A Nonlinear Optical Switchable Sulfate of Ultrawide Bandgap

Abstract

Open AccessCCS ChemistryRESEARCH ARTICLE1 Jul 2021A Nonlinear Optical Switchable Sulfate of Ultrawide Bandgap Yanqiang Li, Congling Yin, Xiaoyan Yang, Xiaojun Kuang, Jie Chen, Lunhua He, Qingran Ding, Sangen Zhao, Maochun Hong and Junhua Luo Yanqiang Li State Key Laboratory of Structural Chemistry, Fujian Institute of Research on the Structure of Matter, Chinese Academy of Sciences, Fuzhou, Fujian 350002 University of Chinese Academy of Sciences, Beijing 100049 Google Scholar More articles by this author , Congling Yin MOE Key Laboratory of New Processing Technology for Nonferrous Metals and Materials, Guangxi Key Laboratory of Optical and Electronic Materials and Devices, College of Materials Science and Engineering, Guilin University of Technology, Guilin, Guangxi 541004 Google Scholar More articles by this author , Xiaoyan Yang MOE Key Laboratory of New Processing Technology for Nonferrous Metals and Materials, Guangxi Key Laboratory of Optical and Electronic Materials and Devices, College of Materials Science and Engineering, Guilin University of Technology, Guilin, Guangxi 541004 Google Scholar More articles by this author , Xiaojun Kuang *Corresponding author: E-mail Address: [email protected] MOE Key Laboratory of New Processing Technology for Nonferrous Metals and Materials, Guangxi Key Laboratory of Optical and Electronic Materials and Devices, College of Materials Science and Engineering, Guilin University of Technology, Guilin, Guangxi 541004 Google Scholar More articles by this author , Jie Chen Beijing National Laboratory for Condensed Matter Physics, Institute of Physics, Chinese Academy of Sciences, Beijing 100190 Songshan Lake Materials Laboratory, Dongguan, Guangdong 523808 Google Scholar More articles by this author , Lunhua He Beijing National Laboratory for Condensed Matter Physics, Institute of Physics, Chinese Academy of Sciences, Beijing 100190 Songshan Lake Materials Laboratory, Dongguan, Guangdong 523808 Google Scholar More articles by this author , Qingran Ding State Key Laboratory of Structural Chemistry, Fujian Institute of Research on the Structure of Matter, Chinese Academy of Sciences, Fuzhou, Fujian 350002 University of Chinese Academy of Sciences, Beijing 100049 Google Scholar More articles by this author , Sangen Zhao *Corresponding author: E-mail Address: [email protected] State Key Laboratory of Structural Chemistry, Fujian Institute of Research on the Structure of Matter, Chinese Academy of Sciences, Fuzhou, Fujian 350002 University of Chinese Academy of Sciences, Beijing 100049 Fujian Science and Technology Innovation Laboratory for Optoelectronic Information of China, Fuzhou, Fujian 350108 Google Scholar More articles by this author , Maochun Hong State Key Laboratory of Structural Chemistry, Fujian Institute of Research on the Structure of Matter, Chinese Academy of Sciences, Fuzhou, Fujian 350002 Google Scholar More articles by this author and Junhua Luo *Corresponding author: E-mail Address: [email protected] State Key Laboratory of Structural Chemistry, Fujian Institute of Research on the Structure of Matter, Chinese Academy of Sciences, Fuzhou, Fujian 350002 University of Chinese Academy of Sciences, Beijing 100049 Fujian Science and Technology Innovation Laboratory for Optoelectronic Information of China, Fuzhou, Fujian 350108 Google Scholar More articles by this author https://doi.org/10.31635/ccschem.020.202000436 SectionsSupplemental MaterialAboutAbstractPDF ToolsAdd to favoritesTrack Citations ShareFacebookTwitterLinked InEmail Nonlinear optical (NLO) switchable materials have attracted intense attention because of their promising applications in optoelectronic devices. However, previous studies are mainly limited to molecular-based compounds that usually exhibit narrow bandgaps. Here, we report all-inorganic Li9Na3Rb2(SO4)7 as an ultrawide-bandgap NLO switchable material. To the best of our knowledge, this sulfate shows the widest bandgap (6.70 eV) among NLO switchable materials. Furthermore, it undergoes a reversible phase transition with a high NLO switching contrast of about 180, which is almost the highest among known NLO switchable materials. Variable-temperature X-ray and neutron powder diffractions reveal that the exceptional NLO switching contrast can be attributed to the NLO-active [SO4]2− anions translating and rotating on the local scale. This work will provide insights into the development of ultrawide-bandgap NLO switchable materials, which may open up unprecedented opportunities for emerging applications, such as ultrawide-bandgap switches, sensors, and optoelectronics. Download figure Download PowerPoint Introduction Nonlinear optical (NLO) switchable materials, in which NLO responses are reversibly converted among diverse stable states under external stimulation, have attracted considerable interest owing to their prospective application in optoelectronic devices, such as switches, sensors, data storage, information processing of smart devices, and optical communication.1–5 Over the past decades, researchers have devoted great effort to studying NLO switchable materials and found a number of NLO switchable materials in the liquid phase.6,7 Meanwhile, a variety of efficient strategies have been put forward to explore solid-state crystalline NLO switchable materials, such as cation exchange, conformational reorganization, and “push–pull” molecules via photochromism.8–10 The resultant materials include metal–organic framework compounds such as [(H2NMe2)2Cd3(C2O4)4]·MeOH·2H2O and NH2-MIL-53 (Al), anil crystals such as N-(4-hydroxy)-salicylidene-amino-4-(methylbenzoate) and N-(3,5-di-tert-butylsalicylidene)-4-aminopyridine, as well as complexes.8,9,11,12 Unfortunately, these materials usually show low NLO switching contrasts with limited reversibility. A new scheme, relying on structural phase transitions, especially those from centrosymmetric (CS) structures to noncentrosymmetric (NCS) structures, has been recently employed to inspire sharp changes of NLO responses. In 2006, Mercier et al.13 first reported the transformation-induced NLO switchable phenomenon in a molecular-based compound, [{H3N(CH2)2SS(CH2)2NH3}PbI5]·H3O. Subsequently, a series of molecular-based NLO switchable materials were discovered, such as [α-(H3N(CH2)2S-S(CH2)2NH3)BiI5,14 (Hdabco+)(CF3COO−),15 bis(imidazolium) L-tartrate,16,17 2-(hydroxymethyl)-2-nitro-1,3-propanediol,18 (C4H10N)(CdCl3),19 dipropylammonium trichloroacetate,20 (R-CTA)2CuCl4,21 and (S-CTA)2CuCl4.21 However, despite excellent NLO switchable performance, their bandgaps are relatively narrow on account of π-conjugated electrons, d-d or f-f transition, such as [α-(H3N(CH2)2S-S(CH2)2NH3]BiI5 (∼1.84 eV),14 bis(imidazolium) L-tartrate (∼5.21 eV),17 and dipropylammonium trichloroacetate (∼3.82 eV).20 As a result, these molecular-based NLO switchable materials are merely used in visible and ultraviolet spectral regions. With respect to the deep ultraviolet spectral region (bandgap more than 6.20 eV), they play a unique and crucial role in a number of modern instruments, but it almost approaches the theoretical bandgap limit of NLO switchable materials.22,23 Hence, to date, there is still a huge opening for the application of ultrawide-bandgap NLO switchable materials. Recently, our group has reported a series of ultrawide-bandgap NLO sulfates, such as Li8NaRb3(SO4)6•2H2O24 and (NH4)2NaLi(SO4)2.25 Particularly, (NH4)2NaLi(SO4)2 exhibits considerable NLO responses aroused by nearly parallel nonbonding O 2p orbitals of NLO-active [SO4]2− groups, but (NH4)2NaLi(SO4)2 is not NLO switchable.25 As a result of our continuous studies on ultrawide-bandgap NLO sulfates, we successfully synthesized a new sulfate from the high-temperature melt, namely Li9Na3Rb2(SO4)7 ( I) with an ultrawide bandgap of 6.70 eV, which presents the widest working spectral range for NLO switchable materials ever known. Uniquely, I undergoes a reversible phase transition from the NCS state (P3) to CS state (P 3 ¯ ) with a high NLO switching contrast of about 180. Experimental Methods Synthesis of I Polycrystalline I was synthesized by the traditional high-temperature solid-state method using the Li2SO4•H2O (99.0%), Rb2SO4 (99.0%), and Na2SO4 (99.0%) in the stoichiometric molar ratio. The mixture was fully ground and transferred to a muffle furnace. After that, the mixture was preheated at 723 K for about 24 h. Subsequently, the products were gradually heated to 823 K and sintered at 823 K for about 96 h with several intermediate mixings and grindings. The purity of as-synthesized polycrystalline I samples was checked by powder X-ray diffraction (XRD) analyses, which were performed on a Rigaku MiniFlex 600 diffractometer with a Cu Kα radiation (λ = 1.5418 Å) in the 2θ range of 10°–70° with the scan rate of 1° min−1 at room temperature. Crystal growth of I Single crystals of I were grown by the top-seeded solution growth method. The polycrystalline I powders were put into a Φ 45 × 45 mm platinum crucible and melted at 973 K in a temperature-programmable furnace. This melt was held at 973 K for more than 24 h to form a transparent solution. Then, the solution was quickly cooled to 900 K, and a platinum wire was dipped into the solution. The solution was rapidly cooled to the saturation temperature, and then I crystals formed on the platinum wire via cooling the solution at a rate of 2 K h−1. After the crystal growth process was completed, I crystals were drawn out of the solution and served as seed crystals. In the following crystal growth procedure, we first determined the crystallized temperature of I samples by the tentative seed crystal strategy using as-grown I crystals. A seed of good optical quality was introduced into the solution at 10 K above the crystallized temperature. This temperature was held for about 60 min and then dropped to the crystallized temperature in 10 min. The centimeter-grade I single crystal was grown with the solution cooling down at a rate of 0.2–0.5 K per day. Finally, the I crystal was pulled out of the solution. Structure characterization The variable temperature (VT)-XRD analyses were carried out on a Panalytical X’Pert PRO diffractometer with an Anton Parr HTK 1200 N high-temperature attachment. The VT-XRD data over 10–80° 2θ range was collected from room temperature to 783 K using a temperature increment of 100 K except for the 733–783 K temperature range close to the phase transition region with an interval of 10 K. Time-of-flight neutron powder diffraction (NPD) data was collected from 10 to 783 K at several temperatures at the general-purpose powder diffractometer at the China Spallation Neutron Source. The VT-XRD data and time-of-flight NPD data were both refined by Rietveld methods using TOPAS-Academic software26 and Jana2006 software.27 The diffraction data of I at 100(2) K was collected by the single-crystal XRD on a Bruker APEX II CCD diffractometer with a monochromatic Mo Kα radiation. The program APEX3 was used to perform cell refinements and data reductions.28 The structure was established by the direct method with the program SHELXS and refined with the least-squares program SHELXL.29 Final refinements included anisotropic displacement parameters. The solved structure was verified by the program PLATON,30 and no higher symmetries were recommended. Property measurements Elemental analyses were performed by using a Jobin Yvon Ultima 2 inductively coupled plasma optical emission spectrometer. The thermal stability was investigated by the differential scanning calorimetric (DSC) and thermogravimetric (TG) analyses on a NETZSCH STA 449C simultaneous thermal analyzer instrument. About 22 mg I samples were placed in Al2O3 crucibles, heating and cooling at a rate of 10 K min−1 from room temperature to 973 K and from 660 to 820 K. The temperature dependence of the dielectric constant measurements performed on a TongHui TH2828 analyzer were in the temperature range of 680–790 K, with a heating and cooling rate of about 10 K min−1. The transmittance spectrum was acquired on the PerkinElmer LAMBDA 950 UV/Vis/NIR spectrophotometer in the wavelength range of 185–800 nm on a single-crystal plate. Powder second harmonic generation (SHG) measurements were carried out by the Kurtz–Perry method with a Q-switched Nd:YAG laser at the wavelength of λ = 1064 nm.31 Polycrystalline I samples were ground and sieved into a series of particle size ranges: 53–75, 75–125, 125–180, 180–250, and 250–300 μm. Infrared spectra were obtained on a LAMBDA 900 infrared instrument with the wavenumber in the range of 400–4000 cm−1. The bands ranging from 1006 to 1223 cm−1 can be attributed to the stretching vibrations of the [SO4]2− tetrahedra ( Supporting Information Figure S2). The absorption from 423 to 646 cm−1 was assigned to the bending modes of sulfate ( Supporting Information Figure S2). Results and Discussion Pure I samples were synthesized by the high-temperature solid-state method (Figure 1a). The inductively coupled plasma element analyses of I gave a molar ratio of Li:Na:Rb:S = 9.16∶2.71∶2.21∶6.78, which is well consistent with the compositions from the structural data. As illustrated in Figure 2a, DSC curves imply that I melts congruently around 874.8 K. As a result, bulk crystals of I were grown from its stoichiometric high-temperature melt by the top-seed technique. The as-grown I crystal was transparent with sizes of about 18 × 10 × 3 mm3 (Figure 1b). Interestingly, in the heating/cooling curves, a pair of relatively small peaks were observed at 771.7 (Tc) and 759.4 K (Figure 2a). At the same time, there was no obvious weight loss in the TG analysis (Figure 2a). Thus, we speculated that I undergoes a reversible phase transition at Tc, and then carried out additional DSC and TG analyses. As indicated by Figure 2b, DSC analyses can be reproducible, proving that a reversible phase transition occurs. For convenience, the phase above Tc was denoted as high-temperature phase (HTP), and the phase below Tc was marked as low-temperature phase (LTP). We also measured the temperature dependence of the dielectric constant of I under various frequencies. As shown in Figures 2c and 2d, the dielectric constant shows evident peak anomalies near Tc. Figure 1 | XRD patterns and crystal photograph of I. (a) Calculated and experimental XRD patterns of I. (b) Photograph of the as-grown I crystal. Scale bar, 10 mm. Download figure Download PowerPoint To further confirm the reversible phase transition, we collected single-crystal XRD data at 100 K and VT-XRD data from room temperature to 783 K. Considering neutrons are more sensitive to O atoms, and Li atoms have negative neutron scattering lengths in contrast with Na, Rb, and S atoms [neutron scattering lengths (fm) are: Li, −1.90; Na, 3.63; Rb, 7.09; O 5.803; S 2.847],32 the possible positions of O and Li atoms can be better determined with NPD data. We further collected time-of-flight NPD data of I at different temperatures. The VT-XRD data and time-of-flight NPD data were refined by the Rietveld method.33 As shown in Figure 3a, no obvious changes were seen in the VT-XRD patterns in the heating process. It indicated that there was no significant structural reorganization in the course of the phase transition. As illustrated in Figure 4a, the in situ time-of-flight NPD plots were similar from 10 to 763 K but changed dramatically at 783 K. Several intense reflections including (361), (171), (161), (251), and (440) nearly disappeared. Also, the background profiles displayed broad peaks below d = 1.5 Å below 763 K, and additional broad peaks formed within d = 2.0–2.7 Å at 783 K, which shows that structural disorder occurs initially at a local range of d = 1.0–1.5 Å and expands to a larger range of d = 2.0–2.7 Å with the increasing temperature. In view of the large neutron scattering length of O and the absence of broad peaks on the VT-XRD patterns, we can conclude that the structural disorder is most probably related to the thermally activated motion of [SO4]2− anions. In addition, the fitted lattice parameters and cell volumes from the VT-XRD (Figure 3b) and NPD (Figure 4b) data have abrupt changes around 743 K, consistent with the phase transition. Supporting Information Tables S1–S12 list the detailed crystallographic information. Figure 3 | VT-XRD data of I. (a) “Film plot” of VT-XRD data of I powders on heating. (b) VT cell parameters against the powder XRD data. The vertical dashed line marks the occurrence of phase transition. Download figure Download PowerPoint I crystallizes in the NCS space group of P3 (No. 143) in the LTP and the CS space group of P 3 ¯ (No. 147) in the HTP. As shown in Figure 5a, the crystal structure of LTP features a three-dimensional (3D) tetrahedral framework, in which tetrahedra have a basal plane nearly parallel to the ab plane and apical bonds parallel to the c-axis, composed of LiO4 and [SO4]2− tetrahedra. The Na+ and Rb+ are distributed in a well-ordered manner in the cavities to maintain the structural balance. There are nine crystallographically independent sites for S atoms in the LTP. The [S4O4]2− and [S6O4]2− tetrahedra are alternatively linked to LiO4 tetrahedra by sharing corners to construct hexagonal pillars (Figure 5b) around the three-fold axis. The [S5O4]2− tetrahedra occupy the central space of the pillars. The HTP can be considered as a locally melted phase of the LTP. In the structure of HTP, almost all atoms are partially occupied on two equivalent positions, disturbing the ordered alignment of [SO4]2− units. However, the local arrangements of [SO4]2− tetrahedra in the ab plane still retain the structural feature in the LTP, as shown in one possible [SO4]2− tetrahedra arrangement in Figure 5c. As compared to LTP (Figure 5d), the HTP structure features tilted [SO4]2− tetrahedra with the basal plane away from the ab plane. In addition, the [SO4]2− anions on the three-fold axis translate along the c-axis. Due to this shift, considerable lattice strains are formed in the ab plane, which leads to noticeable elongated [SO4]2− anions along the c-axis (Figures 5e and 5f). In order to reduce the strain, the hexagonal pillars formed by [SO4]2− anions on the ab plane are tilted locally to form a larger cavity, as reflected by its radius R increasing from 3.87 (LTP) to 4.03 Å (HTP). Owing to the reminiscent anionic arrangements in the LTP, the Li+, Na+, and Rb+ cations in the HTP are confined to occupy similar positions to the LTP, but they are more likely to move around and become more highly disordered on the local scale. Thus, the phase transition of I is similar to the melting behavior and can be called a quasi-melting phase transition. Figure 5 | Crystal structures of I. (a) Crystal structure of the LTP viewing along the c-axis. (b) The hexagonal pillars formed by [SO4]2− tetrahedra, highlighted by dashed yellow circle in (a). A possible local arrangement of [SO4]2− terahedra in the structure of HTP along (c) the c-axis and (e) the b-axis, in comparison with LTP along (d) the c-axis and (f) the b-axis. The green tetrahedra represent the [SO4]2− tetrahedra. R denotes the radius of the hexagonal pillar. For clarify, the Li+, Na+, and Rb+ cations are omitted in Figures 5c–5f. Download figure Download PowerPoint To obtain the accurate bandgap, a polished I crystal of about 5 × 5 × 1 mm3 (the inner picture in Figure 6a) was used for the test. As demonstrated in Figure 6a, I is transparent down to the spectral region below 185 nm, which corresponds to the bandgap of 6.70 eV. As far as we are aware, the bandgap of I is the widest among known NLO switchable materials, which far exceeds that of many molecular-based NLO switchable materials, such as [α-(H3 N(CH2)2 S-S(CH2)2NH3]BiI5 (∼1.84 eV),14 bis(imidazolium) L-tartrate (∼5.21 eV),17 2-(hydroxymethyl)-2-nitro-1,3-propanediol (∼6.20 eV),18 and dipropylammonium trichloroacetate (∼3.82 eV).20 Figure 6 | Optical properties of I. (a) Transmittance spectrum. The inner picture is a single-crystal I sample for measurements. (b) SHG intensity versus particle sizes at λ = 1064 nm. The red curve is drawn to guide the eyes and does not fit the data. (c) The VT-SHG intensity between SHG-on states and SHG-off states at λ = 1064 nm. The inner picture is the SHG signal at the SHG-off state. (d) Completely reversible and recoverable switching of SHG intensity. Download figure Download PowerPoint Given that I is CS in HTP and NCS in LTP, it is expected that I shows NLO switchable properties, that is, SHG-on states in LTP and SHG-off states in HTP. First, the SHG intensities of I were tested by the Kurtz–Perry method31 with a solid-state Q-switched Nd:YAG laser at 1064 nm at room temperature. As illustrated in Figure 6b, SHG intensities of I increase with increasing particle sizes, showing that I is phase-matchable. Besides, in the same particle size, the SHG intensity of LTP is about 1.3 times that of the benchmark NLO material KH2PO4 ( Supporting Information Figure S3), which is larger than that of some reported ultrawide-bandgap NLO sulfates, such as NH4NaLi2(SO4)2 (1.1 × KH2PO4),25 (NH4)2Na3Li9(SO4)7 (0.5 × KH2PO4),25 and Li8NaRb3(SO4)6·2H2O (0.5 × KH2PO4).24 After that, we performed VT-SHG measurements on a I single-crystal sample. As shown in Figure 6c, the SHG intensities stay nearly constant over a wide temperature range until Tc. When the temperature continues to increase, the SHG intensities drop rapidly and then completely disappear except quite weak noise errors. According to previous reports,9,18 the NLO switching contrast is defined as the ratio of SHG intensities at SHG-on states and SHG-off states. Therefore, the NLO switching contrast of I is about 180. Notably, this NLO switching contrast is higher than that of most solid-state crystalline NLO switchable materials, including metal–organic framework compounds such as [(H2NMe2)2Cd3(C2O4)4]·MeOH·2H2O (∼1.75)8 and NH2-MIL-53 (Al) (∼38),9 ionic salts such as (Hdabco+)CF3COO−) (∼35),15 bis(imidazolium hydrochlorate) dihydrate 18-crown-6 (∼12),34 and NH4[(CH3)4N]SO4·H2O (∼24),35 single-component plastic crystals such as 2-(hydroxymethyl)-2-nitro-1,3-propanediol (∼150)18 and [CH3C(NH2)2]ClO4 (∼62),36 as well as organic–inorganic hybrid compounds such as [C6H11NH3]2[CdCl4] (∼1.3)37 and [C4H10N][CdCl3] (∼8).19 The sole NLO switchable material with higher NLO switching contrast than I appears in imidazolium-FCrO3 (∼250).38 Moreover, the NLO responses quickly recover without any attenuation after 10 cycles, indicating an excellent reversibility (Figure 6d). Evidently, I is a fascinating candidate as an ultrawide-bandgap NLO switchable material because of its facile growth, good thermal stability, ultrawide bandgap, high NLO switching contrast as well as excellent reversibility. Figure 4 | Time-of-flight NPD data of I. (a) Time-of-flight NPD patterns of I from 10 to 783 K. (b) Temperature-dependent cell parameters and volumes against the NPD data. The vertical dashed line marks the occurrence of phase transition. Download figure Download PowerPoint Because the HTP structure is highly disordered, it is difficult to carry out popular first-principle calculations. To better understand structure–property relationships for the NLO switching contrast, we calculated dipole moments for I by the bond-valence method instead.39 Based on famous anionic group theory,40 [SO4]2− anions should be the NLO-active groups determining the optical properties of I, which has also been proved by our previous studies,24 so only the dipole moments of [SO4]2− anions were calculated. Detailed results are provided in the Supporting Information Table S13. For LTP, vector summation over the [SO4]2− anions geometry gives rise to an overall dipole moment of 4.60 D in the unit cell along the z direction, indicating that the microscopic NLO susceptibilities of the [SO4]2− anions are constructive to generate the highly SHG-on state. When the temperature increases to 783 K, a quasi-melting phase transition occurs, which makes [SO4]2− anions translate and rotate on the local scale as compared with LTP. As a result, the structure of HTP becomes CS, and overall dipole moments of [SO4]2− anions are canceled out to be zero, leading to the SHG-off state in HTP. Therefore, it can be rationalized that the high NLO switching contrast is ascribed to the quasi-melting phase transition resulting in the translation and rotation of NLO-active [SO4]2− anions on the local scale. Figure 2 | DSC and dielectric measurements of I. DSC and TG curves for I measured (a) from 623 to 973 K and (b) from 660 to 820 K. The black dashed box in the panel indicates where DSC peaks appear. Temperature dependence of the dielectric constant for I (c) under various frequencies and (d) in the heating and cooling run under 100 kHz. Download figure Download PowerPoint Conclusion In summary, we have synthesized a new all-inorganic NLO switchable material of an ultrawide bandgap, namely Li9Na3Rb2(SO4)7. This sulfate not only has the widest bandgap of 6.70 eV in NLO switchable materials, but also exhibits a high NLO switching contrast of about 180 that is almost the highest in known NLO switchable materials. Structural and primary theoretical analyses suggest that the outstanding NLO switching contrast originates the quasi-melting phase transition with NLO-active [SO4]2− anions translating and rotating on the local scale. These results highlight sulfates as a new source of NLO switchable materials. We believe that this work also gives inspiration for exploring ultrawide-bandgap NLO switchable materials with high NLO switching contrasts, which may provide opportunities for potential applications in the future, such as ultrawide-bandgap switches, sensors, and optoelectronics. Supporting Information Supporting Information is available. Conflict of Interest There is no conflict of interest to report. Funding Information This research is supported by the National Nature Science Foundation of China (21833010, 61975207, 21622101, 21921001, 21525104, and 51662013), the Strategic Priority Research Program of the Chinese Academy of Sciences (XDB20000000 and XDB20010200), the Natural Science Foundation of Fujian Province (2018H0047), the Key Research Program of Frontier Sciences of the Chinese Academy of Sciences (ZDBS-LY-SLH024), the Natural Science Foundation of Guangxi Province (2018GXNSFDA281015), the National Key Research and Development Program of China (2019YFA0210400), Key Laboratory of Functional Crystals and Laser Technology, TIPC, CAS (FCLT 202003), and Key Laboratory of New Processing Technology for Nonferrous Metal & Materials, Ministry of Education/Guangxi Key Laboratory of Optical and Electronic Materials and Devices (20KF-11).