Facile access to efficiently luminescent Ln³⁺ phosphonic ester coordination polymers (Ln = Eu, Tb, Dy)

Abstract

Efficiently luminescent coordination polymers containing lanthanide ions (Eu³⁺, Tb³⁺, Dy³⁺) can be easily obtained using a suitable bisphosphonic ester from which a heterometallic version can be elaborated showing white light emission.

---

Owing to their remarkable photophysical properties, lanthanide ions are unique for applications in lighting, sensing and display technologies.^1–3 In recent years coordination polymers turned out to be a suitable scaffold for accommodating isolated lanthanide centers.^4–6 While the most frequently employed binding sites in coordination polymers or metal–organic frameworks are carboxylates or nitrogen bases,^7,8 phosphonates^9,10 and organophosphates have been used as ligands as well.¹¹ By contrast, the use of phosphonic esters for this purpose is less explored and just started to emerge.^12,13 The special benefit of the P^(V)=O units as ligands towards lanthanide ions is their property to increase quantum yields of complexes compared to their carboxylic acid counterparts. The vibrational frequency of e.g. phosphonates is much lower than that of the respective carboxylates, which suppresses vibrational relaxation of excited states at the coordinated metal, thus increasing the quantum yield of emissive processes.¹⁴ Aiming at high rigidity, mainly aryl bonded oxophosphorane units have been employed as ligands so far, which are accessible via metallation reactions under inert conditions only, imposing some synthetic challenges with respect to organofluorine functionalities.^15,16

For alkyl bonded phosphonic acids the Michaelis–Arbuzov rearrangement provides a well-established access offering large structural variety.¹⁷ A disadvantage of this reaction are usually very long reaction times and sometimes very poor yields following classical procedures.¹⁸ By contrast, microwave irradiation allows uniform heating and higher temperatures which has been demonstrated for phosphonates with small organic substituents at phosphorus.^19,20

In combining the beneficial features of the aryl and the alkyl based phosphonate ligands we set out to explore the microwave driven preparation of a biphenyl based bisphosphonate. Here we report a novel synthetic access to (C₆H₄PO(OiPr)₂)₂ under solvent free conditions in high yield, and its use as ligand for the construction of efficiently photoluminescent coordination polymers involving Eu, Tb, Dy cations along with their non-emissive Y analog. The combination of these metals in heterotrimetallic combinations will be addressed as well, with respect to white light emitting coordination polymers. The use of (C₆H₄PO(OiPr)₂)₂ (L) as ligand has been limited by its synthetic availability so far. It has been mentioned in the literature several decades ago, but only with very limited characterization with its ligand properties remaining unexplored till now.²¹ Starting from 4,4′-dichloromethylbiphenyl and triisopropyl phosphite we have been able to prepare compound L on a multigram scale in excellent yield using microwave heating (Scheme 1). The ³¹P NMR spectrum in solution shows a signal at 24 ppm split to a characteristic multiplet with a triplet of triplet like pattern owing to P–H coupling. Solid L shows two IR bands characteristic for phosphonates¹³ at 1232 and 1105 cm⁻¹.

Scheme 1 Synthesis of L and coordination polymers 1a–e containing Eu (a), Tb (b), Dy (c), Y (d) or Eu_0.15Tb_0.70Y_0.15 (e).

Ligand L crystallizes in a P2₁/n space group. In the asymmetric unit one molecule of ligand is present showing twisted phenyl rings (Fig. 1). The P=O bond lengths range between 1.455(5) and 1.476(5) Å and are within standard bond lengths for such systems.^12,22 The angles around P11 cover a range between 103.8(3) and 114.9(3)°. The other phosphorus atom (P12) has angles within a range of 102.5(4) to 115.5(4)°. Both atoms show slight deviation from tetrahedral environment probably owing to steric interactions of the bulky substituents. The arrangement of the biphenyl system deviates from coplanarity with an interplanar angle of 43.7(3)°.

Fig. 1 Solid state structure of ligand L. Ellipsoids are drawn at 30% probability level.

The formation of coordination polymers 1a–e can be achieved by slow evaporation of an ethanol solution of the respective Ln(NO₃)₃ salt (Ln = Eu, Tb, Dy, Y) and an excess of the ligand L at room temperature (Scheme 1). Ethanol proved to be the best solvent to grow crystals suitable for single crystal X-ray analysis while other alcohols usually gave only microcrystalline powders. Polymers 1a–e are all isostructural which was further confirmed by powder diffraction for 1e as bulk material. Besides single crystal X-ray diffraction, compounds 1a–e were characterized by elemental analysis and IR spectroscopy. All polymers 1a–e show two nearly identical bands for the coordinated ligand at 1165 and 1097 cm⁻¹ shifted to lower wave numbers compared with uncoordinated L. Out of the polymers 1a–e only the structure of europium based 1a will be discussed as representative structure of the series. The polymer crystallizes in space group P2₁/n. The asymmetric unit contains one Eu(NO₃)₃ unit with 1.5 eq. of ligand L suggesting a formula for the repeating unit of polymer as [[Eu(NO₃)₃]₂L₃]. This composition of the coordination polymer is further confirmed by elemental analysis. In the solid state structure each europium ion is coordinated by three bidentate nitrate anions and three monodentate ligands adding up to a coordination number of 9 (Fig. 2). The bond lengths from europium to nitrate oxygen atoms are in the range of 2.418(7) to 2.479(7) Å and do not deviate significantly from similar complexes.¹³ The bond lengths of europium to oxygen atoms belonging to the phosphonic ester ligand are slightly shorter compared with those of a nitrate and are in the range of 2.313(5) to 2.347(5) Å. The polymer forms a zig-zag chain in a two dimensional sheet like arrangement (Fig. 3). These sheets are stacked in an alternating AB packing. Interestingly Π–Π stacking between biphenyl rings is not observed, due to the perpendicularity of neighboring phenyl rings. A minor difference between the molecular structures of 1a–e affects the dihedral angles of the phenyl rings of the biphenyl unit which deviates slightly from coplanarity. The deviation for 1a is slightly (ca. 2°) larger than for 1b–e which are identical within the limits of precision. Also the cell volume is smaller in case of 1a (by ca. 200 ų). Nevertheless such tiny structural features can be decisive for physical properties as in related cases.²³

Fig. 2 Coordination sphere around the europium ion in 1a. Thermal ellipsoids are drawn at 30% probability level.

Fig. 3 Repeating unit of 1a. Hydrogen atoms are omitted for clarity.

Owing to the paramagnetic nature of the metal ion, NMR spectroscopy is unsuitable to get an insight into the polymer formation in solution. However, with mass spectrometry some information could be obtained right after mixing the reactants. With MALDI-MS the formation of dimers of the type [(Eu₂(NO₃)₅)L₂]⁺ (m/z = 1634.28) and trimers [(Eu₃(NO₃)₈)L₃]⁺ (m/z = 2438.30) could be observed which suggests a step wise growth mechanism.

The study of the photophysical properties of ligand L in the solid state revealed a remarkable emission quantum yield of 89% at the excitation wavelength of 350 nm with an emission maximum at 410 nm (Fig. 4). Coordination of L to a non-emissive metal center like yttrium such as in 1d leads to similar yet slightly red shifted (11 nm in excitation, 4 nm in emission) excitation and emission spectra with a still high emission quantum yield of 60%. By comparison, the excitation spectrum of coordination polymer 1a shows a hypsochromic shift of the ligand based excitation at 330 nm. The europium based excitation of 1a, however, remains almost unchanged as compared to the spectra obtained for solid Eu(NO₃)₃ used as reference besides being the starting material in the preparation of 1a. Based on these findings, we performed excitation experiments with wavelengths adapted to either the ligand transitions or the metal transitions for comparison.

Fig. 4 Normalized ligand excitation and emission spectrum of solid L.

Using an excitation wavelength at 330 nm (maximum ligand excitation) leads to an increase of quantum yield from 2% in Eu(NO₃)₃ up to 37% for polymer 1a. The emission spectra obtained for ligand-centered excitation at 330 nm can be assigned to seven literature known luminescent transitions of europium (Fig. 5)²⁴ suggesting an antenna like energy transfer from the ligand to the metal. Despite excitation of the ligand, no ligand emission can be observed in turn at any excitation wavelength, suggesting a unidirectional energy transfer from ligand to metal with significant rise of luminescence quantum yield for the metal emission. Using an excitation wavelength at 396 nm (maximum europium excitation) again leads to a drastic increase of quantum yield from 6% in Eu(NO₃)₃ up to remarkable 70% for polymer 1a. The latter finding suggests that apart from an antenna effect the rigid matrix of the coordination polymer along with the enhanced vibrational properties of the oxophosphorane donor sites leads to reduced energy dissipation upon metal excitation of 1a as compared with the corresponding nitrate salt. To exclude the possibility that the quantum yields of the latter are low simply owing to vibrationally active water molecules in the nitrate salt used for comparison, Eu₂O₃ has been used as an alternative reference giving essentially identical results.

Fig. 5 Normalized excitation and emission spectrum of 1a at 330 nm excitation wavelength with the observed ⁵D₀ → ⁷F_n transitions.

For the terbium based polymer 1b similar effects can be observed. The organophosphorus ligand in 1b again causes a strong excitation band located at 318 nm (Fig. 6) resulting in an increased emission quantum yield of 26% for 1b compared to 5% for Tb(NO₃)₃. However, addressing the excitation maximum of terbium at 353 nm leads to an identical quantum yield of 26% with no further increase as in the corresponding europium case. This indicates a less efficient ligand to metal energy transfer for terbium than for europium with ligand L. Again the emission bands can be assigned to seven literature known metal based transitions.²⁵

Fig. 6 Normalized excitation and emission spectrum of 1b at 318 nm excitation wavelength with the observed ⁵D₄ → ⁷F_n transitions.

Moreover, a metal to ligand energy transfer occurs for the Tb based 1b, which depends on the excitation wavelength. This behavior causes an additional ligand emission observable in the spectra.

In dysprosium based polymer 1c the ligand to metal energy transfer effects a highly enhanced intensity in the excitation spectrum of dysprosium. Like for 1b the emission spectrum of 1c is strongly dependent on the excitation wavelength owing to the competition between metal to ligand energy transfer and direct ligand emission. The combination of metal and ligand emissions raises the luminescence quantum yield from below 2% for Dy(NO₃)₃ up to 34% for polymer 1c at 373 nm excitation wavelength. At an excitation wavelength of 399 nm, where no ligand emission occurs, still a quantum yield of 11% can be measured. In the region characteristic for the metal based emission four literature known luminescence transitions can be detected (Fig. 7).¹ The different luminescence behaviour of 1a compared with the other luminescent metal complexes 1b,c may be attributed to the higher ion potential of the Tb and Dy ions owing to their smaller size or the torsional difference of the biphenyl system in the corresponding coordination polymers. Moreover, for simple model compounds of these metals it is known from the literature that the f-splitting is slightly larger for Eu³⁺ than for the two other metal ions.^26,27 A further more specific explanation may arise from the reduced cell volume for 1a compared to 1b,c – despite the larger ion size of Eu³⁺ – which should increase the rigidity of the polymeric framework leading to reduced vibrational energy dissipation and consequently higher emission quantum yields.

Fig. 7 Normalized excitation and emission spectra (399 nm excitation) of dysprosium based 1c with the observed ⁵F_9/2 → ⁶H_n/2 transitions.

Since the quantum yields of the metal ions could be enhanced by the combination with the novel organophosphorus ligand L we wanted to further explore their suitability for light emitting devices. A mixture of the metal complexes should allow to adjust the emission color just by the choice of the excitation wavelength. Thus, a mixed metal heterotrimetallic version of [Ln(NO₃)₃]₂L₃ (1e) containing Eu, Tb and Y in a 0.15 : 0.70 : 0.15 ratio shows under standard UV-light at 366 nm the emission of white light with a quantum yield of 26%, where the red part comes from europium, green from terbium and the blue one from the yttrium-L-complex (Fig. 8).

Fig. 8 Normalized emission spectrum of heterotrimetallic 1e excited with a wavelength of 366 nm with the resulting color of the emitting light shown in the color map (black dot).

As mentioned above, in case of the yttrium complex the metal itself is not emissive but rather ensures a controlled arrangement of the blue emitting ligand. To conclude we have developed a straightforward access to a ditopic bisphosphonate ligand and explored its ligand properties towards selected f-block metals. The ligand is well suited to transfer energy to the metal centers leading to emissions with quantum yields near 70% for Eu and around 30% for Tb and Dy. By combination of metal ions with different emission colors, mixed hetero-metallic versions can be prepared within this polymer scaffold showing white light emission with 26% quantum yield.

The authors would like to acknowledge Dr. Clemens Bruhn and Astrid Pilz for X-ray measurements. We would also like to thank the EU ITN SHINE for financial support and the EU-COST network CM1302 “Smart Inorganic Polymers” (SIPs).

Conflicts of interest

The authors declare no conflicts of interest.

Notes and references

1. O. Guillou, C. Daiguebonne, G. Calvez and K. Bernot, Acc. Chem. Res., 2016, 49, 844–856.
2. J.-C. Bünzli, Nat. Chem., 2010, 2, 696.
3. B. Samson, A. Carter and K. Tankala, Nat. Photonics, 2011, 5, 466–467.
4. S. Kitagawa, R. Kitaura and S. Noro, Angew. Chem., Int. Ed., 2004, 43, 2334–2375.
5. A. Deluzet, W. Maudez, C. Daiguebonne and O. Guillou, Cryst. Growth Des., 2003, 3, 475–479.
6. D. Kustaryono, N. Kerbellec, G. Calvez, S. P. Freslon, C. Daiguebonne and O. Guillou, Cryst. Growth Des., 2010, 10, 775–781.
7. S. L. James, Chem. Soc. Rev., 2003, 32, 276.
8. C. Daiguebonne, N. Kerbellec, O. Guillou, J.-C. Bünzli, F. Gumy, L. Catala, T. Mallah, N. Audebrand, Y. Gérault, K. Bernot and G. Calvez, Inorg. Chem., 2008, 47, 3700–3708.
9. S. M. F. Vilela, A. D. G. Firmino, R. F. Mendes, J. A. Fernandes, D. Ananias, A. A. Valente, H. Ott, L. D. Carlos, J. Rocha, J. P. C. Tome and F. A. Almeida Paz, Chem. Commun., 2013, 49, 6400–6402.
10. F. A. A. Paz, S. M. F. Vilela and J. P. C. Tomé, Cryst. Growth Des., 2014, 14, 4873–4877.
11. S. K. Gupta, S. K. Langley, K. Sharma, K. S. Murray and R. Murugavel, Inorg. Chem., 2017, 56, 3946–3960.
12. J.-W. Zhang, C.-C. Zhao, Y.-P. Zhao, H.-Q. Xu, Z.-Y. Du and H.-L. Jiang, CrystEngComm, 2014, 16, 6635.
13. L. Zhang, W. Dou, W. Liu, C. Xu, H. Jiang, C. Chen, L. Guo, X. Tang and W. Liu, Inorg. Chem. Commun., 2015, 59, 53–56.
14. Y. Hasegawa, Y. Kimura, K. Murakoshi, Y. Wada, J.-H. Kim, N. Nakashima, T. Yamanaka and S. Yanagida, J. Phys. Chem., 1996, 100, 10201–10205.
15. A. Orthaber, F. Belaj, J. H. Albering and R. Pietschnig, Eur. J. Inorg. Chem., 2010, 34–37.
16. A. Orthaber, C. Seidel, F. Belaj, J. H. Albering, R. Pietschnig and U. Ruschewitz, Inorg. Chem., 2010, 49, 9350–9357.
17. A. E. Arbuzov, J. Russ. Phys.-Chem. Soc., 1906, 38, 687.
18. A. K. Bhattacharya and G. Thyagarajan, Chem. Rev., 1981, 81, 415–430.
19. D. Villemin, F. Simeon, H. Decreus and P.-A. Jaffres, Phosphorus, Sulfur Silicon Relat. Elem., 1998, 133, 209–213.
20. J. J. Kiddle and A. F. Gurley, Phosphorus, Sulfur Silicon Relat. Elem., 2000, 160, 195–205.
21. V. S. Abramov and N. A. Moskva, Zh. Obsch. Khim., 1967, 37, 2243–2247.
22. N. Stock, N. Guillou, J. Senker, G. Férey and T. Bein, Z. Anorg. Allg. Chem., 2005, 631, 575–581.
23. M. Cindrić, G. Pavlović, D. Pajić, K. Zadro, D. Cinčić, T. Hrenar, E. Lekšić, A. Belen Pinar Prieto, P. Lazić and D. Šišak Jung, New J. Chem., 2016, 40, 6604–6614.
24. K. Binnemans, Coord. Chem. Rev., 2015, 295, 1–45.
25. S. Cotton, Lanthanide and Actinide Chemistry, John Wiley & Sons, Ltd, 2006.
26. F. S. Richardson, M. F. Reid, J. J. Dallara and R. D. Smith, J. Chem. Phys., 1985, 83, 3813–3830.
27. M. Atanasov, C. Daul, H. U. Güdel, T. A. Wesolowski and M. Zbiri, Inorg. Chem., 2005, 44, 2954–2963.

---

Footnotes

† Dedicated to Prof. Dr. László Nyulászi on the occasion of his 60th birthday.

‡ Electronic supplementary information (ESI) available: Experimental and synthetic details, spectral and analytical data. CCDC 1562260–1562264 and 1569994. For ESI and crystallographic data in CIF or other electronic format see DOI: 10.1039/c7cc06621j

---