1. INTRODUCTION
During the past decades, people have paid great efforts to explore more high-performance nonlinear optical (NLO) materials because of their wide applications in the field of information processing, photoelectric, optical data storage and many others.1-7 Currently, numerous types of molecules with large NLO response have been studied, including structures of the donor-π bridge-acceptor type,8,9the transition metal-ligand structures,10,11and complexes with loosely bound excess electron.12,13In 2004, Li et al. reported that the molecular cluster anions (FH)2{e}(HF) and (H2O)3{e} with excess electron exhibited significantly large nonlinear optical responses.14,15This opens up a new direction for designing novel compounds with considerable NLO properties. These novel compounds are a class of nontraditional ionic salts, named electrides, where the anionic sites are occupied solely by electrons.16 Studies have shown that alkali metal doping is one of the most common methods for designing electrides.12 Recently, the use of superalkalis instead of alkali atoms to design new electrides with larger hyperpolarizabilities (β 0) has attracted much attention.17,18
Superalkalis19 are a well-known class of superatoms20 that possess lower ionization potential (IP) values than alkali metal atoms, thus their valence electrons are more likely to be polarized by ligands and form loosely bound excess electrons. Therefore, the interaction of superalkali with appropriate ligands can generate more diffuse excess electron, leading to new superalkali-doped systems with larger NLO response.
Although the excess electron generated by the polarization of organic ligands can greatly increase the β 0 value of a molecular system, the presence of the loosely bound excess electrons makes the thermal stability of the molecule unsatisfactory. Besides, NLO materials in the infrared and deep-ultraviolet (deep-UV) working wavebands are currently the research hotspots.3 Consequently, finding a system with high stability, large hyperpolarizability and working area in the infrared or deep-ultraviolet regions has become an important research topic in the field of NLO.
Recently, new two-dimensional carbon allotrope-graphynes with special structural features and thermal stability have attracted the interest of many researchers from different areas. As early as 1987, Baughman et al.21 first proposed that graphynes (GYs) are a series of one atom thick carbon allotropes composed of sp- and sp2-hybridized carbon atoms. Two hybridization states of carbon atoms enable GYs to have many excellent properties, including extended π-conjugation, uniformly distributed pores, tunable electronic properties, good chemical stability and large surface area.22,23Four main types of graphyne namely α -, β -, γ -, and 6, 6, 12-graphynes have been identified.21γ -graphyne is the most widely studied form of GYs, especially graphdiyne (GDY), which is formed by the self-assembly of hexagonal rings and acetylenic groups and has a largely delocalized π-conjugated surface. Although many scientists have attempted to prepare GYs, it was not until 2010 that Li’s group successfully synthesized large-area films of graphdiyne on a copper substrate via a cross-coupling reaction.24 Due to the fascinating structures and particular electronic properties, graphyne materials show promising applications in catalysts, hydrogen storage, anode materials, optoelectronic devices, biomedicine and therapy,etc .22,23,25-29
More interestingly, the graphyne molecules have recently aroused extensive attention of theoretical researchers in the field of nonlinear optics. In 2016, Chakraborti30theoretically investigated the NLO properties of donor-acceptor substituted graphyne structures. In the same year, the hyperpolarizabilities of graphdiyne functionalized by the alkali metal atom etc . were investigated.31 Very recently, Li et al. designed a variety of promising novel GDY-based NLO materials.32-35 Instead of replacing hydrogen of GDY with alkali metal atom,31 Li et al. theoretically confirms that alkali atom doping is a viable approach to increase the β 0 value of GDY.32,33Besides, superalkaline-earth metal M3F (M = Li, Na, and K), and superalkali Li3NM and M2X (M = Li, Na, K and X = F, Cl, Br) can also be adsorbed on the GDY surface, respectively, to produce new complexes with largeβ 0values.34-36 Thus, the GDY and superalkaline-earth metal/superalkali would be a good combination for designing novel NLO materials.
Considering characteristics of GYs including thermal stability, deep-ultraviolet transparency, and large pores, we have designed a series of superalkali salts of graphynes, namely OM3+@GYs (M = Li, Na, and K; GYs include GY, GDY, and GTY) in the present work. Figure 1 shows the graphyne structure consists of hexagons connected by acetylenic (–C≡C–) rather than cumulative (=C=C=) linkages.37 According to the number of acetylenic linkers (–C≡C–), GYs can be classified into GY, GDY, and GTY. These representative structural models of delocalized π-conjugated graphynes were chosen here to combine with superalkalis (OLi3, ONa3, and OK3) to generate a new series of complexes. Note that these superalkalis have been experimentally synthesized,38-40 and they can serve as basic building blocks for new complexes.41-44 The evolution of their first hyperpolarizability with varying superalkali atom and pore size of graphyne has been analyzed in order to explore new high-performance NLO molecules. The structures, stability, and static first hyperpolarizabilities of the investigated complexes are explored by employing density functional theory (DFT). The frontier molecular orbital and atomic charge analyses indicate that OM3+@GYs have superalkali salt characteristics. Two influencing factors onβ 0 values, namely the atomic number of alkali atom M and pore size of graphyne, are discussed in detail. Results show that the combination of OLi3 and GTY (with a large pore) forms a planar stable structure possessing the largestβ 0 up to 6.5×105 au. This work proposes a new kind of high-performance deep-UV NLO molecules. It is extremely expected that this research can attract more experimental interest in designing novel carbon-based NLO materials in the near future.