A new synthetic method and solution equilibria for the chlorotitanate(IV) anions - Evidence for the existence of a new species: [Ti2Cl 11]3-

Article abstract of DOI:10.1002/ejic.200400293

TiCl₃ is oxidized by SOCl₂ in the presence of the appropriate amount of [NEt₃Bz]Cl or [PPN]Cl [Bz = benzyl; PPN = bis(triphenylphosphanyl)iminium] to afford the corresponding salts of [Ti ₂Cl₉]^-, [Ti₂Cl₁₀] ^2- or [TiCl₆]^2-. The results of cyclic voltammetric and solution IR studies in the Ti-Cl stretching region are interpreted in terms of a rapid chloride-redistribution equilibrium between [Ti₂Cl₁₀]^2- on one side and a mixture of [Ti₂Cl₉]^- and a previously unreported [Ti ₂Cl₁₁]^3- species on the other side. In the solid state, however, a compound with the [PPN]₃[Ti ₂Cl₁₁] stoichiometry exists as a mixture of [PPN] ₂[Ti₂Cl₁₀] and [PPN]₂[TiCl ₆]. The PPN salt of [Ti₂Cl₉]^- has been structurally characterized by X-ray diffractometry. Wiley-VCH Verlag GmbH & Co. KGaA, 69451 Weinheim, Germany, 2004.

Full text of DOI:10.1002/ejic.200400293

FULL PAPER
A New Synthetic Method and Solution Equilibria for the Chlorotitanate(IV
)
Anions ؊ Evidence for the Existence of a New Species: [Ti2Cl11]3؊
[a]
[b]
[b]
[a]
´
Emmanuel Robe, Jean-Claude Daran, Sandrine Vincendeau, and Rinaldo Poli*
Keywords: Anions / Electrochemistry / Infrared spectroscopy / Titanium
TiCl3 is oxidized by SOCl2 in the presence of the appropriate
amount of [NEt3Bz]Cl or [PPN]Cl [Bz = benzyl; PPN = bis(tri-
cies on the other side. In the solid state, however, a
compound with the [PPN]3[Ti2Cl11] stoichiometry exists as a
phenylphosphanyl)iminium] to afford the corresponding salts mixture of [PPN]2[Ti2Cl10] and [PPN]2[TiCl6]. The PPN salt of
2−
of [Ti2Cl9]−, [Ti2Cl10
]
or [TiCl6]2−. The results of cyclic vol- [Ti2Cl9]− has been structurally characterized by X-ray diffrac-
tammetric and solution IR studies in the Ti−Cl stretching re- tometry.
gion are interpreted in terms of a rapid chloride-redistribu-
2−
tion equilibrium between [Ti2Cl10
]
on one side and a mix- ( Wiley-VCH Verlag GmbH & Co. KGaA, 69451 Weinheim,
ture of [Ti2Cl9]− and a previously unreported [Ti2Cl11
]
spe- Germany, 2004)
3−
Introduction
Homoleptic chlorometallates are a wide class of funda-
mentally important coordination compounds. They are in
general easily accessible by addition of the required amount
of halide ions to the neutral metal halide. For TiIV, four
different anionic complexes are known, the enneachlorodi-
titanate [Ti2Cl9]Ϫ, the decachlorodititanate [Ti2Cl10]2Ϫ, the
pentachlorotitanate [TiCl5]Ϫ, and the hexachlorotitanate
[TiCl6]2Ϫ, formally related to TiCl4 by the addition of one
half, one, or two ClϪ ions per titanium, respectively
(Scheme 1).[1] The choice between mononuclear or di-
nuclear pentachlorotitanate seems to be controlled by the
size of the counterion.[2] Since TiCl3 is used extensively as
a precursor of ZieglerϪNatta polymerization catalysts,[3Ϫ8]
its coordination chemistry has been widely explored. How-
ever, the only reported chlorotitanate() anions are appar-
ently the edge-sharing bioctahedral [Ti2Cl9]3Ϫ and the
Scheme 1
chloride salts in solution. Given the hydrolytic sensitivity of
TiIII, our approach consists of the use of SOCl2 as solvent,
since this is a well-known medium for ensuring anhydrous
conditions for transition metal chlorides.[15] The results of
this interaction, yielding metal oxidation, are reported here.
Furthermore, we present new information on the solution
behavior of the [Ti2Cl10]2Ϫ ion and spectroscopic and elec-
trochemical evidence for the existence of a previously unre-
ported [Ti2Cl11]3Ϫ ion in solution.
mononuclear octahedral [TiCl6]3Ϫ [9Ϫ11]
Furthermore, these
.
ions are prepared at high temperatures in the solid state (no
solvent),[12,13] and therefore little information is available on
their properties in solution.
In an attempt to gain access to soluble versions of these
anionic complexes and perhaps to prepare the as-yet un-
known[14] decachlorodititanate() tetraanion [Ti2Cl10]4Ϫ
,
we have investigated the interaction between TiCl3 and Results and Discussion
Syntheses
[a]
`
`
d’Electrosynthese
Laboratoire
de
Synthese
et
´
´
Organometalliques (LSEO UMR 5632), Universite de
The addition of one equivalent of chloride, either as the
triethylbenzylammonium (Et3BzN) or the bis(triphenylpho-
sphanyl)iminium (PPN) salt, to TiCl3 in SOCl2 results in a
smooth reaction at room temperature accompanied by a
sharp color change from dark purple to yellow. The color is
´
Bourgogne, Faculte de Sciences ‘‘Gabriel’’,
6 boulevard Gabriel, 21000 Dijon, France
Fax: (internat.) ϩ33-5-61553003
E-mail: poli@lcc-toulouse.fr
[b]
Laboratoire de Chimie de Coordination (LCC UPR 8241),
205 Route de Narbonne, 31077 Toulouse Cedex, France
4108
 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
DOI: 10.1002/ejic.200400293
Eur. J. Inorg. Chem. 2004, 4108Ϫ4114

Solution Equilibria for Chlorotitanate(IV) Anions
FULL PAPER
indicative of chloro complexes of TiIV. Indeed, the selective only a reversible reduction that has been attributed to the
product of the reactions are the decachlorodititanate() [TiCl6]2Ϫ ion,[19] whereas Lewis-acidic melts show less re-
salts [Equation (1)]. The products were obtained in a pure versible features that are attributed to less chloride-rich
state and in good yields. They are moderately soluble in anions, whose nature depends on the melt composition.[20]
dichloromethane, the PPN salt being more soluble than the It is reasonable to expect that the reduction process to TiIII
ammonium salt. The dinuclear nature of these salts is dem- correlates with the effective charge at the metal center, a
onstrated by their IR properties (vide infra).
greater electron density leading to a more difficult reduction
process. Therefore, one could reasonably expect a trend of
reduction potentials in the order [TiCl6]2Ϫ Ͻ [Ti2Cl10]2Ϫ
Ͻ [Ti2Cl9]Ϫ.
(1)
A cyclic voltammetric analysis of [Ti2Cl10]2Ϫ in CH2Cl2
shows that the TiIV center undergoes the expected reduction
process, which is irreversible. This irreversibility is probably
linked to loss of chloride ions upon reduction, induced by
the accumulation of too much negative charge and by the
possible rearrangement to a triply bridged, face-sharing bi-
octahedral [Ti2Cl9]2Ϫ species. Such a mixed-valence com-
plex is apparently not yet known. The surprising obser-
vation, however, is the complexity of the reduction wave
(see Figure 1), which shows at least two distinct processes,
the main one having a cathodic peak potential, Ep,c, of
Ϫ1.03 V, with a broad shoulder between aboutϪ0.8 and
Ϫ0.6 V.
By changing the stoichiometry to one or four equivalents
of chloride salt per two of TiCl3, the ennachlorodititan-
ate() [Ti2Cl9]Ϫ, and the hexachlorotitanate() [TiCl6]2Ϫ
were obtained in good yields [Equation (2) and Equa-
tion (3), respectively]. We have also obtained a product with
the stoichiometry [PPN]3[Ti2Cl11] by using a 3:2 PPNCl/
TiCl3 ratio. The solid state and solution nature of these
products will be discussed in the next section. The necessary
chlorine atoms for the stoichiometries of Equation (1),
Equation (2) and Equation (3) are undoubtedly provided by
the thionyl chloride solvent, which must therefore be re-
duced. The oxidizing properties of thionyl chloride relative
to low-valent metal halides do not appear to have been ex-
tensively explored. A metal oxidation process has been wit-
nessed for the dehydration of SnCl2·2H2O (leading to SnCl4
while elemental sulfur was found amongst the reduction
products), whereas the dehydration of TlCl3·xH2O, which
is affected by a spontaneous reduction to TlCl when the
salt is thermally dehydrated, leads to pure TlCl3 when using
SOCl2.[15] Another potentially reducing metal chloride is
CrCl2, although the dehydration of the corresponding hy-
drate with SOCl2 has apparently never been described[16]
(whereas it is known with COCl2).[15] Unlike the above-
mentioned dehydration of SnCl2, we have not witnessed the
formation of any elemental sulfur. It is known that SOCl2
decomposes at high temperatures to yield Cl2, SO2 and
S2Cl2,[17,18] but, remarkably, the Ti oxidation process that
we report here occurs smoothly at room temperature. We
have not explored the nature of the reduced sulfur products
in this reaction further.
Figure 1. Cyclic voltammograms of solutions of [NBzEt3]2[Ti2Cl10]
alone (thicker line) and with two equiv. of [NBzEt3]Cl (thinner
line); both solutions also contain the internal ferrocene standard
(reversible wave at E1/2 ϭ 0); solvent: CH2Cl2
(2)
(3)
The possibility that the complexity of this voltammetric
response could result from sample impurities was soon ex-
cluded by the reproducibility when measuring several
samples obtained from different syntheses. The key to
understanding this complexity comes from the voltam-
metric measurement after the addition of excess chloride
ions, which should quantitatively convert the electroactive
Electrochemical and Spectroscopic Characterization
Still with the goal in mind of exploring the solution species to the [TiCl6]2Ϫ ion (see Scheme 1), according to the
chemistry of the chlorotitanate() complexes, we have car- literature. The single irreversible reduction wave at Ep,c
ϭ
ried out electrochemical investigations of the chlorotitan- Ϫ1.13 V that is obtained under these conditions (see Fig-
ate() anions. Previous electrochemical investigations on ure 1) may therefore be attributed to the reduction of this
TiIV chloride systems appear to be limited to TiCl4 in ion. Note that this wave is irreversible, whereas full reversi-
AlCl3/chloride melts.[19,20] Basic melts (Cl/Al Ͼ 4) exhibit bility is observed in an AlCl3/chloride melt.[19] This obser-
Eur. J. Inorg. Chem. 2004, 4108Ϫ4114
www.eurjic.org
 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
4109

´
E. Robe, J.-C. Daran, S. Vincendeau, R. Poli
FULL PAPER
vation agrees fully with the likely instability of a reduced
[TiCl6]3Ϫ ion towards chloride-ion loss. This wave occurs at
a very similar potential as the main wave of the [Ti2Cl10]2Ϫ
solution in the absence of free chloride ions. The second
species is reduced at a less negative potential, therefore it
must contain a higher effective positive charge on the metal
center (lower Cl/Ti ratio). A logical hypothesis is to assign
this wave to the [Ti2Cl9]Ϫ ion. Indeed, the cyclic voltammo-
gram for a solution prepared from the isolated [Ti2Cl9]Ϫ
salt shows an irreversible wave with Ep,c ϭ Ϫ0.76 V, corre-
sponding closely to the shoulder of the [Ti2Cl10]2Ϫ solution
in the absence of free chloride ions. Therefore, we conclude
that a chloride-dissociation equilibrium exists in solution
starting from the dianionic decachloro species, according to
Equation (4). The chloride ions that are freed up in solution
can then break the bridge bonds of the decachloro species
to yield a more chloride-rich species. It would seem reason-
able to assume that this species is the mononuclear dianion
[TiCl6]2Ϫ. However, the results obtained from a parallel IR
investigation tell a different story.
Figure 2. Infrared spectra of various chlorotitanate() PPN salts
in CH2Cl2 solution (Ti concentration ϭ 2 ϫ 10Ϫ2 ); the starred
peak corresponds to an oxo impurity (see text)
salt).[2] A band in this region is indeed observed in the solu-
tion spectrum of [PPN]2[Ti2Cl10] (maximum at 370 cmϪ1).
The solid-state spectrum of [PPN]2[Ti2Cl10] (see a in Fig-
ure 3), is characterized by a strong band at around 365
cmϪ1, in-between those previously reported for the mono-
and dinuclear ions, but quite close to the major band that
is observed in solution. The overall shape of the two spec-
tra, in dichloromethane solution and in the solid state, is
very similar. We conclude that the compound structure is
most certainly dinuclear both in solution and in the solid
state. Thus, the rule stating that the pentachlorotitanate()
is mononuclear with large cations does not appear to have
general validity, since PPNϩ is perhaps the largest cation
with which this anion has ever been crystallized.
(4)
An independent investigation of the solution equilibria is
complicated by the absence of available spectroscopic pro-
bes such as spin-1/2 nuclei for an NMR investigation or
unpaired electrons for an EPR investigation. The use of 47Ti
and 49Ti NMR spectroscopy is unsatisfactory for com-
pounds where the Ti center has lower than cubic sym-
metry.[20,21] A suitable solution is provided by infrared spec-
troscopy in the low wavenumber region, which is sensitive
to the TiϪCl stretching vibrations. Infrared studies of chlo-
rotitanate ions have been published previously, but these
have mostly been confined to solid samples[2,22Ϫ24] as the
solubility of most available salts (e.g. with Kϩ, Rbϩ, Csϩ,
Et2NH2ϩ, and Et4Nϩ)[23] is too low. The PPN salts re-
ported here are sufficiently soluble in dichloromethane to
yield reasonable intensities (A Ͼ 1 for some bands in
0.5 mm pathlength polyethylene cells, see Figure 2). It
should be noted that the PPNϩ ion does not interfere in the
investigated region, since it shows only a very weak band at
464 cmϪ1
.
Figure 3. Solid state (nujol mull) infrared spectra of: (a)
[PPN]2[Ti2Cl10], (b) [PPN]3[Ti2Cl11], and (c) [PPN]2[TiCl6]
First, we shall comment on the dichotomy between
[Ti2Cl10]2Ϫ and its mononuclear version, the pentachlorotit-
anate() anion [TiCl5]Ϫ,[2,22,25Ϫ27] and then on the species
The spectra of the [Ti2Cl9]Ϫ and [TiCl6]2Ϫ salts, also
with a different Cl/Ti ratio. Previous studies have shown shown in Figure 2, are quite different from that of the
that, in the solid state, the pentachlorotitanate() salts fea- [Ti2Cl10]2Ϫ ion, demonstrating the suitability of IR spec-
ture a dichloro-bridged bioctahedral geometry for the anion troscopy in this spectral region for the analysis of the spec-
in the presence of small cations, whereas salts of larger cat- ies present in solution. Complex [Ti2Cl9]Ϫ has a very strong
ϩ
ions are mononuclear.[2] The PCl4 salt has been shown to band centered at 428 cmϪ1 and a strong band at 390 cmϪ1
.
undergo a temperature-dependent monomer-dimer equilib- The shoulder on the right-hand side of the latter band may
rium.[22] The solid-state spectrum of the mononuclear salt be due to minor amounts of the [Ti2Cl10]2Ϫ salt. Complex
is characterized by a very strong band at 348 cmϪ1, which [TiCl6]2Ϫ, on the other hand, is characterized by a strong
is absent from the solution spectrum of [PPN]2[Ti2Cl10]. On and broad band centered at 327 cmϪ1. This corresponds to
the other hand, the solid-state spectra of dinuclear anions the frequency reported in the literature for the IR-active
are characterized by a very strong band at around 380 cmϪ1 stretching vibration (ν3, T1u) of this ion (ca. 330 cmϪ1 de-
(385 cmϪ1 for the NEt4 salt and 377 cmϪ1 for the PCl4 pending on the counterion).[23,24,28] It is worth noting the
4110
 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjic.org
Eur. J. Inorg. Chem. 2004, 4108Ϫ4114

Solution Equilibria for Chlorotitanate(IV) Anions
FULL PAPER
weak band observed at 392 cmϪ1. Only two IR-active bands from [TiCl6]2Ϫ to [Ti2Cl11]3Ϫ (327 to 355 cmϪ1) suggests a
(the ν3 stretching and the ν4 bending, both of T1u type) are force-constant increase upon reduction of the metal effec-
expected for the Oh [TiCl6]2Ϫ ion. According to the litera- tive charge. Although the weak band at 392 cmϪ1 is ident-
ture, these are located around 330 cmϪ1 and 183 cmϪ1
,
ical to that observed for the [TiCl6]2Ϫ solution, and is there-
respectively. Therefore, the weak band at 392 cmϪ1 cannot fore attributable to minor amounts of a hydrolytic product,
belong to this ion. We believe that this band is due to oxo its attribution to a genuine normal mode of the [Ti2Cl11]3Ϫ
impurities deriving from a small amount of hydrolysis, cannot be excluded. Once again, this spectrum was found
given the extreme moisture sensitivity of the TiIVϪCl to depend highly on the rigor with which the solution was
bonds. In support of this hypothesis, we have observed that prepared, and the band at 392 cmϪ1 could never be com-
the relative intensity of this band changes as a function of pletely eliminated. The reported spectrum was obtained
the care with which the solutions are handled and increases from a solution prepared in silylated glassware and in care-
when undried dichloromethane is used as solvent, but we fully dried dichloromethane, from a solid sample that was
have been unable to completely eliminate it. The reported obtained by direct synthesis in SOCl2, according to the stoi-
spectrum was recorded in carefully dried solvent and all chiometry of Equation (6).
glassware and transfer syringes were silylated by rinsing
with a 10% v/v chloroform solution of Me3SiCl. It is pos-
sible that some hydrolysis occurred during the isolation pro-
cedure (washings and transfer into the storage ampoules),
as suggested by the observation of the same weak band in
(5)
(6)
the IR spectrum of the solid sample (see c in Figure 3).
Following this consideration, the band observed at about
390 cmϪ1 for the [Ti2Cl10]2Ϫ solution may also be due to a
product of partial hydrolysis.
The spectrum of the [TiCl6]2Ϫ salt was also reproduced
(except for the variable relative intensity of the 392 cmϪ1
band as outlined above) upon addition of Ն 2 equivalents
of PPNCl to the solution of [Ti2Cl10]2Ϫ. The addition of
just one equivalent of PPNCl to the same solution, on the
other hand, shows the complete disappearance of the bands
of [Ti2Cl10]2Ϫ and the appearance of a very strong band
centered at 355 cmϪ1 (Figure 4). It should be noted that the
band of [TiCl6]2Ϫ is also absent from this spectrum. This
suggests the formation of a complex having the stoichi-
ometry [Ti2Cl11]3Ϫ. Addition of PPNCl to this solution gen-
An additional interesting feature is shown by the IR spec-
trum of the [PPN]3[Ti2Cl11] compound in the solid state
(see b in Figure 3). This is quite clearly the superposition
of the spectra of the [PPN]2[Ti2Cl10] and [PPN]2[TiCl6] salts
that are shown in the same figure, whereas the band ob-
served for the same compound in solution is absent. This
observation demonstrates that the species [Ti2Cl11]3Ϫ exists
only in solution, and the crystallization process forces a li-
gand redistribution to yield [Ti2Cl10]2Ϫ and [TiCl6]2Ϫ. In
other words, the ligand-redistribution equilibrium shown in
Equation (7) lies on the left in dichloromethane solution
and on the right in the solid state for PPN salts.
erated again the spectrum of [TiCl6]2Ϫ
.
(7)
As a final relevant observation, note that the solution
spectrum of the [Ti2Cl10]2Ϫ salt (Figure 2) does not show a
significant absorption at 355 cmϪ1, thus excluding the pres-
ence of large amounts of [Ti2Cl11]3Ϫ at equilibrium. A
shoulder on the low-frequency side of the 368 cmϪ1 band
signals the possible presence of minor amounts of
[Ti2Cl11]3Ϫ, which could be related to an imprecise stoichi-
ometry during the synthesis, or to a chloride-redistribution
equilibrium as shown in Equation (8). In the latter case, the
equilibrium must necessarily be shifted almost completely
towards the [Ti2Cl10]2Ϫ ion.
Figure 4. Infrared spectrum of
a
CH2Cl2 solution of
[PPN]3[Ti2Cl11] (Ti concentration ϭ 2 ϫ 10Ϫ2 )
The [Ti2Cl11]3Ϫ anion does not appear to have been re-
ported in the literature. Its structure most likely corre-
sponds to a corner-sharing bioctahedron [Equation (5)]. On
the basis of VSEPR theory, the geometry around the bridg-
ing Cl atom should be bent and the IR spectrum can be
predicted to be quite similar to that of mononuclear octa-
(8)
In the light of the IR study, we can now re-interpret the
hedral [TiCl6]2Ϫ if vibrational coupling across the Cl bridge result of the cyclic voltammetric experiment on the
is weak. Thus, the slight shift to higher frequency on going [Ti2Cl10]2Ϫ salt by assigning the wave at Ϫ1.03 V to the
Eur. J. Inorg. Chem. 2004, 4108Ϫ4114
www.eurjic.org
 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
4111

´
E. Robe, J.-C. Daran, S. Vincendeau, R. Poli
FULL PAPER
reduction of the [Ti2Cl11]3Ϫ complex and the shoulder at the more facile electrochemical reduction of the former
about Ϫ0.7 V to the reduction of [Ti2Cl9]Ϫ. The cyclic vol- species; (iii) [PPN]3[Ti2Cl11] is not a stable species in the
tammetric results seem to indicate the presence of large solid state and redistributes to a mixture of [PPN]2[Ti2Cl10]
amounts of the [Ti2Cl9]Ϫ ϩ [Ti2Cl11]3Ϫ species, i.e. the equi- and [PPN]2[TiCl6] upon crystallization.
librium of Equation 8 seems shifted extensively toward the
right-hand side. This, however, contrasts with the IR results
(see above). There is in fact no contradiction between these
two results because of the dynamic nature of the electro-
Experimental Section
General Procedures. All manipulations were carried out under an
atmosphere of dry and oxygen-free argon with standard Schlenk
techniques. CH2Cl2 was purified by reflux and distillation over
P4O10 under argon. Pentane was purified by reflux over sodium
benzophenone ketyl and distilled under argon. Cyclic voltammog-
rams were recorded with an EG&G 362 potentiostat connected to
a Macintosh computer through MacLab hardware/software. The
electrochemical cell was fitted with an Ag-AgCl reference electrode,
a platinum disk working electrode and a platinum wire counter-
electrode. [Bu4N]PF6 (ca. 0.1 ) was used as supporting electrolyte
in CH2Cl2. All potentials are reported relative to the ferrocene
standard, which was added to each solution and measured at the
end of the experiments. The infrared spectra were recorded on a
Bruker Vector 22 instrument equipped with a Globar (MIR)
source. The solution spectra in the far IR (480Ϫ300 cmϪ1) region
were obtained from CH2Cl2 solutions in polyethylene cells (0.5 mm
path length). The elemental analyses were carried out by the ana-
lytical services of LSEO and LCC with Fisons EA 1108 and
PerkinϪElmer 2400 instruments, respectively. Titanium() chlo-
ride (Aldrich), triethylbenzylammonium chloride (Aldrich) and bi-
s(triphenylphosphanyl)iminium (PPN) chloride (Aldrich) were used
as received.
chemical experiment. Since the equilibration process de-
scribed in Equation (8) may be quite fast, and since the
[Ti2Cl9]Ϫ complex is reduced at a less-negative potential, its
consumption by the reduction process shifts the equilibrium
to the right-hand side at the expense of the [Ti2Cl10]2Ϫ ion
in the electrode diffusion layer.
Structural Characterization of [PPN][Ti2Cl9]
Single crystals of the PPN salt of the [Ti2Cl9]Ϫ species
were grown from dichloromethane/pentane solution and
were investigated by X-ray diffraction. The structure con-
firms the expected face-sharing bioctahedral geometry (see
Figure 5). The bond lengths and angles within the anion
are given in the Exp. Sect. The TiϪCl bond lengths involv-
˚
ing the bridging chlorine atoms are longer [2.4775(8) A to
˚
˚
2.5098(8) A] than the terminal ones [2.1955(8) A to
˚
2.2216(8) A] as reported for related compounds containing
the [Ti2Cl9]Ϫ anion.[1,29Ϫ33]
Synthesis of [NEt3Bz]2[Ti2Cl10]: A Schlenk tube was charged with
TiCl3 (0.114 g, 0.74 mmol) and [NEt3Bz]Cl (0.174 g, 0.76 mmol).
Thionyl chloride (20 mL) was added by syringe under magnetic
stirring and the apparatus was connected to an oil bubbler. Warn-
ing: The reaction between SOCl2 and water (potentially present as
a contaminant in the starting salts) produces toxic SO2 and HCl,
therefore the reaction must be carried out under a well-ventilated
fume hood. The dark-purple suspension transformed gradually into
a yellow solution. After four hours of stirring at room temperature,
pentane (50 mL) was added, causing the precipitation of the prod-
uct as a yellow powder. After filtration, the solid was washed with
pentane and dried under vacuum for three hours. Yield: 0.276 g
(45%). C26H44Cl10N2Ti2: calcd. C 37.4, H 5.3, N 3.4; found C 37.8,
H 5.1, N 3.3.
Figure 5. An ORTEP view of the anion of compound
[PPN]2[Ti2Cl9]; ellipsoids are drawn at the 50% probability level
Synthesis of [PPN]2[Ti2Cl10]: Following an identical procedure as
described above for [NEt3Bz]2[Ti2Cl10], 0.834 g (88%) of
[PPN]2[Ti2Cl10] was obtained starting from TiCl3 (0.192 g,
1.24 mmol) and PPNCl (0.715 g, 1.25 mmol). C72H60Cl10N2P4Ti2:
calcd. C 56.6, H 4.0, N 1.8; found C 56.0, H 3.7, N 1.8.
Conclusions
The two main novelties of this new study of the well-
developed coordination chemistry of chlorititanate() com-
plexes are: (i) the oxidation of TiCl3 by thionyl chloride in Synthesis of [PPN][Ti2Cl9]: A Schlenk tube was charged with TiCl3
(0.565 g, 3.67 mmol) and PPNCl (1.050 g, 1.83 mmol). Thionyl
chloride (30 mL) was added by syringe under magnetic stirring and
the apparatus was connected to an oil bubbler. Warning: the reac-
tion between SOCl2 and water (potentially present as a contaminant
in the starting salts) produces toxic SO2 and HCl, therefore the reac-
tion must be carried out under a well ventilated fume hood. The dark-
purple suspension transformed gradually into a yellow solution.
After cooling to room temperature, it was concentrated until ap-
proximately 10 mL and pentane (60 mL) was added, causing the
precipitation of the product as a yellow powder. After filtration,
the presence of chloride salts, and (ii) the complex chloride
dissociation and redistribution equilibria that are estab-
lished by these anions, with the unambiguous identification
in solution of the previously unknown, yet simple,
[Ti2Cl11]3Ϫ species. The combined IR and electrochemical
study has shown that: (i) complex [Ti2Cl10]2Ϫ is dinuclear
in dichloromethane solution and in the solid state as a
PPNϩ salt; (ii) this ion is stable in dichloromethane but a
chloride-redistribution equilibrium to yield a mixture of
[Ti2Cl9]Ϫ and the novel [Ti2Cl11]3Ϫ complex is induced by the solid was washed with pentane and dried under vacuum. Yield:
4112  2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim www.eurjic.org
Eur. J. Inorg. Chem. 2004, 4108Ϫ4114

Solution Equilibria for Chlorotitanate(IV) Anions
1.574 g (87%). C36H30Cl9NP2Ti2: calcd. C 45.3, H 3.2; found C Table 1 and selected bond lengths and angles are listed in Table 2.
FULL PAPER
44.3, H 2.8.
CCDC-233004 contains the supplementary crystallographic data
for this paper. These data can be obtained free of charge at
www.ccdc.cam.ac.uk/conts/retrieving.html [or from the Cambridge
Crystallographic Data Centre, 12 Union Road, Cambridge CB2
1EZ, UK; Fax: ϩ44-1223-336-033; E-mail: deposit@ccdc.cam.ac.uk].
Synthesis of [PPN]2[TiCl6]: Following an identical procedure as de-
scribed above for [PPN][Ti2Cl9], 0.822 g (96%) of [PPN]2[TiCl6] was
obtained starting from TiCl3 (0.099 g, 0.64 mmol) and PPNCl
(0.740 g, 1.29 mmol). C72H60Cl10N2P4Ti2: calcd. C 64.6, H 4.5, N
2.1; found C 63.8, H 4.1, N 2.0. The solid state and solution IR
characterization is presented in the Results and Discussion section.
˚
Table 2. Bond lengths (A) and angles (°) for [PPN][Ti2Cl9]
Synthesis of a Product with Stoichiometry [PPN]3[Ti2Cl11]: Follow-
ing an identical procedure as described above for [PPN][Ti2Cl9],
TiCl3 (0.112 g, 0.73 mmol) and PPNCl (0.628 g, 1.09 mmol) yielded
Ti(1)ϪTi(2)
Ti(1)ϪCl(11)
Ti(1)ϪCl(12)
Ti(1)ϪCl(13)
Ti(1)ϪCl(31)
Ti(1)ϪCl(32)
Ti(1)ϪCl(33)
3.3922(7)
2.2216(8) Ti(2)ϪCl(21)
2.1955(8) Ti(2)ϪCl(22)
2.2046(9) Ti(2)ϪCl(23)
2.4620(8) Ti(2)ϪCl(31)
2.5098(8) Ti(2)ϪCl(32)
2.4775(8) Ti(2)ϪCl(33)
2.2124(9)
2.2175(9)
2.2031(9)
2.4937(8)
2.4785(8)
2.4496(8)
0.712 g (93%) of
a solid product corresponding to the
[PPN]3[Ti2Cl11] stoichiometry. C108H90Cl11N3P6Ti2: calcd. C 61.7,
H 4.3, N 2.0; found C 60.6, H 3.8, N 1.9. The solid state and
solution IR characterization is presented in the Results and Dis-
cussion section.
Cl(12)ϪTi(1)ϪCl(11) 98.02(3) Cl(21)ϪTi(2)ϪCl(22) 97.49(4)
Cl(13)ϪTi(1)ϪCl(11) 97.12(3) Cl(23)ϪTi(2)ϪCl(21) 96.84(3)
Cl(12)ϪTi(1)ϪCl(13) 99.71(3) Cl(23)ϪTi(2)ϪCl(22) 98.94(4)
Cl(11)ϪTi(1)ϪCl(31) 91.45(3) Cl(21)ϪTi(2)ϪCl(31) 91.30(3)
Cl(11)ϪTi(1)ϪCl(32) 90.86(3) Cl(21)ϪTi(2)ϪCl(32) 89.43(3)
Cl(11)ϪTi(1)ϪCl(33)166.67(3) Cl(21)ϪTi(2)ϪCl(33) 166.33(3)
Cl(12)ϪTi(1)ϪCl(31) 91.01(3) Cl(22)ϪTi(2)ϪCl(31) 89.23(3)
Cl(12)ϪTi(1)ϪCl(32)165.52(3) Cl(22)ϪTi(2)ϪCl(32) 165.02(3)
Cl(12)ϪTi(1)ϪCl(33) 91.65(3) Cl(22)ϪTi(2)ϪCl(33) 92.14(3)
Cl(13)ϪTi(1)ϪCl(31)165.17(3) Cl(23)ϪTi(2)ϪCl(31) 167.58(3)
Cl(13)ϪTi(1)ϪCl(32) 90.45(3) Cl(23)ϪTi(2)ϪCl(32) 93.37(3)
Cl(13)ϪTi(1)ϪCl(33) 90.21(3) Cl(23)ϪTi(2)ϪCl(33) 91.19(3)
Cl(31)ϪTi(1)ϪCl(32) 77.29(3) Cl(32)ϪTi(2)ϪCl(31) 77.29(3)
Cl(31)ϪTi(1)ϪCl(33) 79.18(3) Cl(33)ϪTi(2)ϪCl(31) 79.11(3)
Cl(33)ϪTi(1)ϪCl(32) 77.92(3) Cl(33)ϪTi(2)ϪCl(32) 79.04(3)
X-ray Structural Determinations for Compound [PPN][Ti2Cl9]: A
single crystal was mounted under inert perfluoropolyether at the
tip of a glass fiber and cooled in the cryostream of the Oxford-
Diffraction XCALIBUR CCD diffractometer. Data were collected
using monochromatic Mo-Kα radiation (λ ϭ 0.71073). The struc-
tures was solved by direct methods (SIR97)[34] and refined by least-
squares procedures on F2 using SHELXL-97.[35] All H-atoms at-
tached to carbon were introduced in calculation in idealised posi-
˚
tions [d(CH) ϭ 0.96 A] and treated as riding models. The drawing
of the molecules (Figure 1) was realised with the help of OR-
TEP32.[36] Crystal data and refinement parameters are shown in
Table 1 Crystal data and structure refinement for [PPN][Ti2Cl9]
Empirical formula
Molecular mass
Temperature
C36H30Cl9NP2Ti2
953.40
Acknowledgments
We are grateful to the French Ministry of Research, to the Centre
National de la Recherche Scientifique, and to Polimeri Europa for
293(2) K
˚
Wavelength
0.71073 A
Crystal system
Space group
Unit cell dimensions
triclinic
P1
a ϭ 9.9904(10) A
b ϭ 12.0458(12) A
c ϭ 17.4335(14) A
´
support of this work, and to Mr. Sebastien Maria for technical as-
sistance.
¯
˚
˚
˚
[1]
α ϭ 86.057(8)°
β ϭ 82.379(8)°
γ ϭ 83.861(8)°
T. J. Kistenmacher, G. D. Stucky, Inorg. Chem. 1971, 10,
122Ϫ132.
[2]
C. Creaser, J. Creighton, J. Chem. Soc., Dalton Trans. 1975,
3
˚
Volume
Z
Density (calculated)
Absorption coefficient
F(000)
Crystal size
Theta range for data collection
Index ranges
2064.4(3) A
1402Ϫ1405.
[3]
2
G. Natta, Angew. Chem. 1956, 68, 393.
G. Natta, P. Corradini, G. Allegra, J. Polymer Sci. 1961, 51,
1.534 Mg/m3
[4]
1.075 mmϪ1
399.
[5]
960
E. G. M. Tornqvist, J. T. Richardson, Z. W. Wilchinsky, R. W.
Looney, J. Catal. 1967, 8, 189Ϫ196.
Z. W. Wilchinsky, R. W. Looney, E. G. M. Tornqvist, J. Catal.
0.16 ϫ 0.09 ϫ 0.04 mm3
2.95 to 26.37°
Ϫ12 Յ h Յ 9
Ϫ15 Յ k Յ 15
Ϫ21 Յ l Յ 21
16046
[6]
1973, 28, 351Ϫ367.
[7]
H. Sinn, W. Kaminsky, Adv. Organomet. Chem. 1980, 18,
99Ϫ149.
[8]
Reflections collected
P. Pino, R. Mulhaupt, Angew. Chem. Int. Ed. Engl. 1980, 19,
Independent reflections
Completeness to theta ϭ 26.37°
Absorption correction
Max. and min. transmission
Refinement method
Data/restraints/parameters
Goodness-of-fit on F2
Final R(int) [I Ͼ 2σ(I)]
R(int) (all data)
8414 [R(int) ϭ 0.0421]
99.7%
analytical
0.92334 and 0.78155
full-matrix least-squares on F2
8414/0/451
857Ϫ875.
[9]
C. G. Barraclough, A. K. Gregson, J. Chem. Soc., Faraday
Trans. 2 1972, 68, 177Ϫ180.
[10]
O. Kahn, Proc. Int. Conf. Coord. Chem. 16th 1974, 2.28a, 23 pp.
A. Ceulemans, G. A. Heylen, L. F. Chibotaru, T. L. Maes, K.
Pierloot, C. Ribbing, L. G. Vanquickenborne, Inorg. Chim.
Acta 1996, 251, 15Ϫ27.
P. C. Crouch, G. W. A. Fowles, R. A. Walton, J. Chem. Soc. A
[11]
0.795
[12]
R1 ϭ 0.0348, wR2 ϭ 0.0496
R1 ϭ 0.0706, wR2 ϭ 0.0554
0.349 and Ϫ0.428 e·A
1969, 972Ϫ976.
Ϫ3
[13]
˚
Largest diff. peak and hole
B. Briat, O. Kahn, I. Morgensternbadarau, J. C. Rivoal, Inorg.
Chem. 1981, 20, 4193Ϫ4200.
Eur. J. Inorg. Chem. 2004, 4108Ϫ4114
www.eurjic.org
 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
4113

´
E. Robe, J.-C. Daran, S. Vincendeau, R. Poli
FULL PAPER
[14]
[27]
[28]
[29]
[30]
[31]
[32]
[33]
[34]
D. Armstrong, P. Perkins, J. Stewart, Rev. Roum. Chim. 1974,
A. Shiga, H. Kawamurakuribayashi, T. Sasaki, J. Mol. Catal.
1994, 87, 243Ϫ261.
D. M. Adams, J. Chatt, J. M. Davidson, J. Gerratt, J. Chem.
Soc. 1963, 2189Ϫ2194.
E. Solari, C. Floriani, A. Chiesi-Villa, C. Guastini, J. Chem.
Soc., Chem. Commun. 1989, 1747Ϫ1749.
E. Solari, C. Floriani, K. Schenk, A. Chiesi-Villa, C. Rizzoli,
M. Rosi, A. Sgamellotti, Inorg. Chem. 1994, 33, 2018Ϫ2028.
S. Castro, W. Streib, J. Huffman, G. Christou, Chem. Commun.
1996, 2177Ϫ2178.
R. Bettenhausen, W. Milius, W. Schnick, Chem. Eur. J. 1997,
3, 1337Ϫ1341.
Q. Feng, S. Luo, M. Olmstead, T. Rauchfuss, P. Stafford,
Chem. Mater. 1997, 9, 641Ϫ643.
A. Altomare, M. Burla, M. Camalli, G. Cascarano, C. Giacov-
azzo, A. Guagliardi, A. Moliterni, G. Polidori, R. Spagna, J.
Appl. Cryst. 1999, 32, 115Ϫ119.
G. M. Sheldrick, SHELXL-97. Program for Crystal Structure
refinement, University of Göttingen, Göttingen, Germany,
1997.
19, 1695Ϫ1707.
H. Hecht, Z. Anorg. Chem. 1947, 254, 37Ϫ51.
R. Colton, J. H. Canterford, Halides of the First Row Transition
[15]
[16]
Metals, Wiley-Interscience, London, 1969.
H. Heumann, P. Köchlin, Ber. 1883, 16, 1625.
H. Stavdinger, W. Kreis, Helv. Chim. Acta 1924, 8, 71.
H. Linga, Z. Stojek, R. A. Osteryoung, J. Am. Chem. Soc.
[17]
[18]
[19]
1981, 103, 3754Ϫ3760.
R. T. Carlin, R. A. Osteryoung, J. S. Wilkes, J. Rovang, Inorg.
[20]
Chem. 1990, 29, 3003Ϫ3009.
N. Hao, B. Sayer, G. Denes, D. Bickley, C. Detellier, M.
[21]
Mcglinchey, J. Magn. Reson. 1982, 50, 50Ϫ63.
A. F. Demiray, W. Brockner, B. N. Cyvin, S. J. Cyvin, Z. Natur-
[22]
forsch., Teil A 1979, 34, 362Ϫ368.
R. Clark, L. Maresca, R. Puddephat, Inorg. Chem. 1968, 7,
1603.
W. Van Bronswyk, R. Clark, L. Maresca, Inorg. Chem. 1969,
8, 1395.
O. Poleshchuk, J. Latosinska, B. Nogaj, J. Mol. Struct. 1996,
380, 277Ϫ282.
[23]
[35]
[36]
[24]
[25]
L. J. Farrugia, J. Appl. Crystallogr. 1997, 32, 565.
Received April 10, 2004
[26]
C. Ding, Y. Yu, R. Jensen, W. Balfour, C. Qian, Chem. Phys.
Early View Article
Lett. 2000, 331, 163Ϫ169.
Published Online August 17, 2004
4114
 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjic.org
Eur. J. Inorg. Chem. 2004, 4108Ϫ4114