€erence of the Ti ‡ NO reaction, the inser-
�
1
977 cm and assigned to a OTiNNO structure in
2
tion reaction in O presents a small energy barrier,
Ref. [10]. Interestingly, upon UV±visible photo-
excitation, this species could be converted into
OTiO or N -perturbed OTiO [10], while the
amount of NTiO remains constant (Fig. 2). This
shows that this species is formed only by addition
of NO to NTiO and could be compared to the
which is not readily overcome at a low tempera-
ture (ꢀ10 K here).
In this paper, we shall focus our study on ad-
ditional information relative to the oxonitride
primary insertion product, NTiO. Spectra ob-
tained with di€erent concentrations of NO and Ti
in argon matrices show two sharp bands located at
2
secondary reaction of Ti with CO
insertion step leading to OTiCO, Mascetti et al.
showed that further reaction with CO proceeds
2
: after an initial
�
1
718.3 and 900.7 cm (Figs. 2 and 3) previously
observed and assigned by Kushto et al. [10] to
Ti±N and Ti±O stretching modes, respectively, for
the NTiO species. Their relative infrared intensities
remain constant when the sample is photolysed
or after warming the matrix up to 35 K. Sharp
satellite features near these bands are due to the
2
through a complexation channel only to yield
OTi(CO)(CO ) [11,13]. Experiments were run us-
2
ing Ti ‡ O ‡ N ternary mixtures in argon, which
2
2
2
resulted in the production of N -perturbed OTiO
after Ti photoexcitation, but no trace of OTiNNO
could then be detected. From this, we can con-
46
47
natural titanium isotopes ( Ti 7.9%, Ti 7.3%,
clude that the OTiNNO ! N
2
-perturbed OTiO
isomerization is thermodynamically favoured. The
OTiNNO product yield was too low in our ex-
periments to observe more than the three most
intense fundamentals, thus, precluding a serious
structural discussion for this species.
Fig. 3. Comparison of experimental and simulated infrared
2
spectra in the m region of NTiO trapped in argon matrix at 10
K. The simulations assume lorentzian band shapes. For both
the stable (± ± ±) and unstable (Á Á Á) trapping sites, the relative
intensities are supposed to follow the population of naturally
occurring Ti isotopes. The dotted±dashed line represents the
total simulated pro®le.
Fig. 2. Comparison of infrared spectra in the Ti±O stretching
2
region for Ti ‡ NO and Ti ‡ O reactions in solid argon at 10
K, before and after broad band UV±visible photoexcitation.

L. Krim et al. / Chemical Physics 254 (2000) 267±274
271
2
0
Fig. 4. Infrared spectra of m
isotopic precursors at 10 K in solid argon, 690±730 and 200±300
2 3
and m of A NTiO for various
Fig. 5. Infrared spectra in the TiO stretching region (m
1
of
�
1
NTiO and m of OTiO) for various precursors of NO reacting
3
cm regions, Ti/NO/Ar ˆ 0.5/1/100. The dotted lines corre-
spond to blank samples with NO/Ar ˆ 1/100.
with ground state Ti atoms in solid argon.
4
8
49
50
� 1
Ti 74.0%, Ti 5.5% and Ti 5.3%), see Fig. 3.
Near each of these sharp bands, a weaker blue-
evaluated at )2.9 and )5.0 cm only for the Ti±N
and Ti±O modes, respectively.
In the far-infrared region, one absorption band
�
1
shifted band at 726 and 910 cm can be observed,
which correlate with the main bands when the Ti or
NO concentrations vary, and appear as doublets in
the spectra obtained with NO isotopic mixtures
�
1
has been observed at 248.2 cm (Fig. 4) which
correlated with the NTiO bands, and could be
attributed to the bending mode NTiO. The bands
are broadened by inhomogeneous e€ects
1
4
16
15 16
14 16
14 18
(
N O/ N O or N O/ N O). They remain
�
1
una€ected by the photolysis e€ects but disppear
completely when the matrix is heated. These bands
are thus probably due to a site e€ect of NTiO in
the matrix (Figs. 3±5). The absorptions observed
for the di€erent isotopic species of NTiO are listed
in Table 1, for the main, stable trapping site.
Two new, weaker bands at 1430.7 and 1791.5
(FWHM ꢀ 2 cm ), and no titanium isotopic
structure could be observed for this band, even
�
1
with spectra recorded at a 0.1 cm resolution.
When the matrix is annealed, the NO dimerizes
readily, and several bands are associated with ti-
tanium reactions with (NO)
action (Ti ‡ 2NO) products such as TiO
(N ) formed in these experiments do not originate
2
. The secondary re-
2
, TiO
2
�
1
cm can be observed (Fig. 6), which correlate
2
�
1
with the two strong bands at 700 and 900 cm ,
over any change in experimental conditions (tem-
perature, concentrations, NO isotopic e€ects).
From the observed isotopic e€ects and relative
intensities, these bands correspond to overtones of
the bands located in Ti±N and Ti±O stretching
spectral region, respectively. Their integrated in-
tensities are much smaller than those of the fun-
from sample contamination by molecular oxygen,
but from secondary insertion and recombination,
as shown by isotopic labelling of these species
18
when O-labelled NO is used (see the triplet iso-
16
18
topic structure of TiO
ture is used, which is indicated by dotted lines in
Fig. 5). In addition to OTiN O, which has been
discussed earlier, bands located at 1125.1 and
when a N O/N O mix-
2
2
�
1
2
damental bands (ImTi±N/I2mTi±N ˆ 350 and ImTi±O
/
1581.5 cm could not be attributed to Ti±(n -
1
14
15
I
2mTi±O ˆ 200), and their isotopic shifts ( N/ N,
2 2
NO) and Ti±(n -NO) products, respectively, as
1
6
18
O/ O) are nearly twice those observed for the
suggested in Ref. [10]. These bands were only ob-
served as traces in some of our experiments, and
their relative infrared intensity grew markedly
corresponding fundamental bands. From
knowledge of fundamental and overtone frequen-
a
cies, anharmonicity corrections (x x ) have been
2
when O was added voluntarily in the samples.
e
e

2
72
L. Krim et al. / Chemical Physics 254 (2000) 267±274
Table 1
Observed frequencies in cm of NTi O, NTi O and NTi O molecules isolated in argon matrix (relative intensities are in
parentheses)
�
1
14
16
15
16
14
18
a
14
16
15
16
14
18
Ti isotopic structure
NTi
O
NTi
O
NTi
O
Assignment
b
Ti
248.2
718.3
244.8
701.4
243.0
716.1
m
m
3
(0.3) bending mode
(0.96) ``Ti±N'' stretching mode
48
46
49
48
46
49
Ti
2
47
Ti, Ti
50
722.2, 720.2
716.5, 714.7
900.7
705.4, 703.3
699.5, 697.7
898.2
720.1, 718.3
714.1, 712.4
865.8
Ti, Ti
Ti
47
m
1
(1) ``Ti±O'' stretching mode
Ti, Ti
50
Ti, Ti
904.9, 902.7
898.7, 896.9
1430.7
902.4, 900.4
896.3, 894.3
1397.0
870.1, 867.9
863.8, 861.8
1426.4
b
� 3
Ti
Ti
2m
2m
2
(2:7 Â 10
)
b
c
� 3
1791.5
1785.9
±
1
(5:4 Â 10
)
a
b
c
46
47
48
49
50
The natural titanium isotopic distribution is Ti: 7.9%, Ti: 7.3%, Ti: 74.0%, Ti: 5.5% and Ti: 5.3%.
Titanium isotopic e€ects are unresolved.
Overlap with NO dimer absorption.
place without any energy activation, to give the
NTiO species. A similar situation has been ob-
served and discussed in the reaction of Ti atom
2
with CO molecules [9].
The results of the most recent theoretical studies
10], which have calculated the three harmonic
[
2
0
frequencies for NTiO on the ground A state, are
given in Table 2. The agreement with the observed
frequencies is reasonable, but if the isotopic shifts
14
18
15
16
(
N O and N O) are very well reproduced for
m
2
and m modes, for the m mode the error on the
3
1
calculated isotopic shifts exceeds 40% (Table 2).
The computed IR intensities for each mode are in
qualitative agreement with the experimental rela-
tive intensities. The TiN and TiO stretching modes
have comparable infrared intensities, but the low
frequency bending mode is twice as intense as
predicted.
Since the anharmonicity corrections are indeed
very small for this molecule, semi-empirical har-
monic force-®eld calculations were performed to
test the information about the predicted bonding
properties and molecular shape of NTiO. First,
based on the equilibrium geometry calculated with
2
0
1 2
Fig. 6. Infrared spectra of the 2m and 2m regions of A NTiO
at 10 K in solid argon. For all samples, Ti/NO/Ar ˆ 0.5/1/100.
The stronger absorption near 1789 cm corresponds to unre-
�
1
15
18
acted
N O.
�
1
Nevertheless, two bands at 1100 and 1531 cm
increasing on annealing and disappearing with
photolysis could be assigned to Ti(NO)2 com-
plexes. They exhibit triplets with NO isotopic
14
16
15
16
14
16
14
18
mixtures ( N O/ N O or
N O/ N O),
which denotes a vibration of a species involving
two equivalent NO bonds.
ꢁ
TiO
DFT BP86 method in Ref. [10] (r ˆ 1:661 A,
ꢁ
r
TiN ˆ 1:715 A and \NTiO ˆ 105:8°), we searched
the set of force constants which yielded an ac-
ceptable agreement with the experimental results,
better than the ab initio force ®eld. Next, to test
the geometry, we varied the NTiO bond angle
stepwise between 180° and 80° (the bond lengths
are thus kept constant). For each given angle, the
4. Discussion and vibrational analysis
The products of cocondensation of nitric oxide
with Ti, as described previously, showed that the
insertion reaction of Ti into the NO bond takes

L. Krim et al. / Chemical Physics 254 (2000) 267±274
273
Table 2
Observed and calculated frequencies (cm ) of the various isotopic species of NTiO ( A )
a
� 1
b
2
0
m
3
(bending) m (Ti±N stretch)
2
1
m (Ti±O stretch)
a
Observed
Calculated
Observed
Calculated
Observed
Calculated
1
4
5
48
48
16
16
N
N
Ti
Ti
O
O
248.2
244.8
3.4
246.5
243.1
3.4
718.3
701.3
17.0
761.1
744.4
16.7
900.7
898.2
2.5
936.3
932.2
4.1
1
Isotopic shifts
a
DFT calculation BP86 [10].
b
ꢁ
ꢁ
The geometric parameters of the NTiO are rTiO ˆ 1.661 A, rTiN ˆ 1.715 A and \NTiO ˆ 105.8°.
best set of potential constants is searched for,
giving the best overall agreement with the ex-
show that the TiN and TiO coordinates are only
slightly coupled in the m and m modes (roughly a
1
2
14
16
15
16
perimental isotopic shifts ( N O/ N O and
16
90% TiO±10% TiN coordinate combination for
the m mode and a reverse combination for the m2
14
14
18
N O/ N O) on all three m , m and m modes.
1
2
3
1
Fig. 7 shows the standard deviations between the
experimental and calculated isotopic shifts (nor-
malized to 100 cm ), as a function of the bond
angle value. It is clear that the best results are
obtained for a bond angle in the vicinity of that
calculated by quantum chemical method
3
mode) while the m mode involves an almost pure
bending motion. Our results also con®rm that the
titanium isotopic structure is well within the
�
1
3
bandwidth for the m mode.
Interestingly, the TiN bond force constant is
found to be notably (ꢀ)40%) weaker than that of
the TiO bond, thus, providing a clue for the pre-
ferred O-atom abstraction channel in the gas phase
as deduced in the studies of Ritter [5], Campbell [7]
and Naulin [4]:
ꢃ
ꢃ
(
NTiO force constants are of the order of 3.56,
5
\NTiO ˆ 110 Æ 5 ) (Table 3). The TiN, TiO and
ꢁ ꢁ
� 1 � 2
.61 mdyn A and 0.6 mdyn A rad , respectively.
Calculations of the potential energy distribution
Ti ‡ NO$NTiO!TiO ‡ N; E ˆ DH ‡ E :
a
1
2
1
2
As discussed in Refs. [4,7], the total activation
energy E in the gas phase is small and positive
ꢀ3.6 kJ/mol), while the enthalpy change DH is
negative (the ®rst step is exothermic) and E not
a
(
1
2
much larger in absolute value. In the condensed
phase at low temperature, the reaction is quenched
at the ®rst step, most likely the internal energy of
the NTiO insertion product is too small to over-
come the second step activation energy, E which
2
a
is thus notably higher than E , and reach the ab-
straction channel. These results cast a serious
doubt on the likelihood of a harpooning, or elec-
tron-jump mechanism as previously discussed for
the gas phase reaction.
5
. Conclusions
Fig. 7. Standard deviations on the di€erence between experi-
Reaction of ground state Ti atoms, produced
by thermal e€usion, and NO molecules at a
low temperature in solid argon shows that the
1 2 3
mental and calculated isotopic shifts on the m , m and m modes,
�
1
normalized to 100 cm . The error bars represent the experi-
mental uncertainties on the isotope e€ects.

2
74
L. Krim et al. / Chemical Physics 254 (2000) 267±274
Table 3
Observed and calculated frequencies (cm ) of the various isotopic species of NTiO
a
� 1
b
NTiO
m
3
m
2
m
1
Observed
Calculated
Observed
Calculated
Observed
Calculated
14
14
14
14
14
48
46
47
49
50
16
16
16
16
16
N
N
N
N
N
Ti
Ti
Ti
Ti
Ti
O
O
O
O
O
248.2
c
248.1
249.3
248.7
247.6
247.0
718.3
722.2
720.2
716.5
714.7
718.5
722.4
720.4
716.6
714.8
900.7
904.9
902.7
898.7
896.9
900.6
904.7
902.6
898.6
896.9
±
c
±
c
±
c
±
15
15
15
15
15
48
46
47
49
50
16
16
16
16
16
N
N
N
N
N
Ti
Ti
Ti
Ti
Ti
O
O
O
O
O
244.8
c
244.5
245.7
245.1
244.0
243.4
701.4
705.4
703.3
699.5
697.7
701.4
705.4
703.4
699.5
697.7
898.2
902.4
900.4
896.3
894.3
898.1
902.3
900.2
896.1
894.2
±
c
±
c
±
c
±
14
14
14
14
14
48
46
47
49
50
18
18
18
18
18
N
N
N
N
N
Ti
Ti
Ti
Ti
Ti
O
O
O
O
O
243.0
c
243.1
244.3
243.7
242.6
242.1
716.1
720.1
718.3
714.1
712.4
715.9
720.0
717.9
714.0
712.1
865.8
870.1
867.9
863.8
861.8
865.4
869.6
867.3
863.3
861.4
±
c
±
c
±
c
±
a
The force constants are FTiN ˆ 3.56 mdyn A , FTiO ˆ 5.61 mdyn A , FNTiO ˆ 0.60 mdyn A rad , FTiN;TiO ˆ 0.95 mdyn Aꢁ � 1
ꢁ
� 1
ꢁ
� 1
ꢁ
� 2
,
ꢁ
� 1
ꢁ
� 1
.
F
b
TiO;NTiO ˆ 0.0 mdyn A rad , FTiN;NTiO ˆ 0.15 mdyn A rad
ꢁ
ꢁ
The geometric parameters of the NTiO are rTiO ˆ 1.661 A, rTiN ˆ 1.715 A and \NTiO ˆ 110°.
c
Ti isotopic e€ects are unresolved.
[
6] D. Ritter, J.C. Weisshaar, J. Phys. Chem. 94 (1990) 4907.
7] M.L. Campbell, R.E. McClean, J. Phys. Chem. 97 (1993)
942.
8] D.E. Clemmer, K. Homma, I. Koyano, J. Phys. Chem. 97
1993) 11480.
[9] C. Blanchet, H.A. Duarte, D.S. Salahub, J. Chem. Phys.
06 (1997) 8779.
Ti ‡ NO ! NTiO takes place with no or very little
activation energy. No metastable titanium mono-
nitrosyl species could be detected under these
conditions in contrast with late row transition
metal atoms such as Ni. In this study, the detection
of the third, bending fundamental and two over-
tone levels have completed our knowledge of the
IR spectrum of the NTiO species. A normal co-
ordinate analysis based on complete vibrational
[
7
[
(
1
[
10] G.P. Kushto, M.F. Zhou, L. Andrews, C.W. Bauschlicher,
J. Phys. Chem. A 103 (1999) 1115.
[
[
11] J. Mascetti, M. Tranquille, J. Phys. Chem. 92 (1988) 2177.
12] G. Chertihin, L. Andrews, J. Am. Chem. Soc. 117 (1995)
1595.
13] I. P ꢀa pai, J. Mascetti, R. Fournier, J. Phys. Chem. A 101
(1997) 4465.
14
16
data for several isotopic species ( N Ti O,
48
1
5
16
14
18
46
47
[
N Ti O and N Ti O, with the Ti/ Ti/ Ti/
50
49
Ti/ Ti naturally occurring isotopes) con®rms the
result previous ab initio predictions of the molec-
ular shape and spectral properties.
[
14] L. Krim, L. Manceron, M.E. Alikhani, J. Phys. Chem. A
103 (1999) 2592.
[
[
[
15] H.A. Joly, L. Manceron, Chem. Phys. 226 (1998) 61.
16] B. Tremblay, L. Manceron, Chem. Phys. 242 (1999) 235.
17] L. Manceron, M.E. Alikhani, Chem. Phys. 228 (1999) 73.
References
[18] V. Jonas, W. Thiel, J. Phys. Chem. A 103 (1999) 1381.
[19] M. Rosi, C.W. Bauschlicher, G.V. Chertihin, L. Andrews,
[
[
[
1] L.H. Dubois, J.L. Gole, J. Chem. Phys. 66 (1977) 779.
2] J.L. Gole, D.R. Preuss, Chem. Phys. 19 (1977) 59.
3] J.M. Parson, L.C. Geiger, T.J. Conway, J. Chem. Phys. 74
Theor. Chem. Acc. 99 (1998) 106.
[20] Y.K. Lee, L. Manceron, I. P ꢀa pai, J. Phys. Chem. A 101
(1997) 9650.
(
1981) 5595.
4] C. Naulin, I.M. Hedgecock, M. Costes, Chem. Phys. Lett.
66 (1997) 335.
5] D. Ritter, J.C. Weisshaar, J. Phys. Chem. 93 (1989) 1576.
[21] K. Clusius, Angew. Chem. 66 (1954) 497.
[22] G.K. Ruschel, T.M. Nemetz, D.W. Ball, J. Mol. Struct.
384 (1996) 101.
[
[
2
[23] L. Krim, N. Lacome, J. Phys. Chem. A 102 (1998) 2289.