Triplet Emitters and Control of Emission Properties

Hartmut Yersin

Institut für Physikalische Chemie, Universität Regensburg

D-93040 Regensburg, Germany

E-mail: hartmut.yersin@chemie.uni-regensburg.de

Abstract

Essential photophysical properties of organometallic triplet emitters depend systematically on the metal participation in the triplet states and on the effective spin-orbit coupling. These factors control the amount of zero-field splitting (ZFS) of the triplet state into substates. Increase of ZFS corresponds to higher metal character in the triplet state. Higher metal character reduces the energy difference between excited singlet and triplet states, enhances the singlet-triplet intersystem crossing rate, lowers the emission decay time, changes the vibrational satellite structure, decreases the excited state geometry change, leads to a stronger electronic coupling between the ligands etc.. These effects are discussed by referring to well characterized compounds. Based on the new ordering scheme presented for triplet emitter materials, a controlled development of compounds with pre-defined photophysical properties becomes possible.

1. Ordering scheme for organometallic triplet emitters

Many photophysical properties of the lowest excited triplet states and the corresponding transitions of organometallic compounds are essentially determined by the extent of metal participation in the wavefunctions. This metal participation does not only alter the spatial extension of the wavefunctions, but it also induces significant mixtures of singlet and triplet states by spin-orbit coupling (SOC) which is mainly carried by the metal orbitals. It is of interest to develop an understanding how these effects influence the photophysical properties of emitter materials. Indeed, this is possible. Moreover, it can be shown that a simple ordering scheme can be helpful in this respect. [1-5] Specifically, the energy splitting of the triplet state into substates, the zero-field splitting (ZFS) measured in cm^-1, can be utilized, since this parameter displays the importance of metal character and SOC for the respective triplet state.

The amount of ZFS is usually determined spectroscopically. Nevertheless, it is instructive to visualize that it is dominantly controlled by specific interactions. For example, the splitting of a pure ³ππ* state which is not metal-perturbed is only given by spin-spin interactions (e.g. see [6-8]). In this situation, the ZFS is of the order of 0.1 cm^-1. However, if ³MLCT and ¹MLCT states are in proximity (MLCT = metal-to-ligand charge transfer), the substates of these latter states will interact due to SOC with the substates of the ³ππ* term. This can lead to a significant triplet splitting in particular, when 5d orbitals are involved in the ^1,3MLCT states. As consequence, the ZFS can increase by more than three orders of magnitude, for example, up to 200 cm^-1. [1-5] The amount of ZFS can be determined, at least formally, by second order perturbation theory, whereby the perturbation is given by the spin-orbit Hamiltonian. The resulting matrix elements that induce the coupling are essentially controlled by d-orbital participation in the admixing wave­function. (Compare also Refs. [7-12]. )

Fig. 1: Ordering scheme for organometallic triplet emitters according to the amount of zero-field splitting (ZFS) of the emissive triplet state. This splitting reflects the size of metal participation (MLCT and/or d-orbital character) and spin-orbit coupling in the corresponding wavefunctions. The diagrams at the bottom show energy levels of the relevant frontier molecular orbitals for the different compounds. For detailes see text. (Compare [5] and for the Ir(III) compounds [13-15].)

Fig. 1 shows a sequence of a series of compounds which are arranged according to an increasing ZFS of the emissive triplet state. The ZFS values of most of the compounds have been determined from highly resolved optical spectra. [1-5] (References for the individual compounds are given in [5].) Only for Ir(ppy)₃ [13] and Ir(ppy)₂(CO)(Cl) [14] such resolved spectra were not yet obtainable, and therefore the information was determined indirectly from the temperature dependence of the emission decay time.

The low-lying electronic states of the compounds shown in Fig. 1 have to be assigned to different types of frontier orbital transitions. Thus, the lowest triplets of [Rh(bpy)₃]³⁺ and [Pt(bpy)₂]²⁺ are characterized as ligand centered ³LC(³ππ*) states with very small metal admixtures and those of [Ru(bpy)₃]²⁺, Ir(ppy)₃, [Os(bpy)₃]²⁺, and [Os(phen)₃]²⁺ represent ³MLCT states. The cyclometalated Pt(II)-compounds and [Pt(mnt)₂]^2- have to be assigned to intermediate situations with significant ³LC/³MLCT mixtures, whereas the lowest states of the oxinate complexes, such as Pd(qol)₂, Pt(qol)₂, and Pt(qtl)₂, are characterized as ³ILCT states (intra-ligand charge transfer from the phenolic moiety to the pyridine part of the ligand) with relatively small metal-d or MLCT admixtures. Nevertheless, the ordering expressed by the sequence shown in Fig. 1 can in fact be employed in the sense of a controlled variation of metal participation/spin-orbit coupling. In the next section, it will be discussed how photophysical properties vary with the increase of the zero-field splitting parameter.

2. Photophysical properties of triplet emitters controlled by metal participation

The energy splitting (ZFS) of a triplet state into substates displays the importance of metal character and spin-orbit coupling for this state. The ZFS value varies by more than a factor of 2000 (Fig. 1). Thus, essential changes of photophysical properties are expected to occur. Several important trends will be discussed in the following subsections. This is carried out on a qualitative and introductory basis. (Fig. 2) For comparison, properties of typical organic molecules (ππ* emitters) are also referred to.

2.1 Singlet-triplet splitting

Organic molecules exhibit S₁ – T₁ splittings being typically of the order of 10⁴ cm^-1 (Fig. 2) for states which stem from a ππ* configuration and for molecules of similar sizes like those shown as ligands in Fig. 1. The ΔE(S₁-T₁) splitting is essentially given by the exchange interaction and is a consequence of electron-electron interaction. For an organo-metallic compound, such as Pd(thpy)₂, one finds a value of 5418 cm^-1 [5]. The emissive triplet of this compound has been classified experimentally [5, 16-18] and later also by CASPT2 ab initio calculations [19] as being mainly of LC(ππ*) character with a small MLCT(4dπ*) admixture. Therefore, the compound is found at the left hand side of the ordering scheme of Fig. 1. [5] On the other hand, for Pt(thpy)₂ with a significant MLCT(5dπ*) admixture to the LC(ππ*) state [4, 5], the amount of ΔE(S₁-T₁) is reduced to 3278 cm^-1 [5]. (Fig. 2) Obviously, enhancing of metal character in the corresponding wavefunctions increasingly reduces the ΔE(S₁-T₁) value. Due to the higher metal character, the electronic wavefunctions extend over a larger spatial region of the complex. This is connected to an on average larger spatial separation between the interacting electrons. Thus, electron-electron interaction and also exchange interaction are reduced. This explains both the lowering of the S₁ state and the reduction of ΔE(S₁-T₁) for compounds with higher metal character.

Fig. 2: Photophysical properties of a representative organic molecule compared to two organo-transition-metal emitters. The emissive triplets of Pd(thpy)₂ and Pt(thpy)₂ exhibit small and significant MLCT admixtures to the LC(ππ*) states, respectively. The positions of the compounds in the ordering scheme and the molecular structures are found in Fig. 1. Photophysical properties of Pd(thpy)₂ and Pt(thpy)₂ are discussed in detail in [5]. lig. vibr. = ligand vibrations, M-L vibr. = metal-ligand vibration.

2.2. Inter-system crossing

After excitation of the singlet state S₁ either optically or by electron-hole recombination, the organic molecule can exhibit an efficient fluorescence (S₁→ S₀ emission) with a time constant of the order of 1 ns. In competition to this process, ISC can at least principally depopulate the S₁ state. However, the time constant of τ(ISC) is often much larger (order of 10 to 100 ns) than the radiative decay constant of the S₁ state and thus ISC does not effectively quench the fluores­cence. (Fig. 2) On the other hand, in transition metal complexes, the ISC time is drastically reduced due to singlet-triplet mixing induced by SOC and due to the reduction of the energy separation ΔE(S₁-T₁). Further, due to this latter effect, the number of vibrational quanta which are responsible for the deactivation is reduced and thus the ISC process becomes even more probable. For example, already for a compound, such as Pd(thpy)₂, with a relatively small metal participation in the wavefunction of the T₁ state and a halved ΔE(S₁-T₁) value, ISC is by four to five orders of magnitude faster (800 ps [5]) than in usual organic compounds with ¹ππ* and ³ππ* states. For Pt(thpy)₂ with a higher metal participation, the process is again much faster (τ(ISC) ≈ 50 fs [5]). (Fig. 2)

In conclusion, the process of inter-system crossing in organometallic compounds with transition metal ions is fast and efficient for all compounds shown in Fig. 1. The quantum efficiency for this process is often nearly one. Therefore, an emission from the S₁ state is not observable. This property is the basis of the triplet harvesting effect in OLEDs which are based on triplet emitters. [33]

2.3. Phosphorescence decay time

In pure organic molecules, the S₀(π²) ↔ T₁(ππ*) transition is usually strongly forbidden. The radiative decay time can be of the order of 10 s. [20] On the other hand, non-radiative processes are mostly much faster. Thus, phosphorescence from the T₁ state is normally totally quenched at ambient temperature. With increasing SOC, the radiative decay time of the T₁ → S₀ transition is reduced and thus the radiative path can compete with the non-radiative one. Interestingly, already a relatively small ³MLCT/¹MLCT admixture to the ³LC(ππ*) state increases the S₀ ↔ T₁ transition probability drastically. For example, for Pd(thpy)₂ the radia­tive decay time is reduced to be of the order of 1 ms. For Pt(thpy)₂ with a significant MLCT admixture to the lowest ³LC state, it is even reduced by six to seven orders of magnitude as compared to the organic emitters, and one finds a radiative decay time of the order of 1 μs. (Fig. 2) For completeness, it is remarked that the values given refer to an average radiative decay of the triplet, since the different triplet substates exhibit distinctly different decay properties (e.g. see the following sub-section and [5]).

In summary, for organometallic compounds the transition probability between the T₁ and the S₀ states can be tuned by several orders of magnitude as compared to organic emitters. This is mainly induced by an increase of spin-orbit coupling. Thus, the radiative processes can well compete with the non-radiative ones. Consequently, organo-transition-metal compounds can exhibit efficient emissions (phosphorescence) and therefore are well suited as emitter materials for OLEDs.

2.4. Zero-field splitting

Triplet states split generally into substates. This is also valid for purely organic molecules. However, for these one finds only small values of ZFS of the order of 0.1 cm^-1 which result from spin-spin coupling between the electrons in the π and π* orbitals, respectively (e.g. see [6-8]). On the other hand, in organo-transition-metal compounds SOC will modify the triplet states’ properties by mixing in higher lying states of singlet and triplet character. Already small admixtures have drastic consequences. For example, one finds an increase of the ISC rate (see above) and of the phosphorescence decay rate (see previous subsection) by orders of magnitude, although the ZFS is not yet distinctly altered. [5] Well studied examples with this behavior are [Rh(bpy)₃]³⁺ and Pd(qol)₂ (=Pdq₂). [7, 21, 22]

Higher metal participation and larger SOC lead to distinct ZFSs as has already been discussed and depicted in Fig. 1. Importantly, the individual triplet substates can have very different photophysical properties with regard to radiative decay rates, vibronic coupling, coupling to the environment, emission quantum yields, population and relaxation dynamics due to spin-lattice relaxation, sensitivity with respect to symmetry changes, etc.. [1-5, 7, 13, 14, 16-18, 21-25] At ambient temperature, the individual properties are largely smeared out and one finds mostly only an averaged behavior. Nevertheless, the individual triplet substates still determine the overall emission properties.

2.5. Emission band structure and vibrational satellites

At ambient temperature, the phosphorescence of organometallic compounds consists usually of superimposed spectra, which stem from the different triplet substates (see previous subsection, Fig. 2). An individual spectrum which results from one specific substate is composed of a transition at the electronic origin (0-0 transition), a large number of vibrational satellites, and of in part over­lapping low-energy satellites which involve librations of the complex in its environment. Moreover, all of these individual transitions are smeared out by inhomogeneity effects. Thus, at ambient temperature normally only a broad-band emission results. Sometimes residual, moderately resolved structures occur, which often stem from overlapping vibrational satellites (not necessarily from progressions). Interestingly, at low temperature and under suitable con­ditions, these structures can be well resolved and characterized. [1-5, 15-18, 21-32] From this kind of investigations it follows that also the vibrational satellite structure is influenced by metal praticipation in the electronic states. [29] Specifically, the spectra of compounds with electronic transitions of mainly LC(ππ*) character (small metal parti­cipation) are largely determined by satellites corresponding to ligand vibrations (fundamentals, combinations, progressions). However, with increasing metal participation, low-energy vibrational satellites of metal-ligand character grow in additionally (up to about 600 cm^-1 from the electronic origin). [29] (Fig. 2) Thus, the maximum of the unresolved emission spectrum shifts towards the electronic 0-0 transition.

To summarize, metal participation in the emitting state leads to a slight blue shift of the unresolved emission maximum as compared to an LC spectrum with the same 0-0 transition energy. Moreover, the occurrence of additional low-energy metal-ligand vibrational satellites causes a further smearing out of the spectrum. In particular, the slight spectral shift might be of interest for fine-tuning of the emission color.

2.6. Electronic charge distribution and excited state reorganization

The singlet-triplet transitions of compounds with small metal character, situated on the left hand side of Fig. 1, are localized to one of the ligands, even if the formal symmetry of the complex would allow a delocalization over all of the ligands. This has been proven for [Rh(bpy)₃]³⁺ [21] and [Pt(bpy)₂]²⁺ [30] by use of the method of deuteration labeling. [2, 4] On the other hand, for compounds with emissive ³MLCT states such as [Ru(bpy)₃]²⁺ [31] and [Os(bpy)₃]²⁺ [32] it has been shown that the excited state is delocalized over the three ligands and the metal. [2, 4, 31, 32] This is even valid for Pt(thpy)₂, in which the triplet state is largely of LC(ππ*) character, but for which the metal orbital admixture induces suffi­cient coupling between the ligands. (Fig. 1 middle) Thus, one obtains a delocalization in the lowest excited state. [4, 5] It is remarked that these results were determined for compounds doped into rigid matrices.

oreover, since the metal character or MLCT contribution in the triplet state, if suffici­ently large, can induce a ligand-ligand coupling that delocalizes the excited state wavefunc­tion, the metal contribution will also affect the geometry change that follows an excitation. For the localized situation in the compounds mentioned above, one finds a maximum Huang-Rhys parameter [5] of S = 0.3, while for the compounds with delocalized triplets, this cha­racteristic parameter is by a factor of three to four smaller. [2, 5] Geometry changes or reor­ganization effects which occur upon excitation of the emissive triplet states are small for nearly all of the complexes investigated (in rigid matrices). But still, it can be concluded that the reorganization effects are further minimized for those organometallic compounds with high metal participation in the triplet states (complexes at the right hand side of Fig. 1). The emissive states of these compounds exhibit a weaker coupling to the environment and there­fore represent good candidates for OLED emitter materials with high emission quantum yields.

3. Conclusion

On the basis of a detailed knowledge about the photophysical properties of organo-transition-metal emitters, clear trends are elucidated for triplet emitters. This leads to the possibility to control important factors, such as metal participation and spin-orbit coupling in the triplet states, by chemical variation. Thus, the controlled development of compounds with pre­defined photophysical properties will be achievable. Specifically, increase of metal participation and SOC is displayed in a growing zero-field splitting of the triplet into substates. This parameter can be determined experimentally. The corresponding ZFS values of the compounds discussed in this investigation vary by more than a factor of 2000. In parti­cular, increasing metal character lowers the energy of the excited singlet, reduces the energy separation between the excited singlet and triplet states, enhances the intersystem crossing rate from the lowest excited singlet to the lowest triplet, lowers the emission decay time by increasing the corresponding radiative rate, changes the vibrational satellite structure and thus the spectral distribution of the emission band, decreases the excited state reorganization energy, etc.. This promising new concept can, for example, usefully be applied to develop new triplet emitter materials for more efficient OLEDs. [33]

4. References

1.	Yersin H., Humbs W., Strasser J.; Coord. Chem. Rev. 1997, 159, 325
2.	Yersin H., Humbs W., Strasser J.; Top. Curr. Chem. 1997, 191, 153
3.	Yersin H., Strasser J.; J. Luminescence 1997, 72-74, 462
4.	Yersin H., Humbs W.; Inorg. Chem. 1999, 38, 5820
5.	Yersin H., Donges D.;&xnbsp; Top. Curr. Chem. 2001, 214, 81
6.	Lumb M. D. (ed); Luminescence Spectroscopy, Academic Press, London 1978
7.	Glasbeeek M.; Top. Curr. Chem. 2001, 213, 95
8.	Azumi T., Miki H.; Top. Curr. Chem. 1997, 191, 1
9.	Komada Y., Yamauchi S., Hirota N.; J. Phys. Chem. 1986, 90, 6425
10.	Ikeda S., Yamamoto S., Nozaki K., Ikeyama T., Azumi T., Burt J.A., Crosby G.A.; J. Phys. Chem. 1991, 95, 8538
11.	Colombo M.G., Hauser A., Güdel H.U.; Top. Curr. Chem. 1994, 171, 143
12.	Funayama T., Kato M., Kosugi H., Yagi M., Higuchi J., Yamauchi S.; Bull. Chem. Soc. Jpn. 2000, 73, 1541
13.	Finkenzeller W.J., Yersin H.; Chem. Phys. Lett. 2003, 377, 299
14.	Finkenzeller W.J., Stoessel P., Yersin H.; Chem. Phys. Lett. 2004, 397, 289
15.	Hay P.J.; J. Phys. Chem. A 2002, 106, 1634
16.	Yersin H., Schützenmeier S., Wiedenhofer H., von Zelewsky A.; J. Phys. Chem. 1993, 97, 13496
17.	Schmidt J., Wiedenhofer H., von Zelewsky A., Yersin H.; J. Phys. Chem. 1995, 99, 226
18.	Glasbeek M., Sitters R., van Veldhofen E., von Zelewsky A., Humbs W., Yersin H.; Inorg. Chem. 1998, 37, 5159
19.	Pierloot K., Ceulemans A., Merchán M., Serrano-Andrés L.; J. Phys. Chem. A 2000, 104, 4374
20.	Turro N.; Modern molecular photochemistry. Benjamin/Cummings Publ., Menlo Park, California, USA 1978
21.	Humbs W., Yersin H.; Inorg. Chem. 1996, 35, 2220
22.	Yersin H., Donges D., Nagle J.K., Sitters R., Glasbeek M.; Inorg. Chem. 2000, 39, 770
23.	Yersin H., Strasser J.; Coord. Chem. Rev. 2000, 208, 331
24.	Strasser J., Homeier H.H.H., Yersin H.; Chem. Phys. 2000, 255, 301
25.	Yersin H., Donges D., Humbs W., Strasser J., Sitters R., Glasbeek M.; Inorg. Chem. 2002, 41, 4915
26.	Wiedenhofer H., Schützenmeier S., von Zelewsky A.; Yersin H.; J. Phys. Chem. 1995, 99, 13385
27.	Yersin H., Kratzer C.; Coord. Chem. Rev. 2002, 229, 75
28.	Yersin H., Kratzer C.; Chem. Phys. Lett. 2002, 362, 365
29.	Yersin H., Huber P., Wiedenhofer H.; Coord. Chem. Rev. 1994, 132, 35
30.	Humbs W., Yersin H.; Inorg. Chim. Acta. 1997, 265, 139
31.	Braun D., Huber P., Wudy J., Schmidt J., Yersin H.; J. Phys. Chem. 1994, 98, 8044
32.	Huber P., Yersin H.; J. Phys. Chem. 1993, 97, 12705
33.	Yersin H.; Top. Curr. Chem. 2004, 241, 1

Home