Synthesis, structural studies, theoretical calculations, and linear and nonlinear optical properties of terpyridyl lanthanide complexes: New evidence for the contribution of f electrons to the NLO activity

Article abstract of DOI:10.1021/ja063586j

The synthesis and structural, photophysical, and second-order nonlinear optical (NLO) properties of a novel lanthanide terpyridyl-like complex family LLn(NO₃)₃ (Ln = La, Gd, Dy, Yb, and Y) are reported. The isostructural character of

Full text of DOI:10.1021/ja063586j

Published on Web 08/29/2006
Synthesis, Structural Studies, Theoretical Calculations, and
Linear and Nonlinear Optical Properties of Terpyridyl
Lanthanide Complexes: New Evidence for the Contribution of
f Electrons to the NLO Activity
Katell Se´ne´chal-David,† Anne Hemeryck,‡ Nicolas Tancrez,§ Lo¨ıc Toupet,
J. A. Gareth Williams,| Isabelle Ledoux,§ Joseph Zyss,§ Abdou Boucekkine,‡
Jean-Paul Gue´gan,† Hubert Le Bozec,*,† and Olivier Maury*,†,3
Contribution from the Laboratoire de Chimie de Coordination et Catalyse, UMR 6509
CNRS-UniVersite´ Rennes 1, Institut de Chimie, Campus de Beaulieu, 35042 Rennes Cedex,
France, Laboratoire de Chimie du Solide et Inorganique Mole´culaire, UMR 6511
CNRS-UniVersite´ Rennes 1, Institut de Chimie, Campus de Beaulieu, 35042 Rennes Cedex,
France, Groupe de la Matie`re Condense´e et Mate´riau, UMR 6626 CNRS-UniVersite´ de Rennes
1, Campus de Beaulieu, 35042 Rennes Cedex, France, Department of Chemistry, UniVersity of
Durham, South Road, Durham DH1 3LE, U.K., and Laboratoire de Physique Quantique
Mole´culaire, ENS Cachan, 61 aVenue du pre´sident Wilson, 94235 Cachan, France
Received June 1, 2006; E-mail: olivier.maury@ens-lyon.fr
Abstract: The synthesis and structural, photophysical, and second-order nonlinear optical (NLO) properties
of a novel lanthanide terpyridyl-like complex family LLn(NO3)3 (Ln ) La, Gd, Dy, Yb, and Y) are reported.
The isostructural character of this series in solution and in the solid state has been established on the
basis of X-ray diffraction analysis in the cases of yttrium and gadolinium complexes, theoretical optimization
of geometry (DFT), and NMR spectroscopy. The absorption, emission, and solvatochromic properties of
the free terpyridyl-like ligand L were thourougly investigated, and the twist intramolecular charge transfer
(TICT) character of the lowest energy transition was confirmed by theoretical calculation (TDDFT and CIS).
The similar ionochromic effect of the different lanthanide ions was evidenced by the similar UV-visible
spectra of the complete family of complexes. On the other hand, the quadratic hyperpolarizability coefficient
â, measured by the harmonic light scattering (HLS) technique, is clearly dependent on the nature of the
metal, and a careful examination of the particular case of yttrium unambiguously confirms the contribution
of metal f electrons to the NLO activity.
Introduction
of unpaired electrons, metallo-chromophores are very promising
for the design of molecular NLO switches3g,6,7 and multifunc-
tional materials combining NLO and magnetic properties, for
example.8,7e However, despite the large number of studies
reported in the literature, a better rationalization of the role of
Since the pioneering work of Frasier1 and Green,2 organo-
metallic and coordination complexes have been widely studied
for the design of chromophores for second-order nonlinear optics
(NLO).3 At the end of the 1980s, organotransition metal
complexes generally presented low NLO activity and were
mainly academic curiosities, but at present such compounds have
become increasingly important for the design of highly active
multidimensional dipolar4 and octupolar5 NLO-phores. More-
over, due to the intrinsic properties of transition metals, such
as the large number of stable oxidation states or the presence
(3) (a) Long, N. J. Angew. Chem., Int. Ed. Engl. 1995, 34, 21-38. (b) Whittall,
I. R.; McDonagh, A. M.; Humphrey, M. G.; Samoc, M. AdV. Organomet.
Chem. 1998, 42, 291-362; 1999, 43, 349-405. (c) Lacroix, P. G. Eur. J.
Inorg. Chem. 2001, 339-348. (d) Renouar, T.; Le Bozec, H. Eur. J. Inorg.
Chem. 2001, 229-239. (e) Di Bella, S. Chem. Soc. ReV. 2001, 30, 355-
366. (f) Coe, B. J. In ComprehensiVe Coordination Chemistry II; McClev-
erty, J. A., Meyer, T. J., Eds.; Elsevier Pergamon: Oxford U.K., 2004;
Vol 9, pp 621-687. (g) Powell, C. E.; Humphrey, M. G. Coord. Chem.
ReV. 2004, 248, 725-756. (h) Maury, O.; Le Bozec, H. Acc. Chem. Res.
2005, 38, 691-704.
(4) (a) Le Cours, S. M.; Guan, H.-W.; DiMagno, S. G.; Wang, C. H.; Therien,
M. J. J. Am. Chem. Soc. 1996, 118, 1497-1503. (b) Zhang, T.-G.; Zhao,
Y.; Asselberg, I.; Persoons, A.; Clays, K.; Therien, M. J. J. Am. Chem.
Soc. 2005, 127, 9710-9720. (c) Alain, V.; Blanchard-Desce, M.; Chen,
C.-T.; Marder, S. R.; Fort, A.; Barzoukas, M. Synth. Met. 1996, 81, 133-
136. (d) Liao, Y.; Eichinger, B. E.; Firestone, K. A.; Haller, M.; Luo, J.;
Kaminsky, W.; Benedict, J. B.; Reid, P. J.; Jen, A. K.-Y.; Dalton, L. R.;
Robinson, B. H. J. Am. Chem. Soc. 2005, 127, 2758-2766.
† Laboratoire de Chimie de Coordination et Catalyse, Universite´ Rennes
1.
‡ Laboratoire de Chimie du Solide et Inorganique Mole´culaire, Universite´
Rennes 1.
§ ENS Cachan.
Groupe de la Matie`re Condense´e et Mate´riau, Universite´ Rennes 1.
| University of Durham.
3 Current adress: Laboratoire de Chimie, UMR 5182 ENS-Lyon, 46 Alle´e
d'Italie, F-69364 Lyon Cedex 07, France.
(5) (a) Maury, O.; Viau, L.; Se´ne´chal, K.; Corre, B.; Gue´gan, J.-P.; Renouard,
T.; Ledoux, I.; Zyss, J.; Le Bozec, H. Chem.sEur. J. 2004, 10, 4454-
4466. (b) Viau, L.; Bidault, S.; Maury, O.; Brasselet, S.; Ledoux, I.; Zyss,
J.; Ishow, E.; Nakatani, K.; Le Bozec, H. J. Am. Chem. Soc. 2004, 126,
8386-8387.
(1) Frasier, C. C.; Harvey, M. A.; Cokerham, M. P.; Hand, H. M.; Chauchard,
E. A.; Lee, C. H. J. Phys. Chem. 1986, 90, 5703-5706.
(2) Green, M. L. H.; Marder, S. R.; Thompson, M. E.; Bandy, J. A.; Bloor,
D.; Kolinsky, P. V.; Jones, R. J. Nature 1987, 330, 360-362.
9
10.1021/ja063586j CCC: $33.50 © 2006 American Chemical Society
J. AM. CHEM. SOC. 2006, 128, 12243-12255
12243

A R T I C L E S
Scheme 1 a
Se´ne´chal-David et al.
a (i) Solvent-free, 115 °C, 18 h, 90%; (ii) m-CPBA, CH2Cl2, RT, 3 h, 99%; (iii) (TfO)2O, CH2Cl2, RT, 2 h, 30 min, directly followed by K2CO3, MeOH,
RT, 3 h, 87%; (iv) MnO2, CH2Cl2, RT, 3 days, 58% after column chromatography; (v) (a) CHCl3, reflux, NH4OAc, 18 h; (b) NH3 (25% in water), RT, 4
h, 42% after column chromatography.
the metal in the global NLO activity of the chromophore is still
lacking. Transition metals can act (i) as purely donor end groups,
as in the case of metallocenes2,4c,4d or metal acetylides9 or (ii)
as purely inductive acceptor groups enforcing the intraligand
charge-transfer character (ILCT).10 In most cases, d orbitals are
involved in many different low-energy optical charge-transfer
transitions like metal-to-ligand (MLCT), ligand-to-metal (LMCT),
metal-centered (MC), intervalence (IVCT), or more complex
ligand-to-metal-to-ligand (LMLCT) charge-transfer transitions.
The overall NLO activity of the complex results from the
contributions of the transitions centered on both the metal and
the ligands (ILCT). These charge transfers are generally
multidirectionnal, so it is very difficult to dissociate and quantify
the various contributions and clearly define the role of the
metal.3c-e,5a,10d,11
It is worth noting that all the studies on metal-containing
chromophores have been carried out with transition metals, and
almost nothing is known about the potential of lanthanide
complexes in second-order NLO.12 This is very surprising
because the very strong chemical similarities of the lanthanide-
(III) ions (La-Lu) offer a unique opportunity to design an
isostructural series in which only intraligand charge-transfer
transitions occur. These reasons prompted us to undertake the
design of dipolar and octupolar lanthanide complexes in order
to study the influence of the metal on the second-order NLO
activity. Furthermore, we recently established, for the first time,
a clear correlation between the quadratic hyperpolarizabilities,
â, and the electronic f configuration of the rare earth in octupolar
Ln(DPA)3,3Na‚xH2O complexes (DPA ) 2,6-pyridinedicar-
boxylate).13 To confirm the contribution of f electrons to â, we
decided to prepare a second dipolar series based on a function-
alized annelated terpyridine ligand exibiting a higher NLO
activitity.14 In this paper are described the syntheses of the ligand
and the related lanthanide complexes (Ln ) La, Dy, Gd, Yb
and Y); the isostructural character of the familly was established
by means of solution, structural, and theoretical geometry
optimization studies. The photophysical properties (absorption
and luminescence) were carefully investigated and interpreted
using theoretical calculations. Finally the second-order NLO
properties were measured by harmonic light scattering (HLS)
techniques, unambiguously confirming the contribution of metal
f electrons to the complex quadratic hyperpolarizability.
(6) For a review see: Coe, B. J. Chem.sEur. J. 1999, 5, 2464-2471.
(7) (a) Coe, B. J.; Houbrechts, S.; Asselberghs, I.; Persoons, A. Angew. Chem.,
Int. Ed. 1999, 38, 366-368. (b) Weyland, T.; Ledoux, I.; Brasselet, S.;
Zyss, J.; Lapinte, C. Organometallics 2000, 19, 5235-5237. (c) Malaun,
M.; Reeves, Z. R.; Paul, R. L.; Jeffery, J. C.; McCleverty, J. A.; Ward, M.
D.; Asselberghs, I.; Clays; K.; Persoons, A. Chem. Commun. 2001, 49-
50. (d) Asselberghs, I.; Clays, K.; Persoons, A.; McDonagh, A. M.; Ward,
M. D.; McCleverty, J. A. Chem. Phys. Lett. 2002, 368, 408-411. (e) Sporer,
C.; Ratera, I.; Ruiz-Molina, D.; Zhao, Y.; Vidal-Gancedo, J.; Wurst, K.;
Jaitner, P.; Clays, K.; Persoons, A.; Rovira, C.; Veciana, J. Angew. Chem.,
Int. Ed. 2004, 43, 5266-5268.
(8) (a) Lacroix, P. G. Chem. Mater. 2001, 13, 3495-3506. (b) Margeat, O.;
Lacroix, P. G.; Costes, J.-P.; Donnadieu, B.; Lepetit, C.; Nakatani, K. Inorg.
Chem. 2004, 43, 4743-4750.
(9) For an exhaustive review, see ref 3g and references therein.
(10) (a) Bourgault, M.; Baum, K.; Le Bozec, H.; Pucetti, G.; Ledoux, I.; Zyss,
J. New J. Chem. 1998, 517-522. (b) Roberto, D.; Ugo, R.; Bruni, S.; Cariati,
E.; Cariati, F.; Fantucci, P. C.; Invernizzi, I.; Quici, S.; Ledoux, I.; Zyss,
J. Organometallics 2000, 19, 1775-1788. (c) Roberto, D.; Ugo, R.; Tessore,
F.; Lucenti, E.; Quici, S.; Vezza, S.; Fantucci, P. C.; Invernizzi, I.; Bruni,
S.; Ledoux-Rak, I.; Zyss, J. Organometallics 2002, 21, 161-170. (d)
Se´ne´chal, K.; Maury, O.; Le Bozec, H.; Ledoux, I.; Zyss, J. J. Am. Chem.
Soc. 2002, 124, 4561-4562. (e) Tessore, F.; Roberto, D.; Ugo, R.; Mussini,
P.; Quici, S.; Ledoux-Rak, I.; Zyss, J. Angew. Chem., Int. Ed. 2003, 42,
456-459.
2002, 1674-1675. (f) Coe, B. J.; Harris, J. A.; Brunschwig, B. S.;
Asselberghs, I.; Clays, K.; Garin, J.; Orduna, J. J. Am. Chem. Soc. 2005,
127, 13399-13410. (g) Tessore, F.; Roberto, D.; Ugo, R.; Pizzotti, M.;
Quici, S.; Cavazzini, M.; Bruni, S.; De Angelis, F. Inorg. Chem. 2005, 44,
8967-8978.
(12) Only the Kurtz power NLO activity of a europium-containing coordination
polymer was described in: Shi, J.-M.; Xu, W.; Liu, Q.-Y.; Liu, F.-L.; Huang,
Z.-L.; Lei, H.; Yu, W.-T.; Fang, Q. Chem. Commun. 2002, 756-757.
(13) Tancrez, N.; Feuvrie, C.; Ledoux, I.; Zyss, J.; Toupet, L.; Le Bozec, H.;
Maury, O. J. Am. Chem. Soc. 2005, 127, 13474-13475.
(14) A part of this study has been preliminarily published in: Se´ne´chal, K.;
Toupet, L.; Ledoux, I.; Zyss, J.; Le Bozec, H.; Maury, O. Chem. Commun.
2004, 2180-2181.
(11) (a) Kanis, D. R.; Lacroix, P. G.; Ratner, M. A.; Marks, T. J. J. Am. Chem.
Soc. 1994, 116, 10089-10102. (b) Cummings, S. D.; Cheng, L.-T.;
Eisenberg, R. Chem. Mater. 1997, 9, 440-450. (c) Vance, F. W.; Hupp, J.
T. J. Am. Chem. Soc. 1999, 121, 4047-4053. (d) Gaudry, J. B.; Capes, L.;
Langot, P.; Marcen, S.; Kollmansberger, M.; Lavastre, O.; Freysz, E.;
Letard, J.-F.; Kahn, O. Chem. Phys. Lett. 2000, 324, 321-329. (e) Fave,
C.; Hissler, M.; Se´ne´chal, K.; Ledoux, I.; Zyss, J.; Re´au, R. Chem. Commun.
9
12244 J. AM. CHEM. SOC. VOL. 128, NO. 37, 2006

Terpyridyl Lanthanide Complexes
A R T I C L E S
1
Figure 1. Comparison of the aromatic region of the H NMR spectra (CDCl3, 500 MHz, RT) of L-s (top) and L (bottom).
Scheme 2. Protonation of L To Generate the Cation LH+ and Its Isolation as the Triflate Salt
Results and Discussion
drochloride and cyclohexanone. The ketone 4 was synthesized
by a three-step procedure: (a) N-oxidation of 1, (b) Boekelheide
rearrangement, and (iii) oxidation of the resulting alcohol.
Finally, reaction between the iminium chloride 518 and 2 equiv
of 4 in basic media, followed by cyclization in the presence of
ammonia afforded a 1/1 mixture of U- and S-shape annelated
terpyridine isomers L and L-s, which could be separated by
column chromatography.19 Experimental details and character-
ization by NMR, elemental analysis, and high-resolution mass
spectrometry are described in the Experimental Section. The
isomers can be easily distinguished by their NMR spectra
(Figure 1). L features an overall C2 symmetry resulting in more
simple 1H and 13C NMR spectra than those of L-s. For instance,
the more downfield shifted signals, corresponding to quinolyl
protons H8′ and H11′ (see Figure 13 in Experimental Section
for numbering) appear as two doublets in the case of L and
four doublets in the case of L-s. Similarly, the 13C NMR
spectrum of L-s reveals 37 signals compared to 25 signals for
L. A color change from pale yellow to violet occurred
instantaneously upon protonation of L in dichloromethane with
Ligand Synthesis and Protonation Reaction. The target
ligand L (Scheme 1) is a rigidified terpyridine-like ligand whose
central pyridinic ring is functionalized by a p-dibutylamino-
phenyl moiety in order to induce an intraligand charge-transfer
transition from the amino donor to the pyridine acceptor group.
The ligand features also a dimethylene annelation between the
central pyridinic ring and the distal quinoline moieties in order
to force a cisoid conformation, stabilizing its rare-earth com-
plexes against competitive coordination of solvent.15 Generally,
annelated terpyridine can be prepared either from morpholine
enamines derived from R-pyridyl ketones and aldehydes16 or
from R-pyridyl ketones and the corresponding iminium salts.17
The ligand L was prepared in 32% overall yield by employing
the second synthetic procedure described by Rish and co-
workers (Scheme 1).17 The acridine 1 was obtained by a
Friedla¨nder condensation between o-aminoacetophenone hy-
(15) (a) Chapman, R. D.; Loda, R. T.; Riehl, J. P.; Schwartz, R. W. Inorg. Chem.
1984, 23, 1652-1657. (b) Mallet, C.; Thummel, R. P.; Hery, C. Inorg.
Chim. Acta 1993, 210, 223-231.
(16) (a) Hegde, V.; Jahng, Y.; Thummel, R. P. Tetrahedron Lett. 1987, 28,
4023-4026. (b) For a short account, see: Thummel, R. P. Synlett 1992,
1-12.
(18) Schroth, W.; Jahn, U.; Stro¨hl, D. Chem. Ber. 1994, 127, 2013-2022.
(19) Recently a regioselective synthesis leading only to the formation of the
U-shaped isomer was reported in DMSO; see: Sielemann, D.; Winter, A.;
Flo¨rke, U.; Risch, N. Org. Biomol. Chem. 2004, 2, 863-868.
(17) Keuper, R.; Risch, N.; Flo¨rke, U.; Haupt, H.-J. Liebigs Ann. 1996, 705-
715.
9
J. AM. CHEM. SOC. VOL. 128, NO. 37, 2006 12245

A R T I C L E S
Se´ne´chal-David et al.
Table 1. Selected Bond Lengths (Å) and Angles (deg)
Gd
Y
Ln-O
2.437(7)/2.446(9)
2.459(8)/2.460(9)
2.485(7)/2.503(8)
2.445(9)
2.551(8)/2.573(8)
1.516(16)
2.593/2.291
1.384(15)
1.499(14)/1.537(15)
6.829
2.369(4)/2.376(4)
2.428(4)/2.431(4)
2.454(4)/2.465(4)
2.365(9)
2.503(4)/2.513(4)
1.489(7)
2.390/2.456
1.381(7)
1.490(6)/1.493(6)
6.682
Ln-N(central)
Ln-N(distal)
Cpy-Cph
H-bond (intra)
N-Ph
C-C (interPy)
d ) CPh-CPh
(NCCN)
10.0/15.0
2.03/10.5
Θ
τ
79.2
87.9
95.0
86.1
Figure 2. Head-to-tail stacking (top) and ORTEP drawing of [YL(NO3)3‚
1/2CH3CN‚1/2H2O] (bottom), interstitial solvent molecules being removed
for clarity.
Figure 4. CPK representation of the gadolinium complex (from the crystal
structure).14
Synthesis of Lanthanide Complexes. Treatment of ligand
L with 1 equiv of Ln(NO3)3‚xH2O (x ) 6 for Ln ) La, Gd,
and x ) 5 for Dy, Yb, and Y) in dichloromethane/acetonitrile
led to brown-orange LnL(NO3)3 complexes in good yield (64-
92%) after crystallization from hot acetonitrile and further
recrystallization from a dichloromethane/pentane mixture (Scheme
3). All complexes gave satisfactory microanalysis results as
dichloromethane clathrates (see Experimental Section). In the
cases of gadolinium and yttrium, crystallization from acetonitrile
afforded orange needle monocrystals suitable for X-ray diffrac-
tion analysis (Table S1).
Figure 3.
Scheme 3. Preparation of the Dipolar Complexes
Solid-State Structure of [GdL(NO3)3‚CH3CN] and [YL-
(NO3)3‚1/2CH3CN‚1/2H2O]. Both complexes crystallize in a
centrosymmetric space group with interstitial solvent molecules.
Each dipolar complex adopts a head-to-tail configuration
(Figures 2 and S1), the intermolecular interactions are ensured
by hydrogen bonds between noncoordinated oxygen atoms of
nitrato fragments and protons belonging to the NR2-phenyl
moieties (d(O‚‚‚H) ) 2.464 and 2.632 Å). In the case of yttrium,
an additional interaction with the dimethylene bridge (d(O‚‚‚H)
) 2.654 Å) is present. ORTEP drawings of Y (Figure 2) and
Gd14 reveal that both complexes are isostructural: the metal is
nine-coordinate, being bonded to the tridentate ligand L and
three bidentate nitrato ligands. It is worth noting that no
additional solvent molecules are coordinated to the metal,
acetonitrile and water being present only as interstitial com-
pounds. This behavior is in contrast with that generally described
for other Ln3+ complexes of terdentate nitrogen ligands such
as terpyridine L2 or bis-benzimidazol-pyridine L3 (Figure 3).
1 equiv of HPF6, (Scheme 2). The resulting salt [LH]+ was
1
fully characterized by H, 31P NMR, and elemental analysis.
Further addition of HPF6 results in the complete decoloration
of the solution and neutralization with NaHCO3 reforms the
starting material L.
9
12246 J. AM. CHEM. SOC. VOL. 128, NO. 37, 2006

Terpyridyl Lanthanide Complexes
A R T I C L E S
Figure 5. DF optimized geometries of L′ (top) and L′H+ (bottom), with
important bond lengths (Å) and angles (deg). Experimental data are given
in italics.24
Figure 6. DF optimized geometries of the lanthanum (top) and lutetium
(bottom) complexes, showing important bond lengths (Å) and angles (deg).
While the coordination number is generally 10, even 11 in the
case of lanthanum, with additional coordination of water,
acetonitrile, or methanol,20,21 solvent-free nine-coordinated
complexes are normally only found with ytterbium22,20b,21b and
lutetium,23,21b which feature smaller ionic radius. This surprising
result can be explained by the larger steric hindrance induced
by L compared with L2 or L3: L presents a very rigid helicene-
like structure with seven fused six-membered rings, and the
replacement of a five-member imidazole ring in L2 by a pyridine
(L) brings the distal phenyl ring closer to the metal. As a result,
the bite angle τ (CPh-Ncentral-CPh, Figure 3) and the distances
between the two distal phenyl rings (d ) CPh-CPh) of the
coordinated ligand are significantly smaller for LLn(NO3)3 (Y
τ ) 85°, d ) 6.65 Å; Gd τ ) 87°, d ) 6.85 Å) than for L2-
Lu(NO3)3 (Lu τ ) 105°, d ) 8.23 Å).21b In consequence, to
minimize the steric repulsion between the quinolinic protons
and nitrato moieties, the metal is deeply embedded into the
ligand cavity resulting in a smaller Ln-N distance with the
central nitrogen atom than with the two distal ones (Table 1).
This behavior is in marked difference with the coordination of
L2 or L3, where the metal is always closer to the distal nitrogen
atoms.20-23 Therefore, the Ln-Ncentral distance is significantly
shorter for L than for L2 or L3. For example, the Gd-Ncentral
distance is 2.64 Å for L2Gd(NO3)3(H2O)‚L2,20e 2.536 Å for (L2-
Gd(NO3)2(H2O)3)‚NO3,20a and only 2.495 Å for LGd(NO3)3.
Thus, the absence of any coordinated solvent in both structures
is clearly due to a stronger steric hindrance induced by L,
inhibiting coordination of further molecules at the metal. The
CPK representation of the gadolinium complex clearly illustrates
the helicene-like structure of the ligand wrapped around the
metal (Figure 4).
Solution Structure of Lanthanide Complexes. Diamagnetic
(La, Y) and paramagnetic (Yb, Dy) complexes were character-
ized by 1H NMR in deuterated dichloromethane, and the signals
were assigned by means of COSY experiments. Only one set
of signals is observed even at low temperature, indicating that
(20) For L2 type ligand, see: (a) Semenova, L. I.; White, A. H. Aust. J. Chem.
1999, 52, 507-517. (b) Drew, M. G. B.; Iveson, P. B.; Hudson, M. J.;
Liljenzin, J. O.; Spjuth, L.; Cordier, P.-Y.; Enarson, A.;. Hill, C.; Madic,
C. Dalton Trans. 2000, 821-830. (c) Drew, M. G. B.; Hudson, M. J.;
Iveson, P. B.; Madic, C.; Russel, M. L. Dalton Trans. 2000, 2711-2720.
(d) Cotton, S. A.; Noy, O. E.; Liesener, F.; Raithby, P. R. Inorg. Chim.
Acta 2003, 344, 37-42. (e) Drew, M. G. B.; Hill, C.; Hudson, M. J.; Iveson,
P. B.; Madic, C.; Vaillant, L.; Youngs, T. G. A. New J. Chem. 2004, 28,
462-470.
(22) Ahrens, B.; Cotton, S. A.; Feeder, N.; Noy, O. E.; Raithby, P. R.; Teat, S.
J. Dalton Trans. 2002, 2027-2030.
(23) (a) Nozary, H.; Piguet, C.; Rivera, J.-P.; Tissot, P.; Bernardinelli, G.;
Vulliermet, N.; Weber, J.; Bu¨nzli, J.-C. G. Inorg. Chem. 2000, 39, 5286-
5298. (b) Nozary, H.; Piguet, C.; Tissot, P.; Bernardinelli, G.; Bu¨nzli, J.-
C. G.; Deschenaux, R.; Guillon, D. J. Am. Chem. Soc. 1998, 120, 12274-
12288.
(21) For L3 type ligand, see: (a) Petoud, S.; Bu¨nzli, J.-C. G.; Piguet, C. Inorg.
Chem. 1997, 36, 1345-1353. (b) Terazzi, E.; Torelli, S.; Bernardinelli,
G.; Rivera, J.-P.; Be´nech, J.-P.; Bourgogne, C.; Donnio, B.; Guillon, D.;
Imbert, D.; Bu¨nzli, J.-C. G.; Pinto, A.; Jeannerat, D.; Piguet, C. J. Am.
Chem. Soc. 2005, 127, 888-903.
9
J. AM. CHEM. SOC. VOL. 128, NO. 37, 2006 12247

A R T I C L E S
Se´ne´chal-David et al.
Table 2. Selected Distances (Å) and Angles (deg) from the Optimized Geometries Obtained by DFT
L
′
L
′
H+
L
′
La(NO3)3
L
′
Y(NO3)3
L′Lu(NO3)3
Ln-O
2.577/2.577
2.598 /2.597
2.605/2.605
2.683
2.763/2.763
1.499
2.385/2.385
2.428/2.428
2.440/2.440
2.477
2.638/2.638
1.499
2.361/2.361
2.408/2.408
2.421/2.421
2.447
2.624/2.624
1.499
Ln-N(central)
Ln-N(distal)
Cpy-Cph
1.494
1.392
1.498/1.498
23.6/23.6
63.7
1.490
1.366
1.482/1.482
14.3/14.3
65.6
N-Ph
1.397
1.397
1.397
C-C (interPy)
1.505/1.505
20.7/20.7
65.1
1.502/1.502
19.5/19.5
65.5
1.502/1.502
18.8/18.8
65.9
NCCN
Θ
Table 3. UV-Visible and Emission Data of Terpyridine
Derivatives
λmax (ꢀ),
1
1
compd
solvent
nm (L
‚
mol-
‚
cm-
)
λem (nm)
φ
(%)c
τ (ns)
L
CH2Cl2
361 (21800)-
376 (7500)a
518
26
12
toluene
THF
DMF
360b
412
555
587
608
612
629
no
12
26
6.5
0.8
5.5
2.3
358b
357b
EtOH
365b
acetone
CH3CN
CH2Cl2
365b
365b
[LH][PF6]
510 (3300),
396 (21000)
399 (22500)
[LHn]n+
CH2Cl2
416
30
1.8
Figure 7. Plot of the energy vs θ for ligand L′.
a Estimated with the deconvolution. b The ILCT transition is overlapped
with the more intense transition and no shoulder is visible hindering any
deconvolution. c Fluorescence quantum yields were measured using quinine
sulfate in H2SO4 (1 M, aq) as the standard; φ ) 0.546.
the average C2 symmetric architecture of the complexes found
in the solid state is maintained in solution. For the two
paramagnetic complexes studied, the 1H NMR signals are spread
over ca. 150 ppm and, as expected, the signals of the quinolic
protons pointing toward the lanthanide ion are the most shifted
(δ ) -95.4 ppm for Dy and +114 ppm for Yb). In conclusion,
it appears that this family of complexes can be considered as
isostructural both in the solid state and in dichloromethane
solution.
DFT Calculations. A DFT study (see the Experimental
Section for computational details) has been carried out on L′
containing a NMe2 donor group instead of NBu2, as well as on
[L′H]+ and on the three diamagnetic complexes L′Ln(NO3)3
(Ln ) Y, La, Lu). The optimized geometries (Figures 5, 6, and
S2 and Table 2) are in agreement with the X-ray diffraction
analysis data obtained for [LH]+ 24 and LY(NO3)3 (Table 1).
In the case of [L′H]+, the centrally protonated form is found to
be 7.8 kcal‚mol-1 more stable than the lateral one. The small
difference from the crystal structure can be explained by the
presence of a hydrogen bond with an ethanol molecule.24 For
the three lanthanide complexes, the Ln-Ncentral distance is
significantly shorter than the Ln-Ndistal, as observed in the solid
state. In addition, the dihedral angle θ between the dialkylami-
nophenyl and the central pyridinic ring is estimated to be around
65° versus 68°-95° in the solid state (Tables 1 and 2). However,
the calculated energy variation versus θ for L′ (Figure 7) shows
a very flat minimum allowing free oscillations of θ of about
20° for a small energetic cost (<1 kcal‚mol-1); hence the
observed differences are not really significant and can simply
result from the crystal packing.
spectra of L in CH2Cl2 are shown in Figures 8 and S3, and the
data are collected in Table 3. The UV-visible absorption
spectrum exhibits a broad band at 361 nm overlapped with a
sharp band at 376 nm (see deconvolution). The emission
spectrum exhibits a broad structureless band at 518 nm in
dichloromethane with a moderate fluorescence quantum yield
(0.26) and a relatively long fluorescence lifetime (12 ns). The
assignment of the lowest energy absorption transition to an
intramolecular charge transfer (ICT) transition was confirmed
by a protonation experiment (Figure 8): addition of 1 equiv of
acid causes a large red shift (∆λ ) 134 nm) of this band in
accord with the formation of the strongly electron-withdrawing
pyridinium fragment (Scheme 2). The monoprotonated form is
not significantly emissive. A further addition of acid leads to
the complete disappearance of the red color and a large shift of
the absorption band to the blue, a behavior consistent with the
subsequent protonation of the dibutylamino donor group under
more acidic conditions, which leads to the inhibition of charge
transfer (“pull-pull” compound).25 This interpretation is cor-
roborated by a profound blue shift of the emission from 518 to
416 nm (τ ) 1.8 ns) upon addition of an excess of acid. No
significant variation of the absorption spectrum was observed
with increasing solvent polarity (from toluene to dimethylform-
amide), in agreement with a weak ground-state dipole moment
(µg) due to the distorted structure of the conjugated backbone
(θ ) 65°). By contrast, the emission band shows a very strong
positive solvatochromic shift (Table 3). The red shift in emission
with increasing solvent polarity is accompanied by an increase
Photophysical and Nonlinear Optical Data. A. Optical
Properties of L. The room-temperature absorption and emission
(25) For similar experiments on CT chromophores, see: (a) Maury, O.; Gue´gan,
J.-P.; Renouard, T.; Hilton, A.; Dupau, P.; Sandon, N.; Toupet, L.; Le
Bozec, H. New J. Chem. 2001, 1553-1566. (b) Goodall, W.; Williams, J.
A. G. Chem. Commun. 2001, 2514-2515.
(24) Se´ne´chal-David, K.; Toupet, L.; Maury, O.; Le Bozec, H. Cryst. Growth
Des. 2006, 6, 1493-1496.
9
12248 J. AM. CHEM. SOC. VOL. 128, NO. 37, 2006

Terpyridyl Lanthanide Complexes
A R T I C L E S
Figure 8. Absorption spectra of (i) L (black bold), deconvolution (inset), and simulation (black circle); (ii) the monoprotonated species LH+ (red and
simulation in red circle); and (iii) the polyprotonated species (blue).
Figure 9. Molecular orbital energy diagram for L′ (B3LYP/DZP).
of the Stokes shift and the bandwidth. This behavior is typical
of charge-transfer character in the fluorescent excited state,
featuring a large dipole moment (µe), and is quite similar to
that described for nonannelated terpyridyl ligands.26 Density
9
J. AM. CHEM. SOC. VOL. 128, NO. 37, 2006 12249

A R T I C L E S
Se´ne´chal-David et al.
Figure 10. UV-visible absorption spectra of the ligand (blue), protonated species (violet), and Ln complexes (red).
Table 4. f > 0.01 Lowest ExcitationssDFT LB94/DZP Results
functional theoretical (TDDFT) calculations were performed in
order to understand these spectroscopic data. The molecular
orbital energy diagrams of L′ and L′H+ are shown in Figures
9 and S4, respectively. The corresponding electronic transitions
are listed in Table 4, and their energies are superimposed on
the spectrum in Figure 8 (vertical lines). The lowest energy
transition (exptl 376 nm, calcd 510 nm) involves almost
exclusively (99%) the HOMO, localized on the donor part of
the ligand, and the LUMO, delocalized on the three pyridinic
rings (Figure 9). Thus, this transition can be classically described
as an intramolecular charge transfer (ICT) transition. The second
optical transition centered at 361 nm can be assigned to a
“locally excited” state transition because it results from the
overlap of four electronic excitations, each of them being a
mixture of various HOMO(0 to -6) - LUMO(0 to +3)
transitions (Table 4) featuring a main pyridinic π-π* character
[(HOMO - 3, LUMO) or (HOMO - 6, LUMO)] with some
minor ICT character (HOMO, LUMO + 2).
The photoinduced modifications of the electronic density
distribution result in an important increase of the dipole moment
in the Franck-Condon excited state. The effect of the (HOMO,
LUMO) excitation on the geometries and the charge distributions
has been studied using the CIS approach (configuration interac-
tion with singles),27 which starts from a Hartree-Fock (HF)
description of the molecules. Although not very accurate
regarding excitation energies, this method allows the determi-
nation of approximate wave functions and relaxed geometries
of excited states. When the ground and the CIS excited states
are compared, the following trends have been observed: (i) the
magnitude of the dipole moment in the Franck-Condon excited
state is roughly three times higher than that in the ground state;
for instance, in the case of L′, it rises from 6.7 D (HF) to 18.1
D (CIS); (ii) the magnitude of this property diminishes in the
geometry relaxed excited state; in the case of L′, its value is
energy
λ
(nm)
oscillator
compd
transitions
(
∆E (eV))
strength (au)
composition
%
L′
1
2
510 (2.43)
408 (3.04)
0.041
0.061
HOfLU
HO-3fLU
HO-5fLU+1 26
99
61
3
4
391 (3.17)
388 (3.20)
0.027
0.119
HOfLU+2
HO-6fLU
HO-6fLU
HOfLU+2
HO-3fLU
HOfLU+3
69
21
46
30
12
74
5
376 (3.30)
0.101
HO-3fLU+1 10
HO-6fLU+1
HOfLU
9
98
83
8
93
69
26
72
22
99
81
6
L′H+
1
2
790 (1.57)
464 (2.67)
0.229
0.234
HO-3fLU
HOfLU+3
HOfLU+2
HOfLU+3
HO-5fLU
HO-5fLU
HOfLU+3
HOfLU
HO-3fLU
HO-2fLU+1
HOfLU+5
HO-1fLU+1 64
HO-3fLU+1 13
HO-2fLU
HOfLU+5
HO-6fLU
HOfLU+6
HOfLU
HO-3fLU
HO-2fLU+1
HO-7fLU+1
HOfLU+5
HOfLU+6:
3
4
461 (2.69)
458 (2.71)
0.062
0.018
5
456 (2.72)
0.022
L′Y(NO3)3
1
2
667 (1.86)
440 (2.82)
0.069
0.168
3
3
4
431 (2.88)
429 (2.89)
0.015
0.013
12
89
4
96
99
82
7
3
92
96
5
1
2
425 (2.92)
667 (1.86)
441 (2.81)
0.036
0.068
0.169
L′Lu(NO3)3
3
4
429 (2.89)
425 (2.92)
0.012
0.034
equal to 11.6 D, which remains much higher than the ground-
state one; (iii) the θ angle in the geometry relaxed excited state
is lower than in the ground state, thus indicating a more planar
structure allowing a better π delocalization between the two
parts of the molecules; for L′, this angle goes from ∼75° (HF
ground state) to ∼50° (CIS excited state). Such relaxed ICT
(26) (a) Goodall, W.; Wild, K.; Arm, K. J.; Williams, J. A. G. J. Chem. Soc.,
Perkin Trans. 2 2002, 1660-1681. (b) Mutai, T.; Cheon, J.-D.; Tsuchiya,
G.; Araki, K. J. Chem. Soc., Perkin Trans. 2 2002, 862-865.
(27) Foresman, J. B.; Head-Gordon, M.; Pople, J. A.; Frisch, M. J. J. Phys.
Chem. 1992, 96, 137.
9
12250 J. AM. CHEM. SOC. VOL. 128, NO. 37, 2006

Terpyridyl Lanthanide Complexes
A R T I C L E S
Table 5. Linear and Nonlinear Optical Data for the Ligand and the Related Complexes
λILCT
c
d
(
λπ-π
)
ꢀ
ν1/
â1.91
â0
*
2(ILCT)
1
[nm]a
[L‚
mol-
‚
cm-
]
[cm-
]
fb
[10-30 esu]
[10-30 esu]
1
1
L
376 (361)
510 (396)
454 (392)
453 (391)
452 (393)
450 (394)
452 (394)
7500 (21800)
7400 (22400)
4750 (28900)
5400 (24500)
5800 (26900)
5400 (26700)
5150 (28850)
4294
3373
5363
6055
4550
5825
5465
0.14
0.12
0.11
0.14
0.12
0.14
0.12
190
240
191
186
246
260
288
155
159
139
136
180
190
211
[LH][PF6]
LY(NO3)3
LLa(NO3)3
LGd(NO3)3
LDy(NO3)3
LYb(NO3)3
a Measured in dilute dichloromethane solution (1 × 10-5 mol‚L-1). b The oscillator strength is deduced from the UV-visible spectra according to f )
(4.319 × 10-9)ꢀν1/2
.
c Measured by HLS using λ ) 1.91 µm as fundamental wavelength in concentrated dichloromethane solution (1 × 10-2 mol‚L-1).
d Deduced from the two-level model.
Figure 11. Molecular orbital energy diagram for L′Lu(NO3)3 (B3LYP/DZP).
state accompanied by a conformational change is generally
referred to as twisted intramolecular charge transfer (TICT).28
TICT states have already been described in the case of
4-dialkylaminophenylacridines29 and are strongly sensitive to
the solvent polarity and responsible for the solvent dependence
(28) For a review on TICT, see: Grabowski, Z. R.; Rotkiewicz, K.; Rettig, W.
Chem. ReV. 2003, 103, 3899-4031.
(29) Herbich, J.; Kapturkiewicz, A. J. Am. Chem. Soc. 1998, 120, 1014-1029.
9
J. AM. CHEM. SOC. VOL. 128, NO. 37, 2006 12251

A R T I C L E S
Se´ne´chal-David et al.
of the fluorescence quantum yield and of the Stokes shift. A
more polar solvent contributes to a higher stabilization of the
excited-state resulting in a red shift of the emission band and a
decrease of the quantum yield.
B. Optical Properties and DFT Calculations of Lanthanide
Complexes. All lanthanide ions behave similarly in terms of
absorption spectra (Figure 10): the TICT transition is batho-
chromically shifted by ca. 75 nm upon complexation, the Lewis
acidity of Ln3+ enhancing the acceptor strength of the terpyridyl
moieties. It is also important to note that the oscillator strength
of the TICT band (Table 5) is the same for all complexes (0.11-
0.14) and similar to that of the ligand (0.14), the lower extinction
coefficient ꢀ of the complexes vs that of L being balanced by
the broadening of the band. Interestingly, theoretical calculations
performed in the cases of lutetium and yttrium unambiguously
confirm (i) the CT character of the lowest energy transition with
a 99% contribution of the HOMO-LUMO transition (Table 4
and Figure 11), (ii) the strong bathochromic shift induced by
complexation, and (iii) identical behavior for both metals. This
similar “ionochromic” effect has already been described previ-
ously on bipyridyl metal lanthanide complexes30 and thus seems
to be a general tendency of f-block elements.13
C. NLO Properties. The molecular second-order NLO
properties were measured in solution by means of a harmonic
light scattering technique (HLS) with an incident laser beam at
1.91 µm to avoid any two-photon induced fluorescence contri-
bution to the second harmonic, and the results are summarized
in Table 5.31 The terpyridyl ligand L has a remarkably high
and unexpected static â0 value (â0(L) ) 155 × 10-30 esu),
comparable to that of the parent nonannelated 4′-dibutylami-
nophenyl-2,2′:6′,2′′-terpyridine (âH0 LS ) 147 × 10-30 esu).32,33
This high NLO activity is much larger than those of other push-
pull polypyridine derivatives10a,10c featuring classical planar
π-conjugated systems and can be related to the twisted
π-electron system as recently reported by Marks and Ratner.34
Protonation of L does not lead to an increase of the static â0
value (â0(LH+) ) 159 × 10-30 esu), the increase of the dynamic
â value being compensated by the dispersion factor due to the
large red shift of the maximum absorption wavelength (Table
5). Concerning the lanthanide complexes, it is worth noting that
the NLO activity significantly depends on the nature of the
lanthanide. Whereas the lanthanum complex has a â0 value (â0-
(La) ) 136 × 10-30 esu) similar to that of the free ligand (and
to that of the protonated ligand), the other lanthanide (Gd, Dy,
Yb) complexes display a regular increase of their NLO activities
along the f-block element series. Keeping in mind that all of
these complexes are isostructural and show the same UV-
visible absorption spectrum, these NLO enhancements cannot
be explained on the simple basis of the two-level model35 and
strongly suggest the participation of the metallic center. As
Figure 12. Plot of the hyperpolarizability coefficient [ vs ionic radius
(top) and f orbital filling (bottom).
series,13 the NLO study of the corresponding yttrium complex
is particularly useful to rationalize this “metal-induced” NLO
enhancement effect. Yttrium presents strong chemical similari-
ties to lanthanides, and its ionic radius (1.075 Å) is very close
to that of dysprosium (1.083 Å).36 On the other hand, yttrium
is not an f-block element and therefore has a 4f0 electronic
configuration like lanthanum. Experimentally, the hyperpolar-
izability coefficient of yttrium, â(Y) ) 191 × 10-30 esu, is very
similar to that of lanthanum (186 × 10-30 esu) and significantly
lower than that of dysprosium (260 × 10-30 esu): the difference
between â(Y) and â(Dy) is much larger than the experimental
error estimated to (15% (HLS at 1.91 µm). Once again, it
appears that the NLO activity variation along the lanthanide
series can be better correlated to the f configuration rather to
the ionic radius (Figure 12). This new effect has now been
established on the basis of HLS for two isostructural lanthanide
series, a dipolar one, [LLn(NO3)3], and an octupolar one, Ln-
(DPA)3,3 having completely different structures. Moreover, the
measurements were performed with two different fundamental
laser wavelengths (1.91 and 1.06 µm) in different solvents (CH2-
Cl2 and H2O). Therefore, one can conclude from these two
experiments that the direct contribution of f electrons to the
hyperpolarizability seems to be a general property of f-block
complexes.
3-
recently reported in the case of the octupolar Ln(DPA)3
(30) Wang, B.; Wasielewski, M. R. J. Am. Chem. Soc. 1997, 119, 12-21.
(31) The detailed description of the experimental setup is described in: Le Bozec,
H.; Le Bouder, T.; Maury, O.; Bondon, A.; Ledoux, I.; Deveau, S.; Zyss,
J. AdV. Mater. 2001, 13, 1677-1681.
(32) Tancrez, N. Ph.D. Thesis, ENS-Cachan, France, 2005.
(33) The âE0FISH of this molecule was estimated to be 15 × 10-30 esu in ref 11g
and in: Roberto, D.; Tessore, F.; Ugo, R.; Bruni, S.; Manfredi, A.; Quici,
S. Chem. Commun. 2002, 846-847.
(34) (a) Kang, H.; Facchetti, A.; Zhu, P.; Jiang, H.; Yang, Y.; Cariati, E.;
Righetto, S.; Ugo, R.; Zuccaccia, C.; Macchioni, A.; Stern, C. L.; Liu, Z.;
Ho, S.-T.; Marks, T. J. Angew. Chem., Int. Ed. 2005, 44, 7922-7925. (b)
Keinana, S.; Zojerb, E.; Bredas, J.-L.; Ratner, M. A.; Marks, T. J. J. Mol.
Struct. (THEOCHEM) 2003, 633, 227-235.
(35) Oudar, J. L. J. Chem. Phys. 1977, 67, 446.
(36) Shannon, R. D. Acta Crystallogr., Sect. A 1976, 32, 751-767.
9
12252 J. AM. CHEM. SOC. VOL. 128, NO. 37, 2006

Terpyridyl Lanthanide Complexes
A R T I C L E S
Conclusion
In summary, a series of isostructural dipolar lanthanide(III)
complexes featuring a functionalized terpyridine ligand have
been synthesized, and the linear optical and nonlinear optical
properties of the complexes and the ligand have been investi-
gated. This study shows that coordination chemistry of lan-
thanide ions is a powerful tool for the design of efficient NLO-
phores and that the molecular quadratic hyperpolarizability (â)
values are strongly influenced by the nature of the lanthanide
ion. A careful examination of the NLO measurements allows
some conclusions to be drawn: (i) Concerning the ligand, the
unexpected large â response could be attributed to the large
dihedral angle between the donor and acceptor moieties leading
to a twisted intramolecular charge transfer. (ii) Concerning the
metal ions, the most striking result is the clear dependence of
the first hyperpolarizability â on the lanthanide itself. A strong
increase of â is observed on going from lanthanum to ytterbium,
whereas for yttrium, which has no f electron like lanthanum,
there is no evidence for significant increase of â. We propose
that, as in the case of the octupolar series tris(dipicolinato)-
lanthanates,13 this enhancement can be better correlated to the
f configuration rather than to the ionic radius. This intrinsic
property of rare-earth metals is also particularly attractive in
terms of transparency/nonlinearity tradeoff because improved
â values are reached without any cost of transparency, and late
lanthanide ions such as Er3+, Tm3+, Yb3+, and Lu3+ appear to
be very promising for the design of highly efficient NLO-phores
because they are among the strongest Lewis acids of the periodic
table and they feature a strong f electron density. This study
opens the route for the design of multifunctional molecular
materials combining lanthanide luminescence properties and
NLO activity, which could be particularly interesting in the field
of biological imaging.
Figure 13. Atom numbering for L and L-s.
employed. Note that the TDDFT excitation energy calculations (Table
4) have been carried out using the LB94 functional, whereas the
displayed MO diagrams (Figure 13 and S3) are those of B3LYP. We
checked that both functionals led qualitatively to the same energy
ordering and compositions of the MOs under consideration. Molecular
orbital plots were generated using the program MOLEKEL 4.3.43
Crystal Structure Analysis. The sample was studied on a NONIUS
Kappa CCD with graphite monochromatized Mo KR radiation. The
cell parameters are obtained with Denzo and Scalepack with 10 frames
(psi rotation 1° per frame). The structure was solved with SIR-97,44
which reveals the non-hydrogen atoms of structure. After anisotropic
refinement, many hydrogen atoms may be found with a Fourier
difference. The whole structure was refined with SHELXL9745 by the
full-matrix least-squares techniques (use of F magnitude; x, y, z, Rij
for C, N, and O atoms, x, y, z in riding mode for H atoms). ORTEP
views were made with PLATON 98.46 All the calculations were
performed on a Silicon Graphics Indy computer. Crystallographic data
for the structural analysis have been deposited with the Cambridge
Crystallographic Data Centre, CCDC 229762 (Gd) and 222671 (Y).
HLS Measurements. Details concerning the experimental setup are
described in ref 31.
Linear Absorption and Fluorescence Measurements. UV-visible
spectra were recorded on a KONTRON UVIKON 941 spectrophotom-
eter in diluted dichloromethane solution (ca. 10-5 mol‚L-1). Steady-
state emission spectra were recorded in solution in 1 cm path length
quartz cuvettes, using an Instruments S.A. Fluoromax equipped with a
Hamamatsu R928 photomultiplier tube. Fluorescence quantum yields
were measured by the method of continuous dilution using quinine
sulfate as the standard (φ ) 0.546 in 1 M aqueous H2SO4,47 correcting
for the refractive index). The fluorescence lifetimes were measured
using an Instruments S.A. Fluorolog Tau-3 instrument, by global fitting
of the demodulation and phase shift of the emission, following excitation
with sinusoidally modulated light over the frequency range 10-250
MHz. A suspension of Ludox in water was used as the standard, which
acts a scattering sample of τ 0.0 ns.
Experimental Section
Computational Details. Our DFT geometry optimizations have been
done employing the well-known functional B3LYP,37 using a standard
double-ú polarized basis set, namely, the LANL2DZ supplemented by
a single polarization function on heavy atoms. Another polarized
double-ú basis set, CEP-31G*, where core electrons of the heavier
elements are also described by means of an effective core potential,
has been used to treat the complexes, because LANL2DZ is not
available for the metals under consideration. The CIS and DFT
calculations were carried out using the Gaussian 03 package.38 TDDFT39
excitation energies calculations were performed using the Amsterdam
Density Functional program (ADF2004.01)40 and the PBE41 as well as
the LB94 functional.42 The standard DZP basis set of ADF has been
used for all atoms excluding hydrogen for which the DZ one has been
General Procedures. All reactions were routinely performed under
argon using Schlenk techniques. NMR spectra (1H, 13C, 31P) were
recorded at room temperature on BRUKER DPX 200, AC 300, or
DMX500 spectrometers operating at 200.12, 300.13, and 500.13 MHz
1
for H, respectively. NMR data are listed in parts per million (ppm)
(37) Stephens, P. J.; Devlin, F. J.; Chabalowski, C. F.; Frisch, M. J. J. Phys.
Chem. 1994, 98, 11623-11627.
and are reported relative to tetramethylsilane (1H, 13C), residual solvent
peaks being used as internal standard (CD2Cl2 1H 5.25 ppm; 13C 53.45
ppm). Complete attribution of the 1H and 13C spectra required 2D
experiments (COSY, NOESY, H-C correlation (hmqc and hmbc
(38) Frisch, M. J., et al. Gaussian 03, revision B.04, Gaussian, Inc.: Pittsburgh,
PA, 2003.
(39) Van Gisbergen, S. J. A.; Snijders, J. G.; Baerends, E. J.; Comput. Phys.
Commun. 1999, 118, 119.
(40) (a) Te Velde, G.; Bickelhaupt, F. M.; Van Gisbergen, S. J. A.; Fonseca
Guerra, C.; Baerends, E. J.; Snijders, J. G.; Ziegler, T. J. Comput. Chem.
2001, 22, 931-967. (b) Fonseca Guerra, C.; Snijders, J. G.; Te Velde, G.;
Baerends, E. J. Theor. Chem. Acc. 1998, 99, 391. (c) ADF2004.01, SCM,
Theoretical Chemistry, Vrije Universiteit, Amsterdam, The Netherlands,
http://www.scm.com.
(43) Flu¨kiger, P.; Lu¨thi, H. P.; Portmann, S.; Weber, J. MOLEKEL 4.3; Swiss
Center for Scientific Computing: Manno, Switzerland, 2000; http://
www.cscs.ch.
(44) Altomare, A.; Burla, M. C.; Camalli, M.; Carscarano, G.; Giacovazzo, C.;
Guagliardi, A.; Moliterni, A. G. G.; Polidori, G.; Spagna, R. J. Appl.
Crystallogr. 1998, 31, 74-77.
(41) (a) Ernzerhof, M.; Scuseria, G. E. J. Chem. Phys. 1999, 110, 5029-5036.
(b) Adamo, C.; Barone, V. J. Chem. Phys. 1999, 110, 6158-6170. (c)
Adamo, C.; Cossi, M.; Scalmani, G.; Barone, V. Chem. Phys. Lett. 1999,
307, 265-271. (d) Adamo, C.; Scuseria, G. E.; Barone, V. J. Chem. Phys.
1999, 111, 2889-2899.
(45) Sheldrick, G. M. SHELX93. Program for the Refinement of Crystal
Structures; University of Go¨ttingen: Germany, 1993.
(46) Spek, A. L. PLATON, A Multipurpose Crystallographic Tool; Utrecht
University: Utrecht, The Netherlands, 1998.
(42) Van Leeuwen, R.; Baerends, E. J. Phys. ReV. A 1994, 49, 2421-2431.
(47) Meech, S. R.; Phillips, D. J. Photochemistry 1983, 23, 193.
9
J. AM. CHEM. SOC. VOL. 128, NO. 37, 2006 12253

A R T I C L E S
Se´ne´chal-David et al.
sequences)). High-resolution mass spectrometry measurements (FAB)
were performed at the Centre Regional de Mesures Physiques de l’Ouest
(Rennes, France) and elemental analysis by the Service Central
d’Analyse du CNRS (Solaize, France). N,N-Dibutylaminobenzaldehyde
was classically prepared by a Vilsmeier-Haack formylation. THF and
diethyl ether were distilled over Na/benzophenone, DMF was distilled
prior to use, and CH2Cl2 was distilled over CaH2.
9-Methyl-2,3-dihydro-1H-acridin-4-one (4). To a dichloromethane
solution (300 mL) of 5 (9.4 g, 44 mmol) was added MnO2 (23 g, 264
mmol) at room temperature, and the heterogeneous solution was allowed
to stir for 2 days. After filtration over Celite, the solvent was evaporated.
The crude dark solid was purified by column chromatography (neutral
alumina, dichloromethane as eluant). After evaporation of the solvent,
the desired product was recovered as a brownish solid (5.41 g, 58%).
1H NMR (200.131 MHz, CDCl3) δ(ppm): 8.31 (dd, 3J ) 8.1 Hz and 4J
) 0.8 Hz, 1H), 7.94 (dd, J ) 8.0 Hz and J ) 1.4 Hz, 1H), 7.67-
7.51 (m, 2H), 3.08 (t, 3J ) 6.1 Hz, 2H), 2.82 (t, 3J ) 6.4 Hz, 2H), 2.60
(s, 3H), 2.22 (m, 2H). 13C NMR (50.332 MHz, CDCl3) δ(ppm): 198.2,
148.4, 146.9, 143.6, 134.1, 129.4, 132.4, 129.6, 128.9, 123.8, 40.2,
27.2, 22.4, 14.5. HRMS (EI) m/z calcd for [M]+ (found): 211.0997
(211.0989). Anal. Calcd for C14H13NO (found): C 79.59 (79.72); H
6.20 (6.28); N 6.63 (6.10).
9-Methyl-1,2,3,4-tetrahydroacridine (1). In a round-bottom flask,
cyclohexanone (6.1 mL, 58 mmol) was heated at 90 °C, and 2-ami-
noacetophenone hydrochloride (10 g; 58 mmol) was added by small
fractions. The bottom flask was then equipped with a condenser, and
the crude heterogeneous mixture was further heated overnight at 110
°C. After cooling to room temperature, the red-orange solid was
dissolved in ethanol/HCl (12 N) [95/5 v/v]. The solution was then
neutralized with an aqueous NaOH solution. Ethanol was evaporated,
and the product extracted with diethyl ether (2 × 100 mL). The
combined organic layers were washed with water (2 × 100 mL), dried
over magnesium sulfate and filtered, and the solvent was removed under
3
4
4-(4-Dibutylamino-benzylidene)-morpholin-4-ium Chloride (5).
The product was prepared according to ref 18. In a Schlenk flask, N,N-
dibutylaminobenzaldehyde (5 g, 21 mmol) was dissolved in diethyl
ether (5 mL), and this solution was slowly added to a diethyl ether (5
mL) solution of trimethylsilyl chloride (2.97 g, 28 mmol) and silylated
morpholine (4.44 g, 28 mmol). The solution was further stirred under
reflux for 72 h, and a yellow precipitate was formed. After cooling to
room temperature, the precipitate was filtered off under argon, and the
solid was thoroughly washed with diethyl ether (3 × 25 mL). After
removal of residual solvent under reduced pressure, the desired product
was recovered as a yellow solid, very sensitive to air and moisture
reduced pressure. The desired product was finally obtained as a brown-
1
yellow solid (10.2 g, 89%). H NMR (200.131 MHz; CDCl3) δ(ppm)
:
7.94 (dd, 3J ) 8.3 Hz and 4J ) 1.1 Hz, 1H), 7.87 (dd, 3J ) 8.3 Hz and
4J ) 1.3 Hz, 1H), 7.55 (ddd, J ) 8.3 Hz, J ) 8.3 Hz and J ) 1.3
3
3
4
3
3
4
Hz, 1H), 7.38 (ddd, J ) 8.3 Hz, J ) 8.3 Hz and J ) 1.1 Hz, 1H),
3.07 (t br, 3J ) 6.7 Hz, 2H), 2.79 (t br, 3J ) 6.1 Hz, 2H), 2.43 (s, 3H),
1.86 (m, 2 × 2H). 13C NMR (50.332 MHz, CDCl3) δ(ppm): 157.8, 145.5,
140.6, 128.0, 126.4, 128.6, 127.6, 124.8, 122.9, 34.2, 26.5, 22.8, 22.4,
12.9. HRMS (EI) m/z: calcd for [M]+ (found): 197.1204 (197.1198).
Anal. Calcd for C14H15N (found): C 85.24 (85.25); H 7.66 (7.72); N
7.10 (6.78).
1
(5.08 g, 70%). H NMR (200.131 MHz, CDCl3) δ(ppm): 8.42 (s, 1H),
3
3
7.79 (d, J ) 9.3 Hz, 2H), 6.89 (d, J ) 9,3 Hz, 2H), 3.88 (m, 4H),
3
3.49 (t, J ) 7.6 Hz, 4H), 3.22 (m, 4H), 1.58 (m, 4H), 1.35 (m, 4H),
3
0.96 (t, J ) 7.2 Hz, 6H).
N-Oxide-9-methyl-1,2,3,4-tetrahydroacridine (2). A dichloromethane
solution (300 mL) of 3-chloroperbenzoic acid (26 g, 105 mmol) was
slowly added to a dichloromethane solution (100 mL) of 1 (10.2 g, 52
mmol) at 0 °C. The mixture was stirred for 4 h at room temperature
and quenched with an aqueous NaOH solution. The organic layers were
further washed with water (5 × 100 mL) and dried over MgSO4, and
the solvent was removed under reduced pressure giving 2 as a brownish
Ligand L. In a Schlenk flask equipped with a condenser, 9-methyl-
2,3-dihydro-1H-acridin-4-one (0.91 g, 4.3 mmol), 4-(4-dibutylamino-
benzylidene)-morpholin-4-ium chloride 5 (0.73 g, 2.15 mmol), and
freshly sublimed ammonium acetate (1.62 g, 21 mmol) were solubilized
in chloroform (40 mL). The brownish-yellow mixture was refluxed
for 18 h. After cooling to room temperature, the solution was hydrolyzed
with water (20 mL) and aqueous ammonia (25%, 1 mL) and further
stirred for 4 h. The crude was extracted with dichloromethane (2 × 50
mL), the organic layer was washed with water, dried under MgSO4,
and filtered, and the solvents were removed under vacuum. The two
S- and U- isomers were separated by column chromatography {alumina
activity II, gradient elution from dichloromethane to dichloromethane/
acetone v/v 95/5 (Rf ) 0.1)}. The desired product (U-isomer) was
obtained as a yellow microcrystalline powder (0.69 g, 52%). 1H NMR
1
solid. (10.83 g, 98%). H NMR (200.131 MHz, CDCl3) δ(ppm): 8.77
3
4
3
4
(dd, J ) 8.5 Hz and J ) 1.2 Hz, 1H), 7.97 (dd, J ) 8.5 Hz and J
) 0.9 Hz, 1H), 7.70-7.50 (m, 2×1H), 3.19 (t, 3J ) 6.1 Hz, 2H), 2.85
(t, 3J ) 6.2 Hz, 2H), 2.51 (s, 3H), 1.88 (m, 2 × 2H). 13C NMR (50.332
MHz, CDCl3) δ(ppm): 146.7, 139.1, 131.6, 129.9, 127.7, 129.0, 127.3,
123.9, 119.6, 27.1, 26.6, 22.0, 21.4, 13.4. HRMS (EI) calcd for [M]+
(found): 213.1154 (213.1159).
3
3
(CD2Cl2) δ: 8.4 (d, J ) 8.1 Hz, 2H, H8′), 8.1 (d, J ) 8.1 Hz, 2H,
9-Methyl-1,2,3,4-tetrahydroacridin-4-ol (3). In a two-neck round-
bottom flask equipped with a reflux condenser, N-oxide-9-methyl-
1,2,3,4-tetrahydroacridine (11,2 g, 52 mmol) was dissolved in dichlo-
romethane (250 mL). Trifluoroacetic anhydride (17 mL, 120 mmol)
was slowly added at room temperature (the reaction is exothermic).
The solution was stirred for 5 h, and the solvent was evaporated. The
crude solid was dissolved in methanol (50 mL) and saponified by an
aqueous K2CO3 solution (2 M, 150 mL); a brown solid precipitated.
The methanol was removed under reduced pressure, and the product
was extracted with dichloromethane (2 × 150 mL). The combined
organic layers were washed with brine (2 × 50 mL), dried over
magnesium sulfate, and evaporated to dryness. The desired product
was recovered as a brown solid (9.4 g, 84%). Note that this compound
underwent spontaneous oxidation into the coresponding ketone. 1H
H11′), 7.7 (ddd, 2H, H9′), 7.6 (ddd, 2H, H10′), 7.1 (d, J ) 8.5 Hz, 2H,
3
H8), 6.7 (d, 3J ) 8.5 Hz, 2H, H9), 3.3 (t, 3J ) 7.7 Hz, 4H, H11), 3.1 (m,
4H, H16), 2.9 (m, 4H, H15), 2.7 (s, 6H, H7′), 1.6 (m, 4H, H12), 1.4 (m,
4H, H13), 0.9 (t, J ) 7.2 Hz, 6H, H14). 13C NMR (CD2Cl2) δ: 153.0
3
(C2′), 151.0 (C2,C6), 148.7 (C4), 148.2 (C10), 147.1 (C6′), 141.0 (C4′),
135.33 (C3), 130.3 (C3,C5 masked), 130.6 (C8′), 130.4 (C8), 129.1 (C9′),
128.5 (C5′), 126.9 (C10′),124.1 (C11′), 122.99 (C7), 111.5 (C9), 51.13
(C11), 29.8 (C12), 26.0 and 25.8 (C,15C16), 20.8 (C13), 14.2 and 14.3
(C7′,C14). UV-visible (CH2Cl2): λmax ) 375 nm (overlap), λ ) 361
nm (21700). HRMS [M+] calcd for C43H44N4 (found): 616.3566
(616.3576). Elemental analysis calcd for C43H44N4‚2H2O (found): wt
% C 79.00 (79.11); wt % H 7.52 (7.41); wt % N 8.21 (8.58).
1
3
3
L-s. H NMR (CDCl3) δ: 8.4 (d, J ) 7.7 Hz, 1H, H8′), 8.2 (d, J
3
) 7.6 Hz, 1H, H8′′), 8.0 (m, 2H, H11′ H11′′), 7.6 (d, J ) 8.7 Hz, 2H,
3
NMR (200.131 MHz, CDCl3) δ(ppm): 7.96 (d, J ) 8.3 Hz, 1H), 7.91
(d, 3J ) 8.4 Hz, 1H), 7.58 (dd, 3J ) 8.3 Hz and 3J ) 8.1 Hz, 1H), 7.45
(dd, 3J ) 8.1 Hz and 3J ) 8.4 Hz, 1H), 4.95 (s br, 1H), 4.76 (dd, 3J )
10.3 Hz and 3J ) 10.0 Hz, 1H), 2.89 (m, 2H), 2.54 (s, 3H), 2.40-1.92
H8), 7.7-7.5 (m, 4H, H9′ H10′ H9′′ H10′′), 6.8 (d, 3J ) 8.7 Hz, 2H, H9),
3
3
3.8 (tr, J ) 6.7, 2H, H15′), 3.4 (t, J ) 7.5 Hz, 4H, H11), 3.1 (m, 4H,
H16 H16′), 2.9 (m, 2H, H15), 2.7 and 2.6 (s, 6H, H7′ H7′′), 1.7 (m, 4H,
H12), 1.4 (m, 4H, H13), 1.0 (t, 3J ) 7.2 Hz, 6H, H14). 13C NMR (CDCl3)
δ: 156.54, 153.07, 152.63, 151.16, 148.15, 147.10, 145.91, 141.11,
139.38, 139.04, 132.92, 131.68, 131.20, 131.02, 130.91, 130.34, 130.03,
128.35, 128.05, 128.00, 127.44, 126.94, 126.74, 126.19, 123.73, 123.37,
(2×m, 2×1H), 1.82 (m, 2H). 13C NMR (50.332 MHz, CDCl3) δ(ppm)
:
159.2, 145.3, 142.0, 127.7, 127.3, 129.2, 128.5, 126.0, 123.5, 70.2,
30.3, 26.7, 19.6, 13.8. HRMS (EI) calcd for [M]+ (found): 213.1153
(213.1154).
9
12254 J. AM. CHEM. SOC. VOL. 128, NO. 37, 2006

Terpyridyl Lanthanide Complexes
A R T I C L E S
111.03, 50.86, 29.50, 27.15, 25.97 (two signals overlapped), 25.31,
20.42, 14.08 (two signals overlapped), 13.97. HRMS [M + H]+ calcd
for C43H45N4 (found): 617.3644 (617.3650).
O9‚CH2Cl2 (found): wt % C 51.47 (50.18); wt % H 4.52 (4.67); wt %
N: 9.55 (9.71).
For LGd(NO3)3, no NMR measurements could be performed due to
the strong paramagnetism of the metal: λmax ) 452 nm (5400), λ )
393 nm (26900). Elemental analysis calcd for GdC43H44N7O9‚CH2Cl2
(found): wt % C 50.57 (51.38); wt % H 4.44 (4.86); wt % N 9.38
(9.38).
LDy(NO3)3 1H NMR (CD2Cl2) δ: 0.8, -64.11, -76.5, -95.4 (s, 4
× 2H, H8′,9′,10′,11′), -41.8 (s, 6H, H7′), 33.6, -1.7 (s, 2 × 4H, H15,16),
47.0, 24.7 (s, 2 × 2H, H8,9), 12.1, 10.3, 8.0 (s, 3 × 4H, H11,12,13), 6.22
(s, 6H, H14). UV-visible (CH2Cl2): λmax ) 450 nm (5400), λ ) 394
nm (26700). Elemental analysis calcd for DyC43H44N7O9‚CH2Cl2
(found): wt % C 50.32 (50.18); wt % H 4.41 (4.70); wt % N 9.34
(9.71).
LYb(NO3)3 1H NMR (CD2Cl2) δ: 114.4, 27.3, 27.3, 23.6 (m, br, 4
× 2H, H8′,9′,10′,11′), 14.0 (s, 6H, H7′), 2.9, -6.4 (s, br, 2 × 4H, H15,16),
1.96, -4.0 (d, J ) 8.7 Hz, 2 × 2H, H8,9), 0.86, -0.4, -0.7 (s, 3×4H,
H11,12,13), -0.4 (s, 6H, H14). UV-visible (CH2Cl2): λmax ) 452 nm
(5150), λ ) 394 nm (28850). Elemental analysis calcd for YbC43H44N7-
O9‚CH2Cl2 (found): wt % C 49.82 (49.71); wt % H 4.37 (4.41); wt %
N 9.24 (9.82).
[LH][PF6]. To an acetone solution (15 mL) of L (100 mg, 0.16
mmol) was slowly added 1 equiv of a diluted solution of HPF6 in water
(2.8 mL, 0.19 mmol) at room temperature. The initially yellow solution
turned rapidly to deep red and was stirred at room temperature for 4 h.
The solvent was then removed under vacuum; the resulting solid was
dissolved in the minimum amount of dichloromethane and precipitated
upon addition of pentane (50 mL). After filtration, the red solid was
dried under vacuum (96.6 mg, 78%). 1H NMR (CD2Cl2) δ: 8.3 (d, br,
3
3J ) 8.0 Hz, 2H, H8′), 8.2 (d, J ) 8.0 Hz, 2H, H11′), 7.8 (m, 4H,
H9′,10′), 7.2 (d, 3J ) 8.8 Hz, 2H, H8), 6.9 (d, 3J ) 8.8 Hz, 2H, H9), 3.4
3
(t, J ) 7.6 Hz, 4H, H11), 3.3 (m, 4H, H16), 3.2 (m, 4H, H15), 2.8 (s,
3
6H, H7′), 1.7 (m, 4H, H12), 1.5 (m, 4H, H13), 1.0 (t, J ) 7.3 Hz, 6H,
H14). 31P NMR (CD2Cl2) δ: -144.4 (hept, 3J ) 710 Hz). UV-visible
(CH2Cl2): λmax ) 519 nm (9500), λ ) 408 nm (25500). Elemental
analysis calcd for C43H45N4PF6‚H2O (found): wt % C 66.14 (65.93);
wt % H 6.07 (6.04); wt % N 7.18 (7.10).
General Procedure for the Synthesis of LLn(NO3)3. In a Schlenk
flask, Ln(NO3)3‚xH2O (Ln ) Y, La, Nd, Gd, x ) 6; Ln ) Dy, Yb, x
) 5) was dissolved in dichloromethane (5-10 mL) at room temperature.
A dichloromethane solution of ligand L was then added via a syringe,
and the pale yellow mixture turned immediately to orange. After 1 h
stirring at room temperature, the solvent was evaporated under vacuum.
The resulting complex was then dissolved in the minimum amount of
hot acetonitrile and purified by low-temperature crystallization (-20
°C). The microcrystalline powder was finally filtered off, dissolved
with dichloromethane, and precipitated in pentane. After filtration and
evaporation of the solvents, the desired complexes were recovered as
orange powders. In the case of gadolinium and yttrium, the crystal-
lization gave crystals suitable for X-ray diffraction analysis, whereas
only very thin needles could be obtained in the case of lanthanum,
dysprosium, and ytterbium.
LY(NO3)3 1H NMR (CD2Cl2) δ: 8.3 (d, 3J ) 8.5 Hz, 2H, H8′), 8.1
(d, 3J ) 8.1 Hz, 2H, H11′), 7.8 (dd, 2H, H9′), 7.7 (dd, 2H, H10′), 7.0 (d,
3J ) 8.8 Hz, 2H, H8), 6.7 (d, 3J ) 8.8 Hz, 2H, H9), 3.3 (t, 3J ) 7.5 Hz,
4H, H11), 3.2 (m, 4H, H16), 2.9 (m, 4H, H15), 2.7 (s, 6H, H7′), 1.6 (m,
3
4H, H12), 1.4 (m, 4H, H13), 0.9 (t, J ) 7.0 Hz, 6H, H14). λmax ) 454
nm (4750), λ ) 392 nm (28900). Elemental analysis calcd for
YC43H44N7O9 (found): wt % C 57.92 (58.17); wt % H 4.97 (5.15); wt
% N 10.99 (10.40).
Supporting Information Available: Crystallographic packing
of [GdL1(NO3)3‚CH3CN], selected crystallographic data and
collection parameters for gadolinium and yttrium complexes,
crystallographic data in CIF files, optimized geometry of yttrium
complex, solvatochromism in the emission of L at room
temperature, molecular orbital energy diagram for L′H+, and
complete ref 38. This material is available free of charge via
the Internet at http://pubs.acs.org.
LLa(NO3)3 1H NMR (CD2Cl2) δ: 8.5, (d br, 2H, H8′), 8.1 (d br, 2H,
3
H11′) 7 (m br, 2 × 2H, H9′,10′), 7.0 (d, J ) 8.5 Hz, 2H, H8), 6.7 (d,
3
3J ) 8.5 Hz, 2H, H9), 3.3 (t, J ) 7.5 Hz, 4H, H11), 3.0 (m, 4H, H16),
2.8 (m, 4H, H15), 2.7 (s, 6H, H7′), 1.6 (m, 4H, H12), 1.4 (m, 4H, H13),
3
1.0 (t, J ) 7.0 Hz, 6H, H14). UV-visible (CH2Cl2): λmax ) 450 nm
(5400), λ ) 394 nm (26700). Elemental analysis calcd for LaC43H44N7-
JA063586J
9
J. AM. CHEM. SOC. VOL. 128, NO. 37, 2006 12255