Theoretical Studies of the Structural, Electronic, and Optical Properties

Feb 11, 2010 - Theoretical study on the electronic structure and optical properties of carbazole-π-dimesitylborane as bipolar fluorophores for nondop...
14 downloads 26 Views 4MB Size
J. Phys. Chem. A 2010, 114, 3655–3667

3655

Theoretical Studies of the Structural, Electronic, and Optical Properties of Phosphafluorenes Jun Yin, Run-Feng Chen,* Sheng-Lan Zhang, Qi-Dan Ling, and Wei Huang* Jiangsu Key Laboratory for Organic Electronics & Information Displays (KLOEID) and Institute of AdVanced Materials (IAM), Nanjing UniVersity of Posts and Telecommunications (NUPT), 9 Wenyuan Road, Nanjing 210046, China ReceiVed: December 7, 2009; ReVised Manuscript ReceiVed: January 21, 2010

Phosphafluorenes have drawn increasing attention recently in the applications of organic electronic devices due to their particular optoelectronic properties. To reveal their molecular structures, optoelectronic properties, and structure-property relationships of the newly emerged functional materials, an in-depth theoretical investigation was elaborated via quantum chemical calculations. The optimized geometric and electronic structures in both ground and exited states, the mobility of the hole and electron, the absorption and emission spectra, and the singlet exciton generation fraction of these novel phosphors-containing materials have been studied by density functional theory (DFT), single excitation configuration interaction (CIS), time-dependent density functional theory (TDDFT) methods, and the polarizable continuum model (PCM). The results show that the highest occupied molecular orbitals (HOMOs), the lowest unoccupied molecular orbitals (LUMOs), triplet energies (3Eg), energy gaps (Eg), as well as some other electronic properties including ionization potentials (IPs), electron affinities (EAs), reorganization energies (λ), the singlet exciton generation fraction, radiative lifetime, and absorption and emission spectra can be easily tuned by chemical modifications of the phosphorus atom via methyl, phenyl, oxygen, sulfur, or selenium substitution, indicating that the phosphafluorenes are interesting optoelectronic functional materials, which have great potential in the applications of OLEDs, organic solar cells, organic storage, and sensors. 1. Introduction Organic π-conjugated optoelectronic functional materials have attracted much attention in recent years due to their facilities in structure-property modifications and potential applications in various electronic devices including organic light-emitting diodes (OLEDs),1 solar cells,2 nonlinear optical devices (NLOs),3 and sensors.4 The incorporation of heteroatoms into the π-conjugated backbone of these organic functional materials is an effective way to tune their optoelectronic properties.5 Recently, phosphorus has found many applications in chemical modifications of π-conjugated molecules6-9 and shown considerable promise for the development of new functional materials with novel properties for optoelectronic devices.10 For example, the phosphorus-containing poly(N-arylaniline)s are a new type of π-conjugated polymer with low oxidation potentials and electronic delocalization through phosphorus along the polymer chain.11 The phosphorus analogues of PPVs with PdP linkages along the polymer backbone have interesting electronic and spectroscopic properties.12 Phospholes are currently the most widely used phosphorus-containing building units for optoelectronic materials because of their favored exocyclic delocalization along the π-conjugated backbone.13-15 Dithienophospholes have extraordinary optoelectronic properties, and their electronic structures can be easily tuned by simple chemical modication.16-18 These intriguing features of phosphorus-containing (especially phosphole-containing) materials strongly support the incorporation of phosphorus atoms into π-conjugated molecules to obtain new functional materials with desired properties. * To whom correspondence should be addressed. Tel: +86 25 8586 6008. Fax: +86 25 8586 6999. E-mail: [email protected] (W.H.); [email protected] (R.-F.C.).

Phosphafluorenes (dibenzophosphole) have been known for many years as catalyst ligands, but their potential applications in optoelectronics have not been exploited until recently. Compared to the widely investigated dithienophospholes, phosphafluorene also possesses the phosphole core but with benzoannulated structure. From another point of view, phosphafluorene can be considered to be formed by replacing the vulnerable C-9 carbon of fluorene (which will lead to color instability of fluorene-based materials) with a phosphorus atom). Like its analogues of phosphole, dithienophosphole, fluorene, silafluorene, germafluorene, and carbazole (nitrogenafluorenes), phosphafluorene has shown interesting and unique optoelectronic properties for devices based on electro- and photoluminescence. Phosphafluorene exhibits a wide energy gap, and its fluorescence spectra are in the ultraviolet region. Its absorption and emission spectra can be tuned by structural modifications of oxygenation, sulfuration, and metal complexation.5 Phosphafluorene-containing copolymers have been prepared by introducing a phosphafluorene monomer into the backbone of the conjugated polymers. For example, a phosphafluorene-phenyl copolymer could emit intense blue-green fluorescence with high quantum yields at 432 and 465 nm in CHCl3 and the solid state, respectively.19 The incorporation of phosphafluorene into polyfluorene can improve the electron-injection and -transport abilities of the copolymer, and both the blue and the white electroluminescence can be realized, although the homopolymer of phosphafluorene exhibits green emission (516 nm) in the solid state.20 To rationalize these experimentally observed properties of known materials and to predict those of unknown ones, theoretical investigations based on quantum chemical calculations on the structural and electronic properties of phosphafluorenes are indispensable.

10.1021/jp911624v  2010 American Chemical Society Published on Web 02/11/2010

3656

J. Phys. Chem. A, Vol. 114, No. 10, 2010

Yin et al.

Figure 1. Molecular structures of the phosphafluorenes.

In this work, the structural, electronic, and optical properties of phosphafluorenes with different substituents on the phosphorus (see Figure 1) were investigated by using density functional theory (DFT), single excitation configuration interaction (CIS), and time-dependent density functional theory (TDDFT) to provide theoretical understandings of these phosphoruscontaining materials. The highest occupied molecular orbitals (HOMOs), the lowest unoccupied molecular orbitals (LUMOs), triplet energies (3Eg), energy gaps (Eg), as well as some other electronic properties including ionization potentials (IPs), electron affinities (EAs), reorganization energies (λ), singlet exciton generation fractions, radiative lifetimes, and absorption and emission spectra were calculated and compared. It was found that phosphafluorenes have great potential as excellent building blocks for the molecular design of π-conjugated systems, and the chemical modification of the phosphorus atom via methyl, phenyl, oxygen, sulfur, or selenium substitution is a powerful method to tune the properties of phosphafluorenes. 2. Theoretical and Computational Methodology 2.1. Internal Reorganization Energy. At present, two kinds of models have been used to evaluate the charge (hole and electron) mobility.21 One is the coherent band model,22-24 and the other is the incoherent hopping model.25-27 In the former, the charges transfer through valence or conduction bands formed by the overlapping molecular orbitals with strong coupling between neighboring molecules. The latter is more suitable for most organic materials, where coupling between neighboring molecules is small, and is thus chosen in this study. In the second model, the charge transport is the intermolecular process in which the charge hops between two molecules, which can be summarized as follows

M+/- + M f M + M+/where M is the neutral molecule interacting with neighboring oxidized or reduced M+/-. The rates of charge transfer can be approximately described by Macus theory28-30 in the following equation

Kh/e )

( ) π λh/ekT

1/2

×

( )

2 Vh/e λh/e × exp p 4kT

where T is the temperature, k and p refer to the Boltzman and Planck constants, respectively, λh/e is the hole/electron reorganization energy due to geometric relaxation accompanying charge transfer, and Vh/e is the electronic coupling matrix element between the two species M and M+/-. The electron-transfer rate depends on two important parameters, electron coupling Vh/e and the reorganization energy λh/e. However, experimentally determined Vh/e shows a rather narrow range of values,25,31 and an even more limited range of Vh/e is expected due to the intermolecular charge-transfer processes considered in OLEDs

Figure 2. Internal reorganization energy for hole and electron transfer.

involving direct contacts in amorphous solids. Therefore, the hole/electron mobility can be dominated by the reorganization energy without regard for the transfer integral. The hole/electron reorganization energy can be defined as follows32

λhole ) λ+ + λ1 ) [E+(A) - E+(A+)] + [E(A+) E(A)] ) IP(V) - HEP λelectron ) λ- + λ2 ) [E-(A) - E-(A-)] + [E(A-) E(A)] ) EEP - EA(V) As illustrated in Figure 2, E+, E, and E- represent the energies of the cation, neutral, and anion species based on their lowest energy geometries, respectively, while (A+), (A), and (A-) denote their optimized structures at corresponding ion states. E+/-(A) is the energy of a cation/anion calculated with the optimized structure of the neutral molecule A. λ+ is the relaxation energy of a neutral molecule A that captured a hole going toward the A+ optimum geometry on the potential energy surface of A+, and λ1 is the relaxation energy from A+ extracting a hole going toward the A optimum geometry on the potential energy surface of A. The sum of λ+ and λ1 is the hole reorganization energy λhole. Similarly, in the electron-transport process, λelectron ) λ- + λ2. 2.2. Singlet Exciton Generation Fraction. Recently, the experimental maximum internal quantum efficiency (IQE) of OLEDs has exceeded the statistical upper limit of 25%, which is believed to be due to the underestimated singlet exciton generation fraction (χS), which can be described in the following equation33

χS )

σS σS + 3σT

and the ratio σS/σT of the formation cross section of the singlet and triplet is

EbT σS ) σT EbS where σS and σT represent the formation cross sections of singlet and triplet excitons and EbS and EbT are the binding energies of the singlet and triplet excitons, respectively. EbS and EbT can be calculated as

EbS ) Eg - ES1

Theoretical Studies of the Properties of Phosphafluorenes

J. Phys. Chem. A, Vol. 114, No. 10, 2010 3657

EbT ) Eg - ET1 where Eg is the HOMO-LUMO energy gap. ES1 and ET1 are excitation energies from the ground state to the lowest excited singlet state and the lowest excited triplet state, respectively. All computations were carried out with Gaussian03 program package with different parameters for structure optimizations and vibrational analyses. The singlet ground-state geometries were fully optimized by the Becke’s three-parameter exchange functional34 along with the Lee-Yang-Parr correlation functional with the restricted (B3LYP) and the unrestricted formalism (UB3LYP) for neutral and ion state molecules35 at the standard split valence plus polarization function 6-31G(d) basis set. The excited-state geometries were optimized by the ab initio configuration interaction singles method (CIS). These fully optimized stationary points were further characterized by harmonic vibrational frequency analysis to ensure that real local minima had been found without imaginary vibrational frequency. The electronic absorption and emission spectra in vacuum were carried out using the time-dependent density functional theory (TDDFT) method of B3LYP/6-31G(d) on the basis of the optimized ground and excited structures, respectively. The polarizable continuum model (PCM) was adopted in the solvent (THF). The various properties of these phosphafluorenes, such as the HOMO and LUMO energies, energy gap (Eg), triplet energy (3Eg), ionization potential (IP), electron affinity (EA), and reorganization energy (λ), were derived from the computed results according to literature publications.36,37 3. Results and Discussions 3.1. Optimized Ground-State Geometries. The structural parameters of 9-phenylphosphafluorene (PhP) have been computed with semiempirical AM1, ab initio HF, and density functional theory B3LYP using different basis sets as listed in Table 1 and were compared to the experimental results (X-ray crystallographic data38) in order to propose a suitable method for theoretical calculations of phosphafluorenes.

The phosphole core of phosphafluorene and the phosphorusrelated bonds were paid special attention in the geometry optimization. In Table 1, the semiempirical method AM1 substantially underestimates the C-P bond lengths of P-C(1), P-C(4), and P-C(7) but overestimates the C-P-C bond angles of C(1)P-C(4), C(1)-P-C(7), and C(4)-P-C(7) in comparison with the experimental data. The source of these discrepancies is due to the unsatisfied P atom parametrization in the AM1 semiempirical Hamiltonian.39 The C-P bond lengths calculated by the HF/631G(d) method are quite in agreement with the experimental results (∆rmax < 0.02 Å), while by the HF/3-21G or HF/6-31G method, larger C-P bond length variation (∆rmax > 0.05 Å) was observed. The DFT/B3LYP method with 3-21G and 6-31G basis sets dramatically overestimates the C-P bond lengths (∆rmax > 0.07 Å). However, the B3LYP method with 6-31G(d) and 6-31G(d)+ basis sets yield shorter P-C bond lengths, which are very close to the experimental bond lengths. Increasing the basis set from 6-31G to 6-31G+ or 6-311G in B3LYP calculations leads to the lengthened P-C bond lengths, which depart from the experimental data obviously (∆rmax > 0.07 Å). It is interesting that taking into account the polarization function in the HF and B3LYP methods is essential to optimize the geometry of phosphafluorenes. Compared with the HF method, B3LYP yields longer CdC bonds and shorter P-C bonds, suggesting that the electrons are more delocalized in DFT methods due to taking into account the electron correlation. The difference value percentages (DVPs) between the calculated parameters by the HF or B3LYP method with the 6-31G(d) basis set and experimental data are drawn in Figure 3. It can be found that all of the DVP values are less than 2.0%, except for the DVP value of the bond angle C(4)-P-C(7) calculated by the HF method, indicating that the B3LYP method is slightly better than HF with the 6-31G(d) basis set. Considering the better performance of B3LYP in energy calculations, the method of B3LYP 6-31G(d) was selected for the following calculations. Different from the planar geometry of carbazoles, the optimized ground-state geometries of phosphafluorenes have a threedimensional nonplanar molecular structure, which is reflected in a

TABLE 1: Selected Structural Parameters of PhP Calculated with Various Methodsa

HF

B3LYP

PhP

AM1

3-21

6-31

6-31(d)

3-21

6-31

6-31(d)

6-31+

6-31(d)+

6-311

exp38

P-C(1) P-C(4) C(2)-C(3) P-C(7) C(1)-P-C(4) C(1)-C(2)-C(3) C(2)-C(1)-P C(5)-C(1)-P C(3)-C(4)-P C(6)-C(4)-P C(1)-P-C(7) C(4)-P-C(7) P-C(7)-C(8)

1.682 1.684 1.458 1.663 95.38 111.34 110.62 128.07 110.59 128.21 109.58 107.25 123.22

1.870 1.870 1.480 1.879 88.12 113.66 112.21 127.21 112.21 127.21 102.61 102.61 122.88

1.876 1.876 1.478 1.888 88.00 113.77 112.15 127.09 112.15 127.09 103.4 103.4 123.27

1.834 1.834 1.480 1.846 89.09 113.03 112.28 127.36 112.28 127.36 103.97 103.98 123.67

1.880 1.880 1.477 1.893 87.99 113.68 112.23 127.03 112.23 127.03 101.78 101.78 122.75

1.886 1.886 1.475 1.903 87.85 113.78 112.18 127.01 112.18 127.01 102.67 102.67 123.1

1.841 1.841 1.472 1.857 89.11 113.12 112.12 127.43 112.12 127.43 103.49 103.49 123.47

1.885 1.885 1.476 1.903 87.91 113.73 112.21 126.97 112.21 126.97 102.56 102.56 123.01

1.842 1.842 1.473 1.859 89.11 113.12 112.12 127.40 112.12 127.41 103.52 103.53 123.47

1.887 1.887 1.475 1.905 87.82 113.83 112.16 126.97 112.16 126.97 101.77 101.77 123.16

1.838 1.808 1.473 1.846 89.5 113.1 111.8 127.1 112.8 127.1 104.1 101.8 123.1

a

The bond lengths are in Å and bond angles in ° (degrees).

3658

J. Phys. Chem. A, Vol. 114, No. 10, 2010

Figure 3. Difference value percentages (DVPs) between the calculated and experimental geometries.

Figure 4. (a) The optimized ground-state geometries of phosphafluorenes. (b) Geometrical characteristics of the phosphafluorenes; n1: normal to the biphenyl skeleton; n2: normal to the phenyl ring; n3: normal to the Y ) P-C(7) plane.

distorted tetrahedral geometry formed by the P atom and three adjacent C atoms (Figure 4a).40 In Figure 4b, two additional angles were used to describe the geometries of the phosphafluorenes. ζ1 represents the angle formed by n1 and n2, and ζ2 represents the angle formed by n2 and n3, where n1, n2, and n3 are the normals of the corresponding planes. The calculated ζ1 and ζ2 are 90 and 0°, respectively, showing that benzene ring, oxygen (sulfur or selenium) atom, and phosphorus atom of PhOP (PhSP or PhSeP) are in the same plane, which is perpendicular to the planar biphenyl skeleton. This perpendicular dihedral angle is a very important structural characterization of phosphafluorenes and may influence their properties in many ways. The optimized ground-state structural parameters of phosphafluorenes are collected and compared in Table 2 to study the influence of different substitutions on the phosphole core. After the chemical modifications of the P atom with an O, S, or Se atom, geometric changes take places. (1) The P-C(1) and P-C(7) bond lengths decrease by about 0.01 Å, where the oxidation leads to the shortest P-C(7) bond length and the selenation possesses to

Yin et al. the longest. (2) The C(1)-P-C(4) and P-C(1)-C(5) bond angles slightly increase in the following order: O > S > Se. These geometrical differences may be due to (1) the weak interaction between the O, S, or Se and the H atoms of the planar biphenyl, (2) the strong conjugated interaction between the double bond CdY (Y ) O, S or Se) and the planar biphenyl, and (3) the electronaccepting ability decrease of O > S > Se. As compared with the methyl-modified molecules, the phenyl substitution leads to longer P-C(1) and C(2)-C(3) single bond, while the C(1)dC(2) double bond is predicted to be shortened with the variation within 0.003 Å due to the stronger electron-donating ability and larger steric hindrance of phenyl. The O, S, or Se substitution has a significant impact on the C(2)-C(3), C(1)-C(2), and P-C(1) bond lengths, resulting in reduced P-C(1) and C(1)-C(2) but enlarged C(2)-C(3) with bond length changes up to 0.01 Å. The ionization of the phosphafluorenes (MeP+ and PhP+) results in even more decreased P-C(1) bond lengths, with the largest variation of 0.03-0.04 Å. To understand the structural differences between the groundand excited-state geometries, the bond lengths, bond angles, and dihedral angles of the excited structure of phosphafluorenes are listed in Table 3. The excited-state structural parameters of MeSeP and PhSeP were excluded because the optimizations of their excited states have not been converged. Except for PhP, the central phophole of phosphafluorenes changes dramatically after the excitation, where the P-C(1), C(2)-C(3), and C(1)-P-C(4) decrease by ∼0.02 Å, 0.09 Å, and 1° respectively, while the C(1)-C(2) bond increases by ∼0.04 Å. The particular geometry of PhP may be responsible for its particular emission spectra, which will be discussed later. From Tables 2 and 3, it is clear that the molecular geometries of phosphafluorenes in both the ground and excited states can be tuned by chemical modification of the phosphorus atom, leading to different molecular conformations and optoelectronic properties. 3.2. Frontier Molecular Orbitals. 3.2.1. Electronic Properties. Optoelectronic properties of phosphafluorenes were studied and compared with those of carbazoles because the phosphorus atom is in the same subgroup as nitrogen and phosphafluorene is structurally very similar to carbazoles (nitrogenafluorenes), which are currently the wildly investigated optoelectronic materials. As shown in Table 4, carbazoles (MeN and PhN) have the highest HOMOs and LUMOs, which coincides with the fact that carbazole is the core-building unit for hole-transporting materials.41,42 By replacing nitrogen with a phosphorus atom, the resulting phosphafluorenes (MeP and PhP) show both decreased HOMOs and LUMOs but increased Eg due to the stronger electron-donation ability of the P atom and the nonplanar molecular structure of phosphafluorenes. After chemical modification of the P atom with an O, S, or Se atom, the HOMOs gradually increase in the following order, -6.26 (MeOP) < -5.83 (MeSP) < -5.43 eV (MeSeP) and -6.23 (PhOP)< -5.86 (PhSP) < -5.51 eV (PhSeP), while the LUMOs greatly decrease, lying at around -1.55 eV, which is far below the LUMO energies of MeP (-0.98 eV) and PhP (-0.97 eV), indicating that the introduction of strong electronwithdrawing substitutions of an O, S, or Se atom enhances the electron-transport and -injection abilities of the phosphafluorenes. Especially the HOMO energy of MeSeP, lying at -5.43 eV, is even lower than that of MeN (-5.32 eV), indicating that the holetransport and -injection abilities have been improved by selenidation of the phosphole center. Except for MeP+ and PhP+, there is small influence of methyl and phenyl substitutions on the HOMO and LUMO, with the maximal derivation within 0.08 eV. The ionized phosphafluorenes (MeP+ and PhP+) results in many decreased HOMOs (-9.82 and -9.61 eV) and LUMOs (-5.31 and -5.09 eV), indicating that the positively charged phosphorus atom inside

Theoretical Studies of the Properties of Phosphafluorenes

J. Phys. Chem. A, Vol. 114, No. 10, 2010 3659

TABLE 2: Selected Bond Lengths (Å), Bond Angles (°), and Dihedral Angles (°) of Phosphafluorenes in the Ground State

DFT/B3LYP P-C(1) C(1)-C(2) C(2)-C(3) P-C(7) P-Y C(1)-P-C(4) P-C(1)-C(5) P-C(1)-C(2) C(1)-C(2)-C(3) C(1)-P-C(7) C(1)-P-Y C(7)-P-Y Y-P-C(4)-C(3) Y-P-C(1)-C(2) C(3)-C(4)-P-C(7) C(2)-C(1)-P-C(7)

MeP 1.8398 1.4144 1.4719 1.8767

MeP+ 1.7973 1.4159 1.4811 1.8640

PhP+

PhP 1.8406 1.4140 1.4724 1.8565

1.8018 1.4149 1.4810 1.8640

89.18 127.52 112.18 113.13 101.52

94.37 129.33 108.73 114.09

89.11 127.43 112.12 113.12 103.49

93.74 128.97 109.10 113.99

-105.53 105.55

-117.75 117.68

-109.39 109.38

-112.84 113.36

MeOP

PhOP

MeSP

PhSP

MeSeP

PhSeP

1.8283 1.4110 1.4830 1.8344 1.5006 91.23 128.31 110.77 113.61 105.85 118.81 113.46 -125.22 125.22 -105.84 105.84

1.8283 1.4107 1.4835 1.8287 1.5018 91.15 128.15 110.87 113.56 107.29 118.49 111.92 -123.20 123.20 -108.94 108.93

1.8298 1.4107 1.4800 1.8435 1.9689 91.16 128.02 110.72 113.69 105.39 118.98 113.91 -125.77 125.77 -104.95 104.97

1.8312 1.4102 1.4803 1.8416 1.9715 90.98 127.86 110.86 113.65 106.36 117.92 114.39 -122.71 122.69 -107.31 107.30

1.8290 1.4108 1.4796 1.8463 2.1055 91.07 127.87 110.78 113.63 104.62 119.53 114.16 -128.05 128.07 -102.59 102.61

1.8297 1.4102 1.4799 1.8431 2.1051 90.92 127.68 110.94 113.59 105.99 118.59 113.85 -124.87 124.85 -105.72 105.70

TABLE 3: Selected Bond Lengths (Å), Bond Angles (°), and Dihedral Angles (°) of Phosphafluorenes in the Excited State DFT/CIS P-C(1) C(1)-C(2) C(2)-C(3) P-C(7) P-Y C(1)-P-C(4) P-C(1)-C(5) P-C(1)-C(2) C(1)-C(2)-C(3) C(1)-P-C(7) C(1)-P-Y C(7)-P-Y Y-P-C(4)-C(3) Y-P-C(1)-C(2) C(3)-C(4)-P-C(7) C(2)-C(1)-P-C(7)

MeP

PhP

1.8261 1.4586 1.3932 1.8724

1.8735 1.4585 1.3946 1.9045

87.61 127.98 112.63 112.89 101.39

86.97 127.47 112.48 113.95 103.19

-90.77 90.77

-99.25 99.25

MeOP

PhOP

MeSP

PhSP

1.7969 1.4576 1.3971 1.8239 1.4747 89.77 128.27 112.11 113.00 108.55 119.40 109.46 -124.93 124.93 -108.74 108.74

1.7931 1.4579 1.3973 1.8217 1.4754 89.66 128.27 112.17 112.82 109.49 119.06 108.70 -118.27 118.27 -115.87 115.87

1.7957 1.4560 1.3965 1.8315 1.9745 90.11 128.20 111.77 113.16 108.38 119.20 109.81 -125.32 125.32 -108.18 108.19

1.7925 1.4561 1.3968 1.8333 1.978 90.03 128.10 111.88 113.02 108.78 117.86 111.52 -118.05 118.06 -113.76 113.77

TABLE 4: HOMO and LUMO Energies, Band Gaps (Eg), Triplet Energy (3Eg), Excitation Energies in the Singlet (ES1) and Triplet (ET1) States, Formation Cross Section of the Singlet (σS) and Triplet (σT) Exciton Ratios, and Singlet Exciton Generation Fraction (χS) of the Phosphafluorenes (in eV) HOMO LUMO Eg 3 Eg 3 Eg/Eg ES1 ET1 σS/σT χS a

MeN

PhN

MeP

MeP+

PhP

PhP+

MeOP

PhOP

MeSP

PhSP

MeSeP

PhSeP

-5.32 -0.62 4.70 3.18a 0.66a 4.05 3.18 0.43 43.9

-5.33 -0.65 4.68

-5.93 -0.98 4.95 2.95 0.60 4.34 3.04 0.32 51.3

-9.82 -5.31 4.51

-5.91 -0.97 4.94 2.95 0.60 4.37 3.03 0.30 52.7

-9.61 -5.09 4.52

-6.26 -1.46 4.80 2.83 0.59 4.18 2.94 0.33 50.1

-6.23 -1.48 4.75 2.80 0.59 4.11 2.91 0.35 48.9

-5.83 -1.55 4.28 2.83 0.66 3.53 2.93 0.56 37.5

-5.86 -1.53 4.33 2.81 0.66 3.52 2.90 0.57 37.1

-5.43 -1.55 3.88 2.16 0.56 3.19 2.88 0.69 32.6

-5.51 -1.54 3.97 2.20 0.55 3.19 2.85 0.70 32.4

4.04 3.181 0.43 43.7

The calculated 3Eg and 3Eg/Eg of 9H-carbazole because the calculation of MeN is not converged.

of a five-membered phosphole ring greatly reduces the holeinjection and -transport abilities but enhances the electron-accepting ability of MeP+ and PhP+. The energy gap becomes narrower in following order, 4.95 (MeP) > 4.94 (PhP) > 4.80 (MeOP) > 4.75 (PhOP) > 4.52 (PhP+) > 4.51 (MeP+) > 4.33 (PhSP) > 4.28 (MeSP) > 3.97 (PhSeP) > 3.88 (MeSeP) eV, with a maximum variation of 1.07 eV. The triplet energies (3Eg) of phosphafluorenes were also calculated to evaluate their potential applications as host materials. The calculation reveals that the MeP and PhP have 3Eg values of 2.95 eV and 3Eg/Eg percentages of 60%, which are comparable

with those of the famous host material of carbazole with an 3Eg of 3.18 eV and 3Eg/Eg of 66%, indicating that phosphafluorenes can be new building blocks for the host materials. The 3Eg energies decrease after oxidation of the P atom, and the MeSeP possesses the lowest 3Eg of 2.16 eV. The external quantum efficiencies (EQE, ηEL) of OLEDs can be estimated from ηEL ) ηocηI, where ηoc is the out-coupling factor and ηI is the internal quantum efficiency (IQE). Recently, the experimental maximum internal quantum efficiency (IQE) of OLEDs has exceeded the theoretical statistical upper limit of 25%. The high singlet-to-triplet exciton-formation

3660

J. Phys. Chem. A, Vol. 114, No. 10, 2010

Figure 5. The frontier molecular orbital energies of phosphafluorenes.

cross section ratio (σS/σT) in conjugated systems is believed to be a possible reason for the high IQE. For MeP, PhP, MeOP, and PhOP, the calculated formation cross section ratios of singlet over triplet excitons (σS/σT) are 3.16, 3.34, 3.01, and 2.87, respectivley, with corresponding singlet exciton-generation fractions (χS) of 51.3, 52.7, 50.1, and 48.9%, which are higher than those of carbazoles, suggesting their potential as the highly efficient fluorescent-lightemitting materials. After chemical modification of the P atom with S and Se atoms, the χS values decrease slightly within 15-20% in MeSP, PhSP, MeSeP, and PhSeP, which is coincident with the experimental observation in phospholes. The calculated frontier molecular orbital energies and other parameters of the phosphafluorenes are comparable to those of carbazole and fluorene, which are widely used building units for optoelectronic materials, showing the bright future of phosphafluorene-based molecules as optoelectronic materials. From more frontier molecular energy levels (from HOMO-3 to LUMO-3 in Figure 5), it is interesting to find that the energy difference between the HOMO and HOMO-1 of MeSP, PhSP, MeSeP, and PhSeP is very small, and that of PhSeP is nearly degenerate, which implies that it is possible to promote an electron from HOMO to LUMO or from HOMO-1 to LUMO. This particular electronic structure of phosphafluorenes will lead to their particular absorption and emission spectra, which will be discussed in the following section. After chemical modification of the P atom with O, S, or Se, the LUMO drops about 0.5 eV and remains almost constant ((0.05 eV), while the HOMO drops about 0.3 eV after oxidation and then increases gradually, indicating that the LUMOs and HOMOs of phosphafluorene can be independently tuned by chemical modification of the P atom, which will greatly facilitate the molecular design and property control of this kind of material. In Figure 6, the frontier orbitals of phosphafluorenes and carbazoles show π character and spread over the whole molecules. In general, the HOMO possesses bonding character, and the LUMO holds antibonding character. However, the HOMO orbitals of phosphafluorenes and carbazoles represent an antibonding interaction between the bridging atom and its neighbors, and the LUMO orbitals of phosphafluorenes show the bonding interaction between the bridge phosphorus atom and its adjacent subunits, but those of carbazoles still show antibonding character, leading to their high LUMOs. The different electronic cloud distribution of phosphafluorenes and carbazoles can be explained by the different HOMO-LUMO interaction types present in our previous publication.43 To shed light on the chemistry activity of the planar biphenyl skeleton with different bridging atoms and substitutions in the

Yin et al. HOMO and LUMO, the local density of states of these molecules has been compared in Table 5. The planar biphenyl skeleton has the largest contribution to the LUMO in all of the studied molecules. The N atom in carbazoles (MeN and PhN) has much larger contributions to the HOMO as compared with those of the P atom in phosphafluorenes (MeP and PhP), which results in the increased HOMO of carbazoles. The contributions of the methyl and phenyl substituents to the HOMO and LUMO orbitals are very limited. However, after chemical modification of the P atom with a S or Se atom, the HOMO is mainly localized on the PdS or PdSe bond with dominating electron density contribution, while the planar biphenyl skeleton has a small contribution to the HOMO, indicating that S and Se have strong electron-donating ability with an effectively improved HOMO level, and can greatly facilitate the intramolecular charge transfer of the S- and Se-substituted phosphafluorenes. Importantly, because the lowest singlet excited state corresponds almost to the excitation from the HOMO to the LUMO, the difference of the bond lengths between the ground and lowest singlet excited states can be predicted from the MO nodal patterns. For example, the HOMO has a node across the C(1)-P, C(4)-P, and C(2)-C(3) bonds in MeOP, whereas the LUMO is bonding, and the parameters in Tables 2 and 3 confirm the anticipated contraction of these bonds. On the contrary, the HOMO is bonding across C(1)-C(2) and C(3)-C(4) bonds in MeOP, but the LUMO has nodes in these regions; the parameters in Table 3 confirm the anticipated elongation of these bonds. 3.2.2. Ionization Potential, Electron Affinity, and Reorganization Energy. The charge-injection and -transport and their balance are crucial for optoelectronic compounds; therefore, it is important to investigate their ionization potentials (IPs), electronic affinities (EAs), and reorganization energies (λ) to evaluate the energy barrier for injection and transport rates of the holes and electrons. We calculated IPs and EAs, together with the hole extraction potential (HEP), which is the energy difference from M+ (cationic) to M (neutral molecule) using the M+ geometric structure, and the electron extraction potential (EEP), which is the energy difference from M-(anionic) to M using the M- geometric structure. The IPs and EAs can be either for vertical excitations (v, at the geometry of the neutral molecule) or adiabatic excitations (a, at the optimized structures for both the neutral and charged molecule).44 The lower the IP and the higher the EA, the easier the entrance of holes and electrons, respectively.37 The reorganization energy is the energy difference between two molecules which have gone through electron transfer and can provide a qualitative indication of the charge-transport rate (the lower the λ values, the bigger the charge-transport rate). As observed from Table 6, the variety of trends of the IPs and EAs of the phosphafluorenes is similar to their HOMO and LUMO energies, in accordance with the theory anticipations. The correlation between IP values and HOMO energies (EA values and LUMO energies) can be obtained linearly (see Figure 7)

IP ) -0.5325EHOMO + 4.4978

r2 ) 0.7025

EA ) -1.0638ELUMO - 1.6319

r2 ) 0.9645

where r2 is the regression constant. Although the IP and EA values of some other phosphafluorenes are still unknown, these regression equations may provide a simple formula to predict the IP and EA values of phosphafluorenes from their orbital energies of the HOMO and LUMO.45

Theoretical Studies of the Properties of Phosphafluorenes

J. Phys. Chem. A, Vol. 114, No. 10, 2010 3661

Figure 6. Electron density distributions of the HOMOs and LUMOs of phosphafluorenes.

Carbazoles (MeN and PhN) have the lowest IPs, EAs, and λ, which is in accordance with the fact that carbazoles are good hole-transport materials. By substitution of N with P, phosphafluorenes (MeP and PhP) show increased IPs and EAs, λhole and λelectron, indicating that phosphafluorenes have higher electron-acceptance ability and lower but more balanced chargetransport rates since λhole and λelectron are closer to each other.

After chemical modification of the P atom with an O, S, or Se atom, the IPs increase but then decrease as follows, O > S > Se, while the EAs increases gradually as follows, O < S < Se. By changing the substitutions of methyl to phenyl, the IPs slightly decrease, but the EAs increase. Therefore, the abilities to create holes and accept electrons of the phosphafluorenes are effectively improved by the introduction of strong electron-

3662

J. Phys. Chem. A, Vol. 114, No. 10, 2010

Yin et al.

TABLE 5: Contribution of the Electron Density (%) in Phosphafluorenesa BP MeN PhN MeP MeP+ PhP PhP+ MeOP PhOP MeSP PhSP MeSeP PhSeP

P(N)

R

HOMO

LUMO

HOMO

LUMO

HOMO

LUMO

62.24 57.82 99.49 99.56 96.78 98.62 96.30 92.90 10.83 24.70 8.41 8.15

97.87 96.63 79.70 58.87 81.56 58.62 82.39 80.67 74.33 71.81 74.08 70.24

32.40 30.35 0.12 0.11 0.17 0.12 0.50 0.27 8.93 5.61 7.46 7.14

1.01 1.23 6.24 15.29 7.83 14.59 6.48 7.65 8.77 10.53 8.11 10.24

5.36 11.82 0.39 0.33 3.04 1.26 3.19 6.83 80.24 69.68 84.12 84.71

1.12 2.14 14.05 25.84 10.61 26.79 11.12 11.68 16.90 17.66 17.80 19.53

a BP: the planar biphenyl skeleton; P(N): the phosphorus or nitrogen atom; R: the substituents on the phosphorus or nitrogen atom.

TABLE 6: Ionization Potential (IP), Electronic Affinity (EA), and Reorganization Energy (λ) of Phosphafluorenes (in eV) molecule IP(v) IP(a) HEP MeN PhN MeP PhP MeOP PhOP MeSP PhSP MeSeP PhSeP

6.98 6.84 7.71 7.59 7.93 7.75 7.67 7.51 7.53 7.30

6.93 6.78 7.45 7.32 7.77 7.64 7.49 7.31 7.18 7.02

6.88 6.74 7.30 7.19 7.62 7.51 7.21 6.97 6.84 6.62

EA(v)

EA(a)

EEP

-1.03 -0.87 -0.63 -0.55 -0.16 -0.06 -0.01 0.05 0.001 0.07

-0.89 -0.72 -0.45 -0.34 0.04 0.17 0.21 0.29 0.23 0.30

-0.75 -0.56 -0.27 -0.12 0.24 0.39 0.43 0.52 0.45 0.54

λhole λelectron 0.10 0.10 0.41 0.40 0.31 0.24 0.46 0.54 0.69 0.68

0.28 0.31 0.36 0.43 0.40 0.45 0.44 0.47 0.45 0.47

withdrawing substitutions of an O, S, or Se atom or replacement of the substitutions of methyl with phenyl. Especially the PhSeP, which has the lowest IP and highest EA, may have strong holeand electron-injection and -transport abilities. Except for PhSP, MeSeP, and PhSeP, other molecules have a negative calculated EA value, which was raised by the incomplete cancellation of the electronic self-interaction energy due to the use of an inexact density functional and a finite basis set.46,47 The calculated reorganization energies are also listed in Table 6. Carbazoles (MeN and PhN) are good for hole transport since they have a much lower hole-transport reorganization energy (λhole) as compared with the electron-transport reorganization energy (λelectron). The significantly increased λhole and λelectron are observed in the phosphafluorenes. For the MeOP and PhOP, the λhole are smaller than their respective λelectron, suggesting that the hole transfer should be faster than the electron transport, while for the MeSeP and PhSeP, the λhole are bigger than their respective λelectron, suggesting that the hole transfer should be slower than the electron transfer. Although the hole- and electron-transfer rates of phosphafluorenes are smaller than those of carbazoles, the differences between the λhole and λelectron of MeP, PhP, MeSP, and PhSP are 0.05, 0.03, 0.02, and 0.07 eV, respectively, which are small enough that these phosphafluorenes can act as nice ambipolar materials. 3.3. Absorption and Emission Spectra. To understand electronic transitions of phosphafluorenes, TDDFT calculations on the absorption and emission spectra in both vacuum and solvent (THF) were performed. The calculated wavelengths from absorption and emission spectra, main transition configurations, and oscillator strengths for the most relevant singlet excited states of phosphafluorenes are listed in Tables 7 and 8,

respectively. The electronic absorption and emission spectra were simulated by Gaussian functions with a half-width of 3000 cm-1 based on the 10 lowest singlet energies from TDDFT/ B3LYP/6-31(d) calculations. 3.3.1. Absorption Spectra. From Table 7 and the MOs in Figure 6, all of the electronic transitions of phosphafluorenes are of the π f π* type for the absorption spectra, where the electron transitions are from the initial state that is mainly contributed to by the HOMO and HOMO-1 to the final state that is mainly contributed to by the LUMO and LUMO+1. MeP and PhP have the calculated absorption peak at around 283 nm, which is shorter than that for carbazoles. The oxidized phosphafluorenes (MeOP and PhOP) have the peaks at around 300 nm, which are red-shifted near 20 nm. The MeSP and PhSP have the greatly red-shifted (∼60 nm) absorption peak at around 340 nm, while the MeSeP and PhSeP have the largest red shift of more than 100 nm. The difference between methyl- and phenyl-substituted phsophafluorenes is quite small, suggesting the limited influence of the alkyl and the aryl substituents. The calculated maximum absorption wavelengths exhibit a red shift from MeP/PhP to MeSeP/PhSeP of 284/284 (MeP/PhP) < 296/ 302 (MeOP/PhOP) < 347/340 (MeSP/PhSP) < 386/379 nm (MeSeP/PhSeP), which confirms the prediction from the energy gap discussed above. Phosphafluorenes have comparable absorption peaks as carbazoles and great modification feasibilities by O, S, and Se, indicating their bright potential as optoelectronic molecules. Carbazoles (MeN and PhN), MeP, PhP, MeOP, and PhOP have the strongest absorption peaks, with large oscillator strengths assigned to S0 f S1 electronic transition, while for MeSP, PhSP, MeSeP, and PhSeP, the obvious transition is S0 f S2, and the osillatior strength of the S0 f S1 transition is very small, indicating that the S0 f S1 transition is forbidden. It is interesting to note that the initial states of the S- and Semodified phosphafluorenes are mainly related to the MOs that are localized in P-S and P-Se (see Figure 6), while their final states are mainly related to the MOs that are localized in the biphenyl skeleton, suggesting that the absorptions are photoinduced electron-transfer processes and that the excitations generate charge-separated states, which may find applications in photocurrent conversion devices. This particular MO of Sand Se-modified phosphafluorenes may lead to their particular absorption behaviors. The influence of solvent (THF) on the absorption spectra of phosphafluorenes was simulated by using the PCM model. After oxidation of phosphafluorenes, the calculated maximum absorption peaks in THF also exhibit some red shifts of 284/284 (MeP/PhP) < 301/307 (MeOP/PhOP) < 312/310 (MeSP/PhSP) < 348/345 nm (MeSeP/PhSeP). For MeP, PhP, MeOP, and PhOP, the maximum absorption wavelengths in solvent (THF) are larger than those in vacuum, indicating that the solvent effects stabilize the excited state, which induces the red shift of the absorption spectra as compared with that in vacuum. However, for MeSP, PhSP, MeSeP, and PhSeP, the maximum absorption wavelengths in solvent (THF) are smaller than those in vacuum, showing that the solvent effects stabilize the ground state, which induces the blue shift of the absorption spectra probably because the high electron density distribution on the S and Se atoms (see Table 5 and Figure 6) of the HOMO can be stabilized well by solvent (THF), resulting in decreased ground-state energy. From 200 to 420 nm, the phosphafluorenes (MeP and PhP) show comparable absorption spectra as carbazoles, as illustrated in Figure 8. The phenyl-substituted molecules have fewer and red-shifted absorption spectra than the methyl-

Theoretical Studies of the Properties of Phosphafluorenes

J. Phys. Chem. A, Vol. 114, No. 10, 2010 3663

Figure 7. (a) The correlation between the calculated HOMO energies and the IP values. (b) The correlation between the calculated LUMO energies and the EA values.

TABLE 7: Absorption Spectra in Both Vacuum and Solvent (THF) of Phosphafluorenes molecule MeN

PhN

MeP

PhP

MeOP

PhOP

MeSP

states

oscillator strength f

HOMO f LUMO: 0.658 HOMO-1 f LUMO+1: -0.201 HOMO f LUMO: 0.662 HOMO-1 f LUMO+1: 0.187 HOMO f LUMO: 0.661 HOMO-1 f LUMO+3: -0.190 HOMO f LUMO: 0.665 HOMO-1 f LUMO+3: 0.181 HOMO-1 f LUMO: 0.656 HOMO f LUMO+2: -0.204 HOMO-1 f LUMO: 0.660 HOMO f LUMO+2: 0.203 HOMO-1 f LUMO: 0.646 HOMO-1 f LUMO+2: -0.101 HOMO f LUMO+3: -0.150 HOMO-1 f LUMO+4: 0.153 HOMO f LUMO: 0.596 HOMO f LUMO+2: 0.276 HOMO f LUMO: 0.619 HOMO f LUMO+1: 0.206 HOMO-1 f LUMO+1: -0.101 HOMO-4 f LUMO: -0.109 HOMO f LUMO: 0.640 HOMO f LUMO+1: -0.186 HOMO-1 f LUMO+1: 0.110 HOMO f LUMO: 0.634 HOMO f LUMO+1: -0.187 HOMO f LUMO: 0.649 HOMO f LUMO+1: -0.163 HOMO-1 f LUMO: 0.694 HOMO f LUMO: 0.691 HOMO f LUMO: 0.658 HOMO-2 f LUMO: 0.200 HOMO-1 f LUMO: 0.691 HOMO f LUMO: 0.693 HOMO-1 f LUMO: 0.687 HOMO f LUMO: 0.662 HOMO-2 f LUMO: 0.184 HOMO-1 f LUMO: 0.687 HOMO-1 f LUMO: 0.695 HOMO f LUMO: 0.689 HOMO-1 f LUMO: 0.694 HOMO f LUMO: 0.690 HOMO-1 f LUMO: 0.694 HOMO f LUMO: 0.685 HOMO f LUMO: 0.693 HOMO-1 f LUMO: 0.687

0.032

306

THF

S0 f S 1

309

gas-phase

S 0 f S1

307

THF

S0 f S 1

309

gas-phase

S 0 f S1

285

THF

S0 f S1

284

gas-phase

S 0 f S1

284

THF

S0 f S1

284

gas-phase

S 0 f S1

296

THF

S0 f S1

300

gas-phase

S 0 f S1

302

THF

S0 f S1

307

gas-phase

S 0 f S1 S0 f S2 S0 f S1

351 347 321

S0 f S2 S 0 f S1 S0 f S2 S0 f S1

312 352 340 325

S0 f S2 S 0 f S1 S0 f S2 S0 f S1 S0 f S2 S 0 f S1 S0 f S2 S0 f S1 S0 f S2

310 388 386 354 348 389 379 354 345

gas-phase

gas-phase THF

PhSeP

main transition configuration

S0 f S1

THF MeSeP

calculated wavelength (nm)

gas-phase

THF PhSP

electron transition

gas-phase THF

substituted ones due to the rigid and planar nature of phenyl. After modification of O, S, and Se, the spectra gradually red

0.045 0.029 0.048 0.059 0.089 0.067

0.133 0.066

0.111 0.051 0.084 0.000 0.017 0.005 0.028 0.000 0.016 0.004 0.028 0.001 0.018 0.000 0.029 0.001 0.017 0.001 0.028

shift, and the S- and Se-modified phosphafluorenes show the small intramolecular charge-transfer absorptions. The absorp-

3664

J. Phys. Chem. A, Vol. 114, No. 10, 2010

Yin et al.

TABLE 8: Emission Spectra in Both Vacuum and Solvent (THF) of Phosphafluorenes molecule

states

MeN

gas-phase

S1 f S0

313

THF

S1 f S0

318

gas-phase

S1 f S0

311

THF

S1 f S0

315

gas-phase

S1 f S0

333

THF

S1 f S0

338

PhP

gas-phase

S1 f S0

327

MeOP

gas-phase

S1 f S0

355

THF

S1 f S0

365

PhOP

gas-phase

S1 f S0

366

MeSP

gas-phase

S1 f S0

415

THF

S2 f S0 S1 f S0

406 384

gas-phase

S1 f S0

419

S2 f S0

396

PhN

MeP

PhSP

electron transition calculated wavelength (nm)

tion spectra of carbazoles and phosphafluorenes (MeP and PhP) are slightly changed in vacuum and in THF because the solvent effects of THF toward the ground and excited states are similar. However, the O-, S-, and Se-modified phosphafluorenes show quite different absorption spectra in THF because the more polarized the electron distribution of the ground and excited states, the higher the solvent effects of THF. For example, the highly polarized ground states of S- and Se-modified phosphafluorenes illustrated in Figure 6 lead to blue-shifted absorption spectra in THF. 3.3.2. Emission Spectra. The theoretical emission spectra for phosphafluorenes based on optimized excited-state geometries are presented in Table 8. The emission wavelengths of MeSeP and PhSeP were excluded because the optimizations of their excited states were not converged. All chemical modifications of phosphafluorenes result in materials with deep blue light-emitting features (λem ) 327-419 nm). Like that observed in the absorption spectra, the emission wavelengths for these molecules also exhibit red shifts from MeP to MeSP (see Table 8) of 333/327 (MeP/PhP) < 355/366 (MeOP/PhOP) < 415/419 (MeSP/PhSP) < 464/474 nm (MeSeP/PhSeP). The Stoke shifts of MeP and PhP are about 45 nm, and those of MeOP, PhOP, MeSP, and PhSP are all around 65 nm. The larger Stokes shifts of phosphafluorenes in comparison with those of carbazoles, whose Stokes shifts are within 10 nm, suggest the larger rearrangement of phosphafluorenes upon photoexcitation. The maximum emission wavelengths of phosphafluorenes are all assigned to π f π* character arising from an S1, HOMO f LUMO or HOMO-1 f LUMO transition, judged from the MOs in Figure 6. MeP, PhP, MeOP, and PhOP have stronger oscillator strengths than carbazoles, while MeSP and PhSP have very small oscillator strengths, which is coincident with those observed in the absorption. The simulated emission spectra of phosphafluorenes in vacuum (see Figure 9) show that although the phosphafluorenes (MeP and PhP) have shorter absorption wavelengths than carbazoles, they have

main transition configuration

oscillator strength f radiative lifetimes (ns)

HOMO-1 f LUMO: 0.653 HOMO f LUMO+1: 0.217 HOMO-1 f LUMO: 0.659 HOMO f LUMO+1:0.201 HOMO-1 f LUMO: 0.652 HOMO f LUMO+3: 0.219 HOMO-1 f LUMO: 0.658 HOMO f LUMO+3: 0.202 HOMO f LUMO: 0.626 HOMO f LUMO+1: 0.151 HOMO f LUMO: 0.636 HOMO f LUMO+1: 0.125 HOMO f LUMO: 0.623 HOMO f LUMO+2: -0.171 HOMO f LUMO: 0.630 HOMO f LUMO+1: 0.138 HOMO-2 f LUMO: -0.109 HOMO f LUMO: 0.642 HOMO f LUMO+1: 0.124 HOMO f LUMO: 0.638 HOMO f LUMO+1: -0.127 HOMO-1 f LUMO: 0.680 HOMO-2 f LUMO: 0.135 HOMO f LUMO: 0.690 HOMO f LUMO: 0.622 HOMO-2 f LUMO: 0.258 HOMO f LUMO: 0.675 HOMO-2 f LUMO: 0.149 HOMO-1 f LUMO: 0.688

0.036

39.17

0.052

29.27

0.043

32.93

0.064

23.36

0.186

6.58

0.282

6.08

0.175

6.92

0.108

12.22

0.163

12.27

0.085

16.10

0.003

617.80

0.018 0.050

138.88 44.31

0.002

930.22

0.018

132.41

red-shifted emission peaks, indicating a larger Stokes shift of phosphafluorenes. The phenyl-substituted molecules have fewer and red-shifted emission spectra than the methylsubstituted ones due to the rigid, planar, and bulky nature of phenyl. After modification of O and S, the spectra red shifts, and the S-modified phosphafluorenes show a small intramolecular charge-transfer emission at around 415 nm deduced from the transition configuration and the electron density distributions of frontier obitals. The influence of solvent (THF) on the emission spectra of phosphafluorenes was also considered by using the PCM model. The emission wavelengths of PhP, PhOP, and PhSP were not included in Table 8 because the optimizations of their excited states were not converged. The solvent effects lead to larger Stokes shifts of phosphafluorenes in comparison with that of carbazole. For MeN, PhN, MeP, and MeOP, the solvent stabilizes their excited states, inducing the small red shift (4-10 nm) of the emission spectra as compared with that in vacuum, while for MeSP, the solvent effects stabilize the ground state, which induces the blue shift of the emission spectra, which is coincident with that observed in absorption spectra. From Figures 8 and 9, no mirror relation between adsorption and emission spectra was observed, indicating that the geometry varies significantly from the ground state to the excited state, which is confirmed by the geometric data in Tables 2 and 3. The radiative lifetimes have been computed for spontaneous emission by using the Einstein transition probabilities according to the formula (in au)48

τ)

c3 2(EFlu)2f

where c is the velocity of light, EFlu is the excitation energy, and f is the oscillator strength. The short radiative lifetime leads to the high light-emitting efficiency, while the long radiative

Theoretical Studies of the Properties of Phosphafluorenes

J. Phys. Chem. A, Vol. 114, No. 10, 2010 3665

Figure 8. The simulated absorption spectra of phosphafluorenes in vacuum (left) and in solution (THF, right).

lifetime facilitates the electron and energy transfer and the attack of active species. The famous host materials of carbazoles have radiative lifetimes longer than 30 ns, while MeSP and PhSP show much longer radiative lifetimes (longer than 130 ns). In THF, the radiative lifetime becomes shorter with increased oscillator strength, suggesting the increased luminescence in solution. Considering also the high 3Eg and 3Eg/Eg shown in Table 4 and the narrow differences between the λhole and λelectron, the sulfurated phophafluorenes are believed to be a kind of

excellent host material. The short radiative lifetime of MeP, PhP, MeOP, and PhOP indicates that they are good lightemitting materials, as observed experimentally. The reported maximum absorption wavelengths of PhOP and PhSP are 332 and 330 nm, respectively, and their maximum emission wavelengths are both at 366 nm when measured in CH2Cl2.5 The discrepancies between the TDDFT-calculated values and experimental data may be attributed to two aspects.49 The first involves DFT, which generally gives a smaller gap of

3666

J. Phys. Chem. A, Vol. 114, No. 10, 2010

Yin et al.

Figure 9. The simulated emission spectra of phosphafluorenes in vacuum (left) and in solution (THF, right).

materials and thus generates smaller excited energies, especially for larger conjugate systems and charge-transfer complexs in excited states. The other is solvent effects. Solvent, especially a polar solvent such as THF, could affect the geometry and electronic structure as well as the properties of a molecule through the long-range interaction between solute and solvent molecules. 4. Conclusion A comprehensive investigation on the phosphafluorenes has been performed. From the study, the main conclusions can be drawn as follows: (1) The DFT/B3LYP method with the 6-31G(d) basis set is an appropriate method to determine structural and optical properties of phosphafluorenes. (2) The planar biphenyl skeleton has the largest contribution to LUMO orbitals, while the substitutions of O, S, and Se atoms have great influence on the electron density distributions of HOMO orbitals due to their strong electron-withdrawing effect. (3) After the chemical modifications of the P atom with O, S, or Se atoms, the HOMOs increase subsequently,

whereas the LUMOs remains unchanged, enabling independent tuning of HOMOs and LUMOs. (4) The calculated values of the IPs and EAs show that these compounds can be used as electron-transport and hole-transport materials simultaneously according to different substituents on the P atom of phosphafluorene, and the narrow differences between the λhole and λelectron were found in the studied molecules, especially for the MeSP, which can act as an excellent ambipolar material. (5) The phosphafluorenes exhibit blueshifted absorption spectra and red-shifted emission spectra in comparison with carbazoles, suggesting the larger rearrangement of phophafluorenes upon photoexcitation. (6) The methyl and phenyl substitutions have limited influences on the structures and properties of the phosphafluorenes, while the chemical modifications of the P atom with O, S, or Se atoms have great impacts, and phosphafluorenes can be used as blue-light-emitting materials with short radiative lifetimes and also as host materials or sensors with very long radiative times, high 3Eg, and ambipolar character. This study suggests that phosphafluorenes with many interesting properties are

Theoretical Studies of the Properties of Phosphafluorenes good candidates for optoelectronic application and worthy of experimental investigation. Acknowledgment. This work was financially supported by the National Basic Research Program of China (973 Program, 2009CB930601), the National Natural Science Foundation of China (20804020, 60976019, and 20974046), the Program for New Century Excellent Talents in University (NCET-07-0446), the Natural Science Foundation of Jiangsu College Council (Grant No. 08KJB150012), the Scientific and Technological Innovation Teams of Colleges and Universities in Jiangsu Province (TJ207035), and the Nanjing University of Posts and Telecommunications (NY207161). References and Notes (1) Gross, M.; Muller, D. C.; Nothofer, H. G.; Scherf, U.; Neher, D.; Brauchle, C.; Meerholz, K. Nature 2000, 405, 661–665. (2) Kim, J. Y.; Lee, K.; Coates, N. E.; Moses, D.; Nguyen, T. Q.; Dante, M.; Heeger, A. J. Science 2007, 317, 222–225. (3) Senechal-David, K.; Hemeryck, A.; Tancrez, N.; Toupet, L.; Williams, J. A.; Ledoux, I.; Zyss, J.; Boucekkine, A.; Guegan, J. P.; Le, B. H.; Maury, O. J. Am. Chem. Soc. 2006, 128, 12243–12255. (4) Liu, Y.; Miao, Q.; Zhang, S. W.; Huang, X. B.; Zheng, L. F.; Cheng, Y. X. Macromol. Chem. Phys. 2008, 209, 685–694. (5) Su, H. C.; Fadhel, O.; Yang, C. J.; Cho, T. Y.; Fave, C.; Hissler, M.; Wu, C. C.; Reau, R. J. Am. Chem. Soc. 2006, 128, 983–995. (6) Matano, Y.; Imahori, H. Org Biomol Chem 2009, 7. (7) Hissler, M.; Lescop, C.; Reau, R. C. R. Chim. 2008, 11, 628–640. (8) Crassous, J.; Reau, R. Dalton Trans. 2008, 6865–6876. (9) Durben, S.; Nickel, D.; Kruger, R. A.; Sutherland, T. C.; Baumgartner, T. J. Polym. Sci., Part A: Polym. Chem. 2008, 46, 8179–8190. (10) Hobbs, M. G.; Baumgartner, T. Eur. J. Inorg. Chem. 2007, 3611– 3628. (11) Jin, Z.; Lucht, B. L. J. Am. Chem. Soc. 2005, 127, 5586–5595. (12) Smith, R. C.; Protasiewicz, J. D. J. Am. Chem. Soc. 2004, 126, 2268–2269. (13) Sanji, T.; Shiraishi, K.; Tanaka, M. Org. Lett. 2007, 9, 3611–3614. (14) Hay, C.; Hissler, M.; Fischmeister, C.; Rault-Berthelot, J.; Toupet, L.; Nyulaszi, L.; Reau, R. Chem.sEur. J. 2001, 7, 4222–4236. (15) Su, H. C.; Fadhel, O.; Yang, C. J.; Cho, T. Y.; Fave, C.; Hissler, M.; Wu, C. C.; Reau, R. J. Am. Chem. Soc. 2006, 128, 983–995. (16) Neumann, T.; Dienes, Y.; Baumgartner, T. Org. Lett. 2006, 8, 495– 497. (17) Dienes, Y.; Durben, S.; Karpati, T.; Neumann, T.; Englert, U.; Nyulaszi, L.; Baumgartner, T. Chem.sEur. J. 2007, 13, 7487–7500. (18) Baumgartner, T.; Neumann, T.; Wirges, B. Angew. Chem., Int. Ed. 2004, 43, 6197–6201. (19) Makioka, Y.; Hayashi, T.; Tanaka, M. Chem. Lett. 2004, 33, 44– 45. (20) Chen, R. F.; Zhu, R.; Fan, Q. L.; Huang, W. Org. Lett. 2008, 10, 2913–2916.

J. Phys. Chem. A, Vol. 114, No. 10, 2010 3667 (21) Yang, B.; Kim, S. K.; Xu, H.; Park, Y. I.; Zhang, H. Y.; Gu, C.; Shen, F. Z.; Wang, C. L.; Liu, D. D.; Liu, X. D.; Hanif, M.; Tang, S.; Li, W. J.; Li, F.; Shen, J. C.; Park, J. W.; Ma, Y. G. ChemPhysChem 2008, 9, 2601–2609. (22) Yang, Y. T.; Geng, H.; Yin, S. W.; Shuai, Z. G.; Peng, J. B. J. Phys. Chem. B 2006, 110, 3180–3184. (23) Hannewald, K.; Bobbert, P. A. Appl. Phys. Lett. 2004, 85, 1535– 1537. (24) Hannewald, K.; Stojanovic, V. M.; Schellekens, J. M.; Bobbert, P. A.; Kresse, G.; Hafner, J. Phys. ReV. B 2004, 69. (25) Nelsen, S. F.; Trieber, D. A.; Ismagilov, R. F.; Teki, Y. J. Am. Chem. Soc. 2001, 123, 5684–5694. (26) Malagoli, M.; Bredas, J. L. Chem. Phys. Lett. 2000, 327, 13–17. (27) Lin, B. C.; Cheng, C. P.; You, Z. Q.; Hsu, C. P. J. Am. Chem. Soc. 2005, 127, 66–67. (28) Marcus, R. A. J. Chem. Phys. 1965, 43, 679. (29) Macus, R. A.; Sutin, N. Biochim. Biophys. Acta 1985, 811, 265– 322. (30) Marcus, R. A. ReV. Mod. Phys. 1993, 65, 599. (31) Nelsen, S. F.; Blomgren, F. J. Org. Chem. 2001, 66, 6551–6559. (32) Hutchison, G. R.; Ratner, M. A.; Marks, T. J. J. Am. Chem. Soc. 2005, 127, 2339–2350. (33) Yin, S. W.; Chen, L. P.; Xuan, P. F.; Chen, K. Q.; Shuai, Z. J. Phys. Chem. B 2004, 108, 9608–9613. (34) Becke, A. D. J. Chem. Phys. 1993, 98, 5648. (35) Lee, C.; Yang, W.; Parr, R. G. Phys. ReV. B 1988, 37, 785. (36) Liu, Y. L.; Feng, J. K.; Rent, A. M. J. Comput. Chem. 2007, 28, 2500–2509. (37) Zou, L. Y.; Ren, A. M.; Feng, J. K.; Liu, Y. L.; Ran, X. Q.; Sun, C. C. J. Phys. Chem. A 2008, 112, 12172–12178. (38) Alyea, E. C.; Ferguson, G.; F, G. J. Acta Crystallogr., Sect. C 1992, 48, 959–961. (39) Kushto, G. P.; Watkins, N. J.; Makinen, A. J.; Kafafi, Z. H. J. Phys. Chem. B 2007, 111, 5794–5802. (40) Adkine, P.; Cantat, T.; Deschamps, E.; Ricard, L.; Mezailles, N.; Le, F. P.; Geoffroy, M. Phys. Chem. Chem. Phys. 2006, 8, 862–868. (41) Levesque, I.; Bertrand, P. O.; Blouin, N.; Leclerc, M.; Zecchin, S.; Zotti, G.; Ratcliffe, C. I.; Klug, D. D.; Gao, X.; Gao, F. M.; Tse, J. S. Chem. Mater. 2007, 19, 2128–2138. (42) Iraqi, A.; Wataru, I. Chem. Mater. 2004, 16, 442–448. (43) Chen, R. F.; Zheng, C.; Fan, Q. L.; Huang, W. J. Comput. Chem. 2007, 28, 2091–2101. (44) Curioni, A.; Boero, M.; Andreoni, W. Chem. Phys. Lett. 1998, 294, 263–271. (45) Pan, J. H.; Chiu, H. L.; Wang, B. C. J. Mol. Struct.: THEOCHEM 2005, 725, 89–95. (46) Rienstra-Kiracofe, J. C.; Tschumper, G. S.; Schaefer, H. F.; Nandi, S.; Ellison, G. B. Chem. ReV. 2002, 102, 231–282. (47) Li, X. F.; Cai, Z. L.; Sevilla, M. D. J. Phys. Chem. A 2002, 106, 1596–1603. (48) Lukes, V.; Aquino, A.; Lischka, H. J. Phys. Chem. A 2005, 109, 10232–10238. (49) Zhang, C. R.; Liu, Z. J.; Chen, Y. H.; Chen, H. S.; Wu, Y. Z.; Yuan, L. H. J. Mol. Struct.: THEOCHEM 2009, 899, 86–93.

JP911624V