Thermodynamic Programming of Erbium(III) Coordination Complexes for Dual Visible/Near-Infrared Luminescence

Article abstract of DOI:10.1002/chem.201802277

Intrigued by the unexpected room-temperature dual visible/near-infrared (NIR) luminescence observed for fast-relaxing erbium complexes embedded in triple-stranded helicates, in this contribution, we explore a series of six tridentate N-donor receptors L4–L9 with variable aromaticities and alkyl substituents to extricate the stereoelectronic features responsible for such scarce optical signatures. Detailed solid-state (X-ray diffraction, differential scanning calorimetry, optical spectroscopy) and solution (speciations and thermodynamic stabilities, spectrophotometry, NMR and optical spectroscopy) studies of mononuclear unsaturated [Er(Lk)₂]³⁺ and saturated triple-helical [Er(Lk)₃]³⁺ model complexes reveal that the stereoelectronic changes induced by the organic ligands affect inter- and intramolecular interactions to such an extent that 1) melting temperatures in solids, 2) the affinity for trivalent erbium in solution, and 3) optical properties in luminescent complexes can be rationally varied and controlled. With this toolkit in hand, mononuclear erbium complexes with low stabilities displaying only NIR emission can be transformed into molecular-based dual Er-centered visible/NIR emitters operating at room temperature in both solid and solution states.

Full text of DOI:10.1002/chem.201802277

A Journal of
Accepted Article
Title: Thermodynamic Programming of Erbium(III) Coordination
Complexes for Dual Visible-Near Infrared Luminescence
Authors: Claude Piguet, Bahman Golesorkhi, Laure Guénée,
Homayoun Nozary, Alexandre Fürstenberg, Yan Suffren,
Svetlana Eliseeva, Stéphane Petoud, and Andreas Hauser
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.201802277
Link to VoR: http://dx.doi.org/10.1002/chem.201802277
Supported by

10.1002/chem.201802277
Chemistry - A European Journal
1
Thermodynamic Programming of Erbium(III) Coordination Complexes for Dual
Visible-Near Infrared Luminescence
__________________________________________________________________
Bahman Golesorkhi,[a] Laure Guénée,[b] Homayoun Nozary,[a] Alexandre Fürstenberg,[a] Yan
Suffren,[c],[e] Svetlana V. Eliseeva,[d] Stéphane Petoud,*[d] Andreas Hauser*[c] and Claude Piguet*[a]
[a] Mr. B, Golesorkhi, Dr. H. Nozary, Dr A. Fürstenberg, Prof. Dr C. Piguet Department of Inorganic
and Analytical Chemistry University of Geneva, 30 quai E. Ansermet, CH-1211 Geneva 4,
(Switzerland). E-mail: Claude.Piguet@unige.ch
[b] Dr L. Guénée Laboratory of Crystallography, University of Geneva, 24 quai E. Ansermet, CH-
1211 Geneva 4 (Switzerland).
[c] Dr. Y. Suffren, Prof. Dr A. Hauser, Department of Physical Chemistry University of Geneva, 30
quai E. Ansermet, CH-1211 Geneva 4 (Switzerland). E-mail: Andreas.Hauser@unige.ch
[d] Dr S. V. Eliseeva, Prof. Dr S. Petoud, Centre de Biophysique Moléculaire, CNRS UPR 4301, Rue
Charles Sadron, F-45071 Orléans Cedex 2 (France). E-mail : Stephane.Petoud@inserm.fr
[e] Current address: Université Rennes, INSA Rennes, CNRS, ISCR “Institut des Sciences Chimiques
de Rennes”, F-35708 Rennes (France).
__________________________________________________________________
Supporting information and the ORCID identification number(s) for the author(s) of this article can
be found under https://doi.org/10.1002/chemxxxxx
This article is protected by copyright. All rights reserved.

10.1002/chem.201802277
Chemistry - A European Journal
2
Abstract
Intrigued by the unexpected room-temperature dual visible-NIR luminescence observed for fast-
relaxing erbium complexes embedded in triple-stranded helicates, this contribution explores a series
of six tridentate N-donor receptors L4-L9 with variable aromaticities and alkyl substituents for
extricating the stereo-electronic requirements responsible for such scarce optical signatures. Detailed
solid state (X-ray diffraction, differential scanning calorimetry, optical spectroscopy) and solution
(speciations and thermodynamic stabilities, spectrophotometry, NMR and optical spectroscopy)
studies of mononuclear unsaturated [Er(Lk)2]3+ and saturated triple-helical [Er(Lk)3]3+ model
complexes reveal that the stereo-electronic changes induced by the organic ligands affect inter- and
intra-molecular interactions to such an extent that (i) melting temperatures in solids, (ii) the affinity
for trivalent erbium in solution, and (iii) optical properties in luminescent complexes can be rationally
varied and controlled. With this toolkit in hand, mononuclear erbium complexes with low stabilities
displaying only near-infrared (NIR) emission can be transformed into molecular-based dual Er-
centered visible/NIR emitters operating at room temperature in solids and in solutions.
Introduction
Trivalent neodymium (4f3), holmium (4f10), but especially erbium (4f11) are famous for possessing a
series of regularly spaced 2S+1LJ excited spectroscopic levels which cover the entire near-infrared
(NIR), visible (VIS) and ultra-violet (UV) domains (Scheme 1).[1] The radiative relaxation processes
taking place between the different levels, quantified by their radiative rate constants kr and lifetimes
r = 1/kr obey Einstein’s probability AJ’,J (in s-1) of spontaneous J’J emission (Equation (1)), where
h is Planck’s constant,  is the energy gap (in frequency unit) between the two incriminated J and J’
states, c is the speed of light, gJ and gJ’ are the degeneracies of states J and J’, respectively, and BJ,J’
is Einstein’s coefficient giving the probability per unit time and per unit spectral energy density of
the radiation field that an electron in state J absorbs a photon and jumps to state J’.[2]
1
8h 3 gJ
kr   AJ ',J

BJ ,J '
(1)
r
c3 gJ '
This article is protected by copyright. All rights reserved.

10.1002/chem.201802277
Chemistry - A European Journal
3
E/cm-1
rad/ms
2K15/2
4G9/2
UV
27700
27350
26400
1.36
0.10
0.07
28
24
20
16
12
8
4G11/2
2G9/2
24400
0.80
4F3/2
4F5/2
4F7/2
22300
21950
20300
19150
18350
0.78
0.67
0.26
0.36
1.06
VIS
2H11/2
4S3/2
4F9/2
15150
0.59
12250
10100
4I9/2
8.12
7.82
4I11/2
NIR
4I13/2
6500
3.53
4
0
4I15/2
0
Scheme 1. Details of the near-infrared, visible, and near-ultra-violet parts of the energy-level diagram
for Er3+ ions doped into yttrium orthoaluminate YAlO3 including calculated radiative lifetimes.8
2
ˆ
J H p J '
In quantum mechanics, BJ,J’ is proportional to the square of the transition moment
,
ˆ
where J and J’ are the wavefunctions of the J and J’ states and
is the electromagnetic-induced
Hp
perturbation hamiltonian, which involves electric and magnetic dipole components.[3] Within the
framework of Judd and Ofelt theory of forced electric dipole intrashell 4f-4f transitions, Eq. (1)
transforms into Eq. (2) for estimating the radiative rate constants in a lanthanide complex, where SED
and SMD are the electric-dipole and magnetic-dipole line strengths and n is the refractive index of the
medium.[4]
2
2




64 4e2 3
3h 2J 1
n n  2


SED  n3S
kr  AJ ',J

(2)


MD 

9


Since (i) the maximum energy gap between two adjacent energy levels in trivalent erbium compounds
does not exceed
= 4000 cm-1 (except for the maximum energy gap of 6500 cm-1 found between
Er(4I13/2) and the ground state Er(4I15/2), Scheme 1) and (ii) the associated dipole line strengths SED
This article is protected by copyright. All rights reserved.

10.1002/chem.201802277
Chemistry - A European Journal
4
and SMD are weak for f-f transitions,[4] Eq. (2) predicts small radiative rate constants, i.e. long radiative
lifetimes spanning the 0.1-10 ms range (Scheme 1, right column).[3],[4] On the other hand, the
competitive non-radiative relaxation pathways connecting states J’ and J in trivalent lanthanides
imply no significant displacement along any nuclear coordinates: a situation in line with the complete
neglect of the Huang-Rhys factors[5] and the simple use of the energy gap law given in Eq. (3). The
non-radiative rate constant, knr, thus decreases exponentially with increasing energy gaps E = EJ’-
EJ, as measured by the numbers p of available high-vibrational modes max required for filling these
intervals ( is a constant of the material).[6]
knr eE e p
(3)
In erbium(III)-doped ionic solids or nanoparticles, the effective vibrational energy is low and amounts
to eff  40-100 cm-1, from which peff  40-100 can be deduced for matching E  4000 cm-1. In
these conditions, Eq. (3) predicts that non-radiative processes are poorly efficient (knr ≤ kr), thus
producing long experimental excited lifetimes (close to the radiative lifetimes) and high quantum

yields. In erbium(III) coordination complexes, the effective vibrational energy reaches
 2000
eff
cm-1,[6] and peff reduces to  2. Non-radiative relaxation dominates throughout (knr >> kr) with no hope
for detecting luminescence, except for the weak near-infrared Er(4I13/24I15/2) emission at 6500 cm-
1, which benefits from peff  3.[7] Since the latter luminescent transition occurs at ca. 1.54 m, a
wavelength which is little attenuated by silica-based optical fibers and by biological tissues, Er3+ is
[10]
attractive for working as activator in lasers, in optical amplifiers exploiting silica,[9] Al2O3
or
LiNbO3 ionic hosts,[11] in plastic waveguides,[12] in organometallic light-emitting diodes[7],[13] and in
biological probes and sensors.[14] Major efforts were thus made for maximizing Er(4I13/24I15/2
)
emission quantum yield in molecular compounds[15] through the minimization of knr (i.e.
maximization of the Er(4I13/2) excited state lifetime) which accompanies the replacement of high-
energy X-H oscillators (X = C, N, O) located close to the activator with heavier X-D and X-F
analogues.[16] Such improvements open novel perspectives for inducing linear upconversion into
This article is protected by copyright. All rights reserved.

10.1002/chem.201802277
Chemistry - A European Journal
5
molecular complexes, a process during which two near-infrared photons are successively absorbed
by the erbium activator, prior to be eventually emitted as a single photon of higher energy.[17] This
mechanism, referred to as excited state absorption (ESA), was discovered during the late fifties in
ionic erbium-doped solids because the millisecond-lived excited states observed in these low-phonon
materials (see Scheme 1) are compatible with non-negligible probabilities for these intermediate
excited states to catch a second photon leading to doubly excited levels.[18] Linear upconversion was
significantly improved few years later by the demonstration that the indirect sensitization may greatly
help in feeding the pertinent long-lived erbium-centered excited states (energy transfer upconversion:
ETU),[19] and ETU is currently exploited for engineering functional solid materials and
nanoparticles.[20]
Scheme 2. Erbium-based coordination complexes exhibiting linear upconversion processes via the
[23]
ETU mechanism. X-ray crystal structures are shown for [CrErCr(L1)3](CF3SO3)9
and
[25]
[YbEr(L3)6(DME)2]
(color code: C = grey, N = dark blue, O = red, F= light blue). Chemical
structures deduced from spectroscopic data recorded in solution are depicted for [IR-806][Er(tta)4]
[22] and [(L2Er)F(L2Er)]+.[24]
This article is protected by copyright. All rights reserved.

10.1002/chem.201802277
Chemistry - A European Journal
6
To the best of our knowledge, the unambiguous implementation of ESA in molecular complexes
under reasonable incident pump intensities (i.e. below 1 kW/cm2) is currently unknown,[21] and the
few successful molecular erbium-centered upconverters reported so far exploit the ETU mechanism
as found in [IR-806][Er(tta)4] (use of a polyaromatic sensitizer),[22] [CrErCr(L1)3]9+(use of a d-block
sensitizer),[23] [(L2Er)F(ErL2)]+ [24] and [YbEr(L3)6(DME)2] (use of f-block sensitizers, Scheme
2).[25] Beyond some classical optimization of the sensitization in the latter complexes maximizing
NIR absorption cross sections,[22] the ultimate induction of visible Er-centered emission, for instance
the green Er(4S3/24I15/2) signal, while at least one long-lived intermediate excited state relay of lower
energy is available, for instance Er(4I13/2), represents a major impediment for the implementation of
successful linear piling up of photons in these molecular systems. Looking at erbium chemistry, the
latter requirement for multiple (at least dual) emission is commonly fulfilled in solid-state samples
and nanoparticles,[17] and uponversion is therefore common in doped solids.[20],[26] The exact
mechanism responsible for the feeding of the various emissive levels is still subject of lively
debates.[27] On the contrary, molecular erbium complexes, when they are emissive upon ligand-
4
centered UV excitations,[6] usually deliver a single weak near-infrared I13/24I15/2 emission band,
which is diagnostic for the existence of a single long-lived intermediate excited state.[7],[14],[15] In this
context, the dual visible Er(4S3/24I15/2) (545 nm,
4 I13/2
= 40(2) ns) and near-infrared

4S3/2


Er(4I13/24I15/2) (1525 nm,
= 3.6(3) s) emission bands arising from the pseudo-tricapped



trigonal prismatic ErN9 site in [CrErCr(L1)3]9+ (Scheme 2) appears to be particularly intriguing
because no special care was taken for replacing H atoms with D or F atoms in the ligand strand.[23]
We report here on our efforts for identifying, by structural ‘dissection’ in ligands L4-L9 (Schemes 3-
4), the chemical design required for the successful induction of room-temperature dual visible/near
infrared emissions in stable ErN9 coordination sites: a prerequisite for the design of erbium-containing
molecular light-upconverters.
This article is protected by copyright. All rights reserved.

10.1002/chem.201802277
Chemistry - A European Journal
7
Results and Discussions
Tridentate receptors L4-L9: synthesis and substituent effects.
A careful look at the crystal structure of [CrErCr(L1)3]9+ cation (Scheme 2), suggests that the target
central luminescent ErN9 site could be reproduced in a mononuclear model via the helical wrapping
of three 2,6-bis(benzimidazol-2-yl)pyridine ligands around a trivalent lanthanide cation as found in
[Eu(L4)3](ClO4)3,[28] [Eu(L4’)3](ClO4)3 (L4’ is identical to L4, except for the replacement of N-
methyl groups with N-neopentyl analogues)[29] and [Eu(L5’)3](ClO4)3 (L5’ is identical to L5 except
for the replacement of N-methyl groups with N-ethyl analogues, see Scheme 3).[30] However,
repulsive intramolecular inter-strand interactions produced by the tight wrapping of three ligands
around the metallic cation in [Eu(L4)3]3+ is thought to be responsible for the unsuccessful isolation
of related complexes using cations smaller than Eu(III) along the lanthanide series.[31]
NaBH4
NO2
97%
Br
1
2
3
1) I2/PPh3
Imidazole
2) NaBH4
MeNH2
130 °C
99%
55%
MeNH2
130 °C
99%
5b
5c
SOCl2
DMF
4
7
6
: 5a
R = H
: 5b
: 5c
R = CH3
R = CH2CH3
iPrNEt2
77-92%
: 8a
: 8b
: 8c
R = H
R = CH3
R = CH2CH3
Fe/HCl
50-90%
: L4
: L5
: L6
R = H
R = CH3
R = CH2CH3
Scheme 3. Synthesis of 2,6-bis(5,5’-disubstituted-benzimidazol-2-yl)pyridine ligands L4-L6.
This article is protected by copyright. All rights reserved.

10.1002/chem.201802277
Chemistry - A European Journal
8
Since the ionic radius of the target Er3+ cation is significantly smaller than that of Eu3+, we exploited
compact N-methyl substituents for minimizing spatial expansion in ligands L4-L6, together with the
connection of terminal alkyl substitution of increasing size R = H (L4) < R= CH3 (L5) < R = CH2CH3
(L6) for preventing any infiltration of inner-sphere solvent molecules and/or counter-anions into the
core of the target [Ln(Lk)3]3+ triple helices. The syntheses of the bis-benzimidazolpyridine series
(bisbzimpy: L4-L6) rely on the well-established two-step reductive Philips-modified reaction of
activated dipicolinic acid 7 with substituted ortho-nitroaminophenyl derivatives 5a-c.[32] The
synthesis of L4, which used commercially available 5a, was previously reported.[28],[30] Analogous 5-
methyl derivatives 5b and the 5-ethyl derivative 5c can be easily obtained by aryl nucleophilic
substitution of adapted ortho-bromo-nitrophenyls 1 and 4,[33] thus leading to ligands L5 and L6
(Scheme 3).
HBr/Br2
NaNO2
50%
13
9
10
1) BuLi
2) (Bu)3SnCl
1) BuLi/
51%
83%
2) (Bu)3SnCl
15
11
14
Pd(PPh3)4
Toluene
49-51%
: L7
: L8
: L9
R = H
R = CH3
R = CH2CH3
Scheme 4. Synthesis of 5,5’’-disubstituted 2,2’:6’,2”-terpyridine ligands L7-L9.
Interestingly, the closely related, but less extended 2,2’:6’,2”-terpyridine ligand L7 is known to
produce stable triple-helical [Ln(L7)3]3+ along the complete lanthanide series, this without size-
discriminating effects.[34] It would be highly desirable to prepare the analogous terpyridine series
(terpy: L7-L9) for comparison purpose. Initial attempts to prepare the methyl (L8) and ethyl (L9)
derivatives using catalyzed alkyl/aryl cross-coupling reactions with 5,5”-dichloroterpyridine[35] gave
This article is protected by copyright. All rights reserved.

10.1002/chem.201802277
Chemistry - A European Journal
9
only low yields (< 20%) in our hands. A more successful approach used double Stille-type cross-
coupling reactions between 2,6-dibromopyridine 15[36] and stannylated 5-methylpyridine 11[37] and
5-ethylpyridine 14 (Scheme 4).[38]
Figure 1. Superimposition of molecular structures of a) L4 (red), L5 (blue) and L6 (green) and b) L7
(red), L8 (blue) and L9 (green) as observed in their crystal structures. Hydrogen atoms are omitted
for clarity and N-donor atoms are highlighted.
As previously observed in the crystal structures of L4[39] and L7,[40] the three connected heterocyclic
units in L5-L6 and L8-L9 adopt pseudo-twofold transoid-transoid arrangements of the nitrogen
donor atoms (Figure 1, Tables S1-S13 and Figure S1 in the Supporting Inofrmation). Variable
interplanar angles between the connected aromatic heterocycles (7.2 ≤  ≤ 42.8°) result from specific
intermolecular inter-aromatic packing interactions observed in the crystals (Figure S2). A thorough
thermodynamic analysis of the ligand melting processes (Figures S3-S8 and Appendix 1) within the
framework of enthalpy/entropy compensation,[41] combined with cohesive free energy densities,[42]
leads to the conclusion that the HCH3CH2CH3 sequence implemented in L4L5L6 and
L7L8L9 series produces two successive, but opposite changes in intermolecular cohesion in the
solid state. The replacement of small hydrogen atoms with methyl groups improves the cohesion due
the increase in molecular polarizability, while further extension provided by the introduction of
additional methylene rotors to give non-linear ethyl substituents severely disrupts the inter-aromatic
stacking. These three-dimensional substituent-induced packing effects, identified for the free ligand
L4-L9 in the solid state, have their molecular (i.e. zero-dimensional) counterpart in the target triple
This article is protected by copyright. All rights reserved.

10.1002/chem.201802277
Chemistry - A European Journal
10
helices [Ln(Lk)3]3+ where alkyl-modulated intramolecular inter-strand interactions are known to
control metal protection and crystal-field parameters.[28]-[31]
Erbium complexes with receptors L4-L9: speciation, stabilities and solution structures.
Electrospray-ionization mass spectroscopy (ESI-MS) titrations of ligands L4 (bisbzimpy) and L7
(terpy) with Er(CF3SO3)3 in acetonitrile show the stepwise formation of [Er(Lk)n]3+ (n = 3-1) with
some gas-phase adducts [Er(Lk)n(CF3SO3)i](3-i)+ in agreement with Equilibria (4)-(6) (Figures S9-
S10, Supporting Information).
Er Lk 3
Er,Lk
Er3
+
+
+
Lk
(4)
(5)
(6)


1,1
solvated

solvated
3
Er,Lk
Er3
2 Lk
3 Lk
Er Lk



2
1,2
solvated
solvated
3
Er,Lk
Er3
Er Lk



3
1,3
solvated
solvated
Beyond the ESI-MS signals observed for the protonated ligands [Lk+nH]n+, the intensity of which
are biased by their efficient gas-phase transfer coefficients,[43] only few faint ESI-MS signals can be
detected. They correspond to traces of hydrolyzed dimeric complexes [(Lk)2Er(OH)2Er(Lk)2]4+
which will therefore not be considered further in the thermodynamic modeling of the speciation. 1H
NMR titrations of the methyl-substituted ligands L5 and L8, which were selected for the simple first-
order analysis of their spin systems upon complexation with diamagnetic Y(CF3SO3)3 (the size of Y3+
is close to that of Er3+) confirm Equilibria (4)-(6) and the formation of only three complexes
[Y(Lk)n]3+ (n = 3-1) in acetonitrile at millimolar concentrations (Figures 2 and S11-S13). The strong
downfield shifts observed for H3 ( = 0.8-0.9 ppm), and to a lesser extend for the methyl groups H6
( = 0.5 ppm), between the free ligand and the 1:3 complexes [Y(Lk)3]3+ are diagnostic for the
formation of a triple helix (magenta dashed traces in Figures 2 and S11-S13), which puts these protons
in the shielding region of the terminal aromatic groups of the next ligand strands.[28],[30] As expected,
this effect is stepwise relaxed with the successive loss of ligand strands when [Y(Lk)3]3+ transforms
into [Y(Lk)2]3+ and [Y(Lk)]3+ (orange dashed traces in Figures 2 and S11-S13).
This article is protected by copyright. All rights reserved.

10.1002/chem.201802277
Chemistry - A European Journal
11
1
2
N
N
3
N
Y tot / L5 tot
N
N
6
7
x3
x5
x5
x1,2
x1,2
x4
x4
x4
4
5
1.85
x3
x3
1.22
1.05
x1,2
x5
x1,2
x3
x3
x5 y5 x4 y4
y2
y2
y1
y1
y1
y3
y3
0.78
0.60
0.55
0.50
0.45
0.40
y4
y4
y4
y4
y5
y5
y5
x1,2
x4
x5
y3
y3
y2
y2
y2
y2
y1
y1
y1
y5
z2
z4
z4
z5
z5
y3
z1
z1
z3
z3
z2
z2
y3
y4
y5
z3
z3
z5
z4
y3
z1
z1
y4
0.33
0.17
z2
z2
z5
z5
2
3
3
4
5
5
z4
z4
1
1
2
4
4
z3
z1
0.08
0.0
2
3
5
1
8.4 8.2 8.0 7.8 7.6 7.4 7.2 7.0 6.8 6.6 6.4
 ppm
Figure 2. Aromatic parts of the 1H NMR spectra with numbering scheme recorded upon titration of
L5 with Y(CF3SO3)3 in CD3CN at 298 K with numbering scheme. (5×10-3 ≤ L5tot ≤ 9×10-3 M and
6×10-4 ≤ Ytot ≤ 8×10-3 M). The letters x, y and z denote signals arising from the 1:1 (green), 1:2 (red)
and 1:3 (blue) species, respectively.
The associated speciation (dots in Figures 3a and S14a) can be obtained by simple integration of the
1H NMR signals of the same proton in ligands (ILk) and complexes (IY( )n, Eq. (7) middle). Only
Lk
partial binding isotherms could be obtained (occupancy factors Y > 60%; diamonds in Figures 3b
and S14b) because the NMR technique is not sensitive enough to fix total concentrations of metal
and ligands low enough to limit binding-site saturation. Attempts to fit the available occupancy
Er,Lk
1,n

factors Y with Eq. (7) gives only mere estimations of the stability constants
(gathered in the
captions of Figures 3 and S14), which are then used for the calculation of the theoretical speciation
curves and binding isotherms (full traces in Figures 3 and S14).
This article is protected by copyright. All rights reserved.

10.1002/chem.201802277
Chemistry - A European Journal
12
3
n
n Y,Lk Lk



1,n
(IYLk  IY(Lk)  IY(Lk) ) Lk tot
Lk
Lk  Lk
1
3
1
1
3
1
3
bound
tot
2
3
n1
tot
Y 



n
3
(7)
Y
3 (ILk  IYLk  IY(Lk)  IY(Lk) ) Y tot
Y
1  Y,Lk Lk
tot
2
3



1,n
n1
a) % L5
100
80
60
40
20
0
0.0
0.5
1.0
1.5
2.0
Y tot / L5 tot
b)
1.0
Y
0.8
0.6
0.4
0.2
0.0
-8
-6
-4
-2
0
log( L5 )
Figure 3. a) Macroscopic ligand speciations (dots) obtained by integration of 1H NMR signals during
the titration of L5 (1.6×10-2 molꞏdm-3) with Y(CF3SO3)3 in CD3CN at 298 K (black = L5, blue =
[Y(L5)3]3+, red = [Y(L5)2]3+ and green = [Y(L5)]3+) and b) associated binding isotherm (Eq. (7);
diamonds). The full traces correspond to the re-constructed a) distribution and b) binding isotherm
Y,L5
1,1
Y,L5
1,2
Y,L5
1,3
computed using Equilibria (4)-(6) with
= 8.0,
= 13.9 and

= 17.2.


As expected for ligand-metal association reactions (Eqs 4-6) monitored in weakly polar organic
media, the disruption of the chemical potential of the solvent induces a dependence of the quotient of
the reaction on the advance of the complexation process.[44] This prevents a simple determination of
Er,Lk
1,n

the true thermodynamic constants
, which are then obtained as rough average values from the
1H NMR data.[45] The associated binding isotherms depicted in Figures 3b and S14b therefore only
This article is protected by copyright. All rights reserved.

10.1002/chem.201802277
Chemistry - A European Journal
13
Er,Lk
1,n

roughly fit the theoretical curves predicted with Eq. 7. Pertinent values for
are restored when
1
titrations are recorded at much lower concentrations by spectrophotometry (vide infra). Related H
NMR titrations with ethyl-substituted ligands L6 and L9 obviously give very similar results (Figures
S15-S16), but the hydrogen atoms of the methylene probes are systematically enantiotopic (pure
quartets). This results points to the presence of time-averaged symmetry planes on the NMR time
scale for all complexes including, to our surprise, the D3-symmetrical triple helical complexes
[Y(Lk)3]3+. Making the assumption that a dynamic symmetry plane results from fast exchange
between the two D3-helical enantiomers on the NMR time scale, as previously reported for lanthanide
tris-dipicolinates,[46] low-temperature NMR data were collected, but they did not improve the
situation. However, it is worth reminding here that the experimentally accessible dynamic range of
an exchange process as measured by two diastereotopic protons depends on the difference of the
specific chemical shifts in absence of exchange  , which fixes the coalescence temperature Tc for a

given free energy barrier G (Eq. (8), kB and h are respectively the Boltzmann and the Planck
constants).[47] For diamagnetic yttrium complexes [Y(Lk)3]3+, the two hydrogen atoms of the
methylene probe likely display very similar chemical environments (i.e  is small) and Tc lies thus
much below 233 K, the minimum accessible temperature in acetonitrile.


2kB T
h 
G  RT ln


c 
(8)



Replacing Y(III) with paramagnetic Er(III) boosts the difference in chemical shifts between the
methylene hydrogen atoms H6 and H6’. Two signals, separated by  = 220 Hz, are now recorded
for [Er(L9)3]3+ at room temperature, in agreement with the existence of a ‘blocked’ triple-helical
1
arrangement in solution (Figure S17, Supporting Information). High-temperature H NMR data

combined with linear extrapolation give an estimation of Tc = 433(34) K, from which G = 59(4)
kJ/mol can be computed with Eq. (8) for the intramolecular helical interconversion. The latter values
compares well with G(Tm) = 64(1) kJ/mole (Tc = 290 K) and G(Yb) = 60(1) kJ/mole (Tc = 303
This article is protected by copyright. All rights reserved.

10.1002/chem.201802277
Chemistry - A European Journal
14
K) previously reported for related intramolecular helical exchange processes occurring in D3-
symmetrical [Ln(2,6-dipicolinate)3]3- in D2O.[46]
Having established that the speciation in solution obeys Equilibria (4)-(6), pertinent thermodynamic
-
stability constants were obtained by spectrophotometric titrations of Lk with [ErX3] (X = CF3SO3 ,
-
ClO4 and [Al(OC(CF3)3)4]-) collected at low concentration in acetonitrile (Figures 4a,b and Figures
S18-S22). Evolving factor analysis[48] confirms the existence of four absorbing species (Figure 4c),
while non-linear least-square fit[49] to Equilibria (4)-(6) provide stability constants (Table 1) together
with acceptable reconstructed absorption spectra (Figure 4d).
Figure 4. a) Variation of experimental absorption spectra and b) corresponding variation of molar
extinction at different wavelengths observed for the spectrophotometric titration of L4 with
Er(CF3SO3)3 (total ligand concentration: 3.010-4 M in acetonitrile, 298 K). c) Evolving factor
analysis[48] using four absorbing eigenvectors and d) reconstructed individual electronic absorption
spectra.
-
-
Since poorly coordinating triflate (CF3SO3 ) or perchlorate (ClO4 ) anions are known to display
Er,Lk
1,n
similar affinities for Er3+,[50] the stability constants
obtained in the presence of these counter-

anions for complexing L4 and L7 are logically similar (Table 1, entries 1-8). On the contrary, the use
This article is protected by copyright. All rights reserved.

10.1002/chem.201802277
15
Chemistry - A European Journal
[51]
of less-coordinating perfluoroalkoxyaluminate anions [Al(OC(CF3)3)4]-
further increases the
affinity of Er3+ for ligands L4-L9 by one order of magnitude (Table 1, entries 9-14).
Er,Lk
Table 1. Thermodynamic formation constants log(
) (Equations (4)-(6)) and associated
1,n
GaEffri,nLikty  RT ln( f Er,Lk
)
and
ELk,Lk  RT ln(uLk,Lk
microscopic parameters
(Equations
)
-
(9)-(11) obtained for the spectrophotometric titrations of L4-L9 with ErX3 (X= CF3SO3
-
, ClO4 , [Al(OC(CF3)3)4]-) in acetonitrile at 298 K.
Er,Lk
affinity
ELk,Lk
G
Er,Lk
Er,Lk
Er,Lk
)
log(
)
log(
log(
)
Lk
Anion
1,1
1,3
1,2
kJmol-1
-51(1)
kJmol-1
-
L4
L7
L4
L5
L6
L7
L8
L9
L4
L5
L6
L7
L8
L9
ClO4
9.6(1)
9.8(1)
9.2(1)
10.9(4)
11.6(5)
10.5(2)
9.9(1)
9.2(1)
17.4(4)
17.1(1)
16.5(3)
20.2(8)
22(1)
22.7(4)
22.5(1)
20.9(3)
24.8(8)
26(1)
10(1)
11(1)
11(1)
15(4)
19(6)
15(1)
12(1)
10(1)
9(1)
-
ClO4
-50.7(3)
-49(1)
-
-
-
-
-
-
CF3SO3
CF3SO3
CF3SO3
CF3SO3
CF3SO3
CF3SO3
-60(3)
-66(5)
18.5(4)
17.8(2)
16.7(2)
19.3(5)
20.9(4)
22.8(7)
19.9(3)
18.5(2)
18.6(3)
23.0(4)
22.7(2)
21.8(2)
25.7(7)
27.1(4)
29.5(8)
26.8(5)
24.8(3)
24.7(3)
-56(1)
-53(1)
-49(1)
[Al(OC(CF3)3)4]- 10.4(3)
[Al(OC(CF3)3)4]- 12.0(2)
[Al(OC(CF3)3)4]- 12.9(3)
[Al(OC(CF3)3)4]- 11.2(2)
[Al(OC(CF3)3)4]- 10.4(1)
[Al(OC(CF3)3)4]- 10.1(1)
-56(1)
-63.9(1)
-69.6(5)
-59.1(4)
-54.7(2)
-54(1)
14.7(1)
15.7(6)
10.5(4)
9.9(2)
9(1)
A global thermodynamic analysis exploits the site-binding model, which considers the successive
complexation of tridentate ligands to Er3+ within the frame of next-neighbor Ising-like interacting
chains.[52] According to this simple approach, the purely entropic contributions produced by the
This article is protected by copyright. All rights reserved.

10.1002/chem.201802277
Chemistry - A European Journal
16
change in rotational entropies (i.e. the statistical factors) are safely estimated with the help of
symmetry numbers to give the numerical values given in Equations (9)-(11) (see Figure S23 for
calculations).[53] The free energy change accompanying the chemical reaction is then partitioned
between the simple intermolecular affinity (including solvation effects) accompanying the
GaEfrfi,nLikty  RT ln( f Er,Lk
)
), and the modulation of the latter
3+
complexation of Er to a single ligand Lk (
affinity due to successive ligand binding in [Er(Lk)n]3+ (n  2); a thermodynamic correction
ELk,Lk  RT ln(uLk,Lk
)
often referred to as allosteric cooperativity (eqs 9-11).[54]
(9)
Er,Lk

6f Er,Lk
1,1
2
Er,Lk
1,2


12 f Er,Lk uLk,Lk


(10)
(11)
3
3
Er,Lk
1,3
16 f Er,Lk uLk,Lk

 

Multi-linear least-square fits of the experimental successive stability constants log( Er,Lk ) to the

1,n
Er,Lk
Gaffinity  RT ln( f Er,Lk
)
Lk,Lk
 RT ln(uLk,Lk
)
logarithmic forms of Equations (9)-(11) provide
and
E
gathered in Table 1 (columns 6-7) and illustrated in Figure 5. As expected, the intrinsic metal-ligand
affinities are favored when triflate anions are replaced with the less competitive counter-anion
[Al(OC(CF3)3)4]-, but the successive connections of alkyl groups of larger sizes at the termini of the
GEr,Lk
ligands induce surprising opposite trends when plotting
along the bisbzimpy series (L4-L6)
affinity
and terpy (L7-L9) series (Figure 5a). The latter behavior is paralleled by the anti-cooperative
ELk,Lk
interligand interactions
(Figure 5b). The regular dependence of both intrinsic affinities and
interligand interactions on the molecular volumes[55] within each series suggests the prevalence of
solvation effects in controlling the thermodynamic binding properties as previously established for
helicate self-assemblies[56] and for the metal loading of linear polymers.[57] Pertinent Born-Haber
cycles developed in Appendix 2 demonstrate that, whilst the thermodynamic stabilities of the
[Er(Lk)n]3+ (n = 1-3) complexes in acetonitrile are globally comparable for all six ligands, the minor
variations along the series observed in Figure 5 can be rationalized by a fine balance between
This article is protected by copyright. All rights reserved.

10.1002/chem.201802277
Chemistry - A European Journal
17
inductive effect and size-dependent solvation energies. Specific inter-strand interactions can be
highlighted in [Er(L5)3]3+ and [Er(L6)3]3+ when three bulky extended aromatic ligands are wrapped
around small Er3+.
GaEffri,Lnikty /kJmol-1
a)
-45
L4
L9
L8
L7
L5
-55
-65
-75
L6
150
250
Vm /Ǟ3
350
Lk,Lk
/kJmol-1
b)
E
25
20
15
10
5
L7
L8
L9
L6
L5
L4
150
250
Vm /Ǟ3
350
Er,Lk
Gaffinity  RT ln( f Er,Lk
Figure 5. a) Intermolecular microscopic affinities
for inner sphere connection
)
of Er(CF3SO3)3 (red) and Er([Al(OC(CF3)3)4])3 (blue) to the tridentate binding sites in ligands L4-L9
(CH3CN, 298 K). b) Interligand interactions ELk,Lk  RT ln(uLk,Lk ) operating upon successive
binding of L4-L9 to a single metallic center in Er(CF3SO3)3 (red) and Er([Al(OC(CF3)3)4])3 (blue).
Vm is the molecular volume of the ligand taken as its Connolly volume.[55] The dashed trendlines are
only guides for the eyes.
Erbium complexes with receptors L4-L9: isolation and solid-state structures.
According to the stability constants collected in Table 1, we used close to molar concentrations of
ligands and ErX3 salts at 3:1 stoichiometric ratio for producing  98% of [Er(Lk)3]3+ in acetonitrile
This article is protected by copyright. All rights reserved.

10.1002/chem.201802277
Chemistry - A European Journal
18
(Figures S24-S25), prior to exposing resulting mixtures either to slow evaporation or to the diffusion
of poorly polar volatile solvents for inducing crystallization. In the presence of triflate counter-anions
-
(X = CF3SO3 ), only 1:2 complexes could be crystallized (Figure 6a, Tables S14-S20 and Figures
S26-S28). Since the latter stoichiometry accounts for a maximum 2% of the ligand speciation in
solution, these complexes are probably much less soluble than the lipophilic 1:3 adducts. Complexes
[Er(L4)2(CF3SO3)2](CF3SO3) (1) and [Er(L7)2(CF3SO3)2](CF3SO3) (2)[21a] display eight-coordinate
Er(III) centers, the coordination geometry of which can be assigned to slightly distorted square
antiprisms. Trivalent erbium in [Er(L8)2(CF3SO3)3] (3) is nine-coordinated in a pseudo-tricapped
trigonal prismatic coordination sphere (Figure S29).
Figure 6. Molecular structures of a) mononuclear 1:2 complexes and b) dinuclear 2:4 complexes
found
[Er(L7)2(CF3SO3)2](CF3SO3)1.5C2H5CN
[(L4)2Er(OH)2Er(L4)2](ClO4)42C6H5CN4CH3CN
in
the
crystal
structures
of
[Er(L4)2(CF3SO3)2](CF3SO3)2CH3CN
(2),[21a]
[Er(L8)2(CF3SO3)3]
(1),
(3),
(8),
[(L5)2Er(OH)2Er(L5)2](ClO4)4
C6H5CN8CH3CN (9) and [(L4)2Er(CH3NO2)2Er(L4)2](ClO4)6CH3NO2 (10). The counter-anions,
solvent molecules and H atoms are omitted for clarity. Color code: C = grey, blue = N, red = O,
yellow = S, magenta = F, green = Er.
Replacing triflate with perchlorate counter-anions reduces the solubility of the triple helices
[Er(Lk)3]3+, which quantitatively crystallized to give [Er(L6)3](ClO4)3 (4), [Er(L7)3](ClO4)3 (5),
[Er(L8)3](ClO4)3 (6) and [Er(L9)3](ClO4)3 (7, Figure 7, Tables S21-S29 and Figures S30-S33). The
This article is protected by copyright. All rights reserved.

10.1002/chem.201802277
Chemistry - A European Journal
19
nine-coordinate Er(III) sites in these complexes systematically adopt pseudo-tricapped trigonal
prismatic geometries (Figure S34). The detailed analysis of Er-N bond distances within the
framework of bond valence theory (Table S30)[58] logically confirms that the Er-N interactions in
eight-coordinate [Er(L4)2(CF3SO3)2]+ and [Er(L7)2(CF3SO3)2]+ complexes are slightly stronger than
those found in the corresponding nine-coordinated analogues [Er(L8)2(CF3SO3)3], [Er(Lk)3]3+ (Lk =
L6-L9), [Eu(L4)3]3+ [28],[30] and [Lu(L7)3]3+.[34b] We also note that a significant intramolecular inter-
strand packing[59] produced by the tight wrapping of the helical strands in [Eu(L4)3]3+ [28],[30] and
[Er(L6)3]3+ (Figure 7 top left and Figure S36) are lacking for the terpy series.
Figure 7. Molecular structures of mononuclear 1:3 complexes found in the crystal structures of
[Er(L6)3](ClO4)31.5CH3CN
(4),
[Er(L7)3](ClO4)3
(5),
[Er(L8)3](ClO4)3
(6)
and
[Er(L9)3](ClO4)31.5CH3CN (7). The counter-anions, solvent molecules and H atoms are omitted for
clarity. Color code: C = grey, blue = N, green = Er.
In line with previous unsuccessful attempts to isolate Lu-based triple helical complexes with L4 and
L5 in the solid state,[28],[30] we were unable to isolate [Er(Lk)3](ClO4)3 with these two ligands. Slow
evaporation of Er(ClO4)3/Lk (1:3) mixtures (k = 4, 5) over several weeks provided only small
quantities of dinuclear complexes [(L4)2Er(OH)2Er(L4)2](ClO4)4 (8), [(L5)2Er(OH)2Er(L5)2](ClO4)4
(9) and [(L4)2Er(CH3NO2)2Er(L4)2](ClO4)6 (10), the stoichiometry of which is reminiscent of
mononuclear 1:2 complexes (Figure 6b, Tables S31-S37 and Figures S36-S38). According to the fact
that no identified thermodynamic effect specifically destabilizes [Er(Lk)3]3+ complexes with L4 and
This article is protected by copyright. All rights reserved.

10.1002/chem.201802277
Chemistry - A European Journal
20
L5 (Table 1), we conclude that mass transfers at the solid/solution phase interfaces master solubility
during the crystallization processes and the compositions of the resulting solid state samples do not
mirror the thermodynamic speciation.
Erbium complexes with receptors L4-L9: dual visible/near-infrared luminescence.
Wavelength / nm
a)
500
400
350
300
250
4
3
2
1
0
11*  1
L4
L5
L6
L7
L8
L9
20
25
30
35
40
Wavenumber / 103 cm-1
Wavelength / nm
b)
500
400
350
300
250
6
4
2
0
11*b  1
11*a  1
[Er(L4)2]3+
[Er(L7)2]3+
[Er(L8)2]3+
20
25
30
35
40
Wavenumber / 103 cm-1
Wavelength / nm
c)
500
400
11*a  1
350
300
11*b  1
250
6
4
2
0
[Er(L6)3]3+
[Er(L7)3]3+
[Er(L8)3]3+
[Er(L9)3]3+
20
25
30
35
40
Wavenumber / 103 cm-1
Figure 8. Absorption spectra of a) ligands L4-L9 (0.3 mM), b) 1:2 complexes 1-3 (3.0 mM) and c)
1:3 complexes 4-7 (3.0 mM) in acetonitrile solution at 298K.
This article is protected by copyright. All rights reserved.

10.1002/chem.201802277
Chemistry - A European Journal
21
Figure 9. Emission spectra of a) ligands L4-L6 (0.3 mM, exc = 330 nm) and L7-L9 (0.3 mM, exc
=
280 nm), b) 1:2 complexes 1-3 (3.0 mM) and c) 1:3 complexes 4-7 (3.0 mM) in acetonitrile solution
at 298 K (exc = 370 nm for complexes with L4-L6 and exc = 340 nm for complexes with L7-L9; a
cutoff filter at 400 nm is used).
The electronic absorption spectra of the free ligands L4-L9 in acetonitrile solution display the typical
*
1  transitions characteristic for the bisbzimpy (L4)[30],[60] and terpyridine (L7)[61] series (Figure
This article is protected by copyright. All rights reserved.

10.1002/chem.201802277
Chemistry - A European Journal
22
8a). The 4500 cm-1 red-shift combined with the doubling of the absorption coefficient resulting from
the aromatic extension in going from the terpy series (L7-L9, dotted traces in Figure 8) to bisbzimpy
(L4-L6, full traces in Figure 8) makes the latter ligands more efficient for acting as sensitizers for
luminescent lanthanides in [Ln(Lk)n]3+ complexes.[62] The minor additional 400 cm-1 red-shift
induced by the connection of methyl (L5 and L8) or ethyl (L6 and L9) groups to the distal aromatic
rings are in line with the reduction of the HOMO-LUMO gap predicted by Extended Hückel
Molecular Orbital calculations.[30] These global shifts toward lower energies are mirrored by similar
trends observed in the associated emission spectra recorded in acetonitrile (298 K) upon excitation
*
into the 1  transitions (Figure 9a). According to their nanosecond lifetimes (Table S38 and Figure
1
*
S39), these emission bands can be safely assigned to singlet 1 1 transitions, the intensity of
which can be magnified at lower temperature (77 K in frozen solution, Figure S40a) where the
appearance of low-energy weak bands possessing longer lifetimes (> 1 s, Figure S40b) can be
*
attributed to the contribution of weak spin-forbidden 31 1 transitions. Upon complexation to Er3+
in [Er(Lk)n]3+ (n = 2, 3) at 3.0 mM concentration (> 90% of the ligand speciation corresponds to the
*
*
target complex), the 1  transition undergoes the well-known 1a,1b  splitting accompanying
the trans-trans  cis-cis rearrangement of the aromatic tridentate ligand binding units: a spectral
feature diagnostic for their meridional tri-coordination to a central metallic center (Figures 8b-
c).[30],[60],[61] Concomitant measurements in the solid state (i.e. at higher concentration) confirm the
latter splitting process (Figure S41) together with the detection of additional weak Er(2S+1LJ4I15/2
)
absorption transitions covering the 6000-25000 cm-1 spectral window (Figure S42). Interestingly, the
radiative lifetime associated to any emission transition terminating onto the ground Er(4I15/2) level
can be estimated by its oscillator strength calculated with Eq. (12) where
is the integrated


(v)dv

J
spectrum of the incriminated absorption transition recorded in solution, J and
refer to the ground
(J = 15/2) and excited states, respectively, n is the refractive index of the medium, NA is Avogadro’s
is the barycenter of the transition (Table 2).[4]

vm
number, c is the speed of light in vacuum and
This article is protected by copyright. All rights reserved.

10.1002/chem.201802277
Chemistry - A European Journal
23
2
2

1
8cn vm (2J 1)
 
(v)dv
 2303
(12)

rad
NA (2J 1)
Table 2
Radiative lifetimes calculated with Eq. (20) for selected Er(III) excited levels in the
complexes 1-7 (acetonitrile, 298 K).
4 I13/2
4 I11/2
4 S3/2
2 H11/2
4 F7/2
Er,rad Er,rad Er,rad Er,rad Er,rad
State /
Compound
Solvent
/ ms
6.3
9.1
6.1
8.7
5.6
8.0
6.3
9.1
6.4
9.2
6.1
8.8
6.1
8.7
/ ms
4.3
6.1
5.6
8.1
6.3
9.0
5.4
7.8
4.4
6.3
6.8
9.8
6.9
9.9
/ μs
546
785
515
741
453
651
716
/ μs
177
254
301
433
259
372
282
/ μs
246
353
269
387
266
382
405
582
352
506
299
430
299
430
[Er(L4)2(O3SCF3)2](CF3SO3) Solid[a]
CH3CN[b]
[Er(L7)2(O3SCF3)2](CF3SO3) Solid[a]
CH3CN[b]
Solid[a]
Er(L8)2(O3SCF3)3
[Er(L6)3](ClO4)3
[Er(L7)3](ClO4)3
[Er(L8)3](ClO4)3
[Er(L9)3](ClO4)3
CH3CN[b]
Solid[a]
CH3CN[b]
Solid[a]
1029 406
476
684
493
709
495
712
442
636
418
602
435
626
CH3CN[b]
Solid[a]
CH3CN[b]
Solid[a]
CH3CN[b]
[a]
Er,rad
For the solid-state samples,
was calculated assuming that the structures are similar in
different media and taking into account a simple n3 dependence (Eq. (2)) of the refractive index with
nsolid = 1.517 and nCH3CN = 1.344.[4b],[63] [b] c = 3 mM.
Excitation into the ligand-centered Lk(1b ) transition in the unsaturated [Er(Lk)2]3+ complexes
*
(Lk = L4, L7, L8) in solution at room temperature shows the residual emission of the ligand-centered
singlet emission (Figure 9b and Table S39). Upon freezing the solution at 77 K, additional very weak
This article is protected by copyright. All rights reserved.

10.1002/chem.201802277
Chemistry - A European Journal
24
green Er(4S3/24I15/2) emission bands ((4S3/2) = 8-12 ns, Table S40) can be identified around 18450
cm-1 (Figure S43), the intensity of which further increases for solid-state samples measured at 7-11
4 S3/2
Er,rad

K (Figure S44). Compared with natural radiative lifetimes of 540 <
< 750 s calculated for this
transition in the corresponding 1:2 complexes (Table 2, column 5), the observed nanosecond
4
4 S3/2
Er,exp
4 S3/2
Er,rad
QEr S3/2  
Er(4S3/24I15/2) emission corresponds to small Er-centered quantum yields


Er
< 0.002% (Table S40). The latter visible Er-centered luminescence becomes more sizeable in the
saturated triple-stranded helical [Er(Lk)3]3+ complexes (Lk = L6-L9), where it can be now easily
detected at room temperature in solution (Figure 9c) additionally to the residual ligand-centered
singlet emission (Table S41). This observation mirrors the previous and rather surprising report of
detectable upconverted green Er(4S3/24I15/2) emission implemented in [CrErCr(L1)3]9+, a
4 S3/2
Er,exp
coordination complex possessing high-energy oscillators.[23a] Taking
= 40(2) ns measured in

[GaErGa(L1)3](CF3SO3)9 at 3K[23c] as an upper limit of the experimental Er(4S3/2) lifetimes in these
[ErN9] chromophores, the intrinsic quantum yield of the Er(4S3/24I15/2) emission still remains
4
QEr S3/2


limited (
< 0.01%) in mononuclear triple-helical complexes, but still roughly one order of
Er
magnitude larger than in unsaturated complexes. This observation justifies the use of solid-state
samples of [Er(Lk)3](ClO4)3 complexes (Lk = L6-L9), where local concentration and antenna effect
are maximized, for getting well-resolved emission spectra both at room temperature (Figure 10) and
at 77 K (Figure S45). Beyond the uncommon visible Er(4S3/24I15/2) luminescence observed upon
ligand-centered excitation (main plot in Figure 10), the latter triple helical complexes additionally
display standard Er-centered infrared emission associated with Er(4I13/24I15/2) transitions occurring
4 I13/2
Er,exp
-1

around 6300 cm (inset in Figure 10), the room temperature lifetimes of which 1 <
< 6 s
Er
Er
4 I15/2
 0.1% quantum yields for the best protected [Er(L6)3](ClO4)3
corresponds to intrinsic
Q


4
triple helix. It is worth noting here that the induction of dual Er-centered S3/24I15/2 (Visible) and
4I13/24I15/2 (NIR) emission is exceptional in coordination complexes and the detailed investigation
This article is protected by copyright. All rights reserved.

10.1002/chem.201802277
Chemistry - A European Journal
25
of the less-protected hydroxyl-bridged analogous [(L4)2Er(OH)2Er(L4)2](ClO4)4 (8) and
[(L5)2Er(OH)2Er(L5)2](ClO4)4 (9), or solvent-bridged [(L4)2Er(CH3NO2)2Er(L4)2](ClO4)6 (10)
dimers indeed show no trace of visible emission, even in the solid state even at low temperature
(Figure S46).
Figure 10. Dual visible (main plot) and near-infrared (inset) emission spectra recorded for the
complexes 4 (exc = 370 nm) and 5-7 (exc = 340 nm) in the solid-state at room temperature. The dual
emission is assigned using color fonts, while the dips (assigned using black fonts) correspond to Er-
centered re-absorption of the residual ligand-centered 1* emission.
Conclusion
In our attempt to reproduce the intriguing luminescent properties of [ErN9] chromophores found in
the self-assembled trinuclear d-f helicate [CrErCr(L1)3]9+ (Scheme 2), into simple, stable and tunable
mononuclear triple-helical model complexes [Er(Lk)3]3+ (Lk = L4-L9), we discovered that the stereo-
electronic requirements responsible for the uncommon room-temperature dual Er-centered
visible/near-infrared emission essentially depend on the tight wrapping of the three polyaromatic
strands around the metallic center. The Jablonski diagram shown in Figure 11 summarizes the
established mechanisms responsible for the generation of the dual visible/near-infrared luminescence
in the triple helical [Er(Lk)3]3+ complexes (Lk = L6-L9).
This article is protected by copyright. All rights reserved.

10.1002/chem.201802277
Chemistry - A European Journal
26
36
34
32
28
24
20
16
12
8
S1
T1
7/2
4G
4F
9/2
S1
T1
11/2
2H9/2
3/2
5/2
7/2
2H11/2
4S3/2
4F9/2
9/2
11/2
4I
13/2
4
15/2
0
ErIII
L7-L9
L6
Figure 11. Jablonski diagram summarizing the excitation processes (straight upward arrows), energy
transfers (dashed arrows), non-radiative multiphonon relaxation (undulating arrows) and radiative
emission processes (straight downward arrows) operating in the triple-helical complexes [Er(Lk)3]3+
(Lk = 6-9). Ligand-to-metal energy transfers arising from both singlet and triplet states are
considered.
The replacement of one ligand strand with low-frequency oscillators triflate counter-anions in
[Er(Lk)2(CF3SO3)n](3-n)+ chromophores (n = 2, 3) is still compatible with the detection of very weak
green Er(4S3/24I15/2) emission at 542 nm, but only at low temperature. As soon as hydroxide anions
(OH oscillators) or solvent molecules (CH oscillators) are bound in the Er(III) first coordination
sphere in the dimers [(L4)2Er(OH)2Er(L4)2](ClO4)4 (8), [(L5)2Er(OH)2Er(L5)2](ClO4)4 (9) and
[(L4)2Er(CH3NO2)2Er(L4)2](ClO4)6 (10), the visible emission is totally quenched. The extension of
the aromatic units in going from the terpy series (L7-L9, Scheme 4) to the bisbzimpy series (L4-L6,
Scheme 3) shifts the ligand-centered absorption bands toward lower energies, a trend in line with
potential visible sensitization of Er-centered luminescence in [Er(L4)3]3+, a process which is not
accessible for [Er(L7)3]3+. A further red-shift accompanies the peripheral substitution with terminal
methyl (L5, L8) or ethyl (L6, L9) groups, but no spectacular luminescence change results from the
protection of the termini of the [Er(Lk)3]3+ triple-helix with these ligands. However, the systematic
This article is protected by copyright. All rights reserved.

10.1002/chem.201802277
Chemistry - A European Journal
27
(often tedious) preparation and analysis of the Hmethylethyl structural series demonstrates its
pertinence for the rationalization of (i) the driving forces controlling the intermolecular interactions
operating in the solid state, which are responsible for phase transition temperatures, and (ii) the
solvation energies pertinent to thermodynamic association processes occurring in solution. To
summarize, we are now fully equipped for the molecular design of stable erbium coordination
complexes, the ErN9 chromophore of which displays room temperature dual visible/near-infrared
emission in solution and in the solid state at room temperature. Further use as activators for molecular-
based excited state absorption (ESA) and energy-transfer (ETU) linear upconversion are currently
under investigation.
Acknowledgements
Financial support from the Swiss National Science Foundation is gratefully acknowledged. The work
performed in France was supported by La Ligue contre le Cancer, Cancéropôle Grand Ouest, Agence
Nationale de la Recherche (ANR-13-BS08-0011) and INSERM.
Conflicts of interest
The authors declare no conflict of interest
Keywords: erbium complexes, thermodynamics, dual emission, solvation, triple helical
References
[1]
[2]
[3]
a) C. Piguet, Nature Chemistry 2014, 6, 370; b) C.-G. Ma, M. G. Brik, D.-X. Liu, B. Feng, Y.
Tian, A. Suchocki, J. Luminesc. 2016, 170, 369-374.
a) A. Einstein, Phys. Z. 1917, 18, 121–128; b) S. Strickler, S. J. Berg, J. Chem. Phys. 1962,
37, 814–822; c) J. B. Berks, D. J. Dyson, R. London Soc. Ser. A 1963, 275, 135–148.
a) W. T. Carnall, Handbook on the Physics and Chemistry of Rare Earths, K. A. Gschneidner
and L. Eyring, Eds. North-Holland Publishing Company: Amsterdam, New York, Oxford,
1979; Vol. 3, chap 24, pp 171-208; b) J.-C. G. Bünzli, S. V. Eliseeva, Lanthanide
This article is protected by copyright. All rights reserved.

10.1002/chem.201802277
Chemistry - A European Journal
28
Luminescence: Photophysical, Analytical and Biological Aspects, P. Hänninen and H. Härmä,
H., Eds. Springer-Verlag: Berlin Heidelberg, 2010; Vol. 7, pp 1-45.
[4]
[5]
a) L. D. Carlos, R. A. S. Ferreira, V. D. Bermudez, S. J. L. Ribeiro, Adv. Mater. 2009, 21,
509-534; b) J.-C. G. Bünzli, A.-S. Chauvin, H. K. Kim, E. Deiters, S. V. Eliseeva, Coord.
Chem. Rev. 2010, 254, 2623-2633.
T. C. Brunold, H. U. Güdel, Inorganic Electronic Structures and Spectroscopy; E. I. Solomon
and A. B. P. Lever, Eds; Wiley: New York, 1999, pp 259-306.
[6]
[7]
[8]
[9]
C. Reinhard, H. U. Güdel, Inorg. Chem. 2002, 41, 1048-1055.
F. Artizzu, M. L. Mercuri, A. Serpe, P. Deplano, Coord. Chem. Rev. 2011, 255, 2514-2529.
D. K. Sardar, S. Chandrasekharan, K. L. Nash, J. B. Gruber, J. Appl. Phys. 2008, 104, 023102.
T. Feuchter, E. K. Mwarania, J. Wang, L. Reekie, J. S. Williams, IEEE Photonics Technol.
Lett. 1992, 4, 542-544.
[10] G. N. van den Hoven, R. J. I. M. Koper, A. Polman, C. van Dam, J. W. M. Uffelen, K. M.
Smit, Appl. Phys. Lett. 1996, 68, 1886-1888.
[11] P. Becker, R. Brinkmann, M. Dinand, W. Sohler, H. Suche, Appl. Phys. Lett. 1992, 61, 1257-
1259.
[12] K. Kuriki, K. Koike, Y. Okamoto, Chem. Rev. 2002, 102, 2347-2356.
[13] a) J. Kido, Y. Okamoto, Chem. Rev. 2002, 102, 2357-2368; b) P. Martin-Ramos, C. Coya, A.
L. Alvarez, M. Ramos Silva, C. Zaldo, J. A. Paixao, P. Chamorro-Posada, J. Martin-Gil, J.
Phys. Chem. C 2013, 117, 10020-10030; c) H. Wei, G. Yu, Z. Zhao, Z. Liu, Z. Bian, C. Huang,
Dalton Trans. 2013, 42, 8951-8960.
[14] a) J. W. Hofstraat, M. P. Oude Wolbers, F. C. J. M. Van Veggel, D. Reinhoudt, M. H. V.
Werts, J. W. Verhoeven, J. Fluoresc. 1998, 8, 301-308; b) S. Quici, G. Marzanni, A. Forni,
G. Accorsi, F. Barigelletti, Inorg. Chem. 2004, 43, 1294-1301; c) G. M. Davies, H. Adams,
S. J. A. Pope, S. Faulkner, M. D. Ward, Photochem. Photobiol. Sci. 2005, 4, 829-834; d) S.
Comby, D. Imbert, C. Vandevyver, J.-C. G. Bünzli, Chem. Eur. J. 2007, 13, 936-944.
This article is protected by copyright. All rights reserved.

10.1002/chem.201802277
Chemistry - A European Journal
29
[15] a) A. Mech, A. Monguzzi, F. Meinardi, J. Mezyk, G. Macchi, R. Tubino, J. Am. Chem. Soc.
2010, 132, 4574-4576; b) E. R. Triveldi, S. V. Eliseeva, J. Jankolovits, M. M. Olmstead, S.
Petoud, V. L. Pecoraro, J. Am. Chem. Soc. 2014, 136, 1526-1534; c) B. L. Reid, S. Stagni, J.
M. Malicka, M. Cocchi, A. N. Sobolev, B. W. Skelton, E. G. Moore, G. S. Hanan, M. I.
Ogden, M. Massi, Chem. Eur. J. 2015, 21, 18354-18363; d) Y. Peng, J. X. Xu, H. Lu, R. M.
Wilson, M. Motevalli, I. Hernandez, W. P. Gillin, P. B. Wyatt, H. Q. Ye, RSC Adv. 2017, 7,
128-131.
[16] a) G. Mancino, A. J. Ferguson, A Beeby, N. J. Long, T. S. Jones, J. Am. Chem. Soc. 2005,
127, 524-5251; b) P. B. Glover, A. P. Basset, P. Nockemann, B. M. Kariuki, R. Van Deun, Z.
Pikramenou, Chem. Eur. J. 2007, 13, 6309-6320; c) L. Song, J. Hu, J. Wang, X. Liu, Z. Zhen,
Photochem. Photobiol. Sci. 2008, 7, 689-693; d) C. Doffek, N., Alzakhem, M. Molon, M.
Seitz, Inorg. Chem. 2012, 51, 4539-4545.
[17] a) F. Auzel, Chem. Rev. 2004, 104, 139-173; b) B. M. van der Ende, L. Aarts, A. Meijerink,
Phys. Chem. Chem. Phys. 2009, 11, 11081-11095.
[18] N. Bloembergen, Phys. Rev. Lett. 1959, 2, 84-85,
[19] a) F. Auzel, C. R. Acad. Sc. Paris 1966, B262, 1016-1019; b) F. Auzel, C. R. Acad. Sc. Paris
1966, B263, 819-821.
[20] a) G. Chem, H. Qiu, P. N. Prasad, X. Chen, Chem. Rev. 2014, 114, 5161-5214; b) P.
Ramasamy, P. Manivasakan, J. Y. Kim, RSC Adv. 2014, 4, 34873-34895; c) J. Zhou, Q. Liu,
W. Feng, Y. Sun, F. Li, Chem. Rev. 2015, 115, 395-465; d) S. Ye, E.-H. Song, Q.-Y. Zhang,
Adv. Sci. 2016, 3, 1600302; e) R. Medishetty, J. K. Zareba, D. Mayer, M. Samoc, R. A.
Fischer, Chem. Soc. Rev. 2017, 46, 4976-5004; f) X.-Y. Wang, R. R. Valiev, T. Y.
Ohulchanskyy, H. Agren, C. Yang, G. Chen, Chem. Soc. Rev. 2017, 46, 4150-4167.
[21] a) Y. Suffren, B. Golesorkhi, D. Zare, L. Guénée, H. Nozary, S. V. Eliseeva, S Petoud, A.
Hauser, C. Piguet, Inorg. Chem. 2016, 55, 9964-9972; b) L. J. Charbonnière, Dalton Trans.,
2018 DOI 10.1039/c7dt04737a.
This article is protected by copyright. All rights reserved.