Sensitised lanthanide luminescence using a RuIIpolypyridyl functionalised dipicolinic acid chelate

Article abstract of DOI:10.1039/d1dt00982f

A visible light absorbing [Ru^II(tpy)₂]²⁺-type chromophore appended with a dipicolinic acid Ln^IIIchelator has been prepared and complexed with several differing lanthanide cations to form the corresponding heterobimetallic d-f assemblies. The subseqent solution speciation analysed by¹H NMR spectroscopy revealed an unexpected decrease in the Ln^IIIchelate complex stability, in particular for the 1?:?3 complex, when compared to the parent dipicolinic acid. As a result, the desired Ln(ML)₃complexes could not be isolated, and the 1?:?1 Ln^III-ML complexes were instead characterised and investigated using steady state absorption and emission spectroscopy. Sensitised NIR emission from the Yb^III, Nd^IIIand Er^IIIcomplexes was observed upon¹MLCT excitation of the Ru^IIbased metalloligand in the visible region atca.485 nm. Investigations using transient absorption spectroscopy revealed essentially quantitative intersystem crossing to form the³MLCT excited state, as expected, which then acts as the energy donor for the metalloligand based antennae effect, facilitating sensitisation efficiencies of 4.8, 17.0 and 37.4% respectively for the Yb^III, Er^IIIand Nd^IIIcations.

Full text of DOI:10.1039/d1dt00982f

Dalton
Transactions
View Article Online
View Journal | View Issue
PAPER
Sensitised lanthanide luminescence using a RuII
polypyridyl functionalised dipicolinic acid chelate†
Cite this: Dalton Trans., 2021, 50,
7400
Divya Rajah,
Michael C. Pfrunder, Bowie S. K. Chong, Alexander R. Ireland,
Isaac M. Etchells and Evan G. Moore
*
A visible light absorbing [RuII(tpy)2]2+-type chromophore appended with a dipicolinic acid LnIII chelator
has been prepared and complexed with several differing lanthanide cations to form the corresponding
heterobimetallic d–f assemblies. The subseqent solution speciation analysed by 1H NMR spectroscopy
revealed an unexpected decrease in the LnIII chelate complex stability, in particular for the 1 : 3 complex,
when compared to the parent dipicolinic acid. As a result, the desired Ln(ML)3 complexes could not be
isolated, and the 1 : 1 LnIII–ML complexes were instead characterised and investigated using steady state
absorption and emission spectroscopy. Sensitised NIR emission from the YbIII, NdIII and ErIII complexes
1
was observed upon MLCT excitation of the RuII based metalloligand in the visible region at ca. 485 nm.
Received 25th March 2021,
Accepted 26th April 2021
Investigations using transient absorption spectroscopy revealed essentially quantitative intersystem cross-
ing to form the 3MLCT excited state, as expected, which then acts as the energy donor for the metalloli-
gand based antennae effect, facilitating sensitisation efficiencies of 4.8, 17.0 and 37.4% respectively for
the YbIII, ErIII and NdIII cations.
DOI: 10.1039/d1dt00982f
rsc.li/dalton
tor functionalised with a [RuII(bpy)2(phen)]2+ complex can be
Introduction
an effective sensitiser for NIR emission, particularly with NdIII.
The development of chelating d–f metalloligands which can Not surprisingly, the observed LnIII emission is longer-lived in
function as light absorbing antennae for sensitised lanthanide D2O compared to H2O, due to the elimination of solvent oscil-
emission, especially for near infra-red (NIR) emitting cations lators which can otherwise lead to non-radiative quenching of
such as YbIII, ErIII and NdIII, has received considerable litera- the LnIII emission.9 In each of these cases, the distance
ture interest due to their highly desirable luminescence pro- between the transition metal and 4f metal ion was relatively
perties. For example, rapid Intersystem Crossing (ISC) and the short, and non-conjugated alkyl groups were used to link the
typically long-lived triplet excited states of RuII, OsII and IrIII components. Alternately, by utilising an extended aromatic π
complexes makes them well suited for this purpose.1,2 Indeed, linked systems, Ward was able to demonstrate efficient sensit-
a range of d–f complexes using these metals and a variety of isation of NIR emitting LnIII cations at distances of over ca.
polypyridyl type ligands such as 2,2′-bipyridine (bpy) or func- 20 Å via a Dexter type superexchange mechanism using a func-
tionalised derivatives have been shown to exhibit efficient tionalised [RuII(bpy)3]2+ chromophore as the sensitiser.10
energy transfer and sensitisation of LnIII luminescence.3–6
More recently, we reported the use of a functionalised
Considering RuII complexes more specifically, Klink et al. [RuII(tpy)2]2+-type complex (tpy = 2,2′:6′,2″-terpyridine) as a sen-
reported7 several years ago that functionalised lanthanide sitiser for LnIII emission in the NIR region.11 Despite the
complexes utilising m-terphenyl based LnIII chelators much shorter MLCT lifetime of a bis-tridentate [RuII(tpy)2]2+
3
appended with a [RuII(bpy)3]2+ chromophore could be used as complex as an antenna chromophore, which can be attribu-
an antenna to sensitise NIR LnIII luminescence. Sénéchal- ted12 to the structural distortion away from Oh symmetry, and
David et al. also demonstrated8 that a macrocyclic LnIII chela- the corresponding increase in non-radiative deactivation via
low lying 3MC excited states, we nonetheless demonstrated
that relatively efficient sensitisation of NIR emitting 4f metals
can be achieved using this type of antenna.
School of Chemistry and Molecular Biosciences, The University of Queensland,
St Lucia, QLD, 4072, Australia. E-mail: egmoore@uq.edu.au
In this previous work11 one of the tpy ligands bound to RuII
was subsitituted in the 4′-position with an additional pendant
tpy chelator, in order to facilitate formation of the desired d–f
complex. However, it is quite well known that LnIII complexes
with purely N donor ligands generally suffer from lower
†Electronic supplementary information (ESI) available: Experimental and syn-
thetic details, X-ray crystallographic data, additional NMR data, details for
TD-DFT calculations and additional photophysical data. CCDC 2065117. For ESI
and crystallographic data in CIF or other electronic format see DOI: 10.1039/
d1dt00982f
7400 | Dalton Trans., 2021, 50, 7400–7408
This journal is © The Royal Society of Chemistry 2021

View Article Online
Dalton Transactions
Paper
aqueous stability constants (e.g. [LuIII(tpy)]3+, log10 β110 = 2.8),
since the Lewis acidity of the ligand and metal ions are not
well matched.13 As an alternative, carboxylate groups which
provide significantly harder and negatively charged O donor
atoms, can lead to much higher stability constants. As a perti-
nent example, it has been established that pyridine-2,6-dicar-
boxylic acid, also known as dipicolinic acid (H2DPA), is an
efficient LnIII chelator, and shows significantly larger for-
mation constants (e.g. [LuIII(DPA)n](3–2n)+, log10 β1n0 = 9.0, 16.8
and 21.5 for n = 1, 2, 3), allowing for efficient formation of the
tris LnIII(DPA)3 complexes in aqueous solution.14 Upon exci-
tation in the UV region via the π–π* transitions of the co-
ordinated DPA2− anion, several [LnIII(DPA)3]3− complexes are
known15 to exhibit high sensitisation efficiencies (ηsens > 70%)
and relatively good photo-luminescence quantum yields
(PLQY), in particular with TbIII and EuIII of Φ = 21% and 29%
respectively. This can be attributed to preventing vibrational
quenching by inner sphere solvent molecules such as water,9
and these complexes have also been proposed16 as suitable
secondary standards for PLQY measurements of LnIII com-
plexes containing these cations. Notably, the [YbIII(DPA)3]3−
complex also displays sensitised emission in the NIR region at
ca. 980 nm, although the evaluated PLQY is much smaller at
only 0.015%, which has been attributed15 to its lower sensitis-
ation efficiency of ca. 8% due to the large energy gap between
Scheme 1 Design and synthesis of the [(tpy)RuII(tpyPh-DPA)]2+
metalloligand.
excitation via the range of Q-band absorption maxima at ca.
600 nm yielded sensitized NIR emission in aqueous solution,
with potential applications in multiplex fluorescence imaging.
Considering the limitations in terms of the solution stabi-
lity for our previously reported11 [RuII(tpy)2]2+-type antenna
based system, which featured a pendant 2,2′:6′,2″-terpyridine
group to facilitate LnIII chelation for d–f complex formation,
we have redesigned this system as shown in Scheme 1, in
order to incorporate a more LnIII specific dipicolinic acid
based chelator. However, we have retained the bis tridentate
[Ru(tpy)2]2+-type antenna, since this chromophore features a
strong 1MLCT absorption band in the visible region suitable
for sensitisation of NIR emitting LnIII cations. Similarly, ISC to
2
ligand donor states and the 4f* metal centered F5/2 accepting
state of YbIII.
A variety of synthetic approaches have been reported for the
development of functionalised H2DPA derivatives as LnIII sen-
sitisers. For example, exploiting the 4-hydroxy substituted
version (also known as chelidamic acid), covalent attachment
of additional light absorbing organic sensitisers has been
achieved by several groups.14,17,18 For example, this approach
was used to append a variety of different isomeric coumarin
based chromophores via flexible polyether chains of variable
length.19 Interestingly, shortening the chain length resulted in
practically no change in the overall quantum yields for the
EuIII complexes in solution, which was attributed to the
inherent flexibility of the oligoethylene glycol linkage.
Alternately, using a variety of Pd catalysed aryl/ethynyl cross
coupling reactions, D’Aléo et al. have synthetically modified
the H2DPA chelate, again in the 4-position, allowing for a large
family of push–pull π-conjugated ligands to be prepared in an
effort to increase the two photon absorption (TPA) cross sec-
tions for the corresponding EuIII complexes.20 In this family of
compounds, sensitised emission at 613 nm with PLQY’s as
high as 43% were demonstrated using low energy excitation
wavelengths between 700–900 nm.
3
form the MLCT excited state energy donor can be considered
to be essentially quantitative in these complexes,22 and
although their excited state lifetimes are significantly shorter
than corresponding tris bidentate RuII polypyridyl complexes
such as [Ru(bpy)3]2+, we have shown11 that if the rate of energy
transfer for sensitised LnIII emission via the antennae effect is
sufficiently rapid, sizeable sensitisation efficiencies can still be
achieved. Lastly, by utilising a 4-substituted H2DPA chelate, it
was envisaged that formation of the corresponding [Ln(ML)3]3+
complexes in solution would result in improved luminescence
performance, by both excluding inner sphere solvent mole-
cules and increasing the number of energy donors via the for-
mation of tris chelated d–f metallostars.23
However, to the best of our knowledge, only one example of
a modified H2DPA chromophore bearing a pendant transition
metal based chromophore has appeared in the literature.
Results and discussion
Synthesis and characterisation
Quite recently, Xiong et al. reported21 the synthesis of two The desired RuII metalloligand was prepared adopting a
novel H2DPA based chelators, substituted with both free base variety of literature methods24–27 in the stepwise approach
or Pd(II) metalated chlorins via the 3- or 15-position of the detailed in Scheme 1, which involved first constructing the
latter. This series of antennae yielded the corresponding tris appropriately mono substituted RuII based aryl halide, fol-
complexes with both YbIII and NdIII, and it was shown that lowed by an aryl cross coupling ‘on-the-complex’ and sub-
This journal is © The Royal Society of Chemistry 2021
Dalton Trans., 2021, 50, 7400–7408 | 7401

View Article Online
Paper
Dalton Transactions
sequent deprotection. To summarise, starting from chelidamic
acid, which was esterified, brominated and then subjected to
Miyaura borylation, an overall yield of 46% for the required
coupling precursor was obtained. Notably, borylation of the
brominated chelidamic following a literature procedure20 in
DMSO was unsuccessful in our hands, although an alternative
method reported by Rasheed et al.28 using dioxane as the
solvent gave the desired product. In parallel, [RuCl3·H2O] was
complexed with tpy followed by tpyPh-Br in an overall yield of
51% after chromatography. The subsequent Pd-catalysed
Suzuki cross coupling reaction to form [(tpy)RuII(tpyPh-
DPAMe2)]2+ was successful, albeit in a relatively low yield of
10%, which we attribute to loss of the methyl protecting
groups under the slightly basic reaction conditions required,
which necessitated tedious purification by column chromato-
graphy. Nonetheless, the methyl protected metalloligand was
able to be isolated in high purity as a KPF6 adduct, and was
fully characterised using 1H NMR, HRMS, CHN elemental ana-
lysis and X-ray crystallography.
Fig. 1 View of the X-ray structure for the [(tpy)RuII(tpyPh-DPAMe2)]
(PF6)2·KPF6 complex, emphasising molecular packing. Thermal ellipsoids
are shown at the 50% probability level and H atoms are omitted for
clarity. C (grey), N (blue), O (red), P (orange), F (yellow), P (purple), and
Ru (teal).
−
disordered PF6 ions, which form a ribbon along the b axis,
completing the overall charge balance (Fig. 1).
The RuII ion shows a distorted octahedral coordination
sphere, with average Ru–N bond lengths of 2.043 and 2.033 Å
for the capping and bridging tpy ligands respectively. As
expected, the Ru–N bond lengths for the peripheral tpy pyridyl
rings are slightly longer (avg. = 2.073 Å) than the central
pyridyl bond lengths (avg. = 1.968 Å). The tpy ligands are
approximately perpendicular with each other, with an inter-
planar angle of 88.5°. The phenyl linker between the chelated
tpy and DPAMe2 groups shows a torsion angle of 51.2° relative
to tpy, or 32.6° relative to the DPAMe2 moiety, resulting in the
tpy and DPAMe2 moieties being oriented approximately per-
pendicularly at 84.5°.
The 1H NMR spectrum of [(tpy)RuII(tpyPh-DPAMe2)]2+ in
CD3CN (see Fig. S1, ESI†), was fully assigned using 1H–1H
COSY and 1D-NOESY techniques (see Fig. S2–S5, ESI†).
Protons Ha, Hd and Hm, as labelled in Scheme 1, were readily
assigned by comparison to the precursor [(tpy)RuII(tpyPh-Br)]2+
complex, with proton Ha at 8.71 ppm appearing as the obvious
new signal indicating the successful cross coupling. Proton Hd
located at 9.08 ppm was used to identify the adjacent He and
Hc protons via 1D-NOESY and the assignment was further con-
1
firmed by the observed H–1H COSY correlations. The assign-
ments of proton Hb and Hc signals was confirmed by
irradiation of Ha as well as 1H–1H COSY correlations. Proton
To complete the required synthesis, the methyl deprotected
[(tpy)RuII(tpyPh-DPA)] metalloligand was obtained by hydro-
lysis of the corresponding dimethyl ester in the presence of
excess KOH in an acetonitrile/water solvent mixture (6 : 1 v/v)
followed by the addition of HPF6 to precipitate the desired
product in good yield and analytically pure form as deter-
mined by 1H NMR, CHN and HRMS analysis. Notably, hexa-
fluorophosphoric acid was used necessarily in this step rather
H
m was identified by the 1H–1H COSY correlation with Hn, and
subsequently used to identify Hl by an observed 1D-NOESY
signal, allowing the remaining pyridyl protons to be assigned
using 1H–1H COSY correlations.
Slowly evaporating a solution of the methyl protected metal-
loligand in an acetonitrile water mixture allowed crystals suit-
able for X-ray studies to be obtained. Although these were of
generally poor quality and were found to be a two-fold twin
about the [−105] lattice direction, the structure was nonethe-
−
than an anionic PF6 salt, to avoid formation of an otherwise
zwitterionic complex, which we found was resistant to iso-
lation by precipitation.
ˉ
less able to be solved in the P1 space group, and was formu-
lated as the [(tpy)RuII(tpyPh-DPAMe2)](PF6)2·KPF6 adduct,
allowing the connectivity of the complex to be unambiguously
determined. A single RuII metalloligand as its bis hexafluoro-
phosphate salt together with an additional KPF6 moiety were
located in the asymmetric unit, with the complex cations
packed into 1D chains oriented along the c axis. A poorly
1
The H NMR obtained for [(tpy)RuII(tpyPh-DPA)], shown in
Fig. 2 (see also Fig. S6, ESI†), was fully assigned by comparison
to the spectra for the [(tpy)RuII(tpyPh-DPAMe2)]2+ complex,
since, although different solvents were used, the observed
signals display essentially identical splitting patterns and
quite similar chemical shifts, with the exception of a small
upfield shift for the Ha signal (ca. 0.15 ppm), small downfield
shifts for the Hd, Hm, He and Hl signals (ca. 0.1–0.2 ppm), and
the obvious disappearance of the methyl ester signal at
4.03 ppm.
−
ordered PF6 anion is positioned between the adjacent RuII
cations, and shows two non-classical H-bonding interactions
between two fluorine atoms and the protons of two neighbour-
ing complexes (F17–H2 = 2.407 Å, F14–H12 = 2.252 Å). Notably,
the DPAMe2 binding site contains a single potassium cation,
with an overall coordination number of 10, which is bound to
the two O alkoxy atoms and the pyridyl N atom. The coordi-
Speciation and solution stability
nation sphere of the alkali metal is completed by a mixture of Formation of the desired d–f complexes was investigated using
bidentate and monodentate K–F interactions with adjacent 1H NMR spectroscopy to monitor the titration of Lu
7402 | Dalton Trans., 2021, 50, 7400–7408
This journal is © The Royal Society of Chemistry 2021

View Article Online
Dalton Transactions
Paper
also revealed that the desired Ln(ML)3 ‘metallostar’ complexes
could not be isolated with the [(tpy)RuII(tpyPh-DPA)] metalloli-
gand, due to the equilibria involved in tris complex formation.
In order to assess the individual complex formation con-
stants, the corrected integrals for each of the species attributed
to the ML, Ln(ML), Ln(ML)2 and Ln(ML)3 complexes have been
fitted to four overlapping Lorentzian functions (see Fig. S7,
ESI†), and hence the concentration of each species was evalu-
ated. A best fit of the observed versus calculated percentage
speciation data using HySS2009 29 allowed the cumulative
stability constants to be estimated as log10 β110 = 4.8(1),
log10 β120 = 8.3(1), and log10 β130 = 11.4(1) (see Fig. S8, ESI†).
Notably, these values are several orders of magnitude smaller
than those of the DPA2− anion,14 which we attribute to the
electron withdrawing RuII complex that is covalently appended
to the pyridine-2,6-dicarboxylate, significantly decreasing its
chelating ability.
Since the solution thermodynamic model demonstrated
formation of the 3 : 1 d–f complexes would not be possible at
concentrations suitable for photophysical measurements, the
properties of the 1 : 1 d–f complexes were instead investigated.
As an additional verification that the [(tpy)RuII(tpyPh-DPA)]
complex binds other LnIII cations as the 1 : 1 Ln(ML) complex,
addition of an excess of NdIII(OTf)3·6H2O to a solution of the
metalloligand was also undertaken, and was analysed by 1H
NMR spectroscopy, resulting in much larger downfield chemi-
cal shifts for various protons, including Ha, due to the pres-
ence of the paramagnetic metal ion (see Fig. S9, ESI†).
Fig. 2 1H NMR titration for a 0.2 mM solution of [(tpy)RuII(tpyPh-DPA)]
in MES buffered 2 : 1 MeOD : D2O (v/v) solution (0.1 M, pD ∼ 6.4) upon
addition of [Lu(OTf)3·6H2O].
(OTf)3·6H2O into a 0.2 mM solution of the [(tpy)RuII(tpyPh-
DPA)] complex in a MES buffered solution (MES = 2-(N-mor-
pholino)ethanesulfonic acid, 0.1 M, pD
≈ 6.4) of 2 : 1
MeOD : D2O (v/v), with results as summarised in Fig. 2.
Upon addition of 0.1 molar equiv. of Lu(OTf)3·6H2O, the 1H
NMR spectrum shows only minor changes, although particu-
larly noticeable is the appearance of two new singlet peaks at
9.29 and 8.60 ppm (also highlighted in Fig. 2 by dashed
circles), which we attribute to the Ha and Hd protons of the Lu
(ML)3 complex, which should form preferentially in the pres-
ence of an excess of the ML metalloligand (where ML = [(tpy)
RuII(tpyPh-DPA)].
Density functional theory
In order to improve our understanding on the nature of the
excited states for both the [(tpy)RuII(tpyPh-DPA)] metalloligand
and the corresponding 1 : 1 heterobimetallic d–f [(tpy)
RuII(tpyPh-DPA)LnIII(H2O)x]3+ complexes, we have undertaken
both static and time dependent (TD) density functional theory
(DFT) calculations, using an mPW1PW91 functional and
mixed D95 V/LanL2TZ(f)/def2-QZVP basis sets using an SMD
solvation model with methanol (for further details, see the
ESI†).
The resulting optimised geometries and selected frontier
molecular orbitals for the [(tpy)RuII(tpyPh-DPA)] complex and
corresponding [(tpy)RuII(tpyPh-DPA)LuIII(H2O)5]3+ d–f complex
are shown in Fig. 3. Notably, the calculated structure of [(tpy)
RuII(tpyPh-DPA)] is very similar to the structure of the di-
methylester derivative obtained by X-ray crystallography. The
coordination geometry about the RuII centre is a distorted
octahedron, with capping and bridging tpy moieties in a
planar orientation and average Ru–N bond lengths of ca.
2.03 Å, which are almost identical to those found in the X-ray
structure of 2.04 Å. As expected, slightly shorter Ru–N bond
lengths to the central tpy N atom are evident, due to the
strained chelate ring, which also leads to N–Ru–N bond angles
3 ML þ Ln Ð LuðMLÞ3
ð1Þ
Subsequent addition of LuIII shows further changes in the
1H NMR signals. For example, the spectrum obtained after the
addition of 0.3 molar equiv. clearly shows the appearance of
four different singlet peaks at ca. 9.25 ppm, which we have
assigned to the Hd protons of d–f complexes with ML, Ln(ML),
Ln(ML)2, and Ln(ML)3 stoichiometry formed upon the
addition of increasing equivalents of the LuIII metal ion.
LnðMLÞ3 þ Ln Ð LnðMLÞ2 þ LnðMLÞ
ð2Þ
Further addition of Lu(OTf)3·6H2O reveals a continued
decrease in the amount of 3 : 1 and 2 : 1 d–f complexes.
Similarly, this data also shows that at least two species are
present in solution until an excess (4.0 molar equiv.) of LuIII
has been added, resulting in exclusive formation of the 1 : 1 d–
f complex.
LnðMLÞ2 þ Ln Ð 2 LnðMLÞ
ð3Þ
At the end of the titration, formation of the 1 : 1 d–f for each of the chelated tpy moieties which deviate from the
complex was particularly evident for Ha and Hd, which shifted ideal 90° to approximately 79°. The non-coordinated DPA2−
downfield during the titration from 8.57 to 8.62 and 9.26 to moiety is also twisted at an angle of ca. 66° with respect to the
9.28 ppm respectively. More importantly, the 1H NMR data chelated tpy moiety.
This journal is © The Royal Society of Chemistry 2021
Dalton Trans., 2021, 50, 7400–7408 | 7403

View Article Online
Paper
Dalton Transactions
which predominantly involve the HOMO → LUMO (81.4%)
and HOMO−1 → LUMO (73.5%) for the lowest energy S1 and
T1 excited state respectively, with the simulated spectrum
again showing an excellent match to the experimental spec-
trum (see Fig. S14, ESI†).
It is interesting to note that when the LuIII cation is che-
lated to the DPA moiety, the corresponding LUMO involved in
the lowest energy singlet and triplet excited states appears to
be more localised over the pyridine-2,6-dicarboxylate end of
the tpyPh-DPA ligand, an effect which was less pronounced for
the free metalloligand. It is enticing to contemplate whether
the observed localisation in close proximity to the 4f metal ion
may enhance the rate of EET involved in sensitised LnIII
emission.
Fig. 3 Optimised DFT geometries of [(tpy)RuII(tpyPh-DPA)] (left) and
[(tpy)RuII(tpyPh-DPA)LuIII(H2O)5]3+ (right) and selected HOMO and
LUMO molecular orbitals.
Photophysical properties
The steady state UV-Vis absorption and emission spectra of the
[(tpy)RuII(tpyPh-DPA)] metalloligand is shown in Fig. 4, using
a MES buffered (0.1 M, pD 6.4) 2 : 1 MeOD : D2O (v/v) solvent
system. As expected, the absorption spectrum features several
peaks in the UV region at ca. 220, 270 and 310 nm, which
correspond to π → π* transitions of the aromatic DPA2− and
tpy groups, in agreement with previous literature
examples.11,31,32 In the visible region, a strong characteristic
peak at ca. 485 nm is also observed which we attribute to the
RuII 1MLCT absorption band, and is again similar to those
reported for [Ru(tpy)2]2+-type derivatives.11,32 Excitation of the
metalloligand at ca. 485 nm leads to a broad emission band
centred at ca. 662 nm, which can be attributed to 3MLCT phos-
For the heterobimetallic d–f complex, the diamagnetic LuIII
cation was chosen as a general model for the bound LnIII
cations presented in this study. The calculated geometry
around the LuIII cation is as expected, showing tris coordi-
nation to the deprotonated DPA2− moiety of the bridging
ligand, with an average LuIII–O bond length of 2.29 Å, and
LuIII–N bond length of 2.39 Å. The distance between the RuII
donor and LnIII acceptor was estimated to be approximately
15.8 Å (see Fig. S10, ESI†). Notably, we have modelled the
remainder of the coordination sphere for the LuIII metal ion
with five coordinated solvent water molecules, in accordance
with a known literature precedent.30 While this is most likely
to be the case for the smaller LnIII cations (e.g. ErIII, YbIII and
LuIII) due to their smaller ionic radius, it is possible that the
larger LnIII cations (e.g. NdIII) may exist with either five or six
coordinated solvent water molecules bound to the metal ion.
While this difference in solvation will certainly influence many
of the luminescent properties, such as the 4f* centered PLQY,
it is not of particular relevance in terms of the ligand centered
sensitisation efficiencies presented herein (vide infra).
3
phorescence. Although quite weak, MLCT emission has been
observed in similar [Ru(tpy)2]2+ complexes bearing extended
aryl ligands.33,34
In order to investigate their photophysical properties, the
corresponding 1 : 1 d–f complexes with LnIII = Lu, Yb, Er and
Nd were formed in situ by adding an excess (>5 molar equiv.)
of the corresponding Ln(OTf)3 salt to a stock solution of the
free metalloligand, again using the same concentration and
An analysis of the relevant molecular orbitals obtained
from TD-DFT studies reveals that the HOMO and LUMO for
both the metalloligand and corresponding d–f complex have
significant orbital densities located on the RuII metal center
and bridging tpyPh-DPA ligand respectively (see Fig. 3,
Table S1, and Fig. S11 and S12, ESI†). For the [(tpy)RuII(tpyPh-
DPA)] metalloligand, the predominant contributions for the
lowest energy singlet (S1) excited state are from the HOMO →
LUMO (96.4%) orbitals, while the lowest energy triplet (T1)
excited state involves predominantly the HOMO−1 and LUMO
(83.7%) orbitals. As noted, the majority of the electron density
is localised on the RuII center for the HOMO/HOMO−1, while
the LUMO is located primarily on π* type orbitals on the tpy
group of the bridging tpyPh-DPA ligand, consistent with the
lowest energy singlet and triplet excited states having MLCT
excited state character. A comparison of the calculated UV-Vis
absorption spectrum to the experimental spectrum (see
Fig. S13, ESI†) shows very good agreement. The corresponding
Fig. 4 UV-Vis absorption (left) and emission spectra (λex = 485 nm,
right) for 0.2 mM solutions of 1 : 1 ML–Ln complexes where ML = [(tpy)
RuII(tpyPh-DPA)] and Ln = Lu, Nd, Yb or Er in MES buffered (0.1 M, pD ∼
1 : 1 d–f complex with LuIII also displays electronic transitions 6.4) 2 : 1 d4-MeOD : D2O (v/v).
7404 | Dalton Trans., 2021, 50, 7400–7408
This journal is © The Royal Society of Chemistry 2021

View Article Online
Dalton Transactions
Paper
mixed solvent system as that used for the 1H NMR solution
speciation studies.
The UV-Vis spectra for the [(tpy)RuII(tpyPh-DPA)
LuIII(H2O)5]3+ complex is shown in Fig. 4 (see also LnIII = Nd,
Yb, Er complexes in Fig. S15, ESI†). Notably, coordination of
LnIII cations to the DPA2− chelate is accompanied by only a
small increase in the absorbance of the d–f complexes in the
1
UV region at ca. 270 nm, while the MLCT peak in the visible
region appears relatively unchanged upon 1 : 1 d–f complex
formation. This is consistent with previous reports35,36 using
related DPA2− derivatives, which show only small changes
upon spectrophotometric titration with LnIII cations. In the
present case, the change in the UV region is also relatively
small due to the differences expected in the molar extinction
coefficients of the DPA2− (ca. 5000 M−1 cm−1) compared to the
[Ru(tpy)2]2+ fragments (>30 000 M−1 cm−1).26,37 By contrast, the
emission spectra for the 1 : 1 d–f complexes using excitation
via the 1MLCT absorption peak (λex = 485 nm) showed more
significant differences. For [(tpy)RuII(tpyPh-DPA)LuIII(H2O)5]3+,
the observed 3MLCT visible emission band centered at ca.
667 nm was red-shifted by 5 nm compared to the free metallo-
ligand as shown in Fig. 4, and is noticeably sharper, with a
FWHM of ca. 2370 cm−1 compared to ca. 2800 cm−1 for the
latter (see also Fig. S16, ESI†). Moreover, excitation of the Near
Infra-Red (NIR) emitting d–f complexes with LnIII = Nd, Yb and
Er display variable quenching of the 3MLCT visible emission
(Fig. S16, ESI†) compared to an isoabsorbing solution of the
LuIII complex, and instead exhibit NIR emission bands which
are characteristic for each 4f metal ion. For the [(tpy)
RuII(tpyPh-DPA)YbIII(H2O)5]3+ complex, an emission band at
2
2
ca. 980 nm is evident, which corresponds to the F5/2 → F7/2
transition of the YbIII cation. For the 1 : 1 d–f complex with
NdIII, three sharp emission bands at ca. 890 nm, 1055 nm, and
4
1325 nm are observed for the NdIII centered F3/2
→
4IJ tran-
4
sitions (J = 9/2, 11/2, 13/2). Lastly, albeit weak, the I13/2
→
4I15/2 transition of the ErIII cation can also be observed at ca.
1520 nm for the corresponding ErIII complex. Accordingly, we
1
can confirm that selective excitation of the MLCT absorption
band in the [Ru(tpy)2]2+-type chromophore, followed by rapid
ISC to form the 3MLCT excited state leads to sensitised NIR
emission for these cations, as expected, via the antennae
effect.
Fig. 5 (a) fs-TA spectra (λex = 485 10 nm removed), (b) globally fitted
decay kinetics, (c) and decay associated difference spectra (DADS)
including fitted lifetimes for the [(tpy)RuII(tpyPh-DPA)Yb(H2O)5]3+
complex in MES buffered (0.1 M, pD ∼ 6.4) 2 : 1 d4-MeOD : D2O (v/v).
Energy transfer kinetics
To gain a more detailed insight into the excited state dynamics
of the observed antennae effect which leads to sensitised LnIII
emission, we have utilised femtosecond time resolved
Transient Absorption (TA) spectroscopy, with the resulting
spectra and decay kinetics for the 1 : 1 d–f complex with YbIII
shown as an example in Fig. 5 (see also Fig S17–S20 ESI† for
other complexes). Upon 1MLCT excitation at 485 nm, the TA
spectra show three distinct transient features, including two
positive excited state absorption (ESA) signals centered at ca.
tures observed for [(tpy)RuII(tpyPh-DPA)] and the corres-
ponding [(tpy)RuII(tpyPh-DPA)LnIII(H2O)x]3+ complexes with Ln
= LuIII, YbIII, ErIII, NdIII were all essentially identical, and
differed only in terms of the observed decay kinetics which
were obtained by global fitting using the equation;
I
¼ A1 expꢀk t þ A2 expꢀk t
ð4Þ
1
2
ðtÞ
380 nm and 650 nm, together with a strong negative signal where I(t) is the intensity of the observed TA signal, k1 and k2
centered at ca. 485 nm corresponding to ground state bleach are the fitted decay rates (τ1 = 1/k1, etc.), while A1 and A2 are
1
(GSB) of the MLCT absorption band. Importantly, the TA fea- pre-exponential scaling factors. In the present case, based on
This journal is © The Royal Society of Chemistry 2021
Dalton Trans., 2021, 50, 7400–7408 | 7405

View Article Online
Paper
Dalton Transactions
Table 1 Summary of observed decay lifetimes and resulting kinetic parameters obtained by fs-TA spectroscopy (λex = 485 nm) for ML = [(tpy)
RuII(tpyPh-DPA)] and the corresponding 1 : 1 ML–LnIII complexes with Ln = Lu, Yb, Er and Nd in MES buffered (0.1 M, pD ∼ 6.4) 2 : 1 d4-MeOD : D2O
(v/v)
τ1 (vib) (ps)
τ2 (3MLCT) (ps)
kobsa (×108 s−1
)
kEET (×107 s−1
)
ΦEETb,c (%)
ML
9.6 0.1
2301
3412
3247
2831
2136
4
9
9
6
4
4.346 0.008
2.931 0.008
3.079 0.009
3.532 0.008
4.681 0.009
N/A
N/A
1.5 0.1
6.0 0.1
17.5 0.1
N/A
N/A
4.8 0.4
17.0 0.3
37.4 0.3
ML-Lu
ML-Yb
ML-Er
ML-Nd
12.1 0.2
10.0 0.2
10.1 0.2
11.5 0.2
a kobs (3MLCT) = 1/τ2. b ΦEET = kEET/kobs c ΦEET = ηsens (since ΦISC ≈ 100%).
.
kEET ¼ kobsðNIR emitting LnÞ ꢀ kobsðLuÞ
ð6Þ
ð7Þ
the relatively featureless decay associated difference spectrum
(DADS) shown in Fig. 5(c), and several literature precedents,38
we attribute the short-lived lifetime (τ1) to vibrational cooling
3
kEET
and solvent reorganisation of the MLCT excited state formed
ΦEET
¼
kobs
1
by ISC from the initially populated MLCT state upon photo-
excitation. Instead, the longer lived lifetime (τ2) is attributed to
decay of the relaxed 3MLCT excited state.
Using these calculated kEET values, and since ISC can be
assumed22 to be essentially quantitative for the RuII complexes
(ΦISC ≈ 100%), the energy transfer efficiency (ΦEET) and also
the overall sensitisation efficiencies (ηsens = ΦISC × ΦEET) for
each of the NIR emitting LnIII complexes can be evaluated
from the ratio of kEET to kobs as shown in eqn (7). The resulting
rates and efficiencies for the observed 3MLCT → 4f* energy
transfer are shown in Table 1.
A summary of the excited state lifetimes obtained for the
free metalloligand and each of the 1 : 1 d–f complexes is given
in Table 1. For [(tpy)RuII(tpyPh-DPA)], the 3MLCT lifetime
obtained is ca. 2.3 ns, which is consistent with the known
short lifetimes of [Ru(tpy)2]2+-type complexes (e.g. compared to
[Ru(bpy)3]2+ derivatives), due to rapid non-radiative de-
activation via the RuII metal centered (3MC) excited state.32
Upon complexation with LuIII, the 3MLCT lifetime of the
resulting [(tpy)RuII(tpyPh-DPA)LuIII(H2O)5]3+ complex is
slightly longer-lived, which can be rationalised by an increase
in the rigidity of the ensuing d–f complex, and is also consist-
ent with the observed sharpening of the steady state 3MLCT
emission. For the other 1 : 1 d–f complexes with LnIII = Yb, Er,
or Nd, the observed 3MLCT lifetime decreases compared to the
LuIII complex in a manner which reflects the observed quench-
ing of the steady state visible emission spectra. Again, this can
be readily attributed to energy transfer with the 3MLCT excited
state acting as the energy donor, leading to sensitised NIR
emission for these cations.
To summarise, for each of the 1 : 1 d–f complexes with NIR
emitting LnIII cations (Ln = Nd, Yb and Er), the presence of
low energy 4f* excited states for these metals facilitates energy
3
transfer from the MLCT excited state donor, resulting in both
quenching of the emission intensity, and
a significant
decrease in the luminescence lifetime of the 3MLCT state com-
pared to the LuIII model complex. For the NdIII and ErIII com-
plexes, which have a variety of potentially accepting 4f* excited
states and considerable spectral overlap with the RuII 3MLCT
emission, the energy transfer rates (kEET) we obtain are con-
siderably faster than those with YbIII, which has only a single
2F5/2 accepting state. Nonetheless, despite the relatively short
lifetime of the [Ru(tpy)2]2+-type chromophore, the energy trans-
fer (and overall sensitisation efficiencies) are as high as ca.
37% with NdIII.
Using the observed lifetimes, the rate of energy transfer
3
from the MLCT excited state donor to 4f* excited state accep-
tor can also be calculated. Specifically, due to the absence of
low energy 4f* excited states available for the LuIII cation, the
3MLCT lifetime of this complex can be used as a suitable
model for deactivation of the [Ru(tpy)2]2+-type chromophore in
the absence of competing energy transfer to NIR emitting
cations. The sum of the radiative and non-radiative de-
activation rate constants (kobs = kr + knr) is obtained from the
observed lifetime using eqn (5) in the presence and absence of
the NIR emitting LnIII quencher. By making the reasonable
Importantly, these results also agree well with previous
observations we reported11 using a [(toltpy)RuII(tpyPh-tpy)]2+
metalloligand as the antennae chromophore (where toltpy = 4′-
(p-tolyl)-2,2′:6′,2″-terpyridine, and tpyPh-tpy = 1,4-di([2,2′:6′,2′-
terpyridin]-4′-yl)benzene). Moreover, the kEET rates we obtain
herein are also consistent and very similar to those obtained
with this previous system, as we would expect due to the
similar intramolecular separation between metal ions.
Accordingly, the observed energy transfer we observe for the
1 : 1 d–f complexes reported herein can similarly be attributed
to a Dexter type superexchange mechanism, as was the case
previously, and which is also consistent with earlier reports10
by Ward et al. who studied the distance dependence for RuII
(3MLCT) → LnIII energy transfer using a series of structurally
analogous aryl substituted [Ru(bpy)3]2+-type derivatives.
3
assumption that kr and knr for the MLCT excited state energy
donor are the same (i.e. independent of the 4f metal ion), the
rate of electronic energy transfer (kEET) can then be calculated
from the difference in observed kobs using eqn (6) below.10,32,39
1
kobs
τ ¼
ð5Þ
7406 | Dalton Trans., 2021, 50, 7400–7408
This journal is © The Royal Society of Chemistry 2021

View Article Online
Dalton Transactions
Paper
2 A. M. Nonat, C. Allain, S. Faulkner and T. Gunnlaugsson,
Inorg. Chem., 2010, 49, 8449–8456.
3 W. S. Perry, S. J. A. Pope, C. Allain, B. J. Coe,
A. M. Kenwright and S. Faulkner, Dalton Trans., 2010, 39,
10974–10983.
4 N. M. Tart, D. Sykes, I. Sazanovich, I. S. Tidmarsh and
M. D. Ward, Photochem. Photobiol. Sci., 2010, 9, 886–889.
5 F.-F. Chen, Z.-Q. Bian, B. Lou, E. Ma, Z.-W. Liu, D.-B. Nie,
Z.-Q. Chen, J. Bian, Z.-N. Chen and C.-H. Huang, Dalton
Trans., 2008, 5577–5583.
6 S. J. A. Pope, B. J. Coe, S. Faulkner and R. H. Laye, Dalton
Trans., 2005, 1482–1490.
7 S. I. Klink, H. Keizer and F. C. J. M. Van Veggel, Angew.
Chem., Int. Ed., 2000, 39, 4319–4321.
8 K. Sénéchal-David, S. J. A. Pope, S. Quinn, S. Faulkner and
T. Gunnlaugsson, Inorg. Chem., 2006, 45, 10040.
9 A. Beeby, I. M. Clarkson, R. S. Dickins, S. Faulkner,
D. Parker, L. Royle, A. S. de Sousa, J. A. Gareth Williams
and M. Woods, J. Chem. Soc., Perkin Trans. 2, 1999, 493–
504.
Conclusions
A [RuII(tpy)2]2+-type metalloligand functionalised with a dipico-
linic acid chelate has been synthesised and characterised as a
visible light absorbing antennae for sensitised LnIII emission,
since complexes of this type may be useful for the further
development of NIR emitting biological imaging agents.
Unfortunately, the anticipated formation of 3 : 1 d–f complexes
was not successful, due to a decrease in the complex stability
constants, which were found to be orders of magnitude lower
than the DPA2− ligand. This unanticipated result, attributed to
a decrease in chelate strength due to the adjacent RuII
complex, thwarted our attempts to characterise the photo-
physical properties of the desired 3 : 1 metallostar complexes.
Characterisation of the 1 : 1 d–f complexes was undertaken
instead, and sensitised NIR emission using LnIII = Nd, Yb and
Er was demonstrated. Moreover, using steady state emission
and transient absorption techniques, we were able to establish
the thermalised RuII 3MLCT excited state acts as an energy
donor, after initial 1MLCT excitation and rapid ISC. Analysis of
the decay kinetics allowed both the energy transfer rate and 10 T. Lazarides, D. Sykes, S. Faulkner, A. Barbieri and
sensitisation efficiencies to be estimated, and these were M. D. Ward, Chemistry, 2008, 14, 9389–9399.
found to be similar to a series of [(toltpy)RuII(tpyPh-tpy) 11 E. G. Moore, M. Benaglia, G. Bergamini and P. Ceroni,
LnIII(NO3)3] complexes (Ln = Nd, Yb) we reported previously.11
Eur. J. Inorg. Chem., 2015, 2015, 414–420.
Considering their structural similarity, and the almost identi- 12 E. A. Medlycott and G. S. Hanan, Chem. Soc. Rev., 2005, 34,
cal separation between the RuII and 4f metal ions, the mecha-
133–142.
nism for the observed energy transfer has been attributed to a 13 J. M. Hamilton, M. J. Anhorn, K. A. Oscarson,
Dexter type superexchange, which is also consistent with pre-
vious reports.10
J. H. Reibenspies and R. D. Hancock, Inorg. Chem., 2011,
50, 2764–2770.
Considering the results reported herein, it is obvious that 14 A.-L. Gassner, C. Duhot, J.-C. G. Bünzli and A.-S. Chauvin,
decreasing the intramolecular separation between metals ions Inorg. Chem., 2008, 47, 7802–7812.
should lead to a faster rate of energy transfer and improved 15 A. Aebischer, F. Gumy and J. C. Bunzli, Phys. Chem. Chem.
sensitisation efficiencies for LnIII emission, and our efforts
Phys., 2009, 11, 1346–1353.
toward preparing [RuII(tpy)2]2+-type metalloligands with 16 A. S. Chauvin, F. Gumy, D. Imbert and J. C. G. Bünzli,
shorter linkers will be reported in due course. Similarly, fusing
Spectrosc. Lett., 2004, 37, 517–532.
a [RuII(tpy)2]2+ antennae chromophore to a multidentate 17 J. Andres and K. E. Borbas, Inorg. Chem., 2015, 54, 8174–8176.
chelate (e.g. DOTA, 1,4,7,10-tetraazacyclododecane-1,4,7,10- 18 H.-J. Park, S.-B. Ko, I. W. Wyman and S. Wang, Inorg.
tetraacetic acid) will be expected to yield a significant improve-
Chem., 2014, 53, 9751–9760.
ment in the aqueous stability, whilst also saturating the LnIII 19 J. Andres and A. S. Chauvin, Phys. Chem. Chem. Phys., 2013,
coordination sphere as a 1 : 1 d–f complex, and these activities
15, 15981–15994.
are also underway.
20 A. D’Aléo, A. Picot, P. L. Baldeck, C. Andraud and O. Maury,
Inorg. Chem., 2008, 47, 10269.
21 R. Xiong, D. Mara, J. Liu, R. Van Deun and K. E. Borbas,
J. Am. Chem. Soc., 2018, 140, 10975–10979.
22 J. Kim, D.-g. Kang, S. K. Kim and T. Joo, Phys. Chem. Chem.
Phys., 2020, 22, 25811–25818.
Conflicts of interest
There are no conflicts to declare.
23 G. Dehaen, S. V. Eliseeva, P. Verwilst, S. Laurent, L. Vander
Elst, R. N. Muller, W. De Borggraeve, K. Binnemans and
T. N. Parac-Vogt, Inorg. Chem., 2012, 51, 8775–8783.
24 S. Katagiri, R. Sakamoto, H. Maeda, Y. Nishimori, T. Kurita
and H. Nishihara, Chem. – Eur. J., 2013, 19, 5088–5096.
25 C. J. Aspley and J. A. Gareth Williams, New J. Chem., 2001,
25, 1136–1147.
Acknowledgements
Financial support by the Australian Research Council
(DP170101895) is gratefully acknowledged.
26 F. Barigelletti, L. Flamigni, V. Balzani, J.-P. Collin,
J.-P. Sauvage, A. Sour, E. C. Constable and A. M. W. Cargill
Thompson, J. Am. Chem. Soc., 1994, 116, 7692–7699.
Notes and references
1 M. D. Ward, Coord. Chem. Rev., 2007, 251, 1663–1677.
This journal is © The Royal Society of Chemistry 2021
Dalton Trans., 2021, 50, 7400–7408 | 7407

View Article Online
Paper
Dalton Transactions
27 C. Chakraborty, U. Rana, R. K. Pandey, S. Moriyama and 33 F. Barigelletti, L. Flamigni, G. Calogero, L. Hammarström,
M. Higuchi, ACS Appl. Mater. Interfaces, 2017, 9, 13406–
J.-P. Sauvage and J.-P. Collin, Chem. Commun., 1998, 2333–
13414.
2334.
28 O. K. Rasheed, J. J. W. McDouall, C. A. Muryn, J. Raftery, 34 M. Maestri, N. Armaroli, V. Balzani, E. C. Constable and
I. J. Vitorica-Yrezabal and P. Quayle, Dalton Trans., 2017,
46, 5229–5239.
29 L. Alderighi, P. Gans, A. Ienco, D. Peters, A. Sabatini and
A. Vacca, Coord. Chem. Rev., 1999, 184, 311–318.
30 H. S. Wang, B. Zhao, B. Zhai, W. Shi, P. Cheng,
A. M. C. Thompson, Inorg. Chem., 1995, 34, 2759–2767.
35 J. Andres and A.-S. Chauvin, Inorg. Chem., 2011, 50, 10082–
10090.
36 J. Andres and A.-S. Chauvin, Eur. J. Inorg. Chem., 2010,
2010, 2700–2713.
D. Z. Liao and S. P. Yan, Cryst. Growth Des., 2007, 7, 1851– 37 E. G. Moore, J. Grilj, E. Vauthey and P. Ceroni, Dalton
1857.
Trans., 2013, 42, 2075–2083.
38 A. T. Yeh, C. V. Shank and J. K. McCusker, Science, 2000,
289, 935.
31 E. Moore, Dalton Trans., 2012, 41, 5272–5279.
32 E. Jakubikova, W. Chen, D. M. Dattelbaum, F. N. Rein,
R. C. Rocha, R. L. Martin and E. R. Batista, Inorg. Chem., 39 S. Faulkner, L. S. Natrajan, W. S. Perry and D. Sykes, Dalton
2009, 48, 10720–10725.
Trans., 2009, 3890–3979.
7408 | Dalton Trans., 2021, 50, 7400–7408
This journal is © The Royal Society of Chemistry 2021