Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy 137 (2015) 259–266
Contents lists available at ScienceDirect
Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy journal homepage: www.elsevier.com/locate/saa
Dual emission behavior of phenyleneethynylene gold(I) complexes dictated by intersystem crossing: A theoretical perspective Li Wang a, Yuanyuan Li a, Yanxin Zhang a, Hongqing He b, Jinglai Zhang a,⇑ a
Institute of Environmental and Analytical Sciences, College of Chemistry and Chemical Engineering, Henan University, Kaifeng 475004, PR China Wuhan Center for Magnetic Resonance, State Key Laboratory of Magnetic Resonance and Atomic and Molecular Physics, Wuhan Institute of Physics and Mathematics, Chinese Academy of Science, Wuhan 430071, PR China b
h i g h l i g h t s
g r a p h i c a l a b s t r a c t
Calculations for model molecules
Presentation of the photo-induced decay mechanism for (a) 1a0 and (b) 2a0 suggested by the present quantum-chemical calculations. (Abs: Absorption; PF: Prompt fluorescence; DF: Delayed fluorescence; Ph: Phosphorescence; VR: Vibrational relaxation; ISC: Intersystem crossing; IC: Internal conversation; S/T-SC: Surface intersection on the crossing seam of singlet-triplet excited PES).
prove to be reliable by comparison with experimental values. The nature of absorption and dual emission spectra is analyzed. The consecutive decrease of kISC in experiment is correlated with DEST and length of PE ligand. The nonadiabatic decay pathway after photoexcitation is described in detail.
a r t i c l e
i n f o
Article history: Received 13 May 2014 Received in revised form 6 August 2014 Accepted 23 August 2014 Available online 30 August 2014 Keywords: Gold(I) complexes ISC Dual emission Delayed fluorescence TD-DFT CASSCF
a b s t r a c t In commonly studied gold(I) complexes with oligo (o-, p-, or m-phenyleneethynylene) (PE) ligands, an intriguing photophysical behavior is dual emission composed of fluorescence from S1 and phosphorescence from T1 which is dictated by effective intersystem crossing (ISC) process. In order to explore the salient photodynamics of such oligo-PE gold(I) complexes effectively, we have deliberately chosen three model complexes, namely, PhAC„CAAu(PMe3) (1a0 ) and PhAC„CA(1,m)C6H4AC„CAAu(PMe3) (m = 4, 2a0 ; m = 3, 3a0 ) in place of the real system. Firstly, electronic structure methods based on DFT and TD-DFT are utilized to perform optimization calculations for the ground- and lowest-lying excited states, respectively. Next, basic photophysical properties including absorption and emission spectra are investigated by TD-DFT under the optimized geometries. Besides, on the basis of the electronic spectra herein, we succeed in searching for surface intersections as the minima on the seam of singlet–triplet surface crossings (SCs) at the CASSCF level of theory. By integration of the results available, the process of delayed fluorescence of triplet–triplet annihilation (TTA) and phosphorescence was displayed in detail with SCs playing the lead in monitoring the ISC. Ó 2014 Elsevier B.V. All rights reserved.
⇑ Corresponding author. Tel./fax: +86 371 23881589. E-mail address: [email protected] (J. Zhang). http://dx.doi.org/10.1016/j.saa.2014.08.063 1386-1425/Ó 2014 Elsevier B.V. All rights reserved.
260
L. Wang et al. / Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy 137 (2015) 259–266
Introduction Exploitation of high-performance materials for organic lightemitting diodes (OLEDs) has invited a surge of efforts due to their advantages over current technologies in terms of luminescent properties, power consumption, and shape and size variability [1–7]. Wide-ranging applications in organic optoelectronics [8–10] and intriguing photophysical changes by appropriate adjustment of substituent patterns [11–13] for molecular materials based on oligo- or poly-phenyleneethynylene (PE) have rendered them potential candidates for fabricating effective OLEDs. To achieve the maximum device efficiency, an immediate requirement is exploitation of luminescent materials capable of converting excitons, particularly for the spin-forbidden triplet ones, into available light output based on the previous assumption that both spin singlet and triplet excitons account for the overall electroluminescence (EL) efficiency of OLEDs in the ratio of 1:3 [14–16]. However, no triplet emissions happen to arylacetylenes under ambient conditions from the photophysical point of view [17], which constitutes the major obstacle to further exploration for improved EL efficiency. A workable and facile strategy to address this challenge lies in complexation with heavy metal core, which serves as a key element propitious for intersystem crossing (ISC) and subsequent T1 ? S0 radiative deactivation rate. Comparatively speaking, gold-based organometallic complexes are a sagacious choice for the triplet harvesting dopant because of the twofold factors as follows: (1) As a heavy-metal atom, the incorporation of gold(I) moieties into the main chains of p-conjugated oligo- or poly-arylacetylide compounds is supposed to promote spin–orbit coupling, thus breaking the conventional spin selection rules and facilitating singlet ? triplet intersystem crossing (ISC); (2) The absence of low-lying d-d excited states due to d10 closed-shell configuration in gold(I) ion tends to maintain the p-conjugation of PE units and get rid of the undesirable competing nonradiative decay channel. In this regard, Che and coworkers have designed and synthesized a family of mono- and binuclear Cy3Psupported gold(I) complexes coordinated by various conjugated or cross-conjugated oligo-(p-, o-, or m-PE) in succession [17,18].
Characterization results through multiple spectral techniques indicate that most of the complexes strikingly feature dual emission composed of fluorescence from S1 and phosphorescence from T1. What is more, long-lived delayed fluorescence (DF) with TTA mechanism relative to short-lived prompt fluorescence (PF) is observed for complexes bearing extended arylacetylide units. The excited state dynamics analyses further demonstrated the controlled manipulation role by p-conjugation of the singlet excitedstate PE ligand in the exact ISC rate constant (kISC) rather than the heavy-atom effect [19]. Indisputably, the preceding laboratorial efforts have speculated on the formation mechanism of DF and resolved the origin of the dual emissions displayed by the designed PE-based gold(I) complexes using combined methods of steady-state and ultrafast dynamic spectroscopic measurements. However, for one thing, no structural information on ground and excited states involved in absorption and emission spectra is accessible. For another, we are curious about the exhaustive process for the occurrence of delayed fluorescence and normal emission behavior. Consequently, from the theoretical perspective we have in the present work extended our geometrical and spectral study to the gold(I)-PE complexes 1a–3a [19], the schematic structures of which are depicted in Scheme 1, with the aims: (1) to interpret the nature and onset of absorption and dual emission spectra; (2) to locate the surface crossings bridging spin singlet and triplet excited states in the procession of ISC for mechanism analysis of DF and phosphorescence; (3) to rationalize the experimentally observed consecutive decrease of ISC rate constant (kISC) in the sequence of 1a ? 3a ? 2a. What is noteworthy is that we have actually investigated the methyl-substituted models instead of the original systems 1a–3a, namely, PhAC„CAAu(PMe3) (1a0 ) and PhAC„CA(1,m)C6H4 AC„CAAu(PMe3) (m = 4, 2a0 ; m = 3, 3a0 ), which retains the conjugated character of 2a and 3a that is essential for the electronic excitation. This simplification of cutting down one phenyleneethynylene unit saves computational costs and proves to be reasonable by comparison with both experimental spectra and calculations for real systems 1a–3a. Hopefully, the theoretical results and sketched outline for deactivation channels are significant for insight into the
Scheme 1. The schematic structures for real systems 1a–3a and model molecules 1a0 –3a0 .
L. Wang et al. / Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy 137 (2015) 259–266
structure–property relationship of complexes under study and further fabrication of materials with desirable luminescence efficiency. Computational details The density functional theory (DFT) with B3PW91 [20–23] method and time-dependent DFT (TD-DFT/B3PW91) [24–26] calculations have been employed to obtain the equilibrium geometries of complexes 1a0 –3a0 in their ground and excited states, respectively. The ultraviolet absorption spectra and emission spectra have been determined by means of TD-DFT method with the B3PW91 and M06-2X [27] functionals based on the optimized ground-state (S0) and excited-state (S1 for fluorescence and T1 for phosphorescence) structures. The polarized continuum model (PCM) [28] with dichloromethane is employed to assess aqueous solvatochromic effects on the spectra. Owing to the large number of electrons and meanwhile to account for relativistic effects, the double-f quality basis set (LANL2DZ) [29–31] with inner electrons of the gold(I) atom substituted by Hay and Wadt’s effective core potential (ECP) and a 6-31G(d) [32] basis set on all the nonmetal atoms, have been adopted for both geometric optimizations and spectral estimations. The electronic spectra for original real system 1a–3a have been calculated as well. Moreover, the second-order perturbation theory (CASPT2) [33] method with the same LANL2DZ basis set on gold(I) atom and the Dunning’s valence double-f (ccpVDZ) [34] basis sets on carbon, hydrogen, and phosphor have been utilized in the elaborate calculations for vertical excitation energies under the initial B3PW91 equilibrium geometries. A level-shift technique with an imaginary shift of 0.3 au is applied to remove the intruder states in the CASPT2 theory. Regarding the active space, eight electrons in six orbitals are chosen for the complex 1a0 and four electrons in four orbitals chosen for 2a0 –3a0 , referred to as CAS(8,6) and CAS(4,4), respectively. For the purpose of investigating the role played by ISC in the excited-state decay process, we have carefully searched for the singlet–triplet PESs crossing points using the complete active space self-consistent field (CASSCF) [35] approach along with eight electrons distributed in six orbitals, namely, CAS(8,6), as the active space. Herein, the state-averaging CASSCF (SA-CASSCF) calculations over selected states for each species on an individual basis were conducted. And meanwhile a mixed basis set, LANL2DZ for gold(I), 631G [36] for hydrogen, coupled with 6-31G(d) for carbon and phosphor atoms, has been implemented into the CASSCF computation. DFT and TD-DFT calculations were performed with the Gaussian 09 [37] program package. The CASPT2 predictions for excitation energies and SA-CASSCF treatment for SCs were performed by the Molpro 2009 [38] program. Results and discussion The ground-state structures and absorption spectra in the CH2Cl2 solution The optimized ground-state geometries of model system [PhAC„CAAu(PMe3)] (1a0 ), [PhAC„CA(1,4)C6H4AC„CAAu(PMe3)] (2a0 ), and [PhAC„CA(1,3)C6H4AC„CAAu(PMe3)] (3a0 ) at the B3PW91 level of theory are depicted in Fig. 1 with selected atoms labeled. As shown in Table 1, the ligand length (1a0 vs 2a0 /3a0 ) or substituent position (2a0 vs 3a0 ) makes no difference to the key bond distances with CAP, PAAu, and AuAC kept around 1.83, 2.34, and 2.00 Å. The dihedral angles DC3APAC6AC7 in complexes 1a0 and 3a0 are close to 180°, indicating that C3AP stays within the PE plane. Inspection of the other two dihedral angles associated with C1AP and C2AP immediately reveals that both bonds keep mirror images to each other with respect to the symmetry plane passing the C3AP
261
bond and the PE ligand. However, the case in complex 2a0 is slightly different because of less effective coplanarity between C3AP and PE ligand compared with that in 1a0 and 3a0 . In addition, it is found that the torsional angle between two adjacent ethynylene moieties in 2a0 and 3a0 is 0° and 3.35°, respectively, giving evidence that p-conjugated framework is relatively more easily kept in para-PE compounds than that in the meta position. To correlate the ground-state geometrical differences in complexes 1a0 –3a0 with the absorption spectra, we have compiled the lowest and also the most intense singlet absorption bands together with the triplet excited states isoenergetic with the singlet-spin absorption by different methods in Table 2. Even the most intense singlet absorption was assigned as different excitation (S1 or S2) by different methods due to the inherent character of the methodology, they result from the similar electron transitions of HOMO ? LUMO or H1 ? LUMO and display the same character. Direct comparison with the experimental absorption bands [19] and predicted spectra for the real system 1a–3a (See Table S1 in Supplementary Information) attaches credence to the selected theoretical approaches to understand the spectroscopic properties. Additionally, with respect to the TD-DFT results, an overall blue shift in TD-M06-2X calculation compared with the absorptions in TD-B3PW91 is clearly discovered from the simulated spectra depicted in Fig. S1. On account of the fact that one single-electron excitation, namely, HOMO ? LUMO or H1 ? LUMO, contributes to the first singlet (S1 or S2) excited state in general, it is essential to understand the compositions constituting the frontier molecular orbitals (FMOs), which have been displayed in Fig. 2. In terms of population analysis, the HOMOs for three complexes are similar in character with basically commensurate contribution from pC„C and pPh, while, p⁄C„C character has been reduced and p⁄Ph enhanced in LUMOs. As a result, one can easily figure out that the maximum absorption bands should be ascribed to p ? p⁄ transition designated as ligand centered (LC) character, as the composition from gold(I) d orbital (denoted as dAu) is almost zero. Importantly, the calculated lowest-lying transition energies follow the order 1a0 > 3a0 > 2a0 , which is in accordance with the HOMO–LUMO energy gap listed in Fig. 2 as well as the aforementioned conjugation degree of 1a0 < 3a0 < 2a0 determined by geometrical information. It is expected that the relationship between electronic and geometric structures with spectroscopic nature should be adaptable to other similar metal-containing arylacetylide complexes. The excited-state structures and emission spectra in the CH2Cl2 solution In view of the geometric parameters for the lowest-lying singlet states (S1) of title complexes in Table 1, electron promotion obviously has no influence on the bond lengths of PMe3 moiety, while, it does on the degree of geometric distortion with the order of 2a0 < 3a0 < 1a0 by comparison of dihedral angle changes going from S0 to S1 states. In addition, AuAC4 bond of three complexes is slightly enhanced to various extents after photoexcitation. As has been unambiguously illustrated for complexes 1a–3a in experiment [17,19], dual emission including fluorescence and phosphorescence was observed with the exception for 1a, in which the prompt fluorescence (PF) decays rapidly with time evolution. For deep insight into the onset and nature of the emission spectra, vertical transition energies originating from S1 and T1 states, respectively, have been calculated by the TD-DFT level under homologous geometries. Fluorescent spectra The singlet–singlet emission wavelengths, oscillator strengths, configuration interaction (CI) coefficient, transition character, and
262
L. Wang et al. / Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy 137 (2015) 259–266
Fig. 1. Optimized ground-state structures for the model systems [PhAC„CAAu(PMe3] (1a0 ), [PhAC„CA(1,4)C6H4AC„CAAu(PMe3)] (2a0 ), and [PhAC„CA(1,4)C6H4 AC„CAAu(PMe3)] (3a0 ) under B3PW91 level of theory.
Table 1 Selected geometric parameters of the ground and excited states for model systems [PhC„CAu(PMe3] (1a0 ), [PhC„C(C6H4-l, 4-C„C)Au(PMe3)] (2a0 ), and [PhC„C(C6H4-l, 3-C„C)Au(PMe3)] (3a0 ) under B3PW91 and TD-B3PW91 levels of theory, respectively. Parameter
1a0 (S0)
1a0 (S1)
1a0 (T1)
2a0 (S0)
2a0 (S1)
2a0 (T1)
3a0 (S0)
3a0 (S1)
3a0 (T1)
Bond length (Å) C1AP C2AP C3AP PAAu AuAC4 C4AC5 C5AC6 C6AC7
1.83 1.83 1.83 2.34 2.00 1.23 1.43 1.41
1.84 1.84 1.83 2.34 1.96 1.27 1.37 1.45
1.83 1.83 1.83 2.34 1.99 1.27 1.36 1.48
1.83 1.83 1.83 2.34 2.00 1.23 1.42 1.41
1.83 1.83 1.83 2.34 1.98 1.25 1.39 1.44
1.83 1.83 1.83 2.34 1.99 1.25 1.39 1.45
1.83 1.83 1.83 2.34 2.00 1.23 1.43 1.40
1.83 1.83 1.83 2.34 1.99 1.23 1.42 1.39
1.83 1.83 1.83 2.34 2.00 1.23 1.42 1.39
64.4 55.7 176.0
72.2 48.2 167.9
64.3 55.8 176.1
48.3 71.9 168.1 0.02
48.5 71.9 168.1 0.03
48.4 72.0 168.1 0.01
60.3 59.5 180.0
61.1 58.5 179.1
60.9 58.8 179.3
3.35
0.17
0.05
Dihedral angle (°) C1APAC6AC7 C2APAC6AC7 C3APAC6AC7 C8AC9AC10AC11 C7AC8AC9AC10
263
L. Wang et al. / Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy 137 (2015) 259–266
Table 2 Selected absorption wavelengths (k) based on vertical excitation energies of complexes 1a0 –3a0 in dichloromethane under both TD-DFT and CASPT2 levels of theory, together with experimental valuesa. Species
TD-M06-2X k (nm)
f
Configuration
State
k (nm)
f
Configuration
1a0
T4
289
0.0000
T2
267
0.0000
H1 ? L(69%)
S1
283
0.6617
H ? L+2(67%) H2 ? L(25%) H ? L(93%)
S1
266
0.6831
H 1 ? L(57%)
2a0
T2
382
0.0000
T2
329
0.0000
H 1 ? L(48%)
S1
352
1.9252
H 2 ? L(47%) H ? L+1(27%) H ? L(98%)
S1
316
1.9968
H ? L(94%)
S1 T5 S2
313 304 304
0.3186 0.0000 1.0075
H ? L(73%) H ? L+1(47%) H1 ? L(78%)
T4 S1
281 280
0.0000 1.2870
H5 ? L(33%) H ? L(70%)
0
3a
a
TD-B3PW91 State
kExpt (nm)
Species
CASPT2 State
k (nm)
f
Configuration
1a0
T2 S2
280 281
0.0000 0.5104
H 1 ? L(99%) H ? L(91%)
282
2a0
T2
335
0.0000
S1
325
2.0473
H ? L+1 (50%) H1 ? L (46%) H ? L(89%)
342
3a0
T3
301
0.0000
S2
272
1.3410
H1 ? L(54%) H1 ? L+1(31%) 286
Experimental values from Ref. [19].
experimental values are listed in Table 3. The population analysis for frontier molecular orbitals of the complexes 1a0 –3a0 in their S1 state are depicted in Fig. S2(a) of Supplementary material. An inspection of the figure reveals that 2a0 possesses the most balanced orbital contribution from pC„C and pPh in HOMO among three complexes and meanwhile the most narrow HOMO–LUMO energy gap. Naturally, it is easy to understand the fluorescence wavelength and intensity under TD-B3PW91 level of theory in the order of 2a0 (399 nm) > 3a0 (352 nm) > 1a0 (314 nm) based on an overall consideration of p-conjugation monitored by orbital population and geometry variation. Furthermore, the p-featured HOMO and p⁄-featured LUMO orbitals determine the ligandcentered (LC) S1 1pp⁄ electronic transition accompanied by predominant electron promotion and charge migration.
Phosphorescent spectra The structural changes of three complexes in their lowestenergy triplet excited state T1 relative to the S0 state behaves in the similar fashion with that in S1 according to the optimized data shown in Table 1. As a result, geometries might produce influence on the phosphorescence spectra in the same way as S1 structures does on the fluorescence spectra. As can be seen from Table 4, the TD-M06-2X calculated emission spectra for 1a0 –3a0 in CH2Cl2 media occur at 473, 599, and 522 nm, respectively, in better agreement with the experimental measurements than TD-B3PW91 results and all with the one-electron HOMO ? LUMO excitation configuration. Since no spin–orbital coupling (SOC) effect is included in current TD-DFT calculations, the oscillator strengths for phosphorescent emissions are zero. Basically, the triplet
Fig. 2. Molecular orbital compositions, energy gaps, and contour plots for HOMO and LUMO of 1a0 –3a0 at the ground state S0.
264
L. Wang et al. / Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy 137 (2015) 259–266
Table 3 Fluorescence emissions of 1a0 –3a0 in dichloromethane based on different TD-DFT results, together with experimental valuesa. Method
State
kF (nm)
f
Main configuration
Character
kExpt (nm) F
0
1a
TD-B3PW91 TD-M06-2X
S1 S2
314 294
0.7216 0.7391
H ? L(92%) H ? L(92%)
ILCT ILCT
304
2a0
TD-B3PW91 TD-M06-2X
S1 S1
399 363
2.0460 2.0482
H ? L(98%) H ? L(95%)
ILCT ILCT
390
3a0
TD-B3PW91 TD-M06-2X
S1 S1
352 325
1.0783 1.3071
H ? L(95%) H ? L(90%)
ILCT ILCT
340
Species
a
Experimental values from Ref. [19].
Table 4 Phosphorescent emissions of 1a0 –3a0 in dichloromethane based on different TD-DFT results, together with experimental valuesa. Species
Method
State
kPh (nm)
f
Main configuration
Assignment
kExpt (nm) Ph
1a0
TD-B3PW91 TD-M06-2X
T1 T1
574 473
0.0000 0.0000
H ? L(92%) H ? L(95%)
3
ILCT 3 ILCT
420
2a0
TD-B3PW91 TD-M06-2X
T1 T1
748 599
0.0000 0.0000
H ? L(92%) H ? L(92%)
3
550
TD-B3PW91 TD-M06-2X
T1 T1
650 522
0.0000 0.0000
H ? L(92%) H ? L(90%)
3
3a0 a
ILCT ILCT
3
ILCT ILCT
470
3
Experimental values from Ref. [19].
emissions should be described as 3ILCT(3pp⁄) with charge significantly transferring between C„C and phenyl moieties due to the absent composition of d orbital from gold atom (dAu) and p orbital from the PMe3 group (pPMe3) for three complexes, which is determined by combination of Table 4 and Fig. S2(b). It deserves our notification that the composition analysis for transitional orbital suggests none contribution from dAu to both fluorescent and phosphorescent spectra, which otherwise indicates the liganddependant emissive nature for gole(I)-PE complexes. Additionally, our findings demonstrate that proper adjustment of the substitution pattern of the PE ligand serves as an effective strategy to tune the emitting color of similar PE-supported metal complexes. Overall, reasonable agreement between theoretical calculations and experimental values for both fluorescence and phosphorescence spectra effectively authenticate the reliability of the model molecules and methods employed in this contribution. For better evaluation of the general luminescence efficiency of title complexes, we have adopted the concept of relaxation energy (kS and kT) with reference to Ref. [14], which plays the same role in reflecting the degree of geometry relaxation from S1 to S0 state and thereby the photoluminescence quantum efficiency as Stokes shift does. As is denoted in Fig. 3, DES@S0 represents the vertical excitation energy from S0 ? S1 and DES@S1 is the vertical emission energy rising from S1. The smaller kS or kT is indicative of less distortion in the excited-state geometry, thus suppressing the nonradiative decay channel and resulting in the higher PL efficiency. As shown in Table S2, the kS and k⁄S values based on B3PW91/ TD-B3PW91 calculations vary at relatively low level, which may supply a new theoretical clue that PE-based gold(I) complexes may be considered as potential systems for fabrication of effective OLEDs.
excited state Tn in close proximity to the most intense and lowest-lying singlet state in the Franck–Condon region with the calculated energy deference, referred to as DEST hereafter, for 1a0 –3a0 lying at 0.02, 0.48, and 0.44 eV, respectively (For TD-B3PW91/TDM06-2X calculations, the values severally correspond to 0.10/ 0.03, 0.28/0.16, and 0.00/0.01 eV), which greatly favors the search for SCs. Fortunately, we have successfully located the SCs minima corresponding to 1a0 –3a0 on the crossing seam of the singlet–triplet PESs using state-averaging CASSCF methodology. The common points among three SCs shown in Table 5 consist in the averaging tendency of bond lengths with the extension of the acetylene link C4AC5 and contraction of the single bond C5AC6, which might favor the charge transfer between acetylide and phenyl groups. The determinate existence of SCs effectively verifies the prevailing occurrence of intersystem crossing (ISC) on the excited-state PES and hence excess thermal energy is released to the Tn state. With respect to the consecutively decreased ISC rate constant (kISC) by 1a ? 3a ? 2a as detected in experiment, it should be directly
Surface crossings and non-adiabatic deactivation patterns on the excited-state PESs It is well known that a SC usually makes it facile for the photoinduced dynamic process with participation of excited-state PES, and therefore the accurate location of the SC connecting two close PESs is meaningful for insight into correlation between 1pp⁄ and 3 pp⁄ structures in luminescent behavior. An examination of Table 2 once more suggests that there exists an energy-degenerate triplet
Fig. 3. Sketch map for the potential energy surfaces, vertical transition energies (DES@S0 and DES@S1) and relaxation energies (kS and k⁄S) for the S1 state.
L. Wang et al. / Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy 137 (2015) 259–266 Table 5 CASSCF optimized key geometric parameters of the surface intersections on the excited-state PES for complexes 1a0 –3a0 . Parameter
SC-1a0 (S2/T3)
SC-2a0 (S1/T3)
SC-3a0 (S2/T3)
Bond length (Å) C1AP C2AP C3AP PAAu AuAC4 C4AC5 C5AC6 C6AC7
1.84 1.84 1.84 2.35 2.02 1.25 1.39 1.42
1.83 1.83 1.83 2.36 2.04 1.29 1.36 1.43
1.84 1.84 1.83 2.35 2.02 1.25 1.38 1.42
Dihedral angle (°) C1APAC6AC7 C2APAC6AC7 C3APAC6AC7
60.6 60.6 180.0
48.9 71.7 168.4
59.9 62.3 178.2
related to the variation tendency of DEST-value. increasing p-conjugation with the sequence of should also contribute to the variation of kISC, conjugation promotes fluorescence and otherwise phosphorescence.
Besides, the 1a0 < 3a0 < 2a0 as effective weakens the
265
In view of the present theoretical findings and taking the previous experimental observations into account, the nonadiabatic decay pathway composed of radiative transition and nonradiative transition after irradiation by ultraviolet light is schematically proposed. Due to the mechanism similarity between 2a0 and 3a0 , the photo-induced experience for 1a0 and 2a0 are sketched as representatives in Fig. 4. Upon irriadition into the singlet excited state with 1 pp⁄ character, partial excited-state population relaxes rapidly out of the Franck–Condon region toward the equilibrium on the S1 PES and then decays back to the ground state via regular prompt fluorescence (PF) obeying Kasha’s rule. However, owing to the small energy difference DEST, a big proportion possessing enough energy quickly reaches the 1pp⁄/3pp⁄-SC, by which the photogenerated 1 pp⁄ population is promptly funneled into 3pp⁄ by means of ISC with complex dynamics on the excited-state landscape. Successively, a fast internal conversation (IC) on the triplet excited-state PES takes place, causing the electron populated on the energy-lowest T1 state. Hereafter, different decay channels happen to complexes 1a0 and 2a0 . As for the former, direct phosphorescence emission is observed. While for the latter, alternative deactivation pathways occur with the accumulation of T1 molecules: (1) a triplet ? singlet up-conversion via TTA process is largely enhanced with S1 excitons formed again and consequently the delayed
Fig. 4. Presentation of the photo-induced decay mechanism for (a) 1a0 and (b) 2a0 suggested by the present quantum-chemical calculations. (Abs: Absorption; PF: Prompt fluorescence; DF: Delayed fluorescence; Ph: Phosphorescence; VR: Vibrational relaxation; ISC: Intersystem crossing; IC: Internal conversation; S/T-SC: Surface intersection on the crossing seam of singlet–triplet excited PES).
266
L. Wang et al. / Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy 137 (2015) 259–266
fluorescence (DF) as experimentally detected; (2) the rest T1 population finds its way back to the ground state S0 through normal radiative transition and gives rise to the phosphorescence. To sum up, the SC-induced intersystem crossing plays an important role in the intriguing photophysical decay process of the title gold(I) complexes. And meanwhile the kISC variation is correlated with the length of the PE ligand and DEST. Conclusion The gold(I) complexes containing oligo(p- or m-phenyleneethynylene) (PE) fragments represent potential systems for highperformance OLEDs fabrication because of the dual emission by efficiently harvesting both singlet and triplet excitons. By combined methods of DFT, TD-DFT, coupled with CASSCF and CASPT2, we have theoretically investigated the distinctive photophysical behavior that is different from conventional metal complexes. In the following, we briefly address three conclusions derived from the present work. First, the evaluations for spectra agree well with the experimental observations, which in turn certify the predictions for the geometric structures. Equilibrium geometries of the complexes under study are slightly relevant to the PE oligomers with different lengths and different substitution patterns. Second, the S0 ? S1 and S1/T1 ? S0 transitions feature by remarkable PE ligand-centered (LC) (1pp⁄ or 3pp⁄) electron promotion and charge transfer. The relaxation energies (kS/kT) on the basis of vertical excitation and emission are evaluated to be small, which accordingly reduces the competing nonradiative decay probability and enhances the potential of such systems as lightemitting materials. Third, The detected ISC rate constant (kISC) in the order of 1a > 3a > 2a should be correlated to the conjugation-enhanced fluorescent intensity of 2a0 > 3a0 > 1a0 and as well the energy difference DEST in the sequence of 1a0 < 3a0 < 2a0 . The existence of surface intersections linking the excited singlet and triplet PESs favors ISC, which further sheds light on the mechanism manipulating the phosphorescence and TTA-induced delayed fluorescence. Hopefully, the structure–property relationship revealed in our quantum chemical investigation may on the one hand provide useful information on structures and spectra for similar PE-based complexes and on the other hand contribute to improved design and optimization of high-efficiency OLEDs worthy of experimental testing. Acknowledgements We thank the Beijing Key Laboratory of Ionic Liquids Clean Process and State Key Laboratory of Multiphase Complex Systems, Institute of Process Engineering, Chinese Academy of Sciences for providing computational resources. This work was supported by the Natural Science Foundation of He’nan Province of China (134300510008, 144300510032), Science Foundation of Henan Province (14A150034), and the Foundation for University Key Teachers from the He’nan Educational Committee. Appendix A. Supplementary material Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.saa.2014.08.063.
References [1] M.C. Gather, A. Köhnen, K. Meerholz, Adv. Mater. 23 (2011) 233. [2] M. Wohlgenannt, K. Tandon, S. Mazumdar, S. Ramasesha, Z.V. Vardeny, Nature 409 (2001) 494. [3] J.S. Wilson, A.S. Dhoot, A.J.A.B. Seeley, M.S. Khan, A. Köhler, R.H. Friend, Nature 413 (2001) 828. [4] L.C. Lin, H.F. Meng, J.T. Shy, S.F. Horng, L.S. Yu, C.H. Chen, H.H. Liaw, C.C. Huang, K.Y. Peng, S.A. Chen, Phys. Rev. Lett. 90 (2003) 036601. [5] M. Segal, M.A. Baldo, R.J. Holmes, S.R. Forrest, Z.G. Soos, Phys. Rev. B 68 (2003) 075211. [6] T.K. Hatwar, M.E. Kondakova, D.J. Giesen, J.P. Spindler, Inorganic electronics: materials, processing, devices and applications, in: F. So (Ed.), CRC Press, Boca Raton, FL, 2009, p. 433. [7] M.E. Kondakov, J.C. Deaton, T.D. Pawlik, D.J. Giesen, D.Y. Kondakov, R.H. Young, T.L. Royster, D.L. Comfort, J.D. Shore, J. Appl. Phys. 107 (2010). 014515–113. [8] R.E. Martin, F. Diederich, Angew. Chem., Int. Ed. 38 (1999) 1350. [9] P.J. Stang, F. Diederich, Modern Acetylene Chemistry, VCH, Weinheim, Germany, 1995. [10] G.M. Carter, Y.J. Chen, M.F. Rubner, D.J. Sandman, M.K. Thakur, S.K. Tripathy, in: D.S. Chemla, J. Zyss (Eds.), Non-Linear Optical Properties of Organic Molecules and Crystals, vol. 2, Academic, Orlando, FL, 1987. Chapter III-3. [11] A. Beeby, K. Findlay, P.J. Low, T.B. Marder, J. Am. Chem. Soc. 124 (2002) 8280. [12] J. Kim, I.A. Levitsky, D.T. McQuade, T.M. Swager, J. Am. Chem. Soc. 124 (2002) 7710. [13] J. Cornil, Y. Karzazi, J.L. Brédas, J. Am. Chem. Soc. 124 (2002) 3516. [14] H. Uoyama, K. Goushi, K. Shizu, H. Nomura, C. Adachi, Nature 492 (2012) 234. [15] M.A. Baldo, D.F. O’Brien, Y. You, A. Shoustikov, S. Sibley, M.E. Thompson, S.R. Forrest, Nature 395 (1998) 151. [16] A. Köhler, J.S. Wilson, R.H. Friend, Adv. Eng. Mater. 4 (2002) 453. [17] H.Y. Chao, W. Lu, Y. Li, M.C.W. Chan, C.M. Che, K.K. Cheung, N. Zhu, J. Am. Chem. Soc. 124 (2002) 14696. [18] W. Lu, W.M. Kwok, C. Ma, C.T.L. Chan, M.X. Zhu, C.M. Che, J. Am. Chem. Soc. 133 (2011) 14120. [19] C. Ma, C.T.L. Chan, W.M. Kwok, C.M. Che, Chem. Sci. 3 (2012) 1883. [20] A.D. Becke, J. Chem. Phys. 98 (1993) 5648. [21] J.P. Perdew, Y. Wang, Phys. Rev. B 45 (1992) 13244. [22] J.P. Perdew, J.A. Chevary, S.H. Vosko, K.A. Jackson, M.R. Pederson, D.J. Singh, C. Fiolhais, Phys. Rev. B 46 (1992) 6671. [23] J.P. Perdew, in: P. Ziesche, H. Eschrig (Eds.), Electronic Structure of Solids’91, Akademie Verlag, Berlin, 1991, p. 11. [24] R. Bauernschmitt, R. Ahlrichs, Chem. Phys. Lett. 256 (1996) 454. [25] M.E. Casida, C. Jamorski, K.C. Casida, D.R. Salahub, J. Chem. Phys. 108 (1998) 4439. [26] R.E. Stratmann, G.E. Scuseria, M.J. Frisch, J. Chem. Phys. 109 (1998) 8218. [27] Y. Zhao, D.G. Truhlar, Theor. Chem. Acc. 120 (2008) 215. [28] B. Mennucci, J. Tomasi, J. Chem. Phys. 106 (1997) 5151. [29] W.R. Wadt, P.J. Hay, J. Chem. Phys. 82 (1985) 284. [30] P.J. Hay, W.R. Wadt, J. Chem. Phys. 82 (1985) 299. [31] P.J. Hay, W.R. Wadt, J. Chem. Phys. 82 (1985) 270. [32] P.C. Hariharan, J.A. Pople, Theor. Chim. Acta 28 (1973) 213. [33] P. Celani, H.J. Werner, J. Chem. Phys. 112 (2000) 5546. [34] T.H. Dunning, J. Chem. Phys. 90 (1989) 1007. [35] B.O. Roos, The complete active space self-consistent field method and its applications in electronic structure calculations, Advances in Chemical Physics: Ab Initio Methods in Quantum Chemistry, vol. 69, John Wiley & Sons, Inc., New York, 1987, p. 399. [36] W.J. Hehre, R. Ditchfield, J.A. Pople, J. Chem. Phys. 56 (1972) 2257. [37] M.J. Frisch, G.W. Trucks, H.B. Schlegel, G.E. Scuseria, M.A. Robb, J.R. Cheeseman, G. Scalmani, V. Barone, B. Mennucci, G.A. Petersson, H. Nakatsuji, M. Caricato, X. Li, H.P. Hratchian, A.F. Izmaylov, J. Bloino, G. Zheng, J.L. Sonnenberg, M. Hada, M. Ehara, K. Toyota, R. Fukuda, J. Hasegawa, M. Ishida, T. Nakajima, Y. Honda, O. Kitao, H. Nakai, T. Vreven, J.A. Montgomery Jr., J.E. Peralta, F. Ogliaro, M. Bearpark, J.J. Heyd, E. Brothers, K.N. Kudin, V.N. Staroverov, R. Kobayashi, J. Normand, K. Raghavachari, A. Rendell, J.C. Burant, S.S. Iyengar, J. Tomasi, M. Cossi, N. Rega, J.M. Millam, M. Klene, J.E. Knox, J.B. Cross, V. Bakken, C. Adamo, J. Jaramillo, R. Gomperts, R.E. Stratmann, O. Yazyev, A.J. Austin, R. Cammi, C. Pomelli, J.W. Ochterski, R.L. Martin, K. Morokuma, V.G. Zakrzewski, G.A. Voth, P. Salvador, J.J. Dannenberg, S. Dapprich, A.D. Daniels, O. Farkas, J.B. Foresman, J.V. Ortiz, J. Cioslowski, D.J. Fox, Gaussian 09, Revision A.02, Gaussian Inc., Wallingford CT, 2009. [38] Molpro 2009.1 is a package of ab initio programs written by H.-J. Werner, P.J. Knowles with contributions from F.R. Manby, M. Schütz, P. Celani, G. Knizia, T. Korona, R. Lindh, A. Mitrushenkov, G. Rauhut, T.B. Adler, R.D. Amos, A. Bernhardsson, A. Berning, D.L. Cooper, M.J.O. Deegan, A.J. Dobbyn, F. Eckert, E. Goll, C. Hampel, A. Hesselmann, G. Hetzer, T. Hrenar, G. Jansen, C. Köppl, Y. Liu, A.W. Lloyd, R.A. Mata, A.J. May, S.J. McNicholas, W. Meyer, M.E. Mura, A. Nicklaß, P. Palmieri, K. Pflüger, R. Pitzer, M. Reiher, T. Shiozaki, H. Stoll, A.J. Stone, R. Tarroni, T. Thorsteinsson, M. Wang, A. Wolf, see .
Contents lists available at ScienceDirect
Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy journal homepage: www.elsevier.com/locate/saa
Dual emission behavior of phenyleneethynylene gold(I) complexes dictated by intersystem crossing: A theoretical perspective Li Wang a, Yuanyuan Li a, Yanxin Zhang a, Hongqing He b, Jinglai Zhang a,⇑ a
Institute of Environmental and Analytical Sciences, College of Chemistry and Chemical Engineering, Henan University, Kaifeng 475004, PR China Wuhan Center for Magnetic Resonance, State Key Laboratory of Magnetic Resonance and Atomic and Molecular Physics, Wuhan Institute of Physics and Mathematics, Chinese Academy of Science, Wuhan 430071, PR China b
h i g h l i g h t s
g r a p h i c a l a b s t r a c t
Calculations for model molecules
Presentation of the photo-induced decay mechanism for (a) 1a0 and (b) 2a0 suggested by the present quantum-chemical calculations. (Abs: Absorption; PF: Prompt fluorescence; DF: Delayed fluorescence; Ph: Phosphorescence; VR: Vibrational relaxation; ISC: Intersystem crossing; IC: Internal conversation; S/T-SC: Surface intersection on the crossing seam of singlet-triplet excited PES).
prove to be reliable by comparison with experimental values. The nature of absorption and dual emission spectra is analyzed. The consecutive decrease of kISC in experiment is correlated with DEST and length of PE ligand. The nonadiabatic decay pathway after photoexcitation is described in detail.
a r t i c l e
i n f o
Article history: Received 13 May 2014 Received in revised form 6 August 2014 Accepted 23 August 2014 Available online 30 August 2014 Keywords: Gold(I) complexes ISC Dual emission Delayed fluorescence TD-DFT CASSCF
a b s t r a c t In commonly studied gold(I) complexes with oligo (o-, p-, or m-phenyleneethynylene) (PE) ligands, an intriguing photophysical behavior is dual emission composed of fluorescence from S1 and phosphorescence from T1 which is dictated by effective intersystem crossing (ISC) process. In order to explore the salient photodynamics of such oligo-PE gold(I) complexes effectively, we have deliberately chosen three model complexes, namely, PhAC„CAAu(PMe3) (1a0 ) and PhAC„CA(1,m)C6H4AC„CAAu(PMe3) (m = 4, 2a0 ; m = 3, 3a0 ) in place of the real system. Firstly, electronic structure methods based on DFT and TD-DFT are utilized to perform optimization calculations for the ground- and lowest-lying excited states, respectively. Next, basic photophysical properties including absorption and emission spectra are investigated by TD-DFT under the optimized geometries. Besides, on the basis of the electronic spectra herein, we succeed in searching for surface intersections as the minima on the seam of singlet–triplet surface crossings (SCs) at the CASSCF level of theory. By integration of the results available, the process of delayed fluorescence of triplet–triplet annihilation (TTA) and phosphorescence was displayed in detail with SCs playing the lead in monitoring the ISC. Ó 2014 Elsevier B.V. All rights reserved.
⇑ Corresponding author. Tel./fax: +86 371 23881589. E-mail address: [email protected] (J. Zhang). http://dx.doi.org/10.1016/j.saa.2014.08.063 1386-1425/Ó 2014 Elsevier B.V. All rights reserved.
260
L. Wang et al. / Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy 137 (2015) 259–266
Introduction Exploitation of high-performance materials for organic lightemitting diodes (OLEDs) has invited a surge of efforts due to their advantages over current technologies in terms of luminescent properties, power consumption, and shape and size variability [1–7]. Wide-ranging applications in organic optoelectronics [8–10] and intriguing photophysical changes by appropriate adjustment of substituent patterns [11–13] for molecular materials based on oligo- or poly-phenyleneethynylene (PE) have rendered them potential candidates for fabricating effective OLEDs. To achieve the maximum device efficiency, an immediate requirement is exploitation of luminescent materials capable of converting excitons, particularly for the spin-forbidden triplet ones, into available light output based on the previous assumption that both spin singlet and triplet excitons account for the overall electroluminescence (EL) efficiency of OLEDs in the ratio of 1:3 [14–16]. However, no triplet emissions happen to arylacetylenes under ambient conditions from the photophysical point of view [17], which constitutes the major obstacle to further exploration for improved EL efficiency. A workable and facile strategy to address this challenge lies in complexation with heavy metal core, which serves as a key element propitious for intersystem crossing (ISC) and subsequent T1 ? S0 radiative deactivation rate. Comparatively speaking, gold-based organometallic complexes are a sagacious choice for the triplet harvesting dopant because of the twofold factors as follows: (1) As a heavy-metal atom, the incorporation of gold(I) moieties into the main chains of p-conjugated oligo- or poly-arylacetylide compounds is supposed to promote spin–orbit coupling, thus breaking the conventional spin selection rules and facilitating singlet ? triplet intersystem crossing (ISC); (2) The absence of low-lying d-d excited states due to d10 closed-shell configuration in gold(I) ion tends to maintain the p-conjugation of PE units and get rid of the undesirable competing nonradiative decay channel. In this regard, Che and coworkers have designed and synthesized a family of mono- and binuclear Cy3Psupported gold(I) complexes coordinated by various conjugated or cross-conjugated oligo-(p-, o-, or m-PE) in succession [17,18].
Characterization results through multiple spectral techniques indicate that most of the complexes strikingly feature dual emission composed of fluorescence from S1 and phosphorescence from T1. What is more, long-lived delayed fluorescence (DF) with TTA mechanism relative to short-lived prompt fluorescence (PF) is observed for complexes bearing extended arylacetylide units. The excited state dynamics analyses further demonstrated the controlled manipulation role by p-conjugation of the singlet excitedstate PE ligand in the exact ISC rate constant (kISC) rather than the heavy-atom effect [19]. Indisputably, the preceding laboratorial efforts have speculated on the formation mechanism of DF and resolved the origin of the dual emissions displayed by the designed PE-based gold(I) complexes using combined methods of steady-state and ultrafast dynamic spectroscopic measurements. However, for one thing, no structural information on ground and excited states involved in absorption and emission spectra is accessible. For another, we are curious about the exhaustive process for the occurrence of delayed fluorescence and normal emission behavior. Consequently, from the theoretical perspective we have in the present work extended our geometrical and spectral study to the gold(I)-PE complexes 1a–3a [19], the schematic structures of which are depicted in Scheme 1, with the aims: (1) to interpret the nature and onset of absorption and dual emission spectra; (2) to locate the surface crossings bridging spin singlet and triplet excited states in the procession of ISC for mechanism analysis of DF and phosphorescence; (3) to rationalize the experimentally observed consecutive decrease of ISC rate constant (kISC) in the sequence of 1a ? 3a ? 2a. What is noteworthy is that we have actually investigated the methyl-substituted models instead of the original systems 1a–3a, namely, PhAC„CAAu(PMe3) (1a0 ) and PhAC„CA(1,m)C6H4 AC„CAAu(PMe3) (m = 4, 2a0 ; m = 3, 3a0 ), which retains the conjugated character of 2a and 3a that is essential for the electronic excitation. This simplification of cutting down one phenyleneethynylene unit saves computational costs and proves to be reasonable by comparison with both experimental spectra and calculations for real systems 1a–3a. Hopefully, the theoretical results and sketched outline for deactivation channels are significant for insight into the
Scheme 1. The schematic structures for real systems 1a–3a and model molecules 1a0 –3a0 .
L. Wang et al. / Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy 137 (2015) 259–266
structure–property relationship of complexes under study and further fabrication of materials with desirable luminescence efficiency. Computational details The density functional theory (DFT) with B3PW91 [20–23] method and time-dependent DFT (TD-DFT/B3PW91) [24–26] calculations have been employed to obtain the equilibrium geometries of complexes 1a0 –3a0 in their ground and excited states, respectively. The ultraviolet absorption spectra and emission spectra have been determined by means of TD-DFT method with the B3PW91 and M06-2X [27] functionals based on the optimized ground-state (S0) and excited-state (S1 for fluorescence and T1 for phosphorescence) structures. The polarized continuum model (PCM) [28] with dichloromethane is employed to assess aqueous solvatochromic effects on the spectra. Owing to the large number of electrons and meanwhile to account for relativistic effects, the double-f quality basis set (LANL2DZ) [29–31] with inner electrons of the gold(I) atom substituted by Hay and Wadt’s effective core potential (ECP) and a 6-31G(d) [32] basis set on all the nonmetal atoms, have been adopted for both geometric optimizations and spectral estimations. The electronic spectra for original real system 1a–3a have been calculated as well. Moreover, the second-order perturbation theory (CASPT2) [33] method with the same LANL2DZ basis set on gold(I) atom and the Dunning’s valence double-f (ccpVDZ) [34] basis sets on carbon, hydrogen, and phosphor have been utilized in the elaborate calculations for vertical excitation energies under the initial B3PW91 equilibrium geometries. A level-shift technique with an imaginary shift of 0.3 au is applied to remove the intruder states in the CASPT2 theory. Regarding the active space, eight electrons in six orbitals are chosen for the complex 1a0 and four electrons in four orbitals chosen for 2a0 –3a0 , referred to as CAS(8,6) and CAS(4,4), respectively. For the purpose of investigating the role played by ISC in the excited-state decay process, we have carefully searched for the singlet–triplet PESs crossing points using the complete active space self-consistent field (CASSCF) [35] approach along with eight electrons distributed in six orbitals, namely, CAS(8,6), as the active space. Herein, the state-averaging CASSCF (SA-CASSCF) calculations over selected states for each species on an individual basis were conducted. And meanwhile a mixed basis set, LANL2DZ for gold(I), 631G [36] for hydrogen, coupled with 6-31G(d) for carbon and phosphor atoms, has been implemented into the CASSCF computation. DFT and TD-DFT calculations were performed with the Gaussian 09 [37] program package. The CASPT2 predictions for excitation energies and SA-CASSCF treatment for SCs were performed by the Molpro 2009 [38] program. Results and discussion The ground-state structures and absorption spectra in the CH2Cl2 solution The optimized ground-state geometries of model system [PhAC„CAAu(PMe3)] (1a0 ), [PhAC„CA(1,4)C6H4AC„CAAu(PMe3)] (2a0 ), and [PhAC„CA(1,3)C6H4AC„CAAu(PMe3)] (3a0 ) at the B3PW91 level of theory are depicted in Fig. 1 with selected atoms labeled. As shown in Table 1, the ligand length (1a0 vs 2a0 /3a0 ) or substituent position (2a0 vs 3a0 ) makes no difference to the key bond distances with CAP, PAAu, and AuAC kept around 1.83, 2.34, and 2.00 Å. The dihedral angles DC3APAC6AC7 in complexes 1a0 and 3a0 are close to 180°, indicating that C3AP stays within the PE plane. Inspection of the other two dihedral angles associated with C1AP and C2AP immediately reveals that both bonds keep mirror images to each other with respect to the symmetry plane passing the C3AP
261
bond and the PE ligand. However, the case in complex 2a0 is slightly different because of less effective coplanarity between C3AP and PE ligand compared with that in 1a0 and 3a0 . In addition, it is found that the torsional angle between two adjacent ethynylene moieties in 2a0 and 3a0 is 0° and 3.35°, respectively, giving evidence that p-conjugated framework is relatively more easily kept in para-PE compounds than that in the meta position. To correlate the ground-state geometrical differences in complexes 1a0 –3a0 with the absorption spectra, we have compiled the lowest and also the most intense singlet absorption bands together with the triplet excited states isoenergetic with the singlet-spin absorption by different methods in Table 2. Even the most intense singlet absorption was assigned as different excitation (S1 or S2) by different methods due to the inherent character of the methodology, they result from the similar electron transitions of HOMO ? LUMO or H1 ? LUMO and display the same character. Direct comparison with the experimental absorption bands [19] and predicted spectra for the real system 1a–3a (See Table S1 in Supplementary Information) attaches credence to the selected theoretical approaches to understand the spectroscopic properties. Additionally, with respect to the TD-DFT results, an overall blue shift in TD-M06-2X calculation compared with the absorptions in TD-B3PW91 is clearly discovered from the simulated spectra depicted in Fig. S1. On account of the fact that one single-electron excitation, namely, HOMO ? LUMO or H1 ? LUMO, contributes to the first singlet (S1 or S2) excited state in general, it is essential to understand the compositions constituting the frontier molecular orbitals (FMOs), which have been displayed in Fig. 2. In terms of population analysis, the HOMOs for three complexes are similar in character with basically commensurate contribution from pC„C and pPh, while, p⁄C„C character has been reduced and p⁄Ph enhanced in LUMOs. As a result, one can easily figure out that the maximum absorption bands should be ascribed to p ? p⁄ transition designated as ligand centered (LC) character, as the composition from gold(I) d orbital (denoted as dAu) is almost zero. Importantly, the calculated lowest-lying transition energies follow the order 1a0 > 3a0 > 2a0 , which is in accordance with the HOMO–LUMO energy gap listed in Fig. 2 as well as the aforementioned conjugation degree of 1a0 < 3a0 < 2a0 determined by geometrical information. It is expected that the relationship between electronic and geometric structures with spectroscopic nature should be adaptable to other similar metal-containing arylacetylide complexes. The excited-state structures and emission spectra in the CH2Cl2 solution In view of the geometric parameters for the lowest-lying singlet states (S1) of title complexes in Table 1, electron promotion obviously has no influence on the bond lengths of PMe3 moiety, while, it does on the degree of geometric distortion with the order of 2a0 < 3a0 < 1a0 by comparison of dihedral angle changes going from S0 to S1 states. In addition, AuAC4 bond of three complexes is slightly enhanced to various extents after photoexcitation. As has been unambiguously illustrated for complexes 1a–3a in experiment [17,19], dual emission including fluorescence and phosphorescence was observed with the exception for 1a, in which the prompt fluorescence (PF) decays rapidly with time evolution. For deep insight into the onset and nature of the emission spectra, vertical transition energies originating from S1 and T1 states, respectively, have been calculated by the TD-DFT level under homologous geometries. Fluorescent spectra The singlet–singlet emission wavelengths, oscillator strengths, configuration interaction (CI) coefficient, transition character, and
262
L. Wang et al. / Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy 137 (2015) 259–266
Fig. 1. Optimized ground-state structures for the model systems [PhAC„CAAu(PMe3] (1a0 ), [PhAC„CA(1,4)C6H4AC„CAAu(PMe3)] (2a0 ), and [PhAC„CA(1,4)C6H4 AC„CAAu(PMe3)] (3a0 ) under B3PW91 level of theory.
Table 1 Selected geometric parameters of the ground and excited states for model systems [PhC„CAu(PMe3] (1a0 ), [PhC„C(C6H4-l, 4-C„C)Au(PMe3)] (2a0 ), and [PhC„C(C6H4-l, 3-C„C)Au(PMe3)] (3a0 ) under B3PW91 and TD-B3PW91 levels of theory, respectively. Parameter
1a0 (S0)
1a0 (S1)
1a0 (T1)
2a0 (S0)
2a0 (S1)
2a0 (T1)
3a0 (S0)
3a0 (S1)
3a0 (T1)
Bond length (Å) C1AP C2AP C3AP PAAu AuAC4 C4AC5 C5AC6 C6AC7
1.83 1.83 1.83 2.34 2.00 1.23 1.43 1.41
1.84 1.84 1.83 2.34 1.96 1.27 1.37 1.45
1.83 1.83 1.83 2.34 1.99 1.27 1.36 1.48
1.83 1.83 1.83 2.34 2.00 1.23 1.42 1.41
1.83 1.83 1.83 2.34 1.98 1.25 1.39 1.44
1.83 1.83 1.83 2.34 1.99 1.25 1.39 1.45
1.83 1.83 1.83 2.34 2.00 1.23 1.43 1.40
1.83 1.83 1.83 2.34 1.99 1.23 1.42 1.39
1.83 1.83 1.83 2.34 2.00 1.23 1.42 1.39
64.4 55.7 176.0
72.2 48.2 167.9
64.3 55.8 176.1
48.3 71.9 168.1 0.02
48.5 71.9 168.1 0.03
48.4 72.0 168.1 0.01
60.3 59.5 180.0
61.1 58.5 179.1
60.9 58.8 179.3
3.35
0.17
0.05
Dihedral angle (°) C1APAC6AC7 C2APAC6AC7 C3APAC6AC7 C8AC9AC10AC11 C7AC8AC9AC10
263
L. Wang et al. / Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy 137 (2015) 259–266
Table 2 Selected absorption wavelengths (k) based on vertical excitation energies of complexes 1a0 –3a0 in dichloromethane under both TD-DFT and CASPT2 levels of theory, together with experimental valuesa. Species
TD-M06-2X k (nm)
f
Configuration
State
k (nm)
f
Configuration
1a0
T4
289
0.0000
T2
267
0.0000
H1 ? L(69%)
S1
283
0.6617
H ? L+2(67%) H2 ? L(25%) H ? L(93%)
S1
266
0.6831
H 1 ? L(57%)
2a0
T2
382
0.0000
T2
329
0.0000
H 1 ? L(48%)
S1
352
1.9252
H 2 ? L(47%) H ? L+1(27%) H ? L(98%)
S1
316
1.9968
H ? L(94%)
S1 T5 S2
313 304 304
0.3186 0.0000 1.0075
H ? L(73%) H ? L+1(47%) H1 ? L(78%)
T4 S1
281 280
0.0000 1.2870
H5 ? L(33%) H ? L(70%)
0
3a
a
TD-B3PW91 State
kExpt (nm)
Species
CASPT2 State
k (nm)
f
Configuration
1a0
T2 S2
280 281
0.0000 0.5104
H 1 ? L(99%) H ? L(91%)
282
2a0
T2
335
0.0000
S1
325
2.0473
H ? L+1 (50%) H1 ? L (46%) H ? L(89%)
342
3a0
T3
301
0.0000
S2
272
1.3410
H1 ? L(54%) H1 ? L+1(31%) 286
Experimental values from Ref. [19].
experimental values are listed in Table 3. The population analysis for frontier molecular orbitals of the complexes 1a0 –3a0 in their S1 state are depicted in Fig. S2(a) of Supplementary material. An inspection of the figure reveals that 2a0 possesses the most balanced orbital contribution from pC„C and pPh in HOMO among three complexes and meanwhile the most narrow HOMO–LUMO energy gap. Naturally, it is easy to understand the fluorescence wavelength and intensity under TD-B3PW91 level of theory in the order of 2a0 (399 nm) > 3a0 (352 nm) > 1a0 (314 nm) based on an overall consideration of p-conjugation monitored by orbital population and geometry variation. Furthermore, the p-featured HOMO and p⁄-featured LUMO orbitals determine the ligandcentered (LC) S1 1pp⁄ electronic transition accompanied by predominant electron promotion and charge migration.
Phosphorescent spectra The structural changes of three complexes in their lowestenergy triplet excited state T1 relative to the S0 state behaves in the similar fashion with that in S1 according to the optimized data shown in Table 1. As a result, geometries might produce influence on the phosphorescence spectra in the same way as S1 structures does on the fluorescence spectra. As can be seen from Table 4, the TD-M06-2X calculated emission spectra for 1a0 –3a0 in CH2Cl2 media occur at 473, 599, and 522 nm, respectively, in better agreement with the experimental measurements than TD-B3PW91 results and all with the one-electron HOMO ? LUMO excitation configuration. Since no spin–orbital coupling (SOC) effect is included in current TD-DFT calculations, the oscillator strengths for phosphorescent emissions are zero. Basically, the triplet
Fig. 2. Molecular orbital compositions, energy gaps, and contour plots for HOMO and LUMO of 1a0 –3a0 at the ground state S0.
264
L. Wang et al. / Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy 137 (2015) 259–266
Table 3 Fluorescence emissions of 1a0 –3a0 in dichloromethane based on different TD-DFT results, together with experimental valuesa. Method
State
kF (nm)
f
Main configuration
Character
kExpt (nm) F
0
1a
TD-B3PW91 TD-M06-2X
S1 S2
314 294
0.7216 0.7391
H ? L(92%) H ? L(92%)
ILCT ILCT
304
2a0
TD-B3PW91 TD-M06-2X
S1 S1
399 363
2.0460 2.0482
H ? L(98%) H ? L(95%)
ILCT ILCT
390
3a0
TD-B3PW91 TD-M06-2X
S1 S1
352 325
1.0783 1.3071
H ? L(95%) H ? L(90%)
ILCT ILCT
340
Species
a
Experimental values from Ref. [19].
Table 4 Phosphorescent emissions of 1a0 –3a0 in dichloromethane based on different TD-DFT results, together with experimental valuesa. Species
Method
State
kPh (nm)
f
Main configuration
Assignment
kExpt (nm) Ph
1a0
TD-B3PW91 TD-M06-2X
T1 T1
574 473
0.0000 0.0000
H ? L(92%) H ? L(95%)
3
ILCT 3 ILCT
420
2a0
TD-B3PW91 TD-M06-2X
T1 T1
748 599
0.0000 0.0000
H ? L(92%) H ? L(92%)
3
550
TD-B3PW91 TD-M06-2X
T1 T1
650 522
0.0000 0.0000
H ? L(92%) H ? L(90%)
3
3a0 a
ILCT ILCT
3
ILCT ILCT
470
3
Experimental values from Ref. [19].
emissions should be described as 3ILCT(3pp⁄) with charge significantly transferring between C„C and phenyl moieties due to the absent composition of d orbital from gold atom (dAu) and p orbital from the PMe3 group (pPMe3) for three complexes, which is determined by combination of Table 4 and Fig. S2(b). It deserves our notification that the composition analysis for transitional orbital suggests none contribution from dAu to both fluorescent and phosphorescent spectra, which otherwise indicates the liganddependant emissive nature for gole(I)-PE complexes. Additionally, our findings demonstrate that proper adjustment of the substitution pattern of the PE ligand serves as an effective strategy to tune the emitting color of similar PE-supported metal complexes. Overall, reasonable agreement between theoretical calculations and experimental values for both fluorescence and phosphorescence spectra effectively authenticate the reliability of the model molecules and methods employed in this contribution. For better evaluation of the general luminescence efficiency of title complexes, we have adopted the concept of relaxation energy (kS and kT) with reference to Ref. [14], which plays the same role in reflecting the degree of geometry relaxation from S1 to S0 state and thereby the photoluminescence quantum efficiency as Stokes shift does. As is denoted in Fig. 3, DES@S0 represents the vertical excitation energy from S0 ? S1 and DES@S1 is the vertical emission energy rising from S1. The smaller kS or kT is indicative of less distortion in the excited-state geometry, thus suppressing the nonradiative decay channel and resulting in the higher PL efficiency. As shown in Table S2, the kS and k⁄S values based on B3PW91/ TD-B3PW91 calculations vary at relatively low level, which may supply a new theoretical clue that PE-based gold(I) complexes may be considered as potential systems for fabrication of effective OLEDs.
excited state Tn in close proximity to the most intense and lowest-lying singlet state in the Franck–Condon region with the calculated energy deference, referred to as DEST hereafter, for 1a0 –3a0 lying at 0.02, 0.48, and 0.44 eV, respectively (For TD-B3PW91/TDM06-2X calculations, the values severally correspond to 0.10/ 0.03, 0.28/0.16, and 0.00/0.01 eV), which greatly favors the search for SCs. Fortunately, we have successfully located the SCs minima corresponding to 1a0 –3a0 on the crossing seam of the singlet–triplet PESs using state-averaging CASSCF methodology. The common points among three SCs shown in Table 5 consist in the averaging tendency of bond lengths with the extension of the acetylene link C4AC5 and contraction of the single bond C5AC6, which might favor the charge transfer between acetylide and phenyl groups. The determinate existence of SCs effectively verifies the prevailing occurrence of intersystem crossing (ISC) on the excited-state PES and hence excess thermal energy is released to the Tn state. With respect to the consecutively decreased ISC rate constant (kISC) by 1a ? 3a ? 2a as detected in experiment, it should be directly
Surface crossings and non-adiabatic deactivation patterns on the excited-state PESs It is well known that a SC usually makes it facile for the photoinduced dynamic process with participation of excited-state PES, and therefore the accurate location of the SC connecting two close PESs is meaningful for insight into correlation between 1pp⁄ and 3 pp⁄ structures in luminescent behavior. An examination of Table 2 once more suggests that there exists an energy-degenerate triplet
Fig. 3. Sketch map for the potential energy surfaces, vertical transition energies (DES@S0 and DES@S1) and relaxation energies (kS and k⁄S) for the S1 state.
L. Wang et al. / Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy 137 (2015) 259–266 Table 5 CASSCF optimized key geometric parameters of the surface intersections on the excited-state PES for complexes 1a0 –3a0 . Parameter
SC-1a0 (S2/T3)
SC-2a0 (S1/T3)
SC-3a0 (S2/T3)
Bond length (Å) C1AP C2AP C3AP PAAu AuAC4 C4AC5 C5AC6 C6AC7
1.84 1.84 1.84 2.35 2.02 1.25 1.39 1.42
1.83 1.83 1.83 2.36 2.04 1.29 1.36 1.43
1.84 1.84 1.83 2.35 2.02 1.25 1.38 1.42
Dihedral angle (°) C1APAC6AC7 C2APAC6AC7 C3APAC6AC7
60.6 60.6 180.0
48.9 71.7 168.4
59.9 62.3 178.2
related to the variation tendency of DEST-value. increasing p-conjugation with the sequence of should also contribute to the variation of kISC, conjugation promotes fluorescence and otherwise phosphorescence.
Besides, the 1a0 < 3a0 < 2a0 as effective weakens the
265
In view of the present theoretical findings and taking the previous experimental observations into account, the nonadiabatic decay pathway composed of radiative transition and nonradiative transition after irradiation by ultraviolet light is schematically proposed. Due to the mechanism similarity between 2a0 and 3a0 , the photo-induced experience for 1a0 and 2a0 are sketched as representatives in Fig. 4. Upon irriadition into the singlet excited state with 1 pp⁄ character, partial excited-state population relaxes rapidly out of the Franck–Condon region toward the equilibrium on the S1 PES and then decays back to the ground state via regular prompt fluorescence (PF) obeying Kasha’s rule. However, owing to the small energy difference DEST, a big proportion possessing enough energy quickly reaches the 1pp⁄/3pp⁄-SC, by which the photogenerated 1 pp⁄ population is promptly funneled into 3pp⁄ by means of ISC with complex dynamics on the excited-state landscape. Successively, a fast internal conversation (IC) on the triplet excited-state PES takes place, causing the electron populated on the energy-lowest T1 state. Hereafter, different decay channels happen to complexes 1a0 and 2a0 . As for the former, direct phosphorescence emission is observed. While for the latter, alternative deactivation pathways occur with the accumulation of T1 molecules: (1) a triplet ? singlet up-conversion via TTA process is largely enhanced with S1 excitons formed again and consequently the delayed
Fig. 4. Presentation of the photo-induced decay mechanism for (a) 1a0 and (b) 2a0 suggested by the present quantum-chemical calculations. (Abs: Absorption; PF: Prompt fluorescence; DF: Delayed fluorescence; Ph: Phosphorescence; VR: Vibrational relaxation; ISC: Intersystem crossing; IC: Internal conversation; S/T-SC: Surface intersection on the crossing seam of singlet–triplet excited PES).
266
L. Wang et al. / Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy 137 (2015) 259–266
fluorescence (DF) as experimentally detected; (2) the rest T1 population finds its way back to the ground state S0 through normal radiative transition and gives rise to the phosphorescence. To sum up, the SC-induced intersystem crossing plays an important role in the intriguing photophysical decay process of the title gold(I) complexes. And meanwhile the kISC variation is correlated with the length of the PE ligand and DEST. Conclusion The gold(I) complexes containing oligo(p- or m-phenyleneethynylene) (PE) fragments represent potential systems for highperformance OLEDs fabrication because of the dual emission by efficiently harvesting both singlet and triplet excitons. By combined methods of DFT, TD-DFT, coupled with CASSCF and CASPT2, we have theoretically investigated the distinctive photophysical behavior that is different from conventional metal complexes. In the following, we briefly address three conclusions derived from the present work. First, the evaluations for spectra agree well with the experimental observations, which in turn certify the predictions for the geometric structures. Equilibrium geometries of the complexes under study are slightly relevant to the PE oligomers with different lengths and different substitution patterns. Second, the S0 ? S1 and S1/T1 ? S0 transitions feature by remarkable PE ligand-centered (LC) (1pp⁄ or 3pp⁄) electron promotion and charge transfer. The relaxation energies (kS/kT) on the basis of vertical excitation and emission are evaluated to be small, which accordingly reduces the competing nonradiative decay probability and enhances the potential of such systems as lightemitting materials. Third, The detected ISC rate constant (kISC) in the order of 1a > 3a > 2a should be correlated to the conjugation-enhanced fluorescent intensity of 2a0 > 3a0 > 1a0 and as well the energy difference DEST in the sequence of 1a0 < 3a0 < 2a0 . The existence of surface intersections linking the excited singlet and triplet PESs favors ISC, which further sheds light on the mechanism manipulating the phosphorescence and TTA-induced delayed fluorescence. Hopefully, the structure–property relationship revealed in our quantum chemical investigation may on the one hand provide useful information on structures and spectra for similar PE-based complexes and on the other hand contribute to improved design and optimization of high-efficiency OLEDs worthy of experimental testing. Acknowledgements We thank the Beijing Key Laboratory of Ionic Liquids Clean Process and State Key Laboratory of Multiphase Complex Systems, Institute of Process Engineering, Chinese Academy of Sciences for providing computational resources. This work was supported by the Natural Science Foundation of He’nan Province of China (134300510008, 144300510032), Science Foundation of Henan Province (14A150034), and the Foundation for University Key Teachers from the He’nan Educational Committee. Appendix A. Supplementary material Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.saa.2014.08.063.
References [1] M.C. Gather, A. Köhnen, K. Meerholz, Adv. Mater. 23 (2011) 233. [2] M. Wohlgenannt, K. Tandon, S. Mazumdar, S. Ramasesha, Z.V. Vardeny, Nature 409 (2001) 494. [3] J.S. Wilson, A.S. Dhoot, A.J.A.B. Seeley, M.S. Khan, A. Köhler, R.H. Friend, Nature 413 (2001) 828. [4] L.C. Lin, H.F. Meng, J.T. Shy, S.F. Horng, L.S. Yu, C.H. Chen, H.H. Liaw, C.C. Huang, K.Y. Peng, S.A. Chen, Phys. Rev. Lett. 90 (2003) 036601. [5] M. Segal, M.A. Baldo, R.J. Holmes, S.R. Forrest, Z.G. Soos, Phys. Rev. B 68 (2003) 075211. [6] T.K. Hatwar, M.E. Kondakova, D.J. Giesen, J.P. Spindler, Inorganic electronics: materials, processing, devices and applications, in: F. So (Ed.), CRC Press, Boca Raton, FL, 2009, p. 433. [7] M.E. Kondakov, J.C. Deaton, T.D. Pawlik, D.J. Giesen, D.Y. Kondakov, R.H. Young, T.L. Royster, D.L. Comfort, J.D. Shore, J. Appl. Phys. 107 (2010). 014515–113. [8] R.E. Martin, F. Diederich, Angew. Chem., Int. Ed. 38 (1999) 1350. [9] P.J. Stang, F. Diederich, Modern Acetylene Chemistry, VCH, Weinheim, Germany, 1995. [10] G.M. Carter, Y.J. Chen, M.F. Rubner, D.J. Sandman, M.K. Thakur, S.K. Tripathy, in: D.S. Chemla, J. Zyss (Eds.), Non-Linear Optical Properties of Organic Molecules and Crystals, vol. 2, Academic, Orlando, FL, 1987. Chapter III-3. [11] A. Beeby, K. Findlay, P.J. Low, T.B. Marder, J. Am. Chem. Soc. 124 (2002) 8280. [12] J. Kim, I.A. Levitsky, D.T. McQuade, T.M. Swager, J. Am. Chem. Soc. 124 (2002) 7710. [13] J. Cornil, Y. Karzazi, J.L. Brédas, J. Am. Chem. Soc. 124 (2002) 3516. [14] H. Uoyama, K. Goushi, K. Shizu, H. Nomura, C. Adachi, Nature 492 (2012) 234. [15] M.A. Baldo, D.F. O’Brien, Y. You, A. Shoustikov, S. Sibley, M.E. Thompson, S.R. Forrest, Nature 395 (1998) 151. [16] A. Köhler, J.S. Wilson, R.H. Friend, Adv. Eng. Mater. 4 (2002) 453. [17] H.Y. Chao, W. Lu, Y. Li, M.C.W. Chan, C.M. Che, K.K. Cheung, N. Zhu, J. Am. Chem. Soc. 124 (2002) 14696. [18] W. Lu, W.M. Kwok, C. Ma, C.T.L. Chan, M.X. Zhu, C.M. Che, J. Am. Chem. Soc. 133 (2011) 14120. [19] C. Ma, C.T.L. Chan, W.M. Kwok, C.M. Che, Chem. Sci. 3 (2012) 1883. [20] A.D. Becke, J. Chem. Phys. 98 (1993) 5648. [21] J.P. Perdew, Y. Wang, Phys. Rev. B 45 (1992) 13244. [22] J.P. Perdew, J.A. Chevary, S.H. Vosko, K.A. Jackson, M.R. Pederson, D.J. Singh, C. Fiolhais, Phys. Rev. B 46 (1992) 6671. [23] J.P. Perdew, in: P. Ziesche, H. Eschrig (Eds.), Electronic Structure of Solids’91, Akademie Verlag, Berlin, 1991, p. 11. [24] R. Bauernschmitt, R. Ahlrichs, Chem. Phys. Lett. 256 (1996) 454. [25] M.E. Casida, C. Jamorski, K.C. Casida, D.R. Salahub, J. Chem. Phys. 108 (1998) 4439. [26] R.E. Stratmann, G.E. Scuseria, M.J. Frisch, J. Chem. Phys. 109 (1998) 8218. [27] Y. Zhao, D.G. Truhlar, Theor. Chem. Acc. 120 (2008) 215. [28] B. Mennucci, J. Tomasi, J. Chem. Phys. 106 (1997) 5151. [29] W.R. Wadt, P.J. Hay, J. Chem. Phys. 82 (1985) 284. [30] P.J. Hay, W.R. Wadt, J. Chem. Phys. 82 (1985) 299. [31] P.J. Hay, W.R. Wadt, J. Chem. Phys. 82 (1985) 270. [32] P.C. Hariharan, J.A. Pople, Theor. Chim. Acta 28 (1973) 213. [33] P. Celani, H.J. Werner, J. Chem. Phys. 112 (2000) 5546. [34] T.H. Dunning, J. Chem. Phys. 90 (1989) 1007. [35] B.O. Roos, The complete active space self-consistent field method and its applications in electronic structure calculations, Advances in Chemical Physics: Ab Initio Methods in Quantum Chemistry, vol. 69, John Wiley & Sons, Inc., New York, 1987, p. 399. [36] W.J. Hehre, R. Ditchfield, J.A. Pople, J. Chem. Phys. 56 (1972) 2257. [37] M.J. Frisch, G.W. Trucks, H.B. Schlegel, G.E. Scuseria, M.A. Robb, J.R. Cheeseman, G. Scalmani, V. Barone, B. Mennucci, G.A. Petersson, H. Nakatsuji, M. Caricato, X. Li, H.P. Hratchian, A.F. Izmaylov, J. Bloino, G. Zheng, J.L. Sonnenberg, M. Hada, M. Ehara, K. Toyota, R. Fukuda, J. Hasegawa, M. Ishida, T. Nakajima, Y. Honda, O. Kitao, H. Nakai, T. Vreven, J.A. Montgomery Jr., J.E. Peralta, F. Ogliaro, M. Bearpark, J.J. Heyd, E. Brothers, K.N. Kudin, V.N. Staroverov, R. Kobayashi, J. Normand, K. Raghavachari, A. Rendell, J.C. Burant, S.S. Iyengar, J. Tomasi, M. Cossi, N. Rega, J.M. Millam, M. Klene, J.E. Knox, J.B. Cross, V. Bakken, C. Adamo, J. Jaramillo, R. Gomperts, R.E. Stratmann, O. Yazyev, A.J. Austin, R. Cammi, C. Pomelli, J.W. Ochterski, R.L. Martin, K. Morokuma, V.G. Zakrzewski, G.A. Voth, P. Salvador, J.J. Dannenberg, S. Dapprich, A.D. Daniels, O. Farkas, J.B. Foresman, J.V. Ortiz, J. Cioslowski, D.J. Fox, Gaussian 09, Revision A.02, Gaussian Inc., Wallingford CT, 2009. [38] Molpro 2009.1 is a package of ab initio programs written by H.-J. Werner, P.J. Knowles with contributions from F.R. Manby, M. Schütz, P. Celani, G. Knizia, T. Korona, R. Lindh, A. Mitrushenkov, G. Rauhut, T.B. Adler, R.D. Amos, A. Bernhardsson, A. Berning, D.L. Cooper, M.J.O. Deegan, A.J. Dobbyn, F. Eckert, E. Goll, C. Hampel, A. Hesselmann, G. Hetzer, T. Hrenar, G. Jansen, C. Köppl, Y. Liu, A.W. Lloyd, R.A. Mata, A.J. May, S.J. McNicholas, W. Meyer, M.E. Mura, A. Nicklaß, P. Palmieri, K. Pflüger, R. Pitzer, M. Reiher, T. Shiozaki, H. Stoll, A.J. Stone, R. Tarroni, T. Thorsteinsson, M. Wang, A. Wolf, see .
