From zirconium to titanium: The effect of the metal in t-butylacrylate photoinitiated polymerisation

Article abstract of DOI:10.1039/b315482c

We report here the synthesis, photochemistry and photoinitiator activity of some titanocenes (1-4) and compare the results with those obtained for the corresponding zirconocenes (5-8). Analysis of the electronic spectra showed that the energy modulation of the lowest electronic transition, which appears to be LMCT in character, is driven both by the substituent on the cyclopentadienyl moiety and by the metal centre. Furthermore, the excited state resulting from irradiation of the complexes at the wavelength of the LMCT transition undergoes ligand-metal bond dissociation with formation of a radical pair, as evidenced by EPR spectroscopy coupled with spin trapping techniques. All the complexes were very active, compared with known organometallic photoinitiators, for the free-radical photopolymerisation of t-butylacrylate. The titanium complexes 1-4, which can be used with visible light, were more active than the zirconium derivatives 5-8. The better yields in photopolymerisation can be interpreted on the basis of the combined effect of two factors: (i) the photoreactivity of the complexes in solution and (ii) the high persistence of the Ti(III) radical species in solution, which guarantees high concentrations of initiators in the polymerisation process.

Full text of DOI:10.1039/b315482c

View Article Online / Journal Homepage / Table of Contents for this issue
P A P E R
From zirconium to titanium: the effect of the metal in t-butylacrylate
photoinitiated polymerisation
Eleonora Polo,*a Andrea Barbieria and Orazio Traversob
a
`
C.N.R., Istituto per la Sintesi Organica e la Fotoreattivita (I.S.O.F.), sez. Ferrara,
via L. Borsari 46, 44100, Ferrara, Italy. E-mail: tr3@dns.unife.it; Fax: þ39 0532 24 0709;
Tel: þ39 0532 29 1159
b
`
Dipartimento di Chimica dell’Universita di Ferrara, via L. Borsari 46, 44100, Ferrara, Italy.
E-mail: tr2@dns.unife.it; Fax: þ39 0532 24 0709; Tel: þ39 0532 29 1158
Received (in Toulouse, France) 28th November 2003, Accepted 27th January 2004
First published as an Advance Article on the web 19th April 2004
We report here the synthesis, photochemistry and photoinitiator activity of some titanocenes (1–4) and
compare the results with those obtained for the corresponding zirconocenes (5–8). Analysis of the electronic
spectra showed that the energy modulation of the lowest electronic transition, which appears to be LMCT in
character, is driven both by the substituent on the cyclopentadienyl moiety and by the metal centre.
Furthermore, the excited state resulting from irradiation of the complexes at the wavelength of the LMCT
transition undergoes ligand-metal bond dissociation with formation of a radical pair, as evidenced by EPR
spectroscopy coupled with spin trapping techniques. All the complexes were very active, compared with
known organometallic photoinitiators, for the free-radical photopolymerisation of t-butylacrylate. The
titanium complexes 1–4, which can be used with visible light, were more active than the zirconium
derivatives 5–8. The better yields in photopolymerisation can be interpreted on the basis of the combined
effect of two factors: (i) the photoreactivity of the complexes in solution and (ii) the high persistence of the
Ti(III) radical species in solution, which guarantees high concentrations of initiators in the polymerisation
process.
the hydrogen by a phenyl in position 2 of the cyclopentadienyl
Introduction
ring7 we achieved a significant increase in the absorption
coefficients. In this way the light absorption efficiency was
improved and, consequently, the polymerisation yields were
almost doubled, probably due to a combined effect of a higher
persistence of the radical in solution and/or an increased effi-
ciency of the radical in initiating the polymerisation process.
This kind of substitution, unfortunately, failed to shift the
absorption maxima towards higher wavelengths in the visible
area. As a further step we decided to synthesise, study and test
analogous titanium complexes, since the metallocenes of this
metal usually absorb in the visible region.
Photochemical or photoinitiated polymerisations play a cen-
tral role in several commercial technologies1,2 and are largely
employed today in the areas of surface coatings, photoresists,
adhesives and holography. In the growing area of photoinitia-
tors for use in pigmented coatings, compounds3 showing
absorption peaks at wavelengths higher than 350 nm are in
particular required, since they can be used with the visible light
produced by continuous wave lasers. Furthermore, since light
can generate electronic excited states different from those
achieved by thermal activation, it can be used as a clean,
tunable and versatile ‘‘tool’’ to selectively activate a photo-
initiator in the presence of organic solvents and chemical con-
taminants. While organic compounds still play a leading role
among photoinitiators,2,4 several classes of inorganic and
organometallic complexes5 are receiving increasing attention
due to the possibility to tailor the polymer microstructure by
a systematic and rational variation of their ligand structure
and substitution pattern. The versatility of the chemistry of
the cyclopentadienyl ligand makes the metallocene family par-
ticularly interesting from this point of view. Moreover, these
complexes often leave low toxic and clear (or colourless)
residues after exposure to argon lasers.
In the framework of our search for metallocenes suitable for
use as photoinitiators in radical polymerisation processes, we
have recently6,7 started a systematic study of the photochemi-
cal behaviour and of the photopolymerisation activity of
several mono- and bicyclic cyclopentadienyl group IV
complexes. In particular, we have found that bicyclic cyclopen-
tadienyl zirconium complexes containing saturated rings fused
to the cyclopentadienyl unit gave better polymerisation yields
and were more stable in solution compared with the corre-
sponding unsaturated complexes.6 With the substitution of
In this work, several titanium complexes were synthesised
and studied: bis(Z5-indenyl)titanium dichloride [Ind2TiCl2]
(1), bis[Z5-(4,5,6,7-tetrahydro-1H-indenyl)]titanium dichloride
[(C6Cp)2TiCl2] (2), bis[Z5-(2-phenylindenyl)]titanium dichlor-
ide [(IndPh)2TiCl2] (3) and bis[Z5-(2-phenyl-4,5,6,7-tetrahy-
dro-1H-indenyl)titanium dichloride [(C6CpPh)2TiCl2] (4).
Their properties and activities were compared with those
obtained using the corresponding zirconium metallocenes:
bis(Z5-indenyl)zirconium dichloride [Ind2ZrCl2] (5), bis[Z5-
(4,5,6,7-tetrahydro-1H-indenyl)zirconium dichloride [(C6Cp)2-
ZrCl2] (6), bis[Z5-(2-phenylindenyl)]zirconium dichloride
[(IndPh)2ZrCl2] (7) and bis[Z5-(2-phenyl-4,5,6,7-tetrahydro-
1H-indenyl)]zirconium dichloride [(C6CpPh)2ZrCl2] (8).
T h i s j o u r n a l i s Q T h e R o y a l S o c i e t y o f C h e m i s t r y a n d t h e
C e n t r e N a t i o n a l d e l a R e c h e r c h e S c i e n t i f i q u e 2 0 0 4
652
N e w . J . C h e m . , 2 0 0 4 , 2 8 , 6 5 2 – 6 5 6

View Article Online
Table 1 Spectroscopic properties of complexes 1–8 in benzene
Complex
lmax/nm
e
max/Mꢁ1 cmꢁ1
1
2
3
4
5
6
7
8
(Ind)2TiCl2
397
519
415
543
397
503
394
479
310
380
310
350
310
330
320
345
3240
1820
1700
160
(C6Cp)2TiCl2
(IndPh)2TiCl2
a
550
1225
3275
3400
1475
720
(C6CpPh)2TiCl2
(Ind)2ZrCl2
Results and discussion
Synthesis of complexes 1–4
(C6Cp)2ZrCl2
(IndPh)2ZrCl2
(C6CpPh)2ZrCl2
1600
1100
The starting materials necessary for the synthesis of the ligands
were the commercially available indene, 9, and the dienes, 10,8
11,9 and 12,7 prepared according to previously published pro-
cedures. The synthetic approach to the Ti complexes is similar
for all complexes and it is described in the Experimental.
60 500
41 000
23 400
–
a
In CH2Cl2 as 3 is not completely soluble in benzene.
metal, combined with the presence of the phenyl substituent,
seems to accelerate the photoreactivity of the metallocenes.
This effect is due to the fact that the orbitals are less stabilised
with titanium than with zirconium because of a weaker metal-
ligand interaction. Once irradiated, the titanium complexes
decay more easily by nonradiative processes than the corre-
sponding zirconium derivatives.10 After 1 h irradiation the
photodecomposition of the indenyl complexes 1 and 3 was
almost complete, while only 20% of their tetrahydro counter-
parts, 2 and 4, was degraded. In both cases the presence of
the phenyl substituent increased the rate and the extent of
the process.
An interesting feature common to the spectra of the tita-
nium complexes 1–4 is the presence of an isobestic point at
570 nm and the concomitant growth upon irradiation of a
new broad transition at around 635 nm, which can be attribu-
ted to a new Ti(III) species on the basis of the literature data
(Fig. 1).12
We have found that the synthesis by direct reaction of the
lithium salts (a method that works well for the corresponding
zirconium derivatives) of the dienes 9–12 with TiCl4 in THF
(or with TiCl3 , followed by oxidation with PbCl2) gave poor
yields of complexes extremely difficult to purify. This problem
is due to the fact that the Ti complexes 1–4 are easily reduced,
either by visible light in solution (in particular in THF) or by
traces of alkyllithiums from the previous deprotonation step
acting as reducing agents. This problem was overcome by con-
verting the lithium salts into their corresponding air-stable tri-
methylsilyl derivatives and by carrying out the subsequent
metallation reaction in CH2Cl2 , taking care to protect the
reaction vessel from light. The complexes were collected by
elimination under vacuum of the solvent and of the ClSiMe3
produced in the reaction and washing with petroleum ether.
This two-step procedure gave higher yields of easy-to-purify
complexes.
EPR spin-trapping investigation
To study the nature of the species produced upon irradiation,
EPR spin-trapping experiments have been carried out by direct
irradiation of benzene solutions of the titanocenes inside the
spectrometer cavity. The LMCT irradiation (l > 350 nm) of
Photochemistry
UV/Vis spectra of the titanium complexes 1–4 have been
recorded in benzene and compared with those obtained pre-
viously7 for the zirconium complexes 5–8 (Table 1). As
expected, we found the position of the absorption maxima to
be remarkably shifted towards longer wavelengths ( > 400
nm). This feature fulfils the desired condition of requiring
lower energies for the photochemical activation and it is in
agreement with the fact that the energy of the lowest transition
depends strongly on the nature of the metal, an effect that is
particularly enhanced with titanium.10
Since the free ligand shows absorption bands at higher ener-
gies (l < 330 nm), the observed shift supports the hypothesis of
a prevalent ligand-to-metal charge transfer (LMCT) character
for the electronic transition. Both the lmax and the emax values
(Table 1) strongly support the LMCT assignment and are in
agreement with the most recent findings on the excitation
states for similar metallocenes of group IV.10,11 Then, the pri-
mary photochemical act should consist in the homolytic clea-
vage of the ligand-metal bond with formation of ligand- and
metal-centred radicals upon electronic excitation.
The study of the photoreactivity in solution of the com-
plexes 1–4 was performed by photolysis in benzene at
l > 350 nm. The nature of the complexes has a significant
influence on their photoreactivity in solution: the change of
Fig. 1 (C6CpPh)2TiCl2 , 4, irradiated in benzene (l > 350 nm) at
20 ꢀC under argon atmosphere. The spectra were recorded every 5 min.
T h i s j o u r n a l i s Q T h e R o y a l S o c i e t y o f C h e m i s t r y a n d t h e
C e n t r e N a t i o n a l d e l a R e c h e r c h e S c i e n t i f i q u e 2 0 0 4
N e w . J . C h e m . , 2 0 0 4 , 2 8 , 6 5 2 – 6 5 6
653

View Article Online
complexes 1–4 in the presence of the spin trap N-tert-butyl-a-
phenylnitrone (PBN) (metallocene:PBN molar ratio ¼ 1:25)
showed fairly intense EPR signals. We remarked two different
patterns depending on the type of ligand (Table 2). In the case
of the complexes 1 and 2 we observed only one signal, a triplet
of doublets, stable during and after irradiation, that can be
assigned to the spin adduct of the indenyl or tetrahydroindenyl
radical with PBN (Scheme 1).13,14 The aN (14.5–14.6 G) and aH
(2.2–2.3 G) values are in good agreement with those observed
during the photolysis of the corresponding zirconium metallo-
cenes 5 and 6.7
In the case of the spectra of the titanium complexes 3 and 4
(Fig. 2), besides the signals consistent with the presence of a
radical on the cyclopentadienyl moiety,13 we observed also
the presence of another signal due to a species not trapped with
PBN and characterised by a g value typical of a Ti(III) species
(Table 2).15 This result supports the attribution of the new
absorption at 635 nm in the UV/Vis spectrum to a Ti(III) com-
pound. In titanocenes 3 and 4, the phenyl substituent on the
cyclopentadienyl unit seems to provide an additional stabilisa-
tion that allows direct observation of the metal-centred radical,
which was hardly visible in the spectra of the zirconium analo-
gues and only at low temperature.16
Fig. 2 (IndPh)2TiCl2 þ 0.1 M PBN irradiated in benzene (l > 350
nm).
We can conclude that the electronic spectral changes that
occur during the irradiation of benzene solutions of the tita-
nium complexes and the EPR findings are in agreement with
the literature data reported for similar titanium complexes17
and indicate that the first step in the photoprocess is the disso-
ciation of the LMCT state to give ligand and metal-centred
radicals (Scheme 2).
Scheme 2 Primary photochemical act.
for the zirconium complexes 5–8 and at l > 350 nm for the
titanium complexes 1–4. In the absence of the catalysts, the
irradiation in the same conditions of a reference solution of
t-BA in benzene did not produce any polymeric material.
The polymer yields obtained with the titanium complexes 1–
4 were all higher than those obtained with the corresponding
zirconium metallocenes, 5–8, and followed the same trend:
the indenyl complexes, 1 and 3, exhibited an increase of about
the 15%, while the tetrahydroindenyl derivatives 2 and 4
showed an average increase of 21%. Also in this case the tita-
nium complexes bearing a phenyl substituent on the cyclopen-
tadienyl ring (3 and 4) gave polymerisation yields twice as high
as those obtained with the unsubstituted complexes (1 and 2).
This increase in the polymerisation outcome suggests the invol-
vement of a Ti(III) radical species in the polymerisation process
in addition to the species already detected with the zirconium
complexes.
Polymerisation studies
Polymerisation tests were performed in benzene on t-butyl-
acrylate (t-BA) and the results obtained have been compared
with those obtained with the analogous zirconium derivatives
(Table 3). Since this monomer is the least reactive in the acry-
late series, a positive result obtained in t-BA polymerisation
can be considered as proof of the high efficiency of a metallo-
organic complex as a radical initiator. Since this monomer
absorbs only at l < 300 nm, we can exclude the possibility that
the polymer obtained can be produced, in part or completely,
from direct photochemical activation of the monomer.
In a typical experiment, a vigorously stirred 20% v/v solu-
tion of t-butylacrylate in benzene was irradiated at 20 ꢀC for
1 h in the presence of 7 mg of the metallocene at l > 320 nm
The tacticity of the polymer formed was estimated by decon-
volution of the methine peaks of poly(t-BA) at about 42 ppm
in the 13C-NMR spectra. As we could expect, the stereochemi-
cal outcome of the polymerisation reaction was quite indepen-
dent of the metallocene used and the data were quite similar to
those reported for the zirconium complexes, with a content of
Table 2 Hyperfine coupling constants of the radical spin adducts with
PBN in benzene (metallocene:PBN molar ratio ¼ 1:25).
Species I
Species II
Table 3 tert-Butylacrylate polymerisationa data
Complex
aN/G
aH/G
%
%
g
Complex
Yieldb
Mwc /g molꢁ1 PDId
Ze /dL gꢁ1
1
2
3
4
(Ind)2TiCl2
14.6
14.5
14.6
14.6
2.2
2.2
2.3
2.3
100
100
71
–
–
1
2
3
4
5
6
7
8
(Ind)2TiCl2
(C6Cp)2TiCl2
(IndPh)2TiCl2
30
31
65
1825 482
787 203
2077 649
889 753
1299 118
395 451
1316 731
425 691
3.0
4.5
2.8
3.4
2.5
2.3
2.3
2.3
3.95
1.02
4.33
1.43
2.36
0.92
2.94
1.01
(C6Cp)2TiCl2
(IndPh)2TiCl2
(C6CpPh)2TiCl2
–
–
29
37
2.0113
2.0117
63
(C6CpPh)2TiCl2 58
(Ind)2ZrCl2
(C6Cp)2ZrCl2
(IndPh)2ZrCl2
25
24
56
(C6CpPh)2ZrCl2 46
a
Polymerisation conditions: a vigorously stirred 20% v/v monomer
solution in benzene in a 4 ml cuvette was irradiated for 1 h in the pre-
b
sence of 7 mg of the metallocene. Polymerisation degree ¼ (weight
c
of polymer obtained/weight of starting monomer) ꢂ 100. Average
d
molecular weight by GPC based on polystyrene standards. Polydis-
e
persity index ¼ Mw/Mn Intrinsic viscosity measured in CHCl3 at
35 ꢀC.
Scheme 1 Mechanism of action of PBN.
T h i s j o u r n a l i s Q T h e R o y a l S o c i e t y o f C h e m i s t r y a n d t h e
C e n t r e N a t i o n a l d e l a R e c h e r c h e S c i e n t i f i q u e 2 0 0 4
654
N e w . J . C h e m . , 2 0 0 4 , 2 8 , 6 5 2 – 6 5 6

View Article Online
syndiotactic triads, [rr], ranging from 30% to 35% and of
isotactic triads, [mm], ranging from 5% to 10%. This result is
consistent with the radical nature of the process.
Experimental
General remarks
The polymers obtained have been also characterised by gel
permeation chromatography (GPC) and viscosity measure-
ments (Table 3).
All manipulations of air- and/or moisture-sensitive com-
pounds were carried out under inert atmosphere using a dual
vacuum/argon line and standard Schlenk techniques. All sol-
vents were thoroughly dehydrated and deoxygenated under
argon before use. They were dried and purified by refluxing
under argon over a suitable drying agent (petroleum ether,
diethyl ether, benzene, toluene and THF over Na/benzophe-
none ketyl or potassium; C6D6 over K; CH2Cl2 and CDCl3
over CaH2) followed by distillation and storage under argon
in Young’s ampoules. Solvents and solutions were transferred,
using a positive pressure of argon, through stainless-steel can-
nulae (diameter 0.5–2.0 mm) and mixtures were filtered in a
similar way using modified cannulae that could be fitted with
glass-fibre filter disks (Whatman GFC). Unless otherwise spe-
cified, all reagents were purchased from commercial suppliers
(Aldrich and Fluka) and used without further purification.
Indene, 9, was purchased from Fluka and distilled under
reduced pressure before use; the dienes 10,8 11,9 and 12,7 were
prepared as described in the literature.
The polymer analysis showed that we can obtain quite a
wide range of high molecular weight tert-butylacrylates
through small changes in the structure of the photoinitiator.
If we try to draw a correlation between the structure of the
complexes and the polymerisation data, we can remark that:
(i) the indenyl complexes 1, 3, 5 and 7 gave polymers with
average molecular weights (Mw) and intrinsic viscosities (Z)
higher than those prepared with the tetrahydroindenyl com-
plexes 2, 4, 6 and 8, and the titanocenes always gave higher
values than the corresponding zirconocenes in both series.
(ii) The phenyl-substituted complexes 3, 4, 7 and 8 always
gave higher polymerisation yields than the related unsubsti-
tuted compounds 1, 2, 5 and 6 and the indenyl derivatives also
gave polymers with higher Mw .
(iii) The complexes 2, 4, 6 and 8, which decompose more
slowly in solution under irradiation than the corresponding
unsaturated analogues 1, 3, 5 and 7, gave polymers with lower
Mw . Since the polymerisation yields were quite similar when
compared pairwise, this fact suggests the hypothesis that, in
the presence of more long-lived radical species, it is easier to
build a new polymer chain than to add a monomer to an
already existing one.
Elemental analyses were performed using a Carlo Erba 1106
Elemental Analysis apparatus. 1H (200.13 MHz) and13C (50.32
MHz) nuclear magnetic resonance spectra were recorded at
room temperature with a Bruker AC200 spectrometer.
Measurements
(iv) The Ti complexes 1–4 produced polymers with polydis-
persities higher than those obtained with the Zr derivatives 5–8
and the highest values were obtained with the most photoche-
mically stable complexes in solution. The zirconium complexes
did not show significant differences in this regard. Interest-
ingly, the polydispersity index (PDI) is in all cases much higher
than that expected for a typical radical reaction (usually
around 2). The broader polydispersity can be attributed to
the fact that more than one initiator is produced upon irradia-
tion. This hypothesis is confirmed by the fact that in the case of
titanium complexes 1–4 we always obtain PDI values higher
than with the corresponding zirconium derivatives 5–8, a
fact that is consistent with the UV/Vis and EPR findings
(vide supra).
EPR studies. The spin trap PBN was purchased from Sigma-
Aldrich Chemical Co. and used without further purification.
The EPR and spin trapping experiments were performed on
a Bruker EMX spectrometer, operating in the X band (micro-
wave frequency ¼ 9.74 GHz, microwave power ¼ 20 mW,
magnetic field modulation frequency ¼ 100 KHz, magnetic
field modulation amplitude ¼ 1 G). Sample solutions were
freeze-pump-thaw degassed and then irradiated with a 350 W
medium pressure Hg lamp directly in the EPR spectrometer
cavity (Bruker ER-4104OR, TE102). Hyperfine coupling
constants have been calculated by best fit simulation of
experimental spectra using NIEHS WinSim software.19
Photochemical experiments. Irradiation was performed using
an Oriel 500 W high pressure Hg lamp equipped with cut-off
filters for wavelength selection and a water filter to exclude
thermal processes. The experiments were carried at 20 ꢀC
under argon atmosphere in a 4 ml cuvette equipped with a
Young stopper. UV/Vis spectra were recorded on a Perkin
Elmer Lambda 40 double beam spectrophotometer.
Conclusions
We have reported here the synthesis, photochemistry and
photoinitiator activity for the polymerisation of acrylates of
a series of titanocenes (1–4) and we have compared the results
with those obtained with the corresponding zirconocenes (5–
8). The analysis of the electronic spectra showed that the
change of metal shifts (as expected) the absorption maxima
of the complexes into the visible region of the spectrum, thus
expanding the wavelengths available for the photoinitiation
process.
The radical species involved in the photoinduced cleavage of
the metal-ligand bond have been identified and t-butylacrylate
polymerisation has been studied in detail. All the complexes
are highly active compared with similar photoinitiators18 in
the polymerisation of t-BA. The better yields in photopolymer-
isation can be interpreted on the basis of the combined effect of
two factors: (i) the photoreactivity of the complexes in solution
and (ii) the higher persistence of the Ti(III) radical species in
solution (with respect to the corresponding zirconium analo-
gues) that guarantees higher concentrations of initiators in
the polymerisation process.
Polymerisation tests. t-Butylacrylate was dried with CaH2
and vacuum distilled immediately before sample preparation.
In a typical experiment a vigorously stirred 20% v/v mono-
mer solution in benzene in a 4 ml cuvette was irradiated at
20 ꢀC for 1 h (l > 320 nm for the zirconium complexes 5–8
and l > 350 nm for the titanium complexes 1–4) in the
presence of 7 mg of the metallocene. The reaction was
quenched by addition of a small amount of aqueous hydro-
chloric acid and the polymer was precipitated by pouring a
solution of CH3OH–HCl (1:1) into the reaction cuvette. The
polymer was collected by removal of the solvent under reduced
pressure, redissolved in acetone, air dried and collected as a
thin film.
As to the future, in order to determine which structural
modifications in the catalysts are necessary to improve their
performance in the photoinitiated polymerisation of acrylates,
a deeper understanding of the details of the polymerisation
mechanism is needed.
GPC measurements. The molecular mass distribution
(MMD) was measured using a Waters GPCV 2000 high tem-
perature GPC, equipped with a refractive index (RI) detector
and a viscosity detector (differential viscometer). The column
set was composed of three TSKGMHXL-HT columns.
T h i s j o u r n a l i s Q T h e R o y a l S o c i e t y o f C h e m i s t r y a n d t h e
C e n t r e N a t i o n a l d e l a R e c h e r c h e S c i e n t i f i q u e 2 0 0 4
N e w . J . C h e m . , 2 0 0 4 , 2 8 , 6 5 2 – 6 5 6
655

View Article Online
Measurements were carried out in chloroform at 35 ꢀC at a
flow rate of 0.8 mL minꢁ1. Universal calibration with 18 PS
narrow standards ranging from 162 to 5.5 ꢂ 106 g molꢁ1 was
applied. Sample concentration was about 2 mg mLꢁ1, injection
volume 218 mL.
Acknowledgements
Prof. S. Sostero of the University of Ferrara is gratefully
acknowledged for the assistance supplied in the photochemical
studies and Dr. R. Mendichi of the Istituto per lo Studio delle
Macromolecole (ISMAC) of the National Council of Research
(CNR) for polymer characterisation.
Synthesis of the Ti complexes: general procedure
The dienes (10 mmol) were dissolved in 30 mL petroleum ether
(9, 10 and 12) or diethyl ether (11) and cooled to ꢁ10 ꢀC; n-
butyllithium (13 mmol) was added dropwise and the solution
was allowed to reach room temperature and stirred overnight.
The lithium salts of 9, 10 and 12 separated from the solution as
dusty solids, which were filtered via a cannula, washed twice
with petroleum ether, then the residual solvent was pumped
off. In the case of the diene 11 a dark solution was obtained.
The solvent was removed under reduced pressure and the dark
solid obtained was washed twice with petroleum ether to
remove excess n-butyllithium, then the residual solvent was
pumped off.
The lithium salts were dissolved in 30 mL anhydrous THF
and ClSiMe3 (30 mmol) diluted in 10 mL of THF was added
dropwise. The resulting cloudy solution was stirred 6 h at
room temperature, then cooled to 0 ꢀC; a saturated solution
of NH4Cl was slowly added until separation of two phases
occurred. After separation, the aqueous solution was washed
three times with diethyl ether. The collected organic fractions
were dried over anhydrous magnesium sulfate, filtered and
the solvent was removed with a rotary evaporator.
References
1
2
3
(a) Photopolymerization and Photoimaging Science and Technol-
ogy, ed. N. S. Allen, Elsevier, Amsterdam, 1989; (b) Radiation
Curing: Science and Technology, ed. S. P. Pappas, Plenum Press,
New York, 1992.
Photopolymerization: Fundamentals and Applications, ACS
Symposium Series, eds. A. B. Scranton, C. N. Bowman and
R. W. Peiffer, American Chemical Society, Washington, DC,
1997, vol. 673.
Photoinitiators for UV Curing. Key Products Selection Guide.
Additives, Ciba, Basel, 1998, available at http://www.cibasc.com.
R. S. Davidson, J. Photochem. Photobiol., A, 1993, 69, 263.
(a) B. M. Monroe and G. C. Weed, Chem. Rev., 1993, 93, 435; (b)
R. S. Davidson, J. Photochem. Photobiol., A, 1993, 73, 81; (c) C.
Kutal, Coord. Chem. Rev., 2001, 211, 353; (d ) A. Barbieri, A.
Droghetti, S. Sostero and O. Traverso, J. Photochem. Photobiol.,
A, 1999, 129, 137.
4
5
6
7
8
9
E. Polo, A. Barbieri, S. Sostero and M. L. H. Green, Eur. J. Inorg.
Chem., 2002, 8, 405.
E. Polo, A. Barbieri and O. Traverso, Eur. J. Inorg. Chem., 2003,
9, 324.
E. Polo, R. M. Bellabarba, G. Prini, O. Traverso and M. L. H.
Green, J. Organomet. Chem., 1999, 577, 211.
G. W. Coates and R. M. Waymouth, Science, 1995, 267, 217.
The trimethylsilyl derivatives were transferred into
a
10 G. V. Loukova and V. A. Smirnov, Chem. Phys. Lett., 2000,
329, 437.
Schlenk flask, degassed, protected from light and dissolved
in about 30 mL of CH2Cl2 ; a solution of TiCl4 (4.8 mmol)
in 5 mL CH2Cl2 was added dropwise. The reaction mixture
was stirred overnight. The solvent was then removed under
vacuum, the residue washed with petroleum ether, filtered,
and dried under reduced pressure to give a dusty solid.
Although the first synthesis of complexes 1 and 2 has been
previously reported,20 we report here the NMR characterisa-
tion, which was not reported in that paper. Furthermore, with
our procedure the yields have been significantly improved.
11 (a) G. L. Geoffroy and M. S. Wrighton, Organometallic Photo-
chemistry, Academic Press, New York, 1979; (b) J. W. Kenney,
D. R. Boone, D. R. Striplin, Y. H. Chen and K. B. Hamar, Orga-
nometallics, 1993, 12, 3671; (c) G. V. Loukova and V. V. Strelets,
J. Organomet. Chem., 2000, 606, 203; (d ) X. Wang, L. Chen,
A. Endou, M. Kubo and A. Miyamoto, J. Organomet. Chem.,
2003, 678, 156; (e) V. V. Strelets, Coord. Chem. Rev., 1992, 114,
1; ( f ) V. W.-W. Yam, G. Z. Qi and K. K. Cheung, J. Chem.
Soc., Dalton Trans., 1998, 1819; (g) V. W.-W. Yam, G. Z. Qi
and K. K. Cheung, Organometallics, 1998, 17, 5448; (h) R. W.
Harrigan, G. S. Hammond and H. B. Gray, J. Organomet.
Chem., 1974, 81, 79; (i) H. Kunkely and A. Vogler, J. Photochem.
Photobiol., A, 1999, 119, 187.
Bis(g5-indenyl)titanium dichloride (1). Purple volatile solid.
Yield 70% (from indene; 4 steps; average yield per step:
88%). Anal. calcd for C18H14Cl2Ti (349.08): C, 61.9; H, 4.0;
12 W. W. Lukens, Jr., M. R. Smith III and R. A. Anderson, J. Am.
Chem. Soc., 1996, 118, 1719.
1
Cl, 20.3; found C, 61.5; H, 4.2; Cl, 19.8. H-NMR (CDCl3):
d 7.10–7.18 (m, 2H, –CH–Cp), 7.22–7.30 (m, 4H, –CH–Cp),
7.42–7.65 (m, 4H, –ArH), 7.63–7.88 (m, 4H, –ArH).
13 D. R. Duling, J. Magn. Reson., Ser. B, 1994, 104, 105.
14 (a) S. Di Martino, S. Sostero, O. Traverso, D. Rehorek and T. J.
Kemp, Inorg. Chim. Acta, 1990, 176, 107; (b) S. Sostero, D.
Rehorek, O. Traverso and E. Polo, Inorg. Chim. Acta, 1993,
209, 171.
15 (a) E. Samuel and J. Vedel, Organometallics, 1989, 8, 237; (b) K.
Mach and J. B. Raynor, J. Chem. Soc., Dalton Trans., 1992, 683;
(c) H. Deng and K. Soga, Macromolecules, 1996, 29, 1847.
16 (a) G. M. Williams and J. Schwartz, J. Am. Chem. Soc., 1982, 104,
1122; (b) C. S. Bajgur, S. B. Jones and J. L. Petersen, Organome-
tallics, 1985, 4, 1929; (c) F. Soleil and R. Choukroun, J. Am.
Chem. Soc., 1997, 119, 2938.
17 (a) E. Vitz and C. H. Brubaker, Jr., J. Organomet. Chem., 1974,
82, C16; (b) R. W. Harrigan, G. S. Hammond and M. B. Gray,
J. Organomet. Chem., 1974, 81, 79; (c) E. Vitz, P. J. Wagner
and C. H. Brubaker, Jr., J. Organomet. Chem., 1976, 107, 301;
(d ) G. L. Lee and C. H. Brubaker, Jr., Inorg. Chim. Acta, 1977,
25, 181; (e) M. R. M. Bruce and D. R. Tyler, Organometallics,
1985, 4, 528.
18 (a) Y. Li, D. G. Ward, S. S. Reddy and S. Collins, Macromole-
cules, 1997, 30, 1875; (b) P. A. Cameron, V. C. Gibson and
A. J. Graham, Macromolecules, 2000, 33, 4329; (c) M. H. Lee,
J.-W. Hwang, Y. Kim, J. Kim, Y. Han and Y. Do, Organometal-
lics, 1999, 18, 5124; (d ) C. Qian, W. Nie and J. Sun, Organometal-
lics, 2000, 19, 4134; (e) H. Frauenrath, H. Keul and H. Ho¨cker,
Macromolecules, 2001, 34, 14; ( f ) J. Q. Sun, Z. D. Pan, Y. Zhong,
W. Q. Hu and S. L. Yang, Eur. Polym. J., 2000, 36, 2375.
19 A. S. W. Li, K. B. Cummings, H. P. RoethLing, G. R. Buetner
and C. F. Chignell, J. Magn. Reson., 1988, 79, 140.
Bis[g5-(4,5,6,7-tetrahydro-1H-indenyl)titanium dichloride (2).
Red-brown solid. Yield 48% (from bicyclo[4.3.0]non-6-en-8-
one; 5 steps; average yield per step: 88%). Anal. calcd for
C18H22Cl2Ti (357.14): C, 60.5; H, 6.2; Cl, 19.8; found C,
60.7; H, 6.4; Cl, 19.4. 1H-NMR (CDCl3): d 1.60–1.85 (m,
4H), 1.85–2.00 (m, 4H), 2.60–2.90 (m, 4H), 3.10–3.30 (m,
4H), 6.65 (d, J ¼ 3.0 Hz, 4H), 7.05 (t, J ¼ 3.0 Hz, 2H).
Bis[g5-(2-phenylindenyl)]titanium dichloride (3). Purple solid.
Yield 60% (from 2-indanone; 5 steps; average yield per step:
90%). Anal. calcd for C30H22Cl2Ti (501.27): C, 71.9; H, 4.4;
1
Cl, 14.2; found C, 71.5; H, 4.2; Cl, 14.8. H-NMR (CDCl3):
d 7.1–7.5 (m, 4H), 7.6–7.7 (m, 18H).
Bis[g5-(2-phenyl-4,5,6,7-tetrahydro-1H-indenyl)]titanium
dichloride (4). Red-brown solid. Yield 52% (from bicyclo-
[4.3.0]non-6-en-8-one; 5 steps; average yield per step: 87%).
Anal. calcd for C30H30Cl2Ti (509.33): C, 70.7; H, 5.9; Cl,
13.9; found C, 70.8; H, 6.1; Cl, 13.4. 1H-NMR (CDCl3):
d 1.50–1.90 (m, 4H, C–CH2–CH2–), 1.90–2.20 (m, 4H, C–
CH2–CH2–), 2.90–3.20 (m, 8H, –C–CH2–CH2–), 7.15 (s, 4H,
–CH–Cp), 7.30–7.50 (m, 6H, –ArH), 7.60–7.80 (m, 4H, –ArH).
20 E. Samuel, Bull. Soc. Chim. Fr., 1966, 11, 3548.
T h i s j o u r n a l i s Q T h e R o y a l S o c i e t y o f C h e m i s t r y a n d t h e
C e n t r e N a t i o n a l d e l a R e c h e r c h e S c i e n t i f i q u e 2 0 0 4
656
N e w . J . C h e m . , 2 0 0 4 , 2 8 , 6 5 2 – 6 5 6