Discrete Ti?O?Ti Complexes: Visible-Light-Activated, Homogeneous Alternative to TiO2 Photosensitisers

Article abstract of DOI:10.1002/chem.202001678

A series of novel bimetallic Ti^IV amine bis(phenolate) complexes was synthesised and fully characterised. X-ray crystallography studies revealed distorted octahedral geometries around the Ti centres with single or double oxo-bridges connecting the two metals. These robust, air- and moisture-stable complexes were employed as photosensitisers generating singlet oxygen following irradiation with visible light (420 nm) LED module in a commercial flow reactor. All five complexes showed high activity in the photo-oxygenation of α-terpinene and achieved complete conversion to ascaridole in four hours at ambient temperature. The excellent selectivity of these photosensitisers towards ascaridole (vs. transformation to p-cymene) was demonstrated with control experiments using a traditional TiO₂ catalyst. Further comparative studies employing the free pro-ligands as well as a monometallic analogue highlighted the importance of the ‘TiO₂-like’ moiety in the polymetallic catalysts. Computational studies were used to determine the nature of the ligand to metal charge transfer (LMCT) states and singlet–triplet gaps for each complex, the calculated trends in the UV-vis absorption spectra across the series agreed well with the experimental results.

Full text of DOI:10.1002/chem.202001678

Accepted Article
Accepted Article
Title: Discrete Ti-O-Ti complexes: visible light-activated, homogeneous
alternative to TiO2 photosensitisers
Authors: Kira Behm, Eszter Fazekas, Martin Paterson, Filipe Vilela,
and Ruaraidh D. McIntosh
This manuscript has been accepted after peer review and appears as an
Accepted Article online prior to editing, proofing, and formal publication
of the final Version of Record (VoR). This work is currently citable by
using the Digital Object Identifier (DOI) given below. The VoR will be
published online in Early View as soon as possible and may be different
to this Accepted Article as a result of editing. Readers should obtain
the VoR from the journal website shown below when it is published
to ensure accuracy of information. The authors are responsible for the
content of this Accepted Article.
To be cited as: Chem. Eur. J. 10.1002/chem.202001678
Link to VoR: https://doi.org/10.1002/chem.202001678
01/2020

10.1002/chem.202001678
Chemistry - A European Journal
FULL PAPER
Discrete Ti-O-Ti complexes: visible light-activated, homogeneous
alternative to TiO2 photosensitisers
Kira Behm,[a] Eszter Fazekas,[a] Martin J. Paterson,[a] Filipe Vilela[a] and Ruaraidh D. McIntosh*[a]
[a]
K. Behm, Dr. E. Fazekas, Prof. M. J. Paterson, Dr. F. Vilela, Dr. R. D. McIntosh
Institute of Chemical Sciences
Heriot-Watt University
Edinburgh, EH14 4AS
United Kingdom
Abstract: A series of novel bimetallic TiIV amine bis(phenolate)
complexes was synthesised and fully characterised. X-ray
crystallography studies revealed distorted octahedral geometries
around the Ti centres with single or double oxo-bridges
connecting the two metals. These robust, air- and moisture-stable
complexes were employed as photosensitisers generating singlet
oxygen following irradiation with visible light (420 nm) LED in a
commercial flow reactor. All five complexes showed high activity
in the photooxygenation of α-terpinene and achieved complete
conversion to ascaridole in four hours at ambient temperature.
The excellent selectivity of these photosensitisers towards
ascaridole (vs. transformation to p-cymene) was demonstrated
with control experiments using traditional a TiO2 catalyst. Further
comparative studies employing the free pro-ligands as well as a
monometallic analogue highlighted the importance of the ‘TiO2-
like’ moiety in the polymetallic catalysts. Computational studies
were used to determine the nature of the ligand to metal charge
transfer (LMCT) states and singlet-triplet gaps for each complex,
the calculated trends in the UV-vis absorption spectra across the
series agreed well with the experimental results.
POTs for enhanced visible light harvesting include heterometallic
doping or the incorporation of simple functional ligands that
influence the absorption.[12]
While heterogeneous systems usually benefit from higher stability
and easier separation, the application of homogeneous
photocatalysts often allows better fine-tuning that leads to
enhanced activity and selectivity,[13] moreover they enable
convenient, solution-phase reaction monitoring and mechanistic
studies. In the past decades a variety of well-defined, soluble
molecular catalysts have been designed for photochemical
processes comprising photoorganocatalysts (including organic
dyes) and metal complexes.[14-16] The latter group is dominated by
polypyridyl complexes of Ru and Ir for photoredox reactions,[17-19]
while transition metal complexes of porphyrin and
phthalocyanine-derivatives have been frequently applied as
photosensitisers.[20-21]
An important photocatalytic reaction is the generation of singlet
oxygen (1O2), which can be employed as a green oxidising agent
in dye degradation, wastewater disinfection, cancer treatment and
synthetic processes.[22-24] This technique
requires
a
photosensitiser that can exist in a relatively long-lived triplet
excited state, by efficiently undergoing inter-system crossing
following absorption of a photon of the appropriate wavelength.[25-
Introduction
26]
The energy is consequently transferred onto ground-state
triplet oxygen molecules (3O2) converting them into metastable
1O2.[27-29] Most commonly, organic dyes such as methylene blue
have been applied as photosensitisers.[30-31] More recently,
complexes of precious metals (Ru, Re, Os, Ir and Pt) have also
been used, often as bimodal catalysts with large, light-harvesting
moieties incorporated into the ligand framework.[32-37]
Despite the popularity of TiO2 and POT-based systems, the
utilisation of discrete, well-defined complexes of Ti is
underexplored in the field of photocatalysis.[38-40] Our group has
had a longstanding interest the application of this abundant and
non-toxic metal, particularly in combination with amine
bis(phenolate) (ABP) ligands due to their convenient synthesis
and easily tailored stereoelectronic properties.[41-42] The
employment of Ti ABP complexes in photocatalysis is promising,
as they combine the benefits of TiO2-like motifs (via Ti-O and Ti-
O-Ti moieties) with the benefit of adjusting their light absorption
into the visible region (as indicated by their yellow-red colour).
Here, we report the first examples of bimetallic homogeneous
Photocatalytic transformations are increasingly attractive as they
offer environmentally friendly pathways for a range of chemical
processes including water purification, water splitting, CO2
conversion and organic syntheses.[1-4] The rapid rise of flow
photoreactors, coupled with affordable LED light-sources in the
past decade has further promoted the development of both
heterogeneous and homogeneous photocatalytic systems.[5]
Catalysts that display a strong response to visible light (400-700
nm) are desirable, as they are safer to use and enable the efficient
utilisation of natural sunlight whilst preventing potential side-
reactions and catalyst degradation that commonly occur under
high energy UV irradiation (< 400 nm).[2, 6]
Various forms of the ubiquitous titania (TiO2) catalysts have been
successfully applied in classical UV-activated photochemistry due
to their abundance, low toxicity and large bandgap energy.[7] To
extend the absorption into the visible region, however, TiO2 needs
to be doped with additives including metals (Cd, Ce, Mn, Bi etc.)
and non-metallic anions.[8-10] While the incorporation of these
(often expensive or toxic) components significantly increases the
photocatalytic efficiency, it can simultaneously lower the thermal
stability of the catalyst and create undesired electron traps.[10]
Other TiO2-like materials, such as polyoxotitanates (POTs) have
also been investigated as photocatalysts.[11] Modifications of
‘TiO2-like’
complexes
utilised
as
visible-light-activated
photosensitisers in the generation of 1O2.
1
This article is protected by copyright. All rights reserved.

10.1002/chem.202001678
Chemistry - A European Journal
FULL PAPER
Results and Discussion
Synthesis
A series of amine bis(phenol) pro-ligands (L1H2 – L4H2) was
synthesised via double Mannich condensations following
modified literature procedures.[43] Using water as solvent, the
desired compounds formed with high yields (69 - 87%) as white
powders following trituration in methanol (Figure 1).
OH
OH
OH
OH
R
R
N
L1H2: R = tBu (85 %)
L2H2: R = Me (68 %)
L3H2: R = Cl (69 %)
N
L4H2 (87 %)
Figure 3. Molecular structures of C2 (left) and C3 (right) with ellipsoids set at
the 50% probability level. Hydrogen atoms have been omitted for clarity.
Selected bond lengths for C2 (Å): Ti1–O1 1.9249(11), Ti1-O2 1.8827(13), Ti1-
O3 1.7748(12), Ti1-O3’ 1.9550(11), Ti1-N1 2.3261(14), Ti1-N2 2.2280(15).
Selected bond angles for C2 (°): O1-Ti1-N1 83.84(5), O1-Ti1-O3 105.35(5), O3-
Ti1–O3’ 83.54(5), N1-Ti1-N2 73.31(5), N2-Ti1-O2 160.71(5). Selected bond
lengths for C3 (Å): Ti1–O1 1.9332(12), Ti1-O2 1.8965(13), Ti1-O3 1.7743(12),
Ti1-O3’ 1.9495(12), Ti1-N1 2.3169(14), Ti1-N2 2.2063(15). Selected bond
angles for C3 (°): O1-Ti1-N1 85.63(5), O1-Ti1-O3 103.23(5), O3-Ti1–O3’
83.48(5), N1-Ti1-N2 73.82(5), N2-Ti1-O2 158.47(5).
N
R
R
O
Figure 1. Structures of amine bis(phenol) pro-ligands (L1H2-L4H2).
The corresponding TiIV complexes were formed using two
different metal precursors; C1, C2, C4 and C5 were synthesised
through the deprotonation of the pro-ligands with NaH, followed
by the addition of TiCl4, while C3 was formed via the direct
addition of Ti(OiPr)4 at ambient temperature. Coordination of the
ligands to the Ti centre was confirmed via the splitting of N-
methylene resonances (4.93 - 2.57 ppm for C1) in the 1H NMR
spectra. The bimetallic nature of the complexes was corroborated
using ESI mass spectrometry and single crystal X-ray
crystallography.
The
compounds
were
purified
via
recrystallisation from diethyl ether or chloroform to give
complexes C1 – C5 or with moderate to high yields (39 - 64%) as
yellow and red solids (Figure 2).
R
R
N
Ti
R
N
Ti
Figure 4. Molecular structures of C4 (left) and C5 (right) with ellipsoids set at
O
O
O
O
O
N
the 50% probability level. Hydrogen atoms have been omitted for clarity.
Selected bond lengths for C4 (Å): Ti1–O1 1.8555(15), Ti1-O2 1.8543(15), Ti1-
O3 1.8002(4), Ti1-O4 2.3046(16), Ti1-N1 2.2671(18), Ti1-Cl1 2.3312(6).
Selected bond angles for C4 (°): O1-Ti1-N1 83.36(7), O1-Ti1-O3 95.67(5), O3-
Ti1–Cl1 101.20(2), O1-Ti1-O2 163.69(7), O1-Ti1-Cl1 93.68(5). Selected bond
lengths for C5 (Å): Ti1–O1 1.884(4), Ti1-O2 1.889(4), Ti1-O3 1.8066(11), Ti1-
O4 2.305(4), Ti1-O5 1.807(4), Ti1-N1 2.288(5). Selected bond angles for C5 (°):
O1-Ti1-N1 83.26(18), O1-Ti1-O3 96.31(14), O3-Ti1–O5 104.96(14), O1-Ti1-O2
161.29(19), O1-Ti1-O5 94.79(19).
O
O
R
R
O
R
O
O
O
O
Ti
N
Ti
N
O
N
R
R
C1: R = tBu (64 %)
C2: R = Me (56 %)
C3: R = Cl (62 %)
C4: R = Cl (57 %)
C5: R = OMe (39 %)
All complexes exhibit a distorted octahedral geometry around the
metal centres, with N2-Ti-O2 (C2 and C3) and O1-Ti-O2 (C4 and
C5) bond angles ranging from 158.47(5)° to 163.69(7)°. Notably,
the phenolate groups take up a cis arrangement, when ligands
with methylpyridine side-arms were used (L1 – L3, Figure 3),
whereas using L4, the decreased steric bulk of the methoxyethyl
side-arm favours the phenolate groups arranged in a trans
position (Figure 4). Similar behaviour was observed in related Ti
complexes investigated by Mountford and co-workers.[43] The
metal-ligand bond lengths in the double-bridged complexes C2
and C3 are slightly longer than in complexes C4 and C5, which
could be attributed to the extra steric bulk of the methylpyridine
side-arm and the presence of the sterically strained four-
membered ring. For example, in C2, the Ti-phenolate bond
lengths are 1.9249(1) Å and 1.8827(1) Å, whereas in C4, the
corresponding values are 1.855(2) Å and 1.854(2) Å, respectively.
Overall, the bond lengths and bond angles in all four novel
bimetallic complexes are comparable to related Ti ABP
derivatives reported in the literature.[44-46]
Figure 2. Structures of bimetallic Ti-O-Ti complexes (C1-C5).
Crystallography
Single crystals, suitable for X-ray diffraction studies, were grown
via the slow evaporation of diethyl ether (C2 and C3), chloroform
(C4) or methanol (C5) solutions of the compounds. Using non-
dried crystallisation solvents has reliably led to the formation of
bimetallic complexes with Ti-O-Ti oxo-bridges. In C5, the
propoxide groups on the Ti centres were displaced with methoxy
ligands originating from the methanol solvent. Interestingly, for
ligands L1 – L3, double oxo-bridged bimetallic complexes (C1 –
C3) were formed, whereas for ligand L4, only single oxo-bridged
complexes C4 and C5 were obtained. A range of procedures were
explored in an attempt to form equivalent complexes with all
ligands, however, the difference in the number of oxo-bridges
persisted, regardless of the synthetic method or the crystallisation
solvent used.
2
This article is protected by copyright. All rights reserved.

10.1002/chem.202001678
Chemistry - A European Journal
FULL PAPER
UV-vis spectroscopy studies
Solution phase UV-vis spectra of complexes C1 – C5 were
recorded using chloroform as the solvent in order to determine the
absorption profiles of the compounds. The spectra of complexes
(e.g. C4, Figure 5) show a larger absorption band at 280 nm,
corresponding to the absorption maximum of the ligand, which
was confirmed with control experiments (Figure S18). A second,
weaker absorption band was observed at 378 nm, which
corresponds to a ligand metal charge transfer (LMCT). All dimeric
Ti complexes exhibited a similar absorption maximum (332-
378 nm, Table 1), with significant response detectable in the
visible region (>400 nm), which was also indicated by their yellow
to red colour. Importantly, this property allowed the
photosensitisation experiments to be carried out with irradiation at
the wavelength of 420 nm, which offers an advantage over
traditional TiO2 catalysts, such as anatase, that does not absorb
in the visible region.
1O2
O
O
CDCl3
-terpinene
ascaridole
p-cymene
Scheme 1. Synthesis of ascaridole from α-terpinene and competing side
reaction to form p-cymene.
Using 5 mol% catalyst loading, all five complexes showed good
activity in the production of 1O2 at ambient temperature (Figure 6).
α-Terpinene was selectively converted to ascaridole, the
formation of common by-products such as p-cymene was not
detected in the reaction mixtures.[49] On the other hand,
comparative studies using TiO2 (P25) under identical reaction
conditions afforded exclusively transformation into p-cymene as
indicated by the appearance of an aromatic resonance at
7.11 ppm in the 1H NMR spectra of reaction mixtures. The lack of
oxygenated products in these experiments confirms that
traditional TiO2 catalysts are not suitable for 1O2 generation under
visible light irradiation. Notably, control experiments in the
absence of photosensitisers achieved a low conversion of 4% in
three hours, which can be attributed to the commonly observed
self-sensitisation phenomenon.[34] These results highlight that the
presence of the Ti complexes was essential for efficient 1O2
generation.
Ligand
absorption
0.3
0.25
0.2
0.15
0.1
Ligand metal
charge transfer
0.05
0
250
300
350
400
450
500
550
600
Wavelength (nm)
Figure 5. UV-vis spectrum of C4 showing the ligand and complex absorption
maxima.
100
90
80
70
60
50
40
30
20
10
0
Table 1. UV-vis absorption maxima of complexes C1-C5.
Complex
C1

max (nm)[a]
334
C2
334
C3
332
C4
378
0
50
100
150
200
250
C5
338
Time (mins)
[a]Spectra were recorded with 0.0003 M solution of the complexes in CHCl3 at
ambient temperature.
C1
C3
C5
C2
C4
no sensitiser
Figure 6. Conversion of α-terpinene to ascaridole vs. time.
Photocatalytic activity
Complexes C1 – C5 were screened for activity as triplet
photosensitisers in the generation of singlet oxygen using a
commercial flow photoreactor under visible light (420 nm)
irradiation. The activity was monitored via the photooxygenation
reaction of α-terpinene towards ascaridole, which only proceeds
in the presence of 1O2 (Scheme 1).[47] All experiments were
carried out in CDCl3 solvent as deuterated solvents are known to
be beneficial for increased 1O2 lifetime and enable convenient
The most active photosensitiser was complex C5, achieving full
conversion in two hours (Figure 7). Kinetic studies with samples
taken at 15 minute intervals showed a near linear reaction profile,
which revealed a pseudo first order dependence on substrate
concentration, comparable kinetic profiles were previously
observed in the photooxygenation of -terpinene.[50-51] Similar
rates were observed using C1, which reached full conversion of
the α-terpinene in three hours. While these reaction rates
generally fall below those achieved with precious metal (Re and
Ir) photosensitisers,[32, 52-53] it is clearly demonstrated that these
simple and stable Ti ABP complexes are capable to reach 1O2
generation activity levels comparable to TiO2 and POT-based
systems.
sampling for the determination of conversion via 1H NMR
48]
spectroscopy (Figure S20).[27,
The reaction mixtures were
saturated with O2 prior to the experiments and flow rates of both
the substrate solution and a continuous oxygen supply were kept
constant at 1 mL min-1. A schematic of the flow reactor setup can
be found in the supporting information (Figure S19).
3
This article is protected by copyright. All rights reserved.

10.1002/chem.202001678
Chemistry - A European Journal
FULL PAPER
100.00
90.00
80.00
70.00
60.00
50.00
40.00
30.00
20.00
10.00
0.00
6
5
4
3
2
1
0
C5
L4H2
0
20
40
60
80
100
120
0
100
200
300
400
Time (mins)
Time (mins)
Figure 7. α-Terpinene ln(C0/C) vs. time using complex C5.
Figure 9. α-Terpinene conversion in the presence of C5 vs. L4H2.
Complexes C2, C3 and C4 exhibit a slower rate in the first 30
minutes of the reaction, therefore full conversion is reached in up
to four hours. This lag period suggested an initial transformation
of these photosensitisers into a different, catalytically species.
However, the photostability of C4 was investigated under 420 nm
irradiation for 24 hours and the complex remained unchanged
Computational studies
Each of the target molecules (C2 – C5) was optimised using CAM-
B3LYP/6-311G(d,p) starting from the crystal structure coordinates.
A polarizable continuum model (PCM) was used to model a
chloroform solvent (dielectric constant
= 4.7113). Time-
dependent (TD) CAM-B3LYP was used to determine the excited
singlet and triplet electronic states. This has previously been
shown to compare well to high-order correlated wavefunction
response approaches,[54] and perform well in describing neutral
and charged TiO2 clusters.[55] S0 and T1 states were optimised and
the analytical Hessian was confirmed as positive definite to verify
the structures as minima. The nature of the excited electronic
states was determined using the response eigenvectors with the
canonical Kohn-Sham orbitals. All computations were performed
using a local version of Gaussian16.[56]
1
according to H NMR spectroscopy studies. Furthermore, mass
spectrometry analysis of the reaction mixtures after the
photocatalytic experiment verified the presence of the initial
bimetallic complexes. It was therefore concluded that the active
photosensitisers were the dimeric Ti-O-Ti bridged structures. It
appears that there is an acceleration in reaction rate at higher
conversions, which can be attributed to a combination of different
effects, such as some autocatalytic behaviour of the ascaridole
formed, or the increased ratio of 1O2 versus α-terpinene.
With respect to structural properties, tert-butyl groups on the
ligands showed a positive impact on the production of O2, likely
due to the increased solubility of these complexes (C1, C4, C5).
All optimised geometries agree well with the crystal structures.
The calculated absorption spectra agree reasonably well with the
experimental results for clusters of this size (with a general blue-
shift). All excited states are very mixed with many LMCT
transitions contributing to each state. The LMCT transitions are
from  orbitals, localised on the aromatic units, to orbitals of Ti
localised d-character mixed with higher * orbitals (Figure 10).
1
tBu
tBu
N
tBu
O
O
N
tBu
Ti
N
tBu
O
O
N
O
tBu
Ti
Cl
O
Cl
O
O
Ti
N
tBu
N
tBu
tBu
Figure 8. Structures of monometallic vs. bimetallic Ti complexes of ligand L1H2.
Importantly, it was shown that the presence of the Ti-O-Ti moiety
is crucial for the efficient photosensitisation: Control reactions with
a monometallic complex (L1)TiCl2 (Figure 8) only achieved 18%
conversion in four hours cf. 100% was achieved in three hours
with the bimetallic µ-oxo-bridged analogue C1. Moreover, the
remarkable conversion achieved with C5 showed, that the activity
is enhanced by the methoxy groups on the metal centres, which
further extend the oxo-bridged framework. Control experiments
revealed that the free pro-ligands (L1H2, L3H2 and L4H2) also
exhibit some catalytic activity under these reaction conditions,
however, they yielded similarly low conversions (9 - 28% after
4 hrs).
Figure 10. (TD)-CAM-B3LYP calculated absorption spectra for target molecules
C2-C5. Inset are dominant particle-hole transitions for C2 and C3 highlighting
the nature of the bright electronic states for each species through LMCTtod*
transitions. Spectra generated by convoluting a Gaussian (FWHM=20 nm) over
lowest 30 singlet states.
Figure 9 compares the activities of C5 and the corresponding pro-
ligand L4H2. Experiments without oxygen bubbling showed that it
is essential to have a constant supply of O2 to achieve high
conversions.
The dominant particle-hole pair is shown in Table 2 (for each
molecule this is around 40 - 50% of the state). There is a large
density of low-lying electronic states and for all molecules the
lowest excited state is generally a quasi-degenerate pair of one
bright and one dark state (Table 2). For C4 and C5 the brightest
state is not this lowest pair but a (slightly) higher lying one. Each
molecule C2 - C4 has a low-lying triplet state that is well separated
4
This article is protected by copyright. All rights reserved.

10.1002/chem.202001678
Chemistry - A European Journal
FULL PAPER
Table 2. (TD)-CAM-B3LYP calculated optical properties for target molecules.[a]
State
S1/S2
S1/S2
S1/S2
S10
Evert / eV
4.14
Character
LMCT -d*
LMCT -d*
LMCT -d*
LMCT -d*
LMCT -d*
LMCT -d*
p-h
osc. strength (f)
0.3390
ES/T / eV
C2
C3
C4
H-1 ➜ L
H-1 ➜ L
H ➜ L
2.36
2.39
2.38
-
4.23
0.4336
3.17
0.0021
3.87
H-1 ➜ L
H-1 ➜ L
H ➜ L
0.5332
C5
S1/S2
S3
3.82
0.1961
2.41
-
3.98
0.5116
[a]
Vertical excitation energies to lowest and the brightest state in the spectral range up to 250 nm, associated oscillator strengths, character, particle-hole nature
relative to HOMO and LUMO, and (radiative) singlet-triplet energy gaps from lowest triplet T1 state.
from all other excited electronic states. All molecules show a
similar energy gap from this triplet to the ground state singlet. For
each molecule the T1 state has a significant component of HOMO-
LUMO particle-hole character, and like the singlet manifold these
involve LMCT (d*) character. Consistent with this the triplet
states have elongated Ti-O elongated Ti-O bonds relative to the
ground state structures but generally a similar topology.
spectroscopy data was acquired with a Bruker AVIII 300 MHz
instrument or Bruker AVIII 400 MHz instrument at 298 K in CDCl3.
Electrospray ionisation mass spectrometry (ESI) was recorded
using a Bruker micrOTOF II. Solution-state UV-Vis spectra were
recorded on a Shimadzu UV-2550 system with 10 mm quartz
cuvettes.
General procedure for the synthesis of pro-ligands L1H2
L4H2:
–
Pro-ligands L1H2
– L4H2 were reported previously and were
Conclusion
prepared using an adapted literature procedure.[57-60]
Phenol (2,4-di-tert-butylphenol, 2,4-dimethylphenol or 2,4-
dichlorophenol, 24.5 mmol) was suspended in water (50 mL). 2-
picolylamine or 2-methoxyethylamine (12.3 mmol) and aqueous
formaldehyde solution (37% wt., 2.0 mL, 25 mmol) were added
and the mixture was heated at reflux for 48 hours. The suspension
was filtered and the resulting off-white solid was triturated in
methanol to yield the pro-ligands as a white powder.
Data for L1H2
In conclusion, a series of five novel bimetallic Ti-O-Ti bridged
amine bis(phenolate) complexes were synthesised and employed
as photosensitisers in the generation of 1O2. Unlike their
heterogeneous TiO2-based counterparts, these soluble and well-
defined aggregates showed significant response to visible light,
which allowed them to be screened as photosensitisers in a
commercial flow reactor under 420 nm LED irradiation. All
complexes have efficiently converted -terpinene to ascaridole
between two to four hours of residence time, moreover excellent
selectivity (>99%) towards ascaridole (vs. p-cymene) was
achieved as confirmed via control experiments using TiO2.
Comparative studies with the free pro-ligand and a monometallic
analogue showed that the presence of the ‘TiO2-like’ Ti-O-Ti
moiety is crucial to achieve efficient photooxygenation. The LMCT
states and singlet-triplet gaps for each complex were investigated
using computational methods, the calculated UV-vis absorption
spectra agreed reasonably well with the experimental
observations. These results highlight that smaller, modular
complexes of TiIV can be competitive alternatives of the widely
researched TiO2 and POT photocatalyst systems. Moreover, fine-
tuning of the ligand design may extend the absorption further into
the visible region, which would allow the utilisation of this
inexpensive and non-toxic metal to replace Ru- and Ir-based
catalysts in a wider scope of photochemical reactions. The
1
(5.82 g, 85 %). H NMR (300 MHz, CDCl3): δ 10.54 (s, 2H, OH),
8.69 (m, 1H, ArH), 7.69 (td, J = 7.7, 1.8 Hz, 1H, ArH), 7.28 (m, 1H,
ArH), 7.23 (d, J = 2.5 Hz, 2H, ArH), 7.12 (d, J = 7.8 Hz, 1H, ArH),
6.94 (d, J = 2.5 Hz, 2H, ArH), 3.85 (s, 2H, Py-CH2), 3.81 (s, 4H,
Ar-CH2), 1.41 (s, 18H, CCH3), 1.29 (s, 18H, CCH3).
Data for L2H2
1
(3.33 g, 68 %). H NMR (300 MHz, CDCl3): δ 10.44 (s, 2H, OH),
8.71 (m, 1H, ArH), 7.70 (td, J = 7.7, 1.8 Hz, 1H, ArH), 7.28 (m, 1H,
ArH), 7.12 (d, J = 7.8 Hz, 1H, ArH), 6.87 (s, 2H, ArH), 6.71 (s, 2H,
ArH), 3.84 (s, 2H, Py-CH2), 3.75 (s, 4H, Ar-CH2), 2.22 (s, 12H,
CH3).
Data for L3H2
1
(3.88 g, 69 %). H NMR (300 MHz, CDCl3): δ 11.36 (s, 2H, OH),
8.72 (m, 1H, ArH), 7.80 (td, J = 7.7, 1.8 Hz, 1H, ArH), 7.36 (m,
1H, ArH), 7.30 (d, J = 2.5 Hz, 2H, ArH), 7.19 (d, J = 7.9 Hz, 1H,
ArH), 6.97 (d, J = 2.5 Hz, 2H, ArH), 3.87 (s, 2H, Py-CH2), 3.81 (s,
4H, Ar-CH2).
Data for L4H2
apparent
interplay
between
activity
and
ligand
1
(5.39 g, 87 %). H NMR (300 MHz, CDCl3): δ 8.48 (bs, 2H, OH),
substituents/bridges between the Ti centres will be further
investigated along with the breadth of reactivity in photocatalysis
that these complexes can display.
7.21 (d, J = 2.5 Hz, 2H, ArH), 6.88 (d, J = 2.4 Hz, 2H, ArH), 3.74
(s, 4H, Ar-CH2), 3.56 (t, J = 5.2 Hz, 2H, CH2), 3.47 (s, 3H, OCH3),
2.74 (t, J = 5.2 Hz,2H, CH2), 1.41 (s, 18H, CCH3), 1.27 (s, 18H,
CCH3).
Experimental
General procedure for the synthesis of amine bis(phenolate)
Ti-O-Ti complexes (C1
– C5):
General considerations
C1, C2 and C4: Pro-ligands L1H2, L2H2 and L4H2 (1.0 mmol)
were dissolved in dry toluene (30 mL). Sodium hydride (60% wt.
dispersion in mineral oil, 0.08 g, 2.0 mmol) was added and the
solution was stirred for 30 minutes. TiCl4 (1M solution in toluene,
1.0 mL, 1.0 mmol) was added slowly. The mixture was stirred for
three hours, the solvent was removed under reduced pressure
Starting materials were purchased and used as received from
Merck, Acros and Fluorochem. Unless stated otherwise,
experiments were carried out under ambient atmosphere. Dry
solvents were purified in an MBRAUN SPS-800 and stored over
activated 4 Å molecular sieves under a dry N2 atmosphere. NMR
5
This article is protected by copyright. All rights reserved.

10.1002/chem.202001678
Chemistry - A European Journal
FULL PAPER
and the crude product was dissolved in methanol (50 mL). Sodium
methoxide (0.11 g, 2.0 mmol) was added and the solution was
stirred for four hours. The crude product was washed with
deionised water and recrystallised from diethyl ether. (The
addition of sodium methoxide was omitted for C4 and the crude
product was recrystallised from chloroform).
C3 and C5: Pro-ligands L3H2 and L4H2 (1.0 mmol) were dissolved
in THF (20 mL). Ti(OiPr)4 (0.28 g, 0.30 mL, 1.0 mmol) was added
and the solution was stirred for two hours. The solvent was
removed under reduced pressure and the crude product was
recrystallised from diethyl ether or methanol. Structures for
complexes C2 – C5 can be found under CCDC numbers 1992707
– 1992710.
Data for C4
(0.35 g, 57 %). 1H NMR (300 MHz, CDCl3): δ 7.09 (d, J = 2.3 Hz,
4H, ArH), 6.91 (d, J = 2.2 Hz, 4H, ArH), 5.44 (d, J = 14.0 Hz, 4H,
CH2), 3.48 (s, 2H, CH2), 3.44 (s, 2H, CH2), 3.42 (s, 6H, OCH3),
3.19 (t, J = 5.5 Hz, 4H, CH2), 2.86 (t, J = 5.4 Hz, 4H, CH2), 1.27
(s, 36H, CCH3), 1.05 (s, 36H, CCH3). 13C NMR (75.5 MHz, CDCl3):
δ 159.7 (C), 159.5 (C), 140.7 (C), 136.4 (C), 136.2 (C), 134.8 (C),
124.6 (C), 124.3 (C), 124.1 (CH), 123.8 (CH), 123.7 (CH), 123.2
(CH), 66.2 (CH2), 65.9 (CH2), 63.6 (CH2), 58.3 (CH2), 35.4 (CCH3),
35.2 (CCH3), 34.6 (CCH3), 34.4 (CCH3), 31.8 (OCH3), 30.3 (CH3),
30.1 (CH3), 29.9 (CH3), 29.2 (CH3). MS (ESI): m/z
[C33H50ClNO4Ti]+• 608.2923 (monometallic oxo radical fragment).
Data for C5
Data for (L1)TiCl2
(0.22 g, 39 %). 1H NMR (300 MHz, CDCl3): δ 7.16 (d, J = 2.4 Hz,
4H, ArH), 6.71 (d, J = 2.4 Hz, 4H, ArH), 4.91 (d, J = 13.2 Hz, 4H,
CH2), 3.98 (s, 6H, OCH3), 3.28 (s, 6H, OCH3), 3.03 (s, 2H, CH2),
2.99 (s, 2H, CH2), 2.86 (t, J = 5.8 Hz, 4H, CH2), 2.57 (t, J = 5.8 Hz,
4H, CH2), 1.53 (s, 6H, CCH3), 1.48 (s, 30H, CCH3), 1.29 (s, 6H,
CCH3), 1.24 (s, 30H, CCH3). 13C NMR (100.6 MHz, CDCl3): δ
160.3 (C), 159.8 (C), 139.0 (C), 138.5 (C), 135.8 (C), 134.9 (C),
124.7 (C), 124.3 (C), 124.2 (CH), 123.5 (CH), 123.0 (CH), 71.9
(CH2), 64.3 (CH2), 63.2 (CH2), 62.0 (CH2), 35.4 (CCH3), 35.1
(CCH3), 34.2 (CCH3), 34.1 (CCH3), 32.0 (OCH3), 31.9 (OCH3),
30.7 (CH3), 30.2 (CH3), 30.1 (CH3), 30.0 (CH3). HRMS (ESI): m/z
[M+H]+ 1193.7087; calcd [M+H]+ 1193.7087.
(L1)TiCl2 (0.44 g, 69 %) was prepared according to literature
procedures[61] for comparison of photosensitising properties.
1H NMR (400 MHz, CDCl3): δ 9.04 (m, 1H, ArH), 7.52 (td, J = 7.7,
1.6 Hz, 1H, ArH), 7.29 (m, 1H, ArH), 7.05 (d, J = 2.2 Hz, 1H, ArH),
6.90 (d, J = 2.2 Hz, 1H, ArH), 6.79 (d, J = 2.3 Hz, 1H, ArH), 5.39
(d, J = 13.1 Hz, 1H, Py-CH2), 5.01 (d, J = 14.8 Hz, 1H, Py-CH2),
3.91 (m, 2H, Ar-CH2), 3.55 (d, J = 13.1 Hz, 1H, Ar-CH2), 3.08 (d,
J = 13.0 Hz, 1H, Ar-CH2), 1.57 (s, 8H, CCH3), 1.42 (s, 2H, CCH3),
1.33 (s, 8H, CCH3), 1.30 (s, 8H, CCH3), 1.27 (s, 2H, CCH3), 1.11
(s, 8H, CCH3).
Data for C1
(0.39 g, 64 %). 1H NMR (400 MHz, CDCl3): δ 8.65 (m, 2H, ArH),
7.17 (t, J = 7.7 Hz, 2H, ArH), 7.02 (d, J = 2.0 Hz, 2H, ArH), 6.88
(d, J = 2.0 Hz, 2H, ArH), 6.84 (s, 1H, ArH), 6.83 (s, 1H, ArH), 6.67
(m, 4H, ArH), 6.46 (d, J = 7.7 Hz, 2H, ArH), 4.93 (d, J = 13.7 Hz,
2H, CH2), 4.75 (d, J = 12.6 Hz, 2H, CH2), 3.50 (m, 4H, CH2), 3.16
(d, J = 11.8 Hz, 2H, CH2), 2.57 (d, J = 13.0 Hz, 2H, CH2), 1.31 (s,
18 H, CCH3), 1.24 (s, 18H, CCH3), 1.06 (s, 18H, CCH3), 0.90 (s,
18H, CCH3). 13C NMR (75.5 MHz, CD2Cl2): δ 161.2 (C), 159.1 (C),
156.7 (C), 151.7 (CH), 139.8 (C), 138.5 (C), 137.9 (CH), 136.2
(C), 134.4 (C), 126.1 (C), 125.8 (C), 124.9 (CH), 124.1 (CH),
123.7 (CH), 122.8 (CH), 122.5 (CH), 121.5 (CH), 64.1 (CH2), 62.4
(CH2), 59.3 (CH2), 35.2 (CCH3), 35.0 (CCH3), 34.3 (CCH3), 34.0
(CCH3), 31.9 (CH3), 31.8 (CH3), 30.4 (CH3), 30.3 (CH3). HRMS
(ESI): m/z [M+H]+ 1213.6652; calcd [M+H]+ 1213.6675.
Data for C2
General procedure of 1O2 experiments under flow conditions
Photosensitised reactions under flow conditions were carried out
using a commercial photochemical flow reactor equipped with an
LED array emitting a wavelength at 420 nm with 61 W light output
and a light intensity of 10.2 W.cm-2 (easy-Photochem, Vapourtec
Ltd.). A schematic of the flow experiment setup can be found in
the supporting information (Figure S19). Titanium complex (C1 -
C5) (12 mg, 5 mol%) and α-terpinene (0.16-0.26 mmol) were
dissolved in CDCl3 (12 mL). The solution was saturated with O2
for ten minutes. The solution was then pumped through the
photochemical reactor at 1 mL min-1. O2 was pumped through a
second pump at the same flow rate, mixing with the solution at a
T-junction before entering the photochemical reactor. Samples
(450 μL) were taken at regular intervals.
(0.24 g, 56 %). 1H NMR (400 MHz, CDCl3): δ 8.66 (m, 2H, ArH),
7.36 (t, J = 7.6 Hz, 2H, ArH), 6.73 (m, 8H), 6.37 (d, J = 10.0 Hz,
4H, ArH), 4.73 (d, J = 12.6 Hz, 2H, CH2), 4.63 (d, J = 14.0 Hz, 2H,
CH2), 3.78 (d, J = 12.9 Hz, 2H, CH2), 3.55 (d, J = 15.0 Hz, 2H,
CH2), 3.26 (d, J = 12.1 Hz, 2H, CH2), 2.66 (d, J = 12.6 Hz, 2H,
CH2), 2.23 (s, 6H), 1.97 (s, 6H) 1.90 (s, 6H), 1.52 (s, 6H). Due to
low solubility in conventional NMR solvents, a 13C NMR spectrum
with appropriate quality could not be recorded. HRMS (ESI): m/z
[M+H]+ 877.2903; calcd [M+H]+ 877.2919.
Crystallography: Single-crystal X-ray diffraction data were
collected on a Rigaku Oxford diffraction SuperNova diffractometer
or a Bruker venture d8 diffractometer with CCD detector with
Mo-Kα radiation (λ = 0.7107 Å) or Cu-Kα radiation (λ = 1.5418 Å).
The structures were solved by direct methods using SHELXS or
SHELXT and refined by full-matrix least squares on F2 using
SHELXL interfaced through Olex2.[62-63] Molecular graphics for all
structures were generated using Mercury.
Data for C3
(0.36 g, 62 %). 1H NMR (300 MHz, CDCl3): δ 8.57 (m, 2H, ArH),
7.57 (td, J = 7.9, 1.6 Hz, 2H, ArH), 7.19 (d, J = 2.7 Hz, 2H, ArH),
7.01 (d, J = 2.5 Hz, 2H, ArH), 6.98 (s, 1H, ArH), 6.95 (s, 1H, ArH),
6.89 (m, 2H, ArH), 6.82 (d, J = 2.4 Hz, 2H, ArH), 6.68 (d,
J = 2.5 Hz, 2H, ArH), 4.73 (d, J = 15.6 Hz, 2H, CH2), 4.66 (d,
J = 13.0 Hz, 2H, CH2), 3.74 (d, J = 11.7 Hz, 2H, CH2), 3.63 (d,
J = 15.1 Hz, 2H, CH2), 3.33 (d, J = 13.5 Hz, 2H, CH2), 2.82 (d,
J = 12.8 Hz, 2H, CH2). Due to solubility issues, a 13C NMR
spectrum could not be recorded. HRMS (ESI): m/z [M+H]+
1036.8566; calcd [M+H]+ 1036.8549.
Acknowledgements
We gratefully acknowledge the EPSRC (EP/S005781/1) and
Heriot-Watt University for funding. Further, we would like to thank
Mary Jones for her assistance with the flow photoreactor,
Georgina Rosair (Heriot-Watt University) and Gary Nichol
(University of Edinburgh) for crystal structure collections.
Keywords: Titanium • Amine bis(phenolate) • Flow Chemistry •
Photosensitiser • Singlet oxygen
6
This article is protected by copyright. All rights reserved.

10.1002/chem.202001678
Chemistry - A European Journal
FULL PAPER
[40] Z. Zhang, T. Hilche, D. Slak, N. R. Rietdijk, U. N. Oloyede, R. A.
Flowers, A. Gansauer, Angew. Chem. Int. Ed. 2020, 59.
[41] J.-M. E. P. Cols, C. E. Taylor, K. J. Gagnon, S. J. Teat, R. D.
McIntosh, Dalton Trans. 2016, 45, 17729-17738.
[42] J.-M. E. P. Cols, V. G. Hill, S. K. Williams, R. D. McIntosh,
Dalton Trans. 2018, 47, 10626-10635.
References
[1] M. A. Rauf, S. S. Ashraf, Chem. Eng. J. 2009, 151, 10-18.
[2] B. König, Eur. J. Org. Chem. 2017, 2017, 1979-1981.
[3] X. Chang, T. Wang, J. Gong, Energy Environ. Sci. 2016, 9,
2177-2196.
[4] J. Corredor, M. J. Rivero, C. M. Rangel, F. Gloaguen, I. Ortiz, J.
Chem.l Technol. Biotechnol. 2019, 94, 3049-3063.
[5] K. Mizuno, K. Kakiuchi, in Flow Chemistry: Integrated
Approaches for Practical Applications, The Royal Society of
Chemistry, 2020, pp. 129-152.
[43] C. L. Boyd, T. Toupance, B. R. Tyrrell, B. D. Ward, C. R. Wilson,
A. R. Cowley, P. Mountford, Organometallics 2005, 24, 309-330.
[44] A. J. Chmura, M. G. Davidson, M. D. Jones, M. D. Lunn, M. F.
Mahon, A. F. Johnson, P. Khunkamchoo, S. L. Roberts, S. S. F.
Wong, Macromolecules 2006, 39, 7250-7257.
[6] L. Marzo, S. K. Pagire, O. Reiser, B. König, Angew. Chem. Int.
Ed. 2018, 57, 10034-10072.
[7] J. Schneider, M. Matsuoka, M. Takeuchi, J. Zhang, Y. Horiuchi,
[45] S. Barroso, F. Madeira, M. J. Calhorda, M. J. Ferreira, M. T.
Duarte, A. M. Martins, Inorg. Chem. 2013, 52, 9427-9439.
[46] A. J. Nielson, J. M. Waters, Polyhedron 2010, 29, 1715-1726.
M. Anpo, D. W. Bahnemann, Chem. Rev. 2014, 114, 9919-9986. [47] J. M. Tobin, T. J. D. McCabe, A. W. Prentice, S. Holzer, G. O.
[8] M. Pawar, S. Topcu Sendoğdular, P. Gouma, J. Nanomat. 2018,
2018, 1-13.
Lloyd, M. J. Paterson, V. Arrighi, P. A. G. Cormack, F. Vilela,
ACS Catal. 2017, 7, 4602-4612.
[48] M. Bregnhøj, M. Westberg, F. Jensen, P. R. Ogilby, Phys. Chem.
Chem. Phy. 2016, 18, 22946-22961.
[9] R. Chauhan, A. Kumar, R. P. Chaudhary, Spectrochim. Acta A
2012, 98, 256-264.
[10] R. Daghrir, P. Drogui, D. Robert, Ind. Eng. Chem. Res. 2013,
52, 3581-3599.
[49] K. S. Elvira, R. C. R. Wootton, N. M. Reis, M. R. Mackley, A. J.
deMello, ACS Sust. Chem. Eng. 2013, 1, 209-213.
[11] J. Zhang, W. Hu, J. Zhang, S. Liu, J. Tong, X. Hou, W. Liu, J.
Yang, B. Liu, J. Phys. Chem. C 2017, 121, 18326-18332.
[12] N. Li, P. D. Matthews, H.-K. Luo, D. S. Wright, Chem. Commun.
2016, 52, 11180-11190.
[13] C. Xu, P. Ravi Anusuyadevi, C. Aymonier, R. Luque, S. Marre,
Chem. Soc. Rev. 2019, 48, 3868-3902.
[50] J. Shen, R. Steinbach, J. M. Tobin, M. Mouro Nakata, M. Bower,
M. R. S. McCoustra, H. Bridle, V. Arrighi, F. Vilela, Appl. Catal.,
B 2016, 193, 226-233.
[51] J. M. Tobin, J. Liu, H. Hayes, M. Demleitner, D. Ellis, V. Arrighi,
Z. Xu, F. Vilela, Polym. Chem. 2016, 7, 6662-6670.
[52] C. Y. Park, Y. J. Kim, H. J. Lim, J. H. Park, M. J. Kim, S. W. Seo,
C. P. Park, RSC Adv. 2015, 5, 4233-4237.
[53] J. Sun, J. Zhao, H. Guo, W. Wu, Chem. Commun. 2012, 48,
[14] D. Ravelli, M. Fagnoni, A. Albini, Chem. Soc. Rev. 2013, 42, 97-
113.
4169-4171.
[15] I. K. Sideri, E. Voutyritsa, C. G. Kokotos, Org. Biomol. Chem.
2018, 16, 4596-4614.
[54] D. J. Taylor, M. J. Paterson, J. Chem. Phys. 2010, 133, 204302.
[55] D. J. Taylor, M. J. Paterson, Chem. Phys. 2012, 408, 1-10.
[56] R. A. Gaussian 16, M. J. Frisch, G. W. Trucks, H. B. Schlegel,
G. E. Scuseria, M. A. Robb, J. R. Cheeseman, G. Scalmani, V.
Barone, G. A. Petersson, H. Nakatsuji, X. Li, M. Caricato, A. V.
Marenich, J. Bloino, B. G. Janesko, R. Gomperts, B. Mennucci,
H. P. Hratchian, J. V. Ortiz, A. F. Izmaylov, J. L. Sonnenberg, D.
Williams-Young, F. Ding, F. Lipparini, F. Egidi, J. Goings, B.
Peng, A. Petrone, T. Henderson, D. Ranasinghe, V. G.
Zakrzewski, J. Gao, N. Rega, G. Zheng, W. Liang, M. Hada, M.
Ehara, K. Toyota, R. Fukuda, J. Hasegawa, M. Ishida, T.
Nakajima, Y. Honda, O. Kitao, H. Nakai, T. Vreven, K. Throssell,
J. A. Montgomery, Jr., J. E. Peralta, F. Ogliaro, M. J. Bearpark,
J. J. Heyd, E. N. Brothers, K. N. Kudin, V. N. Staroverov, T. A.
Keith, R. Kobayashi, J. Normand, K. Raghavachari, A. P.
Rendell, J. C. Burant, S. S. Iyengar, J. Tomasi, M. Cossi, J. M.
Millam, M. Klene, C. Adamo, R. Cammi, J. W. Ochterski, R. L.
Martin, K. Morokuma, O. Farkas, J. B. Foresman, and D. J. Fox,
Gaussian, Inc., Wallingford CT, 2016.
[57] Y. Shimazaki, S. Huth, A. Odani, O. Yamauchi, Angew. Chem.
Int. Ed. 2000, 39, 1666-1669.
[58] L.-Z. Fu, L.-L. Zhou, Q.-N. Liang, C. Fang, S.-Z. Zhan,
Polyhedron 2016, 107, 83-88.
[59] A. M. Reckling, D. Martin, L. N. Dawe, A. Decken, C. M. Kozak,
J. Organomet. Chem. 2011, 696, 787-794.
[60] E. Y. Tshuva, S. Groysman, I. Goldberg, M. Kol,
Organometallics 2002, 21, 662-670.
[16] M. A. Theodoropoulou, N. F. Nikitas, C. G. Kokotos, Beilstein J.
Org. Chem. 2020, 16, 833-857.
[17] C. K. Prier, D. A. Rankic, D. W. C. MacMillan, Chem. Rev. 2013,
113, 5322-5363.
[18] A. A. Abdel-Shafi, P. D. Beer, R. J. Mortimer, F. Wilkinson, J.
Phys. Chem. A 2000, 104, 192-202.
[19] W. Zhang, B. Li, H. Ma, L. Zhang, Y. Guan, Y. Zhang, X. Zhang,
P. Jing, S. Yue, ACS Appl. Mater. Interfaces 2016, 8, 21465-
21471.
[20] R. C. H. Wong, P.-C. Lo, D. K. P. Ng, Coord. Chem. Rev. 2019,
379, 30-46.
[21] T. Punniyamurthy, S. Velusamy, J. Iqbal, Chem. Rev. 2005, 105,
2329-2364.
[22] P. R. Ogilby, Chem. Soc. Rev. 2010, 39, 3181-3209.
[23] I. Pibiri, S. Buscemi, A. Palumbo Piccionello, A. Pace,
ChemPhotoChem 2018, 2, 535-547.
[24] A. A. Ghogare, A. Greer, Chem. Rev. 2016, 116, 9994-10034.
[25] J. Zhao, W. Wu, J. Sun, S. Guo, Chem. Soc. Rev. 2013, 42,
5323-5351.
[26] F. Strieth-Kalthoff, M. J. James, M. Teders, L. Pitzer, F. Glorius,
Chem. Soc. Rev. 2018, 47, 7190-7202.
[27] C. Schweitzer, R. Schmidt, Chem. Rev. 2003, 103, 1685-1758.
[28] W. Fudickar, T. Linker, J. Phys. Org. Chem. 2019, 32.
[29] E. L. Clennan, Tetrahedron 2000, 56, 9151-9179.
[30] M. C. DeRosa, R. J. Crutchley, Coord. Chem. Rev. 2002, 233-
234, 351-371.
[61] C. L. Boyd, T. Toupance, B. R. Tyrrell, B. D. Ward, C. R. Wilson,
A. R. Cowley, P. Mountford, Organometallics 2005, 24, 309-330.
[62] G. M. Sheldrick, Acta Crystallogr., Sect. C: Cryst. Struct. Chem.
2015, 71, 3-8.
[31] A. Mishra, M. K. Fischer, P. Bauerle, Angew. Chem. Int. Ed.
2009, 48, 2474-2499.
[32] X. Yi, J. Zhao, J. Sun, S. Guo, H. Zhang, Dalton Trans. 2013,
42, 2062-2074.
[63] O. V. Dolomanov, L. J. Bourhis, R. J. Gildea, J. A. K. Howard,
H. Puschmann, J. Appl. Crystallogr. 2009, 42, 339-341.
[33] A. A. Abdel-Shafi, J. L. Bourdelande, S. S. Ali, Dalton Trans.
2007, 2510-2516.
[34] N. Manav, P. E. Kesavan, M. Ishida, S. Mori, Y. Yasutake, S.
Fukatsu, H. Furuta, I. Gupta, Dalton Trans. 2019, 48, 2467-2478.
[35] V. Prusakova, C. E. McCusker, F. N. Castellano, Inorg. Chem.
2012, 51, 8589-8598.
[36] A. A. Abdel-Shafi, D. R. Worrall, A. Y. Ershov, Dalton Trans.
2004, 30-36.
[37] J.-H. Shon, T. S. Teets, ACS Energy Lett. 2019, 4, 558-566.
[38] Y. Zhang, T. S. Lee, J. M. Favale, D. C. Leary, J. L. Petersen,
G. D. Scholes, F. N. Castellano, C. Milsmann, Nat. Chem. 2020,
12, 345-352.
[39] A. Erdoğmuş, M. Durmuş, A. L. Uğur, O. Avciata, U. Avciata, T.
Nyokong, Synth. Met. 2010, 160, 1868-1876.
7
This article is protected by copyright. All rights reserved.

10.1002/chem.202001678
Chemistry - A European Journal
FULL PAPER
Entry for the Table of Contents
Bimetallic titanium-oxo complexes generate singlet oxygen under visible light irradiation, which is used to photooxygenate -terpinene
to ascaridole under flow conditions. The activity of five titanium amine bis(phenolate) complexes was investigated and the compounds
were found to be selective for the production of ascaridole over p-cymene.
Twitter username: @mcintosh_group
8
This article is protected by copyright. All rights reserved.