Article Cite This: Inorg. Chem. XXXX, XXX, XXX−XXX
pubs.acs.org/IC
Ba10In6Zn7S26-nZnS: An Inorganic Composite System with Interface Phase-Matching Tuned for High-Performance Infrared Nonlinear Optical Materials An-Yi Zhou,† Wei-Long Zhang,§ Chen-Sheng Lin,† Fang-Yu Yuan,† Yong-Yu Pang,† Hao Zhang,† Wen-Dan Cheng,*,† Jing Zhu,*,‡ and Guo-Liang Chai*,†
Inorg. Chem. Downloaded from pubs.acs.org by EAST CAROLINA UNIV on 03/03/19. For personal use only.
†
State Key Laboratory of Structural Chemistry, Fujian Institute of Research on the Structure of Matter, Chinese Academy of Sciences, Fuzhou 350002, P. R. China ‡ Yunnan Key Laboratory for Micro/nano Materials & Technology, School of Materials Science and Engineering, Yunnan University, Kunming 650091, P. R. China § College of Electronics and Information Science, Fujian Jiangxia University, Fuzhou 350108, P. R. China S Supporting Information *
ABSTRACT: Mid- and far-infrared nonlinear optical (MFIR NLO) materials are important in modern laser technologies. However, it is very challenging to develop materials that can achieve a subtle balance between the key requirements, such as large NLO response, high laser-induced damage threshold (LIDT), wide IR transparency, and phase-matching. In this work, a new wide IR transparency (0.38−15.3 μm) NLO crystal Ba10In6Zn7S26 (SS26) is synthesized. Further, its composite system Ba10In6Zn7S26-nZnS is synthesized by eutectic reaction. In particular, Ba10In6Zn7S26-14ZnS (SS40) shows excellent balanced NLO performance that includes a large band gap of 3.05 eV, high LIDT (13.3 × AgGaS2), large second harmonic generation (SHG) response (2.1 × AgGaS2 at 2050 nm, 5.2 × KDP at 1064 nm), and wide optical transmission window (0.37− 15.4 μm). Importantly, the phase-matching condition is realized for SS40 by interfaces formed between the crystal face (112) of matrix SS26 and the crystal face (111) of reinforcement cubic ZnS by topological chemical reaction, and the NLO performance can be tuned by different concentrations of ZnS. First-principles simulations are employed to study NLO properties of SS26 and the interfaces. This work demonstrates that SS40 is a promising MFIR NLO material, and tuning components of the composite material system is a useful way to develop new MFIR NLO materials with excellent comprehensive performance.
■
INTRODUCTION With the development of laser technologies, nonlinear optical (NLO) materials are becoming more and more important. NLO materials have many applications, such as the generation of new laser sources by frequency conversion technology, and the preparation of optical switches via different NLO processes. High average power tunable solid-state midfar infrared (IR) lasers can be enabled from high-performance IR NLO materials in the transmission range 3−20 μm.1−13 For second-order NLO processes, an excellent IR NLO material used in solid-state lasers is desired to show a large NLO response (larger than 0.5 times of the commercial AgGaS2), large band gap (Eg > 3.0 eV), high laser-induced damage threshold (LIDT) (greater than AgGaS2), wide optical transparent region (the UV cutoff edge is below 400 nm and the infrared cutoff edge is over 14 μm), phase-matching in a desired area, and excellent thermal and chemical stability. Unfortunately, there are very few materials that can satisfy all the above requirements to date in the mid-far-IR region.10−15 The current commercial IR NLO materials such as AgGaS2, AgGaSe2, and ZnGeP2 show the advantages of large second© XXXX American Chemical Society
order nonlinear coefficients, a wide infrared transmission region, and a single optical axis structure.1−3 However, they suffer from low laser-induced damage threshold (LIDT) and harmful two-photon absorption (TPA) due to the narrow band gaps. These weaknesses limit their applications in high-power lasers. It is challenging to develop materials that can achieve a subtle balance between the above-mentioned requirements for practical applications. This is because some requirements contradict each other. For example, a large Eg always leads to high LIDT but small SHG response.16−18 The materials with wide IR transmission performance always have a weak chemical bond and heavy elements which lead to poor mechanical processing performance, small Eg, and low LIDT. One of the most important properties is to achieve phasematching (PM) for NLO materials.19−21 The NLO response of non-phase-matching materials is too weak to use in practical applications, even if they have a large NLO coefficient. Obviously, it is difficult for a single material to satisfy these key Received: January 10, 2019
A
DOI: 10.1021/acs.inorgchem.9b00094 Inorg. Chem. XXXX, XXX, XXX−XXX
Article
Inorganic Chemistry
Ba10In6Zn7S26. A stoichiometric amount of BaS (211.74 mg, 1.25 mmol), In2S3 (122.188 mg, 0.375 mmol), and ZnS (85.31 mg, 0.875 mmol) was ground in a mortar for about 15 min first. The mixtures were added to a graphite crucible that was placed into a predried silica tube. Then, the tube was evacuated to 10−2 Pa and sealed. The silica tube was put in a vertical oven, heated to 1213 K in 50 h (holding for 50 h), and then slowly cooled to 973 K in 120 h. The cooling rate around the crystallization area from 1213 to 973 K is 2 K per hour. Finally, it took 10 h to get to room temperature. These crystals are stable in the air for over one year. The material obtained under this condition is named SS26. In order to accurately determine the melting point and a consistent melting range for the growth of large crystals, the maximum reaction temperature is increased gradually with a change of 5 K in the range 1123−1213 K, and observe whether it melts after the reaction. Finally, it is found that the complete melting phenomenon appears at 1188 K. By comparing the experimental and the simulated powder patterns before and after melting, we found that the peak position and intensity can be substantially consistent only in the range 1173−1193 K. So, this temperature range can be looked at as a melting interval as shown in Figure 3b. So, the pure phase (powder) of SS26 was synthesized by the same procedure as that of the crystal except that the maximum temperature was set to 1173−1193 K and slow cooling was not required. Growth of Large Sized SS26 Crystals. A stoichiometric amount of BaS (211.74 mg, 1.25 mmol), In2S3 (122.188 mg, 0.375 mmol), and ZnS (85.31 mg, 0.875 mmol) was ground in a mortar for about 15 min first. Next the mixtures were added to a graphite crucible that was placed in a predried silica tube. Then, the silica tube was evacuated to 10−2 Pa and sealed. The silica tube was put in a vertical oven, heated to 1193 K in 10 h (holding for 50 h), and then slowly cooled to 1153 K in 50 h. The cooling rate around the crystallization area from 1193 to 1153 K is 0.8 K per hour. It took 5 h to get to room temperature. Finally, an ellipsoidal red crystal formed with a homogeneous color with size of 6 × 6 × 4 mm3 (Figure 3c). The total mass of the reactants was about 419.238 mg, and the weight of the obtained crystal was 401.20 mg. Microscale loss is due to the grinding and melting processes. After splitting it, light yellow crystals of small size were found (Figure 2a). There were 10 that were selected randomly, and their crystal parameters were almost identical after scanning by a single crystal X-ray diffractometer (SXRD), which demonstrated the homogeneity of the crystals. Synthesis of Composite Materials Ba10In6Zn7S26-nZnS (n = 0, 2, 3, 4, 5, 12, 14, 18, 28). A stoichiometric amount of BaS (211.74 mg, 1.25 mmol), In2S3 (122.188 mg, 0.375 mmol), and ZnS ((7+n) × 12.1871 mg, (7+n) × 0.1250 mmol; i.e., a proximate molar ratio of 10 (BaS):3 (In2S3):(7+n) (ZnS)) was ground in a mortar for about 15 min first. Next, the mixtures were added to a graphite crucible and placed into a predried silica tube that was evacuated to 10−2 Pa and sealed. The tube was placed in a vertical oven, heated to 1193 K in 10 h (holding for 40 h), and cooled to 1173 K in 10 h; this was then cooled to room temperature in 10 h. The materials obtained under this condition are named SS26 + n (n = 0, 2, 3, 5, 12, 14, 18, 28). For an example, a stoichiometric amount of BaS (211.74 mg, 1.25 mmol), In2S3 (122.188 mg, 0.375 mmol), and ZnS (255.93 mg, 2.625 mmol) can obtain Ba10In6Zn7S26-14ZnS when n = 14, which is named SS40. It is worth noting that these obtained crystals were tested by a single crystal diffractometer, and it was found that their parameters are basically the same as those of SS26, indicating that SS26 is a continuum that is a matrix in each material, and ZnS acts as a reinforcement.36 X-ray Diffraction Determination. Single Crystal X-ray Analysis. A suitable single crystal of Ba10In6Zn7S26 was selected for SXRD analysis. The SXRD data were collected on a Mercury70 CCD detector equipped with a graphite-monochromator Mo Kα radiation source (λ = 0.71073 Å) at 293 K. Absorption corrections based on the multiscan method were executed. The crystal structure was resolved by direct methods using ShelXS-201837 and refined by full-matrix least-squares on F2 using ShelXL-201838 in the Olex2 software.39 The initial refinement generated three Ba, six S, and three In atoms with all
conditions at the same time. Therefore, it is very feasible to construct a composite material that satisfies all the conditions by rationally selecting components. Many organic−inorganic composite materials with excellent NLO properties have been obtained, but the presence of organic substances makes them unsuitable for the MFIR. Therefore, pure inorganic composite materials have a very positive significance for furture expansion of the research scope of MFIR NLO materials. In this work, we first synthesized a new MFIR NLO material Ba10In6Zn7S26 via crystal design and engineering. By using commercial materials AgGaQ2 (Q = S, Se) as a template, the IR transmission can be improved by replacing Ga by In and Zn elements. Also, replacing Ag by Ba elements can increase the Eg. These two factors would result in a reasonable Eg that is larger than that of AgGaS2 and an SHG response that is larger than half of that of AgGaS2. As mentioned above, a good NLO crystal also needs high thermal stability. Replacing T1 tetrahedra by T2 supertetrahedra is a good strategy to increase the NLO coefficient and thermal stability, such as the reported chalcogenide IR NLO crystals of BaGa4S7,22 PbGa4S7,23 Pb5ZnGa6S15,24 and Sr5ZnGa6S15.25 On the other hand, the structural entropy should be the highest when the In:Zn ratio value is close. These findings give the most stable phase in these types of compounds, such as AZn 4 In 5 Se 12 , 26 KHg4Ga5Se12,27 and Na6Zn3MIII2Q9 (MIII = Ga, In; Q = S, Se).28 Here, the synthesized new noncentrosymmetric (NCS) chalcogenide with diamond-like29 Ba10In6Zn7S26 is named SS26. SS26 shows a large nonlinear coefficient (0.71 × AgGaS2) and band gap (2.95 eV), a wide crystal optical transmission range (0.38−15.3 μm), a high thermal stability (>1253 K), a low melting point (1183 K), nonhygroscopic behavior, and a high LIDT (13.7 × AgGaS2). It is also easy to grow a large-sized crystal of SS26 (6 × 6 × 4 mm3). Unfortunately, it is a non-phase-matching NLO material due to its rather low birefringence (Δncal < 0.01). So, realizing phasematching by using artificial techniques is important in this case, and previous studies demonstrate that composites with the presence of interfaces with sufficient birefringence may achieve phase-matching, such as GaAs and oxidized AlAs systems.30 SS26 is further engineered by forming composites to realize phase-matching conditions and improve the NLO performance. After some commercial crystal materials were tested, ZnS is selected because it shows the same sphalerite topology as SS26. Also, it is a wide band gap (3.6 eV) semiconductor with a large SHG response (2.7 × AgGaS2 at 2050 nm, 2.1 × KTP at 1064 nm) and wide transmission range (0.35−15 μm). However, it is non-phase-matching (more specifically it is a random quasi-phase-match (RQPM)31−34 due to optical isotropic) and shows disadvantages of a high melting point35 (>1873 K). Thus, the composites Ba10In6Zn7S26-nZnS (n = 2, 3, 4, 5, 12, 14, 18, 28) are synthesized. Among them, we find that the Ba10In6Zn7S26-14ZnS (named SS40) appears with excellent balanced NLO performances: a large SHG response (2.1 × AgGaS2 at 2050 nm, 5.2 × KDP at 1064 nm), high LIDT (13.3 × AgGaS2) at a particle size of 150−210 μm, and a high thermal/chemical stability. Most importantly, phasematching behavior is realized.
■
EXPERIMENTAL SECTION
Synthesis. Reagents. The following reagents were used as received: BaS (99.7%, Alfa Aesar China Co., Ltd.), In2S3 (99.999%, Alfa Aesar China Co., Ltd.), and ZnS (99.99%, Sinopharm Chemical Reagent Co.,Ltd.). B
DOI: 10.1021/acs.inorgchem.9b00094 Inorg. Chem. XXXX, XXX, XXX−XXX
Article
Inorganic Chemistry atoms refined anisotropically and R values of R1= 6.75%, wR2 = 16.68%. Note here that the temperature factor of one Ba is too high (0.105) and shows a large Q peak (5.45 Å) near it. So, the Q peak was also identified as Ba and dealt with the disorder and free refinement of both occupations with a fixed total occupancy (order FVAR[2]). According to the EDX results (Figure S1 of Supporting Information), the compound should consist of Zn as well. Subsequently, In and Zn atoms were constrained to the same site and only limit the total occupancy rate in each position, and the specific proportion allows it to be freely refined (order FVAR[3], FVAR[4], FVAR[5], EADP, and EXYZ). Meanwhile, the In to Zn ratio was close to 6:7 which is also close to the EDX results. According to the above-mentioned free refinement ratio, the occupancy rate of each position is fixed (In1:Zn1 = 0.53:0.47, In2:Zn2 = 0.60:0.40, In3:Zn3 = 0.38:0.62, Ba2:Ba4 = 0.39:0.61). The final refined structure was checked by PLATON,40 and no other missed or higher-symmetry element was found. The structure refinement converged to R1 = 2.46%, wR2 = 4.75%, a Flack parameter of 0.02(4), and a formula of Ba20In12.02Zn13.98S52 (hereafter denoted as Ba10In6Zn7S26). The SHG effect of this compound is also proof of its noncentrosymmetry. Crystallographic data and structural refinement information are presented in Table 1. The atomic coordinates, equivalent isotropic thermal parameters, important bond lengths, and bond angles are given in Tables S1 and S2 of Supporting Information.
mg of a powder sample of SS26 or SS40 was placed in an Al2O3 crucible, which was then heated from 40 to 1000 °C at 20 °C min−1 under a N2 atmosphere at a flow rate of 150 mL/min. TGA can be used to evaluate the thermal stability of a material as shown in Figure 3d. Infrared Transmittance and Solid-State UV−Vis−NIR Diffuse Reflectance Spectra. The air was used as IR transmission background. The IR transmission ranges of SS26, SS40, and AgGaS2(AGS) crystal (irregular block crystals with at least one face larger than 3 × 3 mm2) were measured on a PerkinElmer Spectrum One FT-IR Spectrometer in the range 2.5−25 μm (Figure 3f). The UV−vis−NIR transmission range of the SS26 and SS40 powders was measured on a PerkinElmer Lambda950 UV−vis−NIR spectrometer in the range 0.25−2.5 μm (Figure 5c) with a BaSO4 plate used as a UV−vis−NIR transmission background. The UV−vis−NIR diffuse reflectance spectra of the powder samples SS26 and SS40 were also carried out on the PerkinElmer Lambda950 UV−vis−NIR spectrometer in the range 250−2500 nm at room temperature. A BaSO4 plate was used as a 100% reflectance criterion material. The optical absorption spectra were figured out from diffuse reflectance spectra using the Kubelka−Munk function,41 α/S = (1 − R)2/2R, where α is the absorption coefficient, S is the scattering coefficient, and R is the reflectance. The band gap can been determined on the basis of the following formula: αhν = B(hν − Eg)n/2 (n = 1 for direct band gap, and n = 4 for indirect band gap). Here, the α can be replaced by F(R) in diffused reflectance spectra; the plotting curve of (F(R)hν)2 vs hν is extrapolated to the hν axis to obtain the band gap (Figure 5a,b). Powder Second Harmonic Generation (SHG) Measurement. The SHG measurements were carried out on microcrystal samples of SS26, ZnS, SS30, SS54, and SS40 by using the Kurtz and Perry method with a 2.05 and 1.064 μm Q-switch laser.42 First, SS26, ZnS, SS30, SS54, and SS40 microcrystal samples were ground and sieved in the ranges 25−45, 45−74, 74−106, 106−150, and 150−210 μm. The SS26 crystals and ZnS crystals were simply physically mixed at a molar ratio of 1:14 for each particle size and were denoted as S40M. Then, SS26, ZnS, S40M, SS30, SS54, and SS40 of each particle size were put into a silicone gasket which is 0.5 mm thick with a diameter of 8 mm and pressed between two glass slides, respectively. Then, they were fixed in an aluminum box as tightly as possible. The AgGaS2, KTP, and KH2PO4 (KDP) crystals were crushed, ground, and sieved into the same size range with the identical laser settings (AgGaS2 is used with 2.05 μm, KDP, KTP is used with 1.064 μm) as references (2 × 2 × 4 cm3 single crystal of AgGaS2 was supplied by Anhui Institute of Optics and Fine Mechanics, Chinese Academy of Sciences). At last, the fundamental pulsed light of 2.05 and 1.064 μm which is the light from the Q-switched Nd:YAG laser passed through the monochromator, and then passed through every sample, and the resulting doubled frequency signal data (1025 and 532 nm) collected is the voltage displayed on the DS1052E 50-MHz oscilloscope (RIGOL). Powder LIDT Measurement. The LIDTs of ground powders of SS26 and SS40 crystals were measured by using a 1064 nm Q-switch laser with AgGaS2 serving as a reference. The samples were surrounded by a 0.5 mm thick silicone gasket with an 8 mm diameter hole and pressed between two glass slides. Then, they were fixed in an aluminum box as tight as possible with the plane surface explored under a pulsed laser beam (1064 nm, 10 ns, 1 Hz). It is irradiated with sequentially increasing energy until damage occurs at this point for a point on the plane surface. Electronic Structure Calculation. Band structure of Ba10In6Zn7S26 was calculated by using the first-principles plane-wave method by using the CASTEP code in the Material Studio package.43,44 The generalized gradient approximation (GGA) was applied for exchange and correlation. The interaction of the electrons with ion cores was represented by norm-conserving pseudopotentials,45,46 and the valence electronic configurations for the component elements were treated as Ba 5s25p66s2, In 5s25p1, Zn 3d104s2, and S 3s23p4, respectively. Cutoff energy and converge precision were set to 650 eV and 1.0 × 10−6 eV/atom, respectively. The first Brillouin zone
Table 1. Crystal Data and Structure Refinement for Ba10In6Zn7S26 formula fw wavelength temp cryst syst space group unit cell dimensions
V Z, calcd density abs coeff F(000) θ range for data collection limiting indices refinement method reflns collected/unique completeness to θ = 27.47 abs correction data/restraints/params GOF on F2 final R indices [I > 2σ(I)] R indices (all data) largest diff peak and hole
Ba10In6Zn7S26 3353.5 0.71073 Å 293(2) K tetragonal I4̅2m (No. 121) a = b = 11.0190(2) Å c = 21.1853(8) Å α = β = γ = 90 °C 2572.28(12) Å3 2, 4.330 g cm−3 14.394 mm−1 2960 2.08−27.45° −14 ≤ h ≤ 14; −10 ≤ k ≤ 14; −27 ≤ l ≤ 27 full-matrix least-squares on F2 9747/1579 [R(int) = 0.0412] 99.3% multiscan 1579/0/108 1.100 R1 = 0.0246, wR2 = 0.0475 R1 = 0.0233, wR2 = 0.0478 0.807 and −0.732 e Å−3
Powder X-ray Diffraction. The melting point obtained from multiple experiments is about 1188 K. So, the Ba10In6Zn7S26 obtained under reaction temperatures of 1193 K and 1173K was taken as samples. The powder X-ray diffraction (PXRD) patterns of them were measured on a Rigaku MiniFlex II diffractometer (Cu Kα radiation, λ = 1.5406 Å) at room temperature. The range of the 2θ angle is 10− 80° with a step size of 0.02° and a scan speed of 2° min−1. The measurement results show that the PXRD patterns are in good agreement with the simulated pattern generated by the CIF refined structure, and no phase change or decomposition occurs before and after the melting point. The powder samples of SS26 and SS40 were measured by the same processes. Thermal Analysis. Thermal gravimetric analysis (TGA) was performed with a NETZSCH STA449C thermal analyzer. About 15.0 C
DOI: 10.1021/acs.inorgchem.9b00094 Inorg. Chem. XXXX, XXX, XXX−XXX
Article
Inorganic Chemistry
Figure 1. (a) Three-dimensional structure of SS26 viewed toward the (110) plane. (b) The ∞[M34S8]n− layers along the [011] direction (M3 = In3/Zn3), which are formed by corner-shared [M34S10] T2-supertetrahedra. (c) Zero-dimensional [M1S4] unit and T2-supertetrahedra [M24S10] that are between ∞[M34S8]n− layers (M1 = In1/Zn1; M2 = In2/Zn2). (d) Basic repeating building unit (In/Zn)9S20. of the unit cell was sampled as 1 × 1 × 1 in the Monkhorst−Pack scheme.47 Due to the co-occupied position of In and Zn, the MS4 (M = In or Zn) tetrahedra are composited by 12 [InS4] and 14 [ZnS4] randomly. We composed tens of models with different [InS4] and [ZnS4] occupation positions and calculated their total energy, respectively. The Ba atom position was chosen as the average of the three partially occupied Ba atom coordinations. It is found that the more uniform the distribution of [InS4] and [ZnS4] in the unit cell is, the lower the total energy is. At last, the model with the lowest energy (Figure S8) is selected for a further electronic properties calculation. The NLO coefficients was calcuated using the Sipe formula rewritten with the δfunction as shown in Supporting Information.
■
RESULTS AND DISCUSSION Synthesis, Structure, and NLO Properties of Ba10In6Zn7S26. Ba10In6Zn7S26 (SS26) is synthesized via structure design as described above. The detailed synthesis procedure is shown in the Experimental Section. The SS26 crystal is in the tetragonal I4̅2m space group, and its unit cell parameters are a = b = 11.0190(2) Å, c = 21.1853(8) Å, as determined by the single crystal X-ray diffraction (SXRD) determination. There are three unequal Ba atoms, three M atoms (M = In or Zn), and six unequal S atoms in the SS26 crystal. Note that the bond length range for M1−S is 2.4302, the bond length range for M2−S is 2.4023−2.4680 Å, and the bond length range for M3−S is 2.3588−2.4754 Å. Consider that Zn2+ is the boundary acid, In3+ is the hard acid, and S2− is the soft base. It is obvious that the bond energy of the Zn−S bond should be higher than the In−S bond in the same case. So, the bond length of the Zn−S bond should tend to be shorter than the In−S bond. Also, it is obvious that the ratio of Zn2+ in M1, M2, and M3 can be presumed accordingly as M3, M1, and M2 from high to low, which is consistent with the analytical results of free refinement at all positions. A threedimensional network structure of the SS26 crystal that is similar to that of chalcopyrite is shown in Figure 1. The crystals, powders, and SEM images of SS26 are shown in Figure 2. The structure of SS26 is composed of [M1S4] T1clusters, [M24S10] T2-clusters, [M34S10] T2-clusters, and
Figure 2. (a) Crystals, (b) powders, and (c) electron backscattered diffraction images of (top) SS26 and (bottom) SS40.
cations Ba2+ (Figure 1a). The T2-clusters of [M34S10] connected to each other in the same direction by S3 atoms and constructed a ∞[M34S8]n− layer (Figure 1b). T2 clusters of [M24S10] are located between ∞[M34S8]n− layers (Figure 1a,c). The [M34S10] and [M24S10] are connected by S3 atoms (Figure 1d) in the same direction. Then, an anionic framework [M28M316S44]29− with a zinc-blende topological structure is formed. The vacancies of the [M28M316S44]29− 3D framework are filled by the anionic tetrahedra [M1S4], and cations Ba2+ balanced the charge (Figure 1a). [M1S4] is a zero-dimensional structural unit which is contrary to the direction of T2-clusters [M24S10] and [M34S10]. Each [M1S4] is surrounded by four [M24S10] T2-clusters. [M1S4] and [M24S10] are coplanar but not connected to each other (Figure 1c). The SS26 crystals are very stable and easy to reproduce. The powder X-ray diffraction (PXRD) and Raman spectra of SS26 are shown in Figure 3a,b,e. The thermal gravimetric analysis (TGA) curve (Figure 3d) indicates that SS26 shows high thermal stability and the thermal stability of the composite D
DOI: 10.1021/acs.inorgchem.9b00094 Inorg. Chem. XXXX, XXX, XXX−XXX
Article
Inorganic Chemistry
Figure 3. (a) PXRD patterns of SS26 and SS40. (b) PXRD patterns of SS26 powder before and after melting. (c) Large SS26 crystal with a size of 6 × 6 × 4 mm3. (d) TGA curves of SS26 and SS40. (e) Raman spectra of SS26 and SS40. (f) IR transmission spectra of AGS, ZnS, SS26, and SS40.
Figure 4. (a) SHG response for SS26, SS30, SS40, and SS54 with AgGaS2 as a reference at 2050 nm. (b) SHG response for SS30, SS40, and SS54 with KTP as a reference at 1064 nm. (c) SHG signals of SS26, SS30, SS40, and SS54 with AgGaS2 as a reference at a particle size of 150−210 μm at 2050 nm. (d) SHG signals of SS30, SS40, and SS54 with KDP as a reference at a particle size of 150−210 μm at 1064 nm. (e) SHG performance of ideal composite material SS26-nZnS (n = 2, 3, 4, 14, 24) with particle size of 150−210 μm at 2050 nm. (f) Currently reported IR NLO chalcogenides with wide band gap (Eg > 3.0 eV) and phase-matching behavior.
which indicates that it has a consistent melting point. These characteristics make it is easy to grow large single crystals of SS26. A SS26 crystal with a size of 6 × 6 × 4 mm3 is shown in
SS40 is better than that of SS26. In addition, the PXRD (Figure 3b) of SS26 is basically unchanged before and after melting and is consistent with the simulated powder pattern, E
DOI: 10.1021/acs.inorgchem.9b00094 Inorg. Chem. XXXX, XXX, XXX−XXX
Article
Inorganic Chemistry
Figure 5. Ultraviolet diffuse reflection curves and band gaps of (a) SS26 and (b) SS40. (c) Transmission spectra of SS26 and SS40 in the range 0.35−2.5 μm.
Figure 3d. Its crystal IR transmission range is 2.5−15.3 μm (Figure 3f) which was measured by using a 3 × 5 × 4 mm3 piece that was cut from the bulk crystal. The SHG response versus particle size for SS26, ZnS, S40M, KTP, KDP, SS30, SS54, and SS40 is plotted in Figure 4a,b according to the Kurtz and Perry method.42 The commercial AGS is used as reference for SHG at 2050 nm. The KDP and KTP are used as references for SHG at 1064 nm. The SHG response of SS26 is only about one-third of KDP at the particle size of 150−210 μm at 1064 nm. The SHG response of SS26 is approximately 0.5 × AGS (Figure 4c) at the particle size of 150−210 μm at 2050 nm, which indicates that the d14 is approximately 0.71 times that of AGS. Note that the d14 of SS26 should be 9.73 pm/V as the value of the d36 of the reference AGS is 13.7 pm/V.48 However, the SS26 powder shows an increased Eg of 2.95 eV (Figure 5a) compared with AGS of 2.64 eV, which should give a large LIDT of SS26. The powder LIDTs of SS26, SS40, and benchmark AGS are systemically evaluated using the single pulse powder LIDT method,49 and the results are illustrated in Table 2. The results indicate that the LIDT of SS26 is 13.7 × AGS.
pattern (Figure 3a) shows the peaks of SS26 and ZnS. This indicates that SS26 and ZnS are present in the composites, for which the matrix is SS26 and ZnS is reinforcement. Considering that SS26 has a very large unit cell volume and molecular weight, it can be looked at as an inorganic polymer. This makes it easy to be the matrix in the composite. At the same time, the PXRD shows that ZnS and SS26 overlap at some peak positions, and the small half-width of SS40 and SS26 indicates that both of them present good crystal quality. According to the overlapping peaks, we can see that they are composed of the (112) crystal plane of SS26 (Figure S6 and Figure S7a in Supporting Information) and the (111) crystal plane of ZnS (Figure S7b in Supporting Information), respectively. The two crystal planes are the most densely packed crystal plane of the respective crystals, which are also the surfaces of the crystals. The number of chemical bonds at the (112) crystal plane of SS26 is significantly less than that of the other crystal faces (Figure S6 in Supporting Information). It is also possible to realize the lattice match between the SS26 (112) and ZnS (111) crystal faces (Figure S7a,b in Supporting Information). Both factors make it easy for them to form a new chemical interface by topological chemical reaction. The properties of the interfaces are studied by the first-principles simulations as shown in Figure S7c in the Supporting Information. Its birefringence is large enough (Δncal > 0.04) for the composites to achieve phase-matching. A comparison of the SHG particle size distribution curves of pure SS26, pure ZnS, simple mixture S40M, and composite SS40 also proves that the interface plays a key role in achieving phase-matching (Figure 4a,b). The ZnS and SS26 are readily formed composites over a large range of concentrations of ZnS. The SEM images of SS26 (S26 + 0ZnS), SS28 (S26 + 2ZnS), SS30 (S26 + 4ZnS), and SS40 (S26 + 14ZnS) are shown in Figure S2 of the Supporting Information. It can be seen from Figure S2 that the ZnS domains at the surface of the composites increase as the concentrations of ZnS increase. The TGA curve of SS40 (Figure 3d) indicates that it shows excellent thermal stability below 1253 K (980 °C). The consistency of the peak positions in the Raman spectrum (Figure 3e) illustrates the similarity between the structures of SS26 and SS40. This does not show an obvious Raman peak in the range 380−4000 cm−1 (2.5−26 μm). Since compounds containing heavy elements generally do not have Raman peaks in the range of 500−4000 cm−1, they are not drawn in Figure 3e. Although ZnS in SS40 has a very high molar ratio of 14:1 compared with that of SS26, the volume ratio of composites is often the most key factor in performance and VSS26:VZnS = 18:7. The bright part is SS26, and the dark part is ZnS in the SEM back scattering images of
Table 2. LIDTs of SS26 and SS40 with AgGaS2 as the Referencea compd
max energy
damage area
LIDT
AgGaS2 SS26 SS40
11 66 65
0.2826 0.12566 0.12566
3.89 52.52 51.72
a
Unit: energy (mJ), damage area (cm2), LIDT (MW/cm2).
Structure and NLO Properties of Ba10In6Zn7S26-nZnS Composite System. In order to realize phase-matching behavior to improve the NLO performance of SS26, the composite systems of Ba10In6Zn7S26-nZnS (n = 2, 3, 4, 5, 12, 14, 18, 28) were further synthesized. The detailed synthesis procedure is described in the Experimental Section. It will be shown below that the engineered Ba10In6Zn7S26-14ZnS composites (named SS40) present excellent balanced NLO performance among the studied materials. Thus, we mainly focus on discussing the SS40 for the composites. The crystals, powders, and SEM images of SS40 are shown in Figure 2. It can be seen from the SEM image that the ZnS domains are well-distributed in the SS26 substrate (Figure 2c). The EDX data of SS40 is close to the chemical reaction ratio and proves that the raw materials are completely used to form composite materials (Figure S1 of Supporting Information). The SXRD and Raman spectroscopy show that SS40 maintains the same crystal symmetry of I4̅2m as SS26, and the PXRD F
DOI: 10.1021/acs.inorgchem.9b00094 Inorg. Chem. XXXX, XXX, XXX−XXX
Article
Inorganic Chemistry
(0.37−15.4 μm), phase-matching, and high thermal/chemical stability. Up to now, MFIR NLO chalcogenide materials can have a band gap larger than 3.0 eV; the IR cutoff edge of the crystal can be over 14 μm, and the PM characteristics are summarized in Table S3 of Supporting Information and Figure 4e. Among the listed 26 compounds, it can be found that only six materials have a conversion efficiency larger than of AgGaS2 and the SS40 is the only material with a conversion efficiency larger than 2 times that of of AgGaS2. Note that some metal thiophosphates meet the conditions, but their IR transmission range is usually narrow (
