Structural, Electronic and Nonlinear Optical Properties of Novel Derivatives of 9,12-Diiodo-1,2-dicarba-closo-dodecaborane: Density Functional Theory Approach

1 Introduction

Iodine (I) substituted at boron (B) atoms in carboranes has been intensively studied as precursors of a large number of B-carboranes [1], [2], [3]. Structural properties of some B-iodo-carboranes were also investigated [4], [5], [6]. Previously, boron-based 9,12-diiodo-1,2-dicarba-closo-dodecaborane(12) (C₂H₁₀B₁₀I₂) was reported [7] as an individual compound for the structural properties of this fascinating compound. Several iodinated ortho-carborane crystal structures were comprehensively investigated during the last few years, and the diversity of the intermolecular interactions was revealed, which these materials may adopt in the solid state [8]. Furthermore, these fascinating complexes have been reported as stable compounds to several chemical conditions and also illustrate a huge potential to incorporate with several substituents. The ortho-carboranes (1,2-dicarba-closo-dodecaboranes) had been described as a prospective building blocks in supramolecular chemistry [9], [10]. Although a large number of iodinated carboranes and decaboranes are available, still they have a lot of potentials to be explored, especially the nonlinear optical (NLO) response of these compounds. Previously, it has been shown that because of better phase matching and increased stability, two-dimensional chromophores having large off-diagonal β tensor components have advantage over one-dimensional chromophores [11], [12]. Nowadays, compounds with great off-diagonal β tensor components are being recommended as they can offer large macroscopic NLO responses. Hitherto, NLO properties of numerous two-dimensional compounds were studied by substituting donor and/or acceptor in V-shaped molecules [12], [13]. It was shown that the angle between the intramolecular charge transfer (ICT) axis is an important parameter to check the relative magnitude of off-diagonal and diagonal components. In the current study of dicarba-closo-dodecaborane derivatives, the CT comprises both sides of dicarba-closo-dodecaborane. Additionally, the work of Coe et al. showed that large off-diagonal β tensor components are intensely interconnected to oscillator strength, low-lying energy excited states with the electron transition dipole moment between two states that are perpendicular to the dipolar axis [13].

Earlier, ICT has been studied through bridging effect on first hyperpolarizability [14]. In some other studies, NLO response properties of various ICT compounds were explored [15], [16], [17]. There is great demand to improve new organic-inorganic hybrid compounds that can associate benefits of organic materials (high NLO efficiency) as well as inorganic materials (good stability, wide transparency range). Owing to good thermal stability of dicarba-closo-dodecaborane and its versatile nature, they are being used in various functional materials [18]. It is anticipated that dicarba-closo-dodecaborane derivatives might lead to better NLO properties because of their V-shaped structures. In the current study, we have designed organic-inorganic hybrid proficient NLO-phores. The present investigations aim to explore the structure-property relationship of V-shaped dicarba-closo-dodecaborane derivatives and their potential as NLO-phores.

The nonlinear response properties are characterized by second- and third-order polarizabilities. The NLO materials have important applications in many fields of modern hi-tech society including optical data processing and frequency doubling. Over the past few years, our group has reported many compounds with efficient NLO response [19], [20]. Similar to our several previous studies [21], we have calculated the static total second-order polarizabilities (β_tot) of compounds 1–3 to check their possible NLO response. Furthermore, a comparison of β_tot and β_vec has been done and presented in Table 5 of Section 3.4 of the Results and Discussion.

It is also important to mention that the amplitude of β_tot is always positive irrespective of the sign of individual tensorial components, which cannot be measured experimentally, whereas the amplitude of β_vec may be positive or negative depending on the direction of dipole moment and charge transfer process for typical molecular system [15], [22]. The relation between β_vec and β_tot presents essential evidence about the direction of charge transfer in the molecules [15], [23] and specified by equation:

In the above equation, θ is the angle between the vector formed by the β_vec components and the dipole moment vector. Usually, this ratio if near to one indicates a unidirectional charge transfer within the molecular systems where β_tot and β_vec carry similar amplitudes.

In the present study, we have selected 9,12-diiodo-1,2-dicarba-closo-dodecaborane(12) (C₂H₁₀B₁₀I₂) [7] as parent molecule (named as 1 in the text) and derived two new compounds ( 2 and 3, respectively) with the aim to tune the optoelectronic and NLO properties. In compound 2, 1-nitro-4-propynyl-benzene group was substituted at the X-positions, while in compound 3, 1-amino-4-propynyl-benzene has been attached at the X-positions of the parent compound ( 1); see Figure 1. The purpose of substituting 1-nitro-4-propynyl-benzene group (strong electron acceptor) and 1-amino-4-propynyl-benzene (strong electron donor) at the X-positions of the parent compound is to explore the effect of acceptor and donor ligands on the various properties of interest, respectively. Both the 1-nitro-4-propynyl-benzene group and 1-amino-4-propynyl-benzene groups are representative of the best electron donor and electron acceptor groups, respectively, which are often used as prototypes to check the electron donor-acceptor effects in push-pull molecules [24]. The aforementioned groups are used as general representative of acceptor and donor moieties of compounds to establish the electronic communication with 9,12-diiodo-1,2-dicarba-closo-dodecaborane(12). We have explored the structural, frontier molecular orbitals, optical spectra [absorption (λ_a) and emission (λ_e)], and NLO properties of the targeted compounds by density functional theory (DFT) [25] and time-dependent DFT (TD-DFT) [26], [27], [28], [29]. Moreover, the β_tot and β_vec values of the studied compounds 1–3 have been evaluated in detail to understand the direction of charge transfer in the molecules.

Figure 1:

The optimized structures and schematic diagram of all derivatives investigated in the current study.

A contemporary survey of literature showed that there is no such computational study carried out heretofore on these studied compounds. This is the first time that we are going to have an in-depth study of 9,12-diiodo-1,2-dicarba-closo-dodecaborane(12) (C₂H₁₀B₁₀I₂) and two of its novel derivatives; see Figure 1.

2 Computational Details

In the present study, B3LYP method was adopted to optimize the ground state S₀ geometries of molecules because the hybrid functional B3LYP of DFT has been widely used to reproduce the experimental electronic [30], [31] and geometric parameters [32], [33], [34], [35], [36], [37], [38]. The basis set Gen is a hybrid basis set including 6-31G** for hydrogen, carbon, boron, and oxygen atoms, while LANL2DZ is for iodine atoms. Preat et al. revealed that the B3LYP method is suitable for the ground state (S₀) geometry optimizations [39], [40]. Moreover, Huong and co-workers confirmed that the B3LYP/6-31G** level is a decent approach to reproduce the experimental crystal data as well as rational for the electronic and charge transport properties [41]. The excited state (S₁) geometries were optimized at TD-DFT [42] using the TD-B3LYP/6-31G** (LANL2DZ) level of theory. Then the same TD-DFT level was applied to evaluate the absorption and emission spectra, which was already verified to be a proficient method [29]. It is important to note down here that many critical studies indicated that B3LYP is not appropriate for the calculation of NLO polarizabilities as reported by Champagn et al. [43], [44]. Recently, Misra et al. [15] reported that the CAM-B3LYP and MP2 are suitable to calculate the NLO response, whereas the CAM-B3LYP is considered more appropriate for NLO response without compromising the computing cost and accuracy. Previously, the hybrid functional CAM-B3LYP has been reported as useful for evaluating the NLO response of several charge transfer base molecule [45], [46], [47]. To overcome such discrepancies, we have also optimized compounds 1–3 at CAM-B3LYP method along with relatively larger basis set 6-31+G** (LANL2DZ) for ground state. The static second-order polarizabilities at CAM-B3LYP/6-31+G** (LANL2DZ) have been evaluated for compounds 1–3 initially optimized at the B3LYP/6-31G** (LANL2DZ) level of theory and are tabulated in Tables 5 and S2 of the Supporting Information. Furthermore, we also compute the static second-order polarizabilities at CAM-B3LYP/6-31+G** (LANL2DZ) for compounds 1–3 optimized at the same level of theory and presented in Table 5.

It is found that there are noticeable differences among the amplitudes of β_tot and β_vec at different levels of theory. At two different levels of theory (B3LYP versus CAM-B3LYP), the approximate percentage differences among the amplitudes of β_tot and β_vec are found to be 32 %, 50 %, and 14 % for compounds 1, 2, and 3, respectively. A detailed analysis is given in Table S1 of the Supporting Information.

A well-known finite field approach combined with DFT method is used for the calculation of NLO response. In finite field method, a static electric field (F) is applied, and the energy (E) of the molecule is given by following equation:

where i, j, and k label the x, y, and z components, respectively. The total energy of the molecule is represented by E⁽⁰⁾ without any field, while μ and α are the dipole moment and linear polarizability. Similarly, the β and γ are the second- and third-order nonlinear polarizabilities, respectively. It is obvious from (1) that differentiating E with respect to F attains the 27 components of β. According to Kleinman symmetry, β_xyy = β_yxy = β_yyx, β_yyz = β_yzy = β_zyy, …, likewise other permutations also take the same value. So the 27 components of second-order nonlinear polarizability (β) can be reduced to 10 components.

The values of β_tot and β_vec are calculated by the following equations:

where

All these quantum chemical calculations were performed using Gaussian09 package [48].

3 Results and Discussion

3.2 Electronic Properties

The ground state and excited state energies of the frontier molecular orbitals and energy gaps of borane-based derivative compounds 1–3 have been tabulated in Table 3. The ground state and excited state highest occupied molecular orbital (HOMO) energies (E_HOMO), lowest unoccupied molecular orbital (LUMO) energies (E_LUMO), and HOMO-LUMO energy gaps (E_g) of compounds 1–3 were calculated at the B3LYP/6-31G** and TD-B3LYP/6-31G** levels of theory, respectively, which are tabulated in Table 3. The trend in the E_HOMO is 3 > 1 > 2, while in the E_LUMO, it is 1 > 3 > 2, for the ground and excited states. The tendency in the E_g was observed as 1 > 3 > 2 at ground state, while it is 1 > 2 > 3 at excited state. Here electron and hole injection energies were calculated to understand the charge injection behavior of the studied derivatives. The electron injection energy was calculated as [= −E_LUMO − (−work function of metal)]. The work function of aluminum (Al) is −4.08 eV. The electron injection energy was found to be 3.24 eV [= −0.84 − (−4.08)], 0.75 eV [= −3.33 − (−4.08)], and 2.19 eV [= −1.89 − (−4.08)] from compounds 1–3 to Al electrode, respectively. It is expected that 2 might be a better electron transport contender having the smallest electron injection barrier to Al electrode. Moreover, the smaller and low-lying E_LUMO of 2 than the other counterparts revealed that the prior compound might be thermodynamically more stable and charge transport cannot be quenched by losing the electrons. The hole injection energy of compounds 1–3 has been predicted, i.e. 2.55 eV [= −4.08 − (−6.63)], 2.71 eV [= −4.08 − (−6.79)], and 1.70 eV [= −4.08 − (−5.78)] from compounds 1–3 to the Al electrode, respectively. It is expected that 3 would be a better hole transport material. The charge distribution patterns of the HOMOs and LUMOs of boron-based derivatives 1–3 have been illustrated in Figure 2 for ground states, while for excited states, the HOMO and LUMO formations have been shown in Figure S2 of the supporting information.

Table 3:

The HOMO energies (E_HOMO), LUMO energies (E_LUMO), and HOMO-LUMO energy gaps (E_g) in eV for ground and first excited states computed at the B3LYP/Gen^a and TD-B3LYP/Gen^a levels of theory, respectively.

Complexes	Ground states			First excited states
	EHOMO	ELUMO	Eg	EHOMO	ELUMO	Eg
1	−6.63	−0.84	5.78	−5.73	−3.61	2.12
2	−6.79	−3.33	3.47	−6.10	−4.37	1.73
3	−5.78	−1.89	3.89	−5.27	−3.65	1.62

1. ^aGen = 6-31G** for H, C, B, and O atoms, while Gen = LANL2DZ for I atoms.

Figure 2:

Distribution patterns of the HOMOs and LUMOs of compounds 1–3 at the ground states.

3.3 Optical Properties

The maximum absorption wavelengths (λ_a), emission wavelengths (λ_e), oscillator strengths (f), and major transitions involved at the TD-B3LYP/6-31G** (LANL2DZ) level of theory have been tabulated in Table 4 and Figure 3. It has been observed that introducing the 1-nitro-4-propynyl-benzene at the X-positions of the parent compound leads to a 92 nm red shift as in 2 in λ_a. Moreover, the introduction of 1-amino-4-propynyl-benzene at the X-positions of the parent compound leads to a 35 nm red shift ( 3) in λ_a. The maximum λ_a peaks of compounds 1–3 were observed at 241, 333, and 276 nm with the major transitions H-1 ⟶ L, H-3 ⟶ L+1, and H-1 ⟶ L, respectively. It was also found that the introduction of 1-nitro-4-propynyl-benzene at the X-positions of the parent compound leads to a 106 nm blue shift as in 2 in λ_e. Moreover, the introduction of 1-amino-4-propynyl-benzene at the X-positions of the parent compound leads to a 33 nm red shift ( 3) in λ_e. The maximum λ_e peaks of compounds 1–3 were observed at 532, 426, and 499 nm with the major transitions L ⟶ H-1, L ⟶ H-4, and L ⟶ H-5, respectively. The Stokes shift from the absorption to the emission wavelengths was observed as 291, 93, and 223 nm for compounds 1–3, respectively.

Table 4:

The absorption (λ_a) and emission wavelengths (λ_e) (nm), oscillator strengths (f), and major transitions of compounds 1–3 calculated at TD-B3LYP/Gen^a level of theory.

Complexes	f	λa	Transition	f	λe	Transition
1	0.0314	241	H-1 ⟶ L	0.0317	532	L ⟶ H-1
2	0.0545	333	H-3 ⟶ L+1	0.3169	426	L ⟶ H-4
3	0.2649	276	H-1 ⟶ L	0.333	499	L ⟶ H-5

1. Gen = 6-31G** for H, C, B, and O atoms, while Gen = LANL2DZ for I atoms.

Figure 3:

The absorption and emission spectra of compounds 1–3 calculated using the TD-B3LYP/6-31G** (LANL2DZ) level of theory.

4 Conclusions

Hence, the current research study highlights numerous insights into the structure-property relationship of the above-entitled three novel borane-iodine compounds 1–3. The TD-DFT calculations show that 1-nitro-4-propynyl-benzene and 1-amino-4-propynyl-benzene at top terminal positions of the parent compound lead to a (92 and 35 nm) red shift and a (106 and 33 nm) blue shift for maximum absorption and emission wavelengths as in compounds 2 and 3, respectively. Likewise, frontier molecular orbitals study reveals the comprehensive intramolecular charge transfer from HOMO to LUMO. The smaller electron injection energy barrier (0.75 eV) of 2 as compared with the other compounds reveals that it would be a better n-channel contestant. Though the smaller hole injection energy barrier of compound 3 as compared with the other counterparts showed that the aforementioned derivative might be a better p-channel. The calculation of the second-order NLO polarizabilities (β_tot) shows that compounds 2 and 3 possess the β_tot amplitudes of 3029 and 4069 a.u., respectively, with the CAM-B3LYP method, which are reasonably larger than similar prototype molecules. The NLO anisotropies are 0.63, 0.34, and 0.44 for compounds 1–3, respectively, at the CAM-B3LYP level of theory, which revealed that the charge transfer process is not unidirectional because of the V-shape of the studied molecular systems. Thus, the NLO amplitudes and NLO anisotropies indicate that the above-entitled compounds might have significant potential to use them in NLO applications. Our quantum chemical exploration of the structural, electro-optical, and NLO properties demonstrates that these materials with trustworthy properties of interests might be prospective competitors for the electro-optical device as well as NLO applications.