Recent Research Progress of Mn4+-Doped A2MF6 (A = Li, Na, K, Cs, or Rb; M = Si, Ti, Ge, or Sn) Red Phosphors Based on a Core–Shell Structure

White light emitting diodes (WLEDs) are widely used due to their advantages of high efficiency, low electricity consumption, long service life, quick response time, environmental protection, and so on. The addition of red phosphor is beneficial to further improve the quality of WLEDs. The search for novel red phosphors has focused mainly on Eu2+ ion- and Mn4+ ion-doped compounds. Both of them have emissions in the red region, absorption in blue region, and similar quantum yields. Eu2+-doped phosphors possess a rather broad-band emission with a tail in the deep red spectral range, where the sensitivity of the human eye is significantly reduced, resulting in a decrease in luminous efficacy of WLEDs. Mn4+ ions provide a narrow emission band ~670 nm in oxide hosts, which is still almost unrecognizable to the human eye. Mn4+-doped fluoride phosphors have become one of the research hotspots in recent years due to their excellent fluorescent properties, thermal stability, and low cost. They possess broad absorption in the blue region, and a series of narrow red emission bands at around 630 nm, which are suitable to serve as red emitting components of WLEDs. However, the problem of easy hydrolysis in humid environments limits their application. Recent studies have shown that constructing a core–shell structure can effectively improve the water resistance of Mn4+-doped fluorides. This paper outlines the research progress of Mn4+-doped fluoride A2MF6 (A = Li, Na, K, Cs, or Rb; M = Si, Ti, Ge or Sn), which has been based on the core–shell structure in recent years. From the viewpoint of the core–shell structure, this paper mainly emphasizes the shell layer classification, synthesis methods, luminescent mechanism, the effect on luminescent properties, and water resistance, and it also gives some applications in terms of WLEDs. Moreover, it proposes challenges and developments in the future.


Introduction
Since the advent of the white light emitting diode (WLED), it has been studied and applied in backlight displays, lighting devices, fast imaging, and other aspects based on its advantages of high efficiency, environmental protection, durability, low energy, and so on [1][2][3][4][5][6][7][8]. The combination of a blue InGaN chip and yellow Y 3 Al 5 O 12 : Ce 3+ (YAG: Ce 3+ ) phosphors was applied to the first commercial phosphor-converted white LED (pc-WLED) to produce white light and achieve a wide range of commercial applications [9][10][11]. It is known that adding red phosphors is beneficial to the white light emission of this WLED with high CRI and low CCT [12,13]. A great deal of research has focused on discovering new types of red phosphors. The search for novel red phosphors has been mainly focused on two alternative activator ions, namely, Eu 2+ and Mn 4+ , doped in various hosts [14]. The overall characteristics of the two activators are very similar; Eu 2+ possesses emission peaks in the range of 600-650 nm, and Mn 4+ produces emission peaks at 630 nm. In addition, the lowest energy excitation peak of Eu 2+ ions is located at~475 nm, and the long wavelength absorption edge extends to~650 nm, while the same eigenvalues of Mn 4+ ions are~450

Classification of the Shell Layer in A2MF6: Mn 4+
In practice, [MnF6] 2− in Mn 4+ -doped fluoride is easily hydrolyzed to brown MnO2, which deteriorates the luminescence and generates HF, leading to the packaging corrosion and failure of WLED. Therefore, it is of extraordinary value to enhance the water resistance of fluorides. In general, coating the surface of A2MF6: Mn 4+ with hydrophobic materials to form a core-shell structure can improve their water resistance. The shell layer can be divided into heterogeneous and homogeneous shell layers.

Heterogeneous Shell Layer
The heterogeneous shell layer means that the shell material is different from the matrix material of the core. It connected with the core through chemical bonds or other interactions. Common heterogeneous shell layer materials are alkyl phosphates, octadecyl trimethoxy silanes (ODTMS), silane coupling agents, oleic acid (OA), CaF2, SrF2, Al2O3, TiO2, SiO2, GQDs (graphene quantum dots), and nano-carbon, which can act as water repellents to improve the chemical stability or the water resistance of A2MF6: Mn 4+ .
Nguyen et al. [35] successfully coated the K2SiF6: Mn 4+ surface with an alkyl phosphate layer to enhance its chemical stability. Zhou et al. [36] used ODTMS to significantly improve the moisture resistance and thermal stability of K2TiF6: Mn 4+ . Kim et al. [37] modified the surface of K2SiF6: Mn 4+ with silane coupling agents. The formation of a hydrophobic shell increased its water resistance, and the removal of surface quench sites enhanced its emission efficiency. The results of these studies suggested that the surface modification of hydrophobic silane coupling agent was an effective method to improve the humidity of fluoride phosphor, which has practical application potential. Arunkumar et al. [38] modified the K2SiF6: Mn 4+ surface with OA, and Fang et al. [39] used SiO2 and OA to form a double-coated KTF@OA@SiO2 phosphors. Both improved the water resistance of K2SiF6: Mn 4+ . In addition, Luo et al. [40] treated K2SiF6: Mn 4+ (KSFM) with pyruvate in one step to construct an impermeable dual-shell-stabilized fluoride phosphor, namely, KSFM-98PA.
CaF2 has good chemical stability and is also suitable for improving the chemical stability of A2MF6: Mn 4+ . Dong et al. [41] and Yu et al. [42] modified the surface of K2TiF6: Mn 4+ and K2SiF6: Mn 4+ with CaF2, which strengthened the humidity resistance and luminescence properties. Fang et al. [43] constructed a water-resistant SrF2 coating on the

Classification of the Shell Layer in A 2 MF 6 : Mn 4+
In practice, [MnF 6 ] 2− in Mn 4+ -doped fluoride is easily hydrolyzed to brown MnO 2 , which deteriorates the luminescence and generates HF, leading to the packaging corrosion and failure of WLED. Therefore, it is of extraordinary value to enhance the water resistance of fluorides. In general, coating the surface of A 2 MF 6 : Mn 4+ with hydrophobic materials to form a core-shell structure can improve their water resistance. The shell layer can be divided into heterogeneous and homogeneous shell layers.

Heterogeneous Shell Layer
The heterogeneous shell layer means that the shell material is different from the matrix material of the core. It connected with the core through chemical bonds or other interactions. Common heterogeneous shell layer materials are alkyl phosphates, octadecyl trimethoxy silanes (ODTMS), silane coupling agents, oleic acid (OA), CaF 2 , SrF 2 , Al 2 O 3 , TiO 2 , SiO 2 , GQDs (graphene quantum dots), and nano-carbon, which can act as water repellents to improve the chemical stability or the water resistance of A 2 MF 6 : Mn 4+ .
Nguyen et al. [35] successfully coated the K 2 SiF 6 : Mn 4+ surface with an alkyl phosphate layer to enhance its chemical stability. Zhou et al. [36] used ODTMS to significantly improve the moisture resistance and thermal stability of K 2 TiF 6 : Mn 4+ . Kim et al. [37] modified the surface of K 2 SiF 6 : Mn 4+ with silane coupling agents. The formation of a hydrophobic shell increased its water resistance, and the removal of surface quench sites enhanced its emission efficiency. The results of these studies suggested that the surface modification of hydrophobic silane coupling agent was an effective method to improve the humidity of fluoride phosphor, which has practical application potential. Arunkumar et al. [38] modified the K 2 SiF 6 : Mn 4+ surface with OA, and Fang et al. [39] used SiO 2 and OA to form a double-coated KTF@OA@SiO 2 phosphors. Both improved the water resistance of K 2 SiF 6 : Mn 4+ . In addition, Luo et al. [40] treated K 2 SiF 6 : Mn 4+ (KSFM) with pyruvate in one step to construct an impermeable dual-shell-stabilized fluoride phosphor, namely, KSFM-98PA. CaF 2 has good chemical stability and is also suitable for improving the chemical stability of A 2 MF 6 : Mn 4+ . Dong et al. [41] and Yu et al. [42] modified the surface of K 2 TiF 6 : Mn 4+ and K 2 SiF 6 : Mn 4+ with CaF 2 , which strengthened the humidity resistance and luminescence properties. Fang et al. [43] constructed a water-resistant SrF 2 coating on the surface of K 2 TiF 6 : Mn 4+ . SrF 2 was uniformly covered on K 2 TiF 6 : Mn 4+ to eliminate lattice defects and improve the emission efficiency. Verstraete et al. [44] deposited a TiO 2 or Al 2 O 3 layer on the surface of K 2 SiF 6 : Mn 4+ and obtained K 2 SiF 6 : Mn 4+ -TiO 2 and K 2 SiF 6 : Mn 4+ -Al 2 O 3 , respectively. Kate et al. [45] deposited ultrathin Al 2 O 3 on the surface of K 2 SiF 6 : Mn 4+ with trimethyl aluminum (TMA) to improve its optical properties, chemical stability, and thermal stability. Quan et al. [46] modified a layer of SiO 2 on the surface of K 2 SiF 6 : Mn 4+ to further improve the water resistance of the material.
In addition, carbon materials can also be applied to improve the water resistance of A 2 MF 6 : Mn 4+ . GQDs have the characteristics of large specific surface area, conjugated large π bonds, multiple uses, environmental friendliness, and good thermal stability. Yu et al. [47] constructed a double shell structure on the surface of K 2 SiF 6 : Mn 4+ , Na + to form K 2 SiF 6 : Mn 4+ , Na + @GQDS@K 2 SiF 6 phosphors with water-resistance. Liu et al. [48] modified carbon nanoparticles on the surface of K 2 SiF 6 : Mn 4+ , and C atoms combined with F atoms in K 2 SiF 6 : Mn 4+ to form carbon-fluorine (C-F) covalent bonds. Carbon has low polarizability and excellent hydrophobicity, which are beneficial for improving the water resistance of K 2 SiF 6 : Mn 4+ .
As summarized above, the matrix materials involved in the construction of heterogeneous core-shell structures are mainly K 2 SiF 6 and K 2 TiF 6 , probably due to their excellent optical properties, good luminescence efficiency, and high quantum efficiency, which can effectively reduce the luminescence loss caused by the shell. When the phosphor was coated with organic material, a film layer on the surface could be clearly seen by transmission electron microscopy (TEM) at different magnifications. Figure 2 shows the TEM images of typical organic coating layers on the surface of K 2 SiF 6 /K 2 TiF 6 with a heterogeneous core-shell structure. The surface shell layer of K 2 SiF 6 @MOPAl and K 2 SiF 6 @OA were affected under the electron beam of a TEM system, leading to decomposition. Moreover, the desorption on the surface leads to the formation of KF and other corrosive substances. They corrode the core-shell interface and destroy the good adhesion between the core and shell, which may result in fluorescence quenching. In addition, carbon materials can also be applied to improve the water resistance of A2MF6: Mn 4+ . GQDs have the characteristics of large specific surface area, conjugated large π bonds, multiple uses, environmental friendliness, and good thermal stability. Yu et al. [47] constructed a double shell structure on the surface of K2SiF6: Mn 4+ , Na + to form K2SiF6: Mn 4+ , Na + @GQDS@K2SiF6 phosphors with water-resistance. Liu et al. [48] modified carbon nanoparticles on the surface of K2SiF6: Mn 4+ , and C atoms combined with F atoms in K2SiF6: Mn 4+ to form carbon-fluorine (C-F) covalent bonds. Carbon has low polarizability and excellent hydrophobicity, which are beneficial for improving the water resistance of K2SiF6: Mn 4+ .
As summarized above, the matrix materials involved in the construction of heterogeneous core-shell structures are mainly K2SiF6 and K2TiF6, probably due to their excellent optical properties, good luminescence efficiency, and high quantum efficiency, which can effectively reduce the luminescence loss caused by the shell. When the phosphor was coated with organic material, a film layer on the surface could be clearly seen by transmission electron microscopy (TEM) at different magnifications. Figure 2 shows the TEM images of typical organic coating layers on the surface of K2SiF6/K2TiF6 with a heterogeneous core-shell structure. The surface shell layer of K2SiF6@MOPAl and K2SiF6@OA were affected under the electron beam of a TEM system, leading to decomposition. Moreover, the desorption on the surface leads to the formation of KF and other corrosive substances. They corrode the core-shell interface and destroy the good adhesion between the core and shell, which may result in fluorescence quenching.  [36]. Copyright from American Chemical Society, 2018); (c) K2SiF6@OA (Reprinted from Ref [38]. Copyright from American Chemical Society, 2017); (d) K2TiF6@OA@SiO2 (Reprinted from Ref [39]. Copyright from American Chemical Society, 2018).

Homogeneous Shell Layer
A homogeneous shell layer can be defined as the material of the shell that is the same as the material of the core. Because the material of the core and shell are the same, and due to the high matching degree between them, the fluorescence performance may be enhanced. The homogeneous core-shell structure of A2MF6: Mn 4+ can be constructed by different methods, removing the Mn 4+ ions on surface, and leaving the layer matrix material that is the same as the core.

Homogeneous Shell Layer
A homogeneous shell layer can be defined as the material of the shell that is the same as the material of the core. Because the material of the core and shell are the same, and due to the high matching degree between them, the fluorescence performance may be enhanced. The homogeneous core-shell structure of A 2 MF 6 : Mn 4+ can be constructed by different methods, removing the Mn 4+ ions on surface, and leaving the layer matrix material that is the same as the core.
Wan et al. [29] obtained KSFM-RSRC with good water resistance by using reductionassisted surface recrystallization (RSRC) to treat K 2 SiF 6 : Mn 4+ , remove surface Mn 4+ ions, and construct their own encapsulated shell structure. Li et al. [59] synthesized K 2 SiF 6 : Mn 4+ @K 2 SiF 6 with good water resistance, luminescent thermal stability, and high efficiency by the coating K 2 SiF 6 on the core surface. Huang et al. [28] constructed K 2 TiF 6 : Mn 4+ @K 2 TiF 6 with a self-coating structure. The outer shell not only prevented the hydrolysis of internal [MnF 6 ] 2− groups in the air but also effectively blocked the path of energy transfer to surface flaws and further increased emission efficiency. Zhou et al. [30] and Zhong et al. [60] synthesized K 2 SiF 6 : Mn 4+ and LiNaSiF 6 : Mn 4+ with almost no Mn 4+ surfaces by gradually reducing [MnF 6 ] 2− on the crystal surface over time based on the dynamic equilibrium between dissolution and crystallization. Jiang et al. [61] synthesized K 2 SiF 6 : Mn 4+ @K 2 SiF 6 by the ethanol-induced epitaxial deposition growth of fluoride. These constructed homogeneous shells were effective in enhancing the water resistance of the cores. Figure 3 shows the model diagram of homogeneous core-shell structure formation. Mn 4+ ions are mainly distributed in the core and on the surface with the Mn 4+ -free layer. The matrix materials for the construction of homogeneous core-shell structure are more diverse, including K 2 SiF 6 , K 2 TiF 6 , K 2 GeF 6 , and cationic iso-alkali double-doped matrix materials, which enrich the types of homogeneous shell matrix materials. The homogenous shell generated in situ can be uniformly wrapped on the surface of the core, avoiding lattice mismatch and further improving the emission efficiency of fluoride.  [29] obtained KSFM-RSRC with good water resistance by using reductionassisted surface recrystallization (RSRC) to treat K2SiF6: Mn 4+ , remove surface Mn 4+ ions, and construct their own encapsulated shell structure. Li et al. [59] synthesized K2SiF6: Mn 4+ @K2SiF6 with good water resistance, luminescent thermal stability, and high efficiency by the coating K2SiF6 on the core surface. Huang et al. [28] constructed K2TiF6: Mn 4+ @K2TiF6 with a self-coating structure. The outer shell not only prevented the hydrolysis of internal [MnF6] 2− groups in the air but also effectively blocked the path of energy transfer to surface flaws and further increased emission efficiency. Zhou et al. [30] and Zhong et al. [60] synthesized K2SiF6: Mn 4+ and LiNaSiF6: Mn 4+ with almost no Mn 4+ surfaces by gradually reducing [MnF6] 2− on the crystal surface over time based on the dynamic equilibrium between dissolution and crystallization. Jiang et al. [61] synthesized K2SiF6: Mn 4+ @K2SiF6 by the ethanol-induced epitaxial deposition growth of fluoride. These constructed homogeneous shells were effective in enhancing the water resistance of the cores. Figure 3 shows the model diagram of homogeneous core-shell structure formation. Mn 4+ ions are mainly distributed in the core and on the surface with the Mn 4+ -free layer. The matrix materials for the construction of homogeneous core-shell structure are more diverse, including K2SiF6, K2TiF6, K2GeF6, and cationic iso-alkali double-doped matrix materials, which enrich the types of homogeneous shell matrix materials. The homogenous shell generated in situ can be uniformly wrapped on the surface of the core, avoiding lattice mismatch and further improving the emission efficiency of fluoride.  [60]. Copyright from American Chemical Society, 2022); (b) T-KGF (Reprinted from Ref [53]. Copyright from Elsevier, Amsterdam, The Netherlands, 2018); (c) KSFM-RSRC (Reprinted from Ref [29]. Copyright from Elsevier, 2021); (d) K 2 SiF 6 : Mn 4+ (Reprinted from Ref [61]. Copyright from Royal Society of Chemistry, 2021).

Preparation Methods of Shell Layer in A 2 MF 6 : Mn 4+
There are usually two strategies to construct the core-shell structure. One is to nucleate first and then construct the shell, while the other is to form a core-shell structure at one time. For A 2 MF 6 : Mn 4+ , the former is suitable for the synthesis of both heterogeneous shells and homogeneous shells, while the latter is mainly suitable for the synthesis of homogeneous shells. There are mainly three preparation methods including the coating construction method, the surface passivation method, and the saturated crystallization method. The materials that making up the shell structure are different, so the properties of the fluoride also change [62].

Coating Construction Method
Nguyen et al. [35] prepared red K 2 SiF 6 : Mn 4+ (KSFM) phosphors by the two-step coprecipitation method and added KSFM to the solution of Al(NO 3 ) 3 , P 2 O 5 , and CH 3 OH (M). Figure 4 shows the synthesis diagram of KSFM-MOPAl. Al 3+ ions act as the cross-linking agent between alkyl groups, and the P-O bond is broken and partially replaced by alkyl groups, forming a M-O-P-O-M bond and generating an organophosphorus layer. Through the adsorption mechanism, a network is formed on the surface of KSFM, and KSFM-MOPAl is obtained. Zhou et al. [36] added K 2 TiF 6 : Mn 4+ to octadecyl trimethoxysilane and collected K 2 TiF 6 @ODTMS after stirring vigorously for 2 h. Kim et al. [37] introduced different silane materials via mixing with steam of isopropyl alcohol (IPA) and ammonium hydroxide (NH 4 OH) into the plasma reactor, respectively. After 20 min of plasma treatment, modified K 2 SiF 6 : Mn 4+ (KSFM) was obtained. Arunkumar et al. [38] dissolved oleic acid (OA) in absolute ethanol, added K 2 SiF 6 : Mn 4+ , dispersed it in the above solution for 1 h, then heated the mixed solution in a reaction kettle at 140 • C for 6 h. Finally, KSF-OA with a shell structure was formed. Fang et al. [37] used HF to remove impurities and small particles on the surface of K 2 TiF 6 : Mn 4+ (KTF) by the surface etching method, so that the surface of KTF was smooth. OA and SiO 2 coatings were constructed by the hydrothermal method and the stirring method at room temperature to obtain a double-coated structural material, namely, KTF@OA@SiO 2 . Luo et al. [40] added K 2 SiF 6 : Mn 4+ to pyruvate (PA, 98%), stirred it for 6 h, washed it with ethanol, and dried it to obtain KSRM-98PA. PA first reacted with Mn 4+ on the surface to construct a Mn 4+ -free layer on the surface. In addition to reduction, PA could also form a soft shell on the fluoride surface through chemical bonds, which could prevent Mn 4+ from reacting with water molecules on the surface. homogeneous shells. There are mainly three preparation methods including the coating construction method, the surface passivation method, and the saturated crystallization method. The materials that making up the shell structure are different, so the properties of the fluoride also change [62].

Coating Construction Method
Nguyen et al. [35] prepared red K2SiF6: Mn 4+ (KSFM) phosphors by the two-step coprecipitation method and added KSFM to the solution of Al(NO3)3, P2O5, and CH3OH (M). Figure 4 shows the synthesis diagram of KSFM-MOPAl. Al 3+ ions act as the cross-linking agent between alkyl groups, and the P-O bond is broken and partially replaced by alkyl groups, forming a M-O-P-O-M bond and generating an organophosphorus layer. Through the adsorption mechanism, a network is formed on the surface of KSFM, and KSFM-MOPAl is obtained. Zhou et al. [36] added K2TiF6: Mn 4+ to octadecyl trimethoxysilane and collected K2TiF6@ODTMS after stirring vigorously for 2 h. Kim et al. [37] introduced different silane materials via mixing with steam of isopropyl alcohol (IPA) and ammonium hydroxide (NH4OH) into the plasma reactor, respectively. After 20 min of plasma treatment, modified K2SiF6: Mn 4+ (KSFM) was obtained. Arunkumar et al. [38] dissolved oleic acid (OA) in absolute ethanol, added K2SiF6: Mn 4+ , dispersed it in the above solution for 1 h, then heated the mixed solution in a reaction kettle at 140 °C for 6 h. Finally, KSF-OA with a shell structure was formed. Fang et al. [37] used HF to remove impurities and small particles on the surface of K2TiF6: Mn 4+ (KTF) by the surface etching method, so that the surface of KTF was smooth. OA and SiO2 coatings were constructed by the hydrothermal method and the stirring method at room temperature to obtain a doublecoated structural material, namely, KTF@OA@SiO2. Luo et al. [40] added K2SiF6: Mn 4+ to pyruvate (PA, 98%), stirred it for 6 h, washed it with ethanol, and dried it to obtain KSRM-98PA. PA first reacted with Mn 4+ on the surface to construct a Mn 4+ -free layer on the surface. In addition to reduction, PA could also form a soft shell on the fluoride surface through chemical bonds, which could prevent Mn 4+ from reacting with water molecules on the surface. Dong et al. [41] added K2TiF6: Mn 4+ to a KF and Ca(NO3)2 solution, and Yu et al. [63] added K2SiF6: Mn 4+ to a HF and Ca(NO3)2 solution via mixing and stirring, resulting in the obtaining of KTF@CaF2 and KSF: Mn 4+ @CaF2, respectively. Fang et al. [43] used K2TiF6: Mn 4+ , KHF2, and Sr(NO3)2 as raw materials to obtain K2TiF6: Mn 4+ @SrF2. KHF2 played a bridging role in the synthesis of the shell structure, which could alleviate lattice mismatch. When K2TiF6: Mn 4+ (hexagonal phase) powder was immersed in a KHF2(cubic phase) Dong et al. [41] added K 2 TiF 6 : Mn 4+ to a KF and Ca(NO 3 ) 2 solution, and Yu et al. [63] added K 2 SiF 6 : Mn 4+ to a HF and Ca(NO 3 ) 2 solution via mixing and stirring, resulting in the obtaining of KTF@CaF 2 and KSF: Mn 4+ @CaF 2 , respectively. Fang et al. [43] used K 2 TiF 6 : Mn 4+ , KHF 2 , and Sr(NO 3 ) 2 as raw materials to obtain K 2 TiF 6 : Mn 4+ @SrF 2 . KHF 2 played a bridging role in the synthesis of the shell structure, which could alleviate lattice mismatch. When K 2 TiF 6 : Mn 4+ (hexagonal phase) powder was immersed in a KHF 2 (cubic phase) solution, the free [HF 2 ] − could easily replace F − ions in K 2 TiF 6 and form a KHF 2 thin layer on the surface of K 2 TiF 6 . After adding Sr 2+ ions, KHF 2 provided F − ions for the nucleation growth of SrF 2 particles on the KHF 2 surface. Based on the chemical precipitation reaction of heterogeneous nucleation, the modification formed a denser and homogeneous coating on the surface of K 2 TiF 6 : Mn 4+ . Verstraete et al. [44] used atomic layer deposition (ALD) of TiO 2 or Al 2 O 3 to form a hydroxyl terminated functional seed layer on the K 2 SiF 6 : Mn 4+ surface. The functionalized particle surface provided enhanced adhesion properties with conventional hydrophobic or moisture-resistant shells compared to the fluorine-terminated surface of the untreated phosphor. K 2 SiF 6 :Mn 4+ -Al 2 O 3 and K 2 SiF 6 :Mn 4+ -TiO 2 core-shell phosphors were obtained by this method. Kate et al. [45] used the vapor deposition method to deposit ultra-thin Al 2 O 3 on the surface of K 2 SiF 6 : Mn 4+ with trimethyl aluminum, O 3 , and N 2 as raw materials. Quan et al. [46] synthesized K 2 SiF 6 : Mn 4+ @SiO 2 (KSF@SiO 2 ) by using tetraethyl orthosilicate (TEOS), isopropanol, H 2 O 2 , K 2 MnO 4 , KF·2H 2 O, sodium dodecyl benzene sulfonate (SDBS), and HF as raw materials. Li et al. [59] used KF·2H 2 O, KMnO 4 , HF (40%), and K 2 SiF 6 as raw materials, and the nuclear material K 2 SiF 6 : Mn 4+ was added to the HF solution of K 2 SiF 6 and stirred for 2 h to obtain the final product K 2 SiF 6 : Mn 4+ @SiO 2 (KSF@SiO 2 ).
Yu et al. [47] mixed K 2 SiF 6 : Mn 4+ , Na + , GQD, and HF solution and transferred them to the reactor. K 2 SiF 6 : Mn 4+ and Na + @GQDs were obtained by holding at 120 • C for 3 h. the K 2 SiF 6 : Mn 4+ , Na + @GQDs@KSF double shell layer structure was finally achieved by stirring K 2 SiF 6 in HF solution. On the basis of improving water resistance, the GQDs coating material not only improved the luminescence intensity of the phosphor, but it also has a negative thermal quenching effect (NTQ) at higher temperatures. Liu et al. [48] synthesized K 2 SiF 6 : Mn 4+ @C by using chemical vapor deposition to decompose acetylene at a high temperature to generate a nano-carbon layer and form a hydrophobic protective layer on the surface of phosphor.
The experimental results showed that the coating construction method can successfully assemble the core-shell structure and enhance the water resistance of the fluoride. However, its process is complex, and there are some special requirements on the reaction conditions and the equipment operated in the experiment. The experiments may be carried out under high temperature and pressure conditions, while even needed to use toxic and volatile HF, which increases the safety hazards. Moreover, the equipment used to deposit the shell is usually expensive. Though constructing a double-layer coating could further enhance the water resisting property, the experimental steps are more complicated, and the difficulty of the experimental is also increased.

Surface Passivation Method
The surface passivation method is another strategy to improve the water resistance of fluoride. A homogeneous shell usually adopts the surface passivation method, in which Mn 4+ ions are removed from the surface by a reducing agent, leaving the matrix material in situ to form a protective shell. This method prevents the formation of crystal defects at the core-shell interface caused by non-uniform coating and reduces fluorescence quantum yields (PLQYs).
Huang et al. [49] synthesized K 2 GeF 6 :Mn 4+ (KGFM) red phosphors by the threestep chemical coprecipitation method. KGFM was dissolved in the mixed solution of DL-mandelic acid and ethanol, and it was stirred for 2 h to obtain the final product KGFM@MA. Huang et al. [50] also synthesized K 2 SiF 6 : Mn 4+ @K 2 SiF 6 by the coating and surface passivation methods, respectively, as shown in Figure 5a,b. Among them, the surface passivation method was use to dissolve K 2 SiF 6 : Mn 4+ in the mixed aqueous solution of H 3 PO 4 /H 2 O 2 , which promoted the release and decomposition of [MnF 6 ] 2− on the surface, so that the K 2 SiF 6 :Mn 4+ surface was converted into K 2 SiF 6 and the product WR-KFSM-8 was obtained. Zhou et al. [51] added K 2 XF 6 : Mn 4+ (KXF, X = Ti, Si, Ge) to H 2 O 2 (30 wt%) solution and stirred it to reduce the distribution of Mn 4+ ions on the surface of K 2 XF 6 : Mn 4+ . After washing with acetic acid and ethanol, the final product (P-KXF) was collected. The luminescence properties of hydrolyzed fluorine phosphors can be repaired by adding H 2 O 2 . Liu et al. [52] synthesized Cs 2 SiF 6 : Mn 4+ phosphor (P-CSFM) by the same method. [MnF 6 ] 2− existed on the surface of P-CSFM phosphor and could be effectively reduced by reacting with H 2 O 2 in an acidic environment. luminescence properties of hydrolyzed fluorine phosphors can be repaired by adding H2O2. Liu et al. [52] synthesized Cs2SiF6: Mn 4+ phosphor (P-CSFM) by the same method.
[MnF6] 2− existed on the surface of P-CSFM phosphor and could be effectively reduced by reacting with H2O2 in an acidic environment. Yu et al. [53] [54] added Rb2SnF6: Mn 4+ to the oxalic acid solution, and they stirred and passivated Rb2SnF6: Mn 4+ with oxalic acid to construct a passivation protective layer on the surface of. Liu et al. [55] employed the "good from bad" method to prepare R-KSFM by hydrolyzing the commercial K2SiF6: Mn 4+ (O-KSFM) phosphors in deionized water for 10 min and then poured the oxalic acid solution into the above solution and stirred for 1 h. The collected R-KSFM not only fully recovered the luminescence properties but also possessed high moisture resistance. Li et al. [56] synthesized bicentric ions CsNaGe0.5Sn0.5F6: Mn 4+ (CNGSF) phosphors by the ethanol crystallization method. CNGSF was added into the weak reducing agent solution composed of citric acid or oxalic acid, and then stirring was carried out for the production of P1-CNGSF and P2-CNGSF samples. Zhong et al. [57] added the prepared Na2SiF6: Mn 4+ , Li + (LNSF: Mn 4+ ) into the KF-HF solution and stirred it; then they added aqueous citric acid solution and maintained stirring for 12 h to obtain the product of LNSF: Mn 4+ with a shell layer (LNSF: Mn 4+ -CA). A protective layer with low solubility was in situ-formed on the surface, which could effectively isolate Mn 4+ from the aqueous layer and help to improve the waterproof performance of LNSF: Mn 4+ -CA. In order to achieve the in situ formation of K2SiF6 on the surface of the K2SiF6: Mn 4+ red phosphor particle, Cai et al. [58] put K2SiF6: Mn 4+ phosphor into FeCl2 solution and utilized Fe 2+ as a reducing agent to produce T-KSF phosphor with excellent water resistance. Wan et al. [29] proposed the new idea of reconstructing Mn 4+free shells of fluoride by reduction-assisted surface recrystallization (RSRC). The synthesized K2SiF6: Mn 4+ was added to a reducing agent containing α-hydroxyl groups, such as L-tartaric acid (TA), DL-malic acid (MA), citric acid (CA), ascorbic acid (AA), or DL-lactic acid solution (LA), forming saturated K2SiF6 solution and stirring for 2 h to obtain serial phosphors. During the dissolution-crystallization equilibrium process of fluoride crystallization in HF, the reducing agent removed Mn 4+ ions from the solution to prevent Mn 4+ from re-entering the surface of the fluoride crystals to construct a Mn 4+ -free shell. Specifically, Huang et al. [28] constructed K2TiF6: Mn 4+ @K2TiF6 phosphors with a uniform coating core-shell structure based on the reverse cation exchange reaction. K2TiF6: Mn 4+ microcrystals were synthesized by replacing Ti 4+ ions in K2TiF6 crystals with Mn 4+ ions in HF Yu et al. [53] added K 2 GeF 6 : Mn 4+ into oxalic acid solution and stirred for 12 h. The interface between K 2 GeF 6 : Mn 4+ and the solution was ionized, releasing K + ions, [GeF 6 ] 2− , and [MnF 6 ] 2− , where [MnF 6 ] 2− was reduced to Mn 2+ . The free [GeF 6 ] 2− anion was combined with K + ions to form K 2 GeF 6 , which was deposited on the surface of K 2 GeF 6 : Mn 4+ to form K 2 GeF 6 : Mn 4+ @K 2 GeF 6 (T-KGF). Jiang et al. [54] added Rb 2 SnF 6 : Mn 4+ to the oxalic acid solution, and they stirred and passivated Rb 2 SnF 6 : Mn 4+ with oxalic acid to construct a passivation protective layer on the surface of. Liu et al. [55] employed the "good from bad" method to prepare R-KSFM by hydrolyzing the commercial K 2 SiF 6 : Mn 4+ (O-KSFM) phosphors in deionized water for 10 min and then poured the oxalic acid solution into the above solution and stirred for 1 h. The collected R-KSFM not only fully recovered the luminescence properties but also possessed high moisture resistance. Li et al. [56] synthesized bicentric ions CsNaGe 0.5 Sn 0.5 F 6 : Mn 4+ (CNGSF) phosphors by the ethanol crystallization method. CNGSF was added into the weak reducing agent solution composed of citric acid or oxalic acid, and then stirring was carried out for the production of P1-CNGSF and P2-CNGSF samples. Zhong et al. [57] added the prepared Na 2 SiF 6 : Mn 4+ , Li + (LNSF: Mn 4+ ) into the KF-HF solution and stirred it; then they added aqueous citric acid solution and maintained stirring for 12 h to obtain the product of LNSF: Mn 4+ with a shell layer (LNSF: Mn 4+ -CA). A protective layer with low solubility was in situ-formed on the surface, which could effectively isolate Mn 4+ from the aqueous layer and help to improve the waterproof performance of LNSF: Mn 4+ -CA. In order to achieve the in situ formation of K 2 SiF 6 on the surface of the K 2 SiF 6 : Mn 4+ red phosphor particle, Cai et al. [58] put K 2 SiF 6 : Mn 4+ phosphor into FeCl 2 solution and utilized Fe 2+ as a reducing agent to produce T-KSF phosphor with excellent water resistance. Wan et al. [29] proposed the new idea of reconstructing Mn 4+ -free shells of fluoride by reduction-assisted surface recrystallization (RSRC). The synthesized K 2 SiF 6 : Mn 4+ was added to a reducing agent containing α-hydroxyl groups, such as L-tartaric acid (TA), DL-malic acid (MA), citric acid (CA), ascorbic acid (AA), or DL-lactic acid solution (LA), forming saturated K 2 SiF 6 solution and stirring for 2 h to obtain serial phosphors. During the dissolution-crystallization equilibrium process of fluoride crystallization in HF, the reducing agent removed Mn 4+ ions from the solution to prevent Mn 4+ from re-entering the surface of the fluoride crystals to construct a Mn 4+ -free shell. Specifically, Huang et al. [28] constructed K 2 TiF 6 : Mn 4+ @K 2 TiF 6 phosphors with a uniform coating core-shell structure based on the reverse cation exchange reaction. K 2 TiF 6 : Mn 4+ microcrystals were synthesized by replacing Ti 4+ ions in K 2 TiF 6 crystals with Mn 4+ ions in HF solution. Notably, the cation exchange process was reversible in nanocrystals. This reverse process was also applicable to K 2 TiF 6 : Mn 4+ crystals. Using K 2 TiF 6 as raw material, Ti 4+ ions replaced Mn 4+ of K 2 TiF 6 : Mn 4+ in HF solution, leaving the K 2 TiF 6 shell to protect the internal K 2 TiF 6 : Mn 4+ core from ambient moisture and to prevent hydrolysis of [MnF 6 ] 2− .
The surface passivation method can complete the construction of the shell at room temperature, which is an effective method to construct the core-shell structure because of its simple operation and easily available raw materials. By using an acidic substance containing an α-hydroxy group as a reducing agent, the content of Mn 4+ ions on the surface of the material can be appropriately decreased, and a protective layer can be formed without reducing the original luminescence. The protective layer can enhance the water resistance of A 2 MF 6 : Mn 4+ by preventing the hydrolysis of the nuclear luminescence center.

Saturated Crystallization Method
In general, the crystallization process of crystals includes nucleation and crystal growth. The crystallization process of A 2 MF 6 : Mn 4+ does not need to add additional heterogeneous materials and reducing agents. The saturated solution will precipitate crystals with the passage of time. During this process, some Mn 4+ ions will generate Mn compounds with other valence states, making their concentrations change. The concentration difference between [MnF 6 ] 2− and matrix ions can precipitate crystals with fewer Mn 4+ ions on the surface. What is more, the single crystal phosphor has the advantages of high crystallinity, few defects, and high thermal conductivity.
Zhou et al. [30] mixed H 2 SiF 6 , KF, K 2 SiF 6 , HF, and K 2 MnF 6 and stirred vigorously to form K 2 SiF 6 /K 2 MnF 6 saturated solution. They filtered the mixed solution with a filter and placed it in a fume hood. After volatilizing it at room temperature for a period of time, K 2 SiF 6 : Mn 4+ (KSFM) crystals could be grown from the saturated solution, as shown in Figure 6a. During the process of crystal growth, [SiF 6 ] 2− groups did not hydrolyze in acidic solution, while its concentration decreased with the precipitation of KSFM single crystals. However, [MnF 6 ] 2− were prone to disproportionation reactions to produce other compounds (Mn 2+ , [MnO 4 ] − ) of manganese. With the growth time prolonging, the concentration of [MnF 6 ] 2− decreased more rapidly than that of [SiF 6 ] 2− , and as a result, it was much lower than that of [SiF 6 ] 2− after a few days. The [SiF 6 ] 2− drove the crystal growth, which promoted the growth of KSF on the surface of KSFM crystals, forming a surface with low Mn 4+ content. As shown in Figure 6b, the KSFM single crystal has a cubic shape and six chamfers. As the growth time was prolonged from 6 h to 4 days, the average size of KSFM crystals increased from 200 µm to 1 mm. Zhong et al. [50] used a similar method to add a certain molar mass of Na 2 SiF 6 : Li + (LNSF) powder to the mixed solution of KMnO 4 , KF·2H 2 O, and HF and then stirred it for 48 h at room temperature. With the passing of time, the [MnF 6 ] 2− on the crystal surface gradually decreased, forming a shell almost free of Mn 4+ . LNSF: Mn 4+ crystals were obtained by washing. The water resistance of single crystals was positively correlated with crystallization time. In particular, Jiang et al. [61] introduced K 2 SiF 6 : Mn 4+ to saturate the HF solution of K 2 SiF 6 and stirred. After stirring for a duration of time, K 2 SiF 6 : Mn 4+ @ K 2 SiF 6 could be synthesized by adding the appropriate amount of ethanol to cause the epitaxial growth of the K 2 SiF 6 : Mn 4+ shell and then synthesized K 2 SiF 6 : Mn 4+ @ K 2 SiF 6 . The addition of ethanol could also enhance the emission intensity of K 2 SiF 6 : Mn 4+ . When the concentration of Mn 4+ was 4% mol, the PL intensity of the product with the addition of ethanol was 5.03 times higher than that without the addition of ethanol. The saturated solution crystallization method can construct the core-shell structure at one time, which reduces the experimental operation steps, and yields phosphor with fewer surface defects. Compared with the other two methods, it has unique advantages and provides a new idea for constructing the core-shell structure. However, if the crystallization process is subjected to external forces, the influence on the experimental results is particularly obvious. When the saturated solution is impacted or stirred, the quality and size of the crystal will be affected. This is caused by the balance between nucleation and growth during crystallization, which may even degrade the optical properties. Moreover, this method requires more time to ensure the formation of the core-shell structure.

Luminescent Mechanism of A2MF6: Mn 4+
In A2MF6: Mn 4+ , each M 4+ ion is surrounded by six F − ions, forming a [MF6] 2− octahedral structure. Furthermore, the A + ions are located in the center of 12 adjacent F − ions, forming a regular polyhedron. Figure 7 shows the schematic crystal structure of K2SiF6: Mn 4+ . The overall and local structure of A2MF6: Mn 4+ are similar. There is also a clear overlap in excitation and emission spectra. Mn 4+ has two broad excitations at 300-500 nm and a series of narrow-band emissions in the 600-650 nm range. The optical properties of some transition metal ions with 3 d3 structures (such as Mn 4+ , Cr 3+ , and V 2+ ) can all be explained by the Tanabe-Sugano energy diagram [64]. The electronic energy levels of 3 d3 transition metal ions are affected by Dq/B, where Dq is the crystal field strength, and B is the Racah parameter. The energies of most multiparticle are strongly dependent on the crystal field strength, except for the 2 T1g and 2 Eg energy levels parallel to the ground state 4 A2g. The strength of Dq depends on the distance between the ions and the ligand R. High energy states such as 4 T2g, 4 T1g, and 2 A1g can be modulated by R and the type of ligand. The saturated solution crystallization method can construct the core-shell structure at one time, which reduces the experimental operation steps, and yields phosphor with fewer surface defects. Compared with the other two methods, it has unique advantages and provides a new idea for constructing the core-shell structure. However, if the crystallization process is subjected to external forces, the influence on the experimental results is particularly obvious. When the saturated solution is impacted or stirred, the quality and size of the crystal will be affected. This is caused by the balance between nucleation and growth during crystallization, which may even degrade the optical properties. Moreover, this method requires more time to ensure the formation of the core-shell structure.

Luminescent Mechanism of A 2 MF 6 : Mn 4+
In A 2 MF 6 : Mn 4+ , each M 4+ ion is surrounded by six F − ions, forming a [MF 6 ] 2− octahedral structure. Furthermore, the A + ions are located in the center of 12 adjacent F − ions, forming a regular polyhedron. Figure 7 shows the schematic crystal structure of K 2 SiF 6 : Mn 4+ . The overall and local structure of A 2 MF 6 : Mn 4+ are similar. There is also a clear overlap in excitation and emission spectra. Mn 4+ has two broad excitations at 300-500 nm and a series of narrow-band emissions in the 600-650 nm range. The optical properties of some transition metal ions with 3 d 3 structures (such as Mn 4+ , Cr 3+ , and V 2+ ) can all be explained by the Tanabe-Sugano energy diagram [64]. The electronic energy levels of 3 d 3 transition metal ions are affected by D q /B, where D q is the crystal field strength, and B is the Racah parameter. The energies of most multiparticle are strongly dependent on the crystal field strength, except for the 2 T 1g and 2 E g energy levels parallel to the ground state 4 A 2g . The strength of D q depends on the distance between the ions and the ligand R. High energy states such as 4 T 2g , 4 T 1g , and 2 A 1g can be modulated by R and the type of ligand. The saturated solution crystallization method can construct the core-shell structure at one time, which reduces the experimental operation steps, and yields phosphor with fewer surface defects. Compared with the other two methods, it has unique advantages and provides a new idea for constructing the core-shell structure. However, if the crystallization process is subjected to external forces, the influence on the experimental results is particularly obvious. When the saturated solution is impacted or stirred, the quality and size of the crystal will be affected. This is caused by the balance between nucleation and growth during crystallization, which may even degrade the optical properties. Moreover, this method requires more time to ensure the formation of the core-shell structure.

Luminescent Mechanism of A2MF6: Mn 4+
In A2MF6: Mn 4+ , each M 4+ ion is surrounded by six F − ions, forming a [MF6] 2− octahedral structure. Furthermore, the A + ions are located in the center of 12 adjacent F − ions, forming a regular polyhedron. Figure 7 shows the schematic crystal structure of K2SiF6: Mn 4+ . The overall and local structure of A2MF6: Mn 4+ are similar. There is also a clear overlap in excitation and emission spectra. Mn 4+ has two broad excitations at 300-500 nm and a series of narrow-band emissions in the 600-650 nm range. The optical properties of some transition metal ions with 3 d3 structures (such as Mn 4+ , Cr 3+ , and V 2+ ) can all be explained by the Tanabe-Sugano energy diagram [64]. The electronic energy levels of 3 d3 transition metal ions are affected by Dq/B, where Dq is the crystal field strength, and B is the Racah parameter. The energies of most multiparticle are strongly dependent on the crystal field strength, except for the 2 T1g and 2 Eg energy levels parallel to the ground state 4 A2g. The strength of Dq depends on the distance between the ions and the ligand R. High energy states such as 4 T2g, 4 T1g, and 2 A1g can be modulated by R and the type of ligand. According to the d-d transition rule, the 4 A 2g → 4 T 1g and 4 A 2g → 4 T 2g of Mn 4+ are spinallowed transitions, which are located in the near ultraviolet (UV) region of 300-400 nm and the blue region of 400-500 nm, respectively. When they are excited to the 4 T 1g or 4 T 2g level, the excited ions usually relax non-radiatively to 2 E g , and the 2 E g → 4 A 2g transition emission is inhibited by both spin and equivalence. However, due to the coupling between electrons and phonons, the emission is partially unlocked, and there are many sharp narrow band emission peaks in the range of 610-650 nm. The optical transition of Mn 4+ ions is sensitive to its local coordination. In a symmetric low-host crystal, Mn 4+ ions can partially break the odd-even forbidden transition rule, resulting in a strong zero phonon line (ZPL), which is generally located at about 620 nm. The emission peaks on the left and right sides of ZPL are the anti-Stokes peak and the Stokes peak, respectively. Strong ZPL can improve the red-light purity of Mn 4+ -doped red phosphors, which is beneficial for WLED applications.

Fluorescent Properties and Water Resistance
It was found that A 2 MF 6 : Mn 4+ was very sensitive to humidity, and the Mn 4+ dopant on the surface was easy to hydrolyze into mixed manganese oxides and hydroxides, which darkened the color and extremely weakened its red emission intensity. For moisturesensitive materials, coating a waterproof layer on the surface was an effective method to raise its stability. Mn 4+ -doped fluoride was unstable in the water environment, and its water resistance was improved by constructing a core-shell structure. Generally, fluoride is exposed to high temperature (80 • C, HT) and high humidity (80%, HH), water, or boiling water for a period of time, and the effect of water resistance is measured by the remaining photoluminescence (PL) intensity. In addition, the water resistance could also be measured by the contact angle between the phosphor and water. Moreover, compared with A 2 MF 6 : Mn 4+ without the shell, some of the A 2 MF 6 : Mn 4+ with the core-shell structure not only improved the water resistance but also had a certain degree of improvement in fluorescence performance and thermal stability.

A 2 MF 6 : Mn 4+ with Heterogeneous Shell Layer
Nguyen et al. [35] coated K 2 SiF 6 : Mn 4+ (KSFM-MOPAl) with an alkyl phosphate layer. The yellow KSFM phosphor turned brown after immersion in water for 30 min, while KSFM-MOPAl remained yellow after immersion in water for 60 min. The non-radiative transition activation energy of the coated KSFM-MOPAl phosphor (0.92 eV) was 0.09 eV higher than that of the uncoated KSFM phosphor (0.83 eV). As a result, the temperature of thermal quenching was improved, and the comprehensive PL intensity of KSFM-MOPAl at 523K was 100% of that at room temperature. However, with the increase of shell thickness, the luminescence intensity of phosphor decreased slightly. Kim et al. [37] modified K 2 SiF 6 : Mn 4+ (KSFM) with an alkyl coupling agent, as shown in Figure 8a,b. The surface of the unmodified phosphor had hydrophilic -OH groups, which combined with water molecules in the air, leading to structure destruction and optical property degradation of the phosphor. Silane particles reacted with -OH bunches on the surface of the fluoride to create a thin layer on its surface, which effectively protected the phosphor from external moisture. In Figure 8c,f, the initialization value of the water contact angle was 6.64 • , while after surface modification, the angle expanded to 122.47 • after surface modification. Superhydrophobic silane coupling agents (SCAs) were employed by Zhou et al. [36] to modify coating surfaces and increase the moisture resistance of Mn 4+ -activated fluoride phosphors. As shown in Figure 9, the fluorescence intensity of K 2 TiF 6 almost did not weaken after surface modification. After surface modification with OTMS, DTMS, HDTMS, and ODTMS solutions, the water contact angle of K 2 TiF 6 significantly increased from nearly 0 • to 36.8 • , 52.7 • , 147.9 • , and 155.4 • , respectively. Compared with the experimental results of Kim et al. [37], both of them formed hydrophobic organic layers on the surface of phosphors, which significantly reduced the -OH or H 2 O molecules on the surface of the phosphors and improved the water-resistance of the phosphor, and the hydrophobicity increased as the carbon chain of silane became longer. However, the experiment of Zhou et al. had higher feasibility because it was simple to operate and was carried out at room temperature, and it also caused the phosphor to have better moisture resistance. Arunkumar et al. [38] used oleic acid (OA) as a hydrophobic layer, and the hydrophobic tail of OA formed a thin layer to protect phosphors from atmospheric moisture and prevent the hydrolysis of [MnF 6 ] 2− . K 2 SiF 6 -OA and DenKa-based K 2 SiF 6 phosphors were aged at HT and HH for 450 h and then packaged into WLEDs. The emission intensity of K 2 SiF 6 -OA-WLED decreased by 15%, while the commercial DenKa-based K 2 SiF 6 -WLED decreased by 23% under the device current of 120 mA. The water stability and efficient red emission of the encapsulated K 2 SiF 6 @OA phosphors indicated that employing OA to treat fluorine phosphors was a feasible method and could be apply in WLED. However, the synthesis temperature of this method was high, and there was a risk of toxic by-products. Luo et al. [40] used pyruvate to prepare KFSM-98PA with a bilayer shell, and its internal quantum efficiency (IQE) value was as high as 99.71%. However, due to the influence of the shell, the PL intensity was 90% of the original KSFM. After soaking in water for 360 h, the PL intensity was maintained at 88.5% of the original intensity, while the shell-free KSFM was only 51.6%. Moreover, the color of KSFM-98PA did not change, even after boiling in water for 1 h. In addition, PA could recover the fluorescence of hydrolyzed fluoride, and the IQE of hydrolyzed h-KSFM could recover up to 95.24%. Kim et al. [37], both of them formed hydrophobic organic layers on the surface of phosphors, which significantly reduced the -OH or H2O molecules on the surface of the phosphors and improved the water-resistance of the phosphor, and the hydrophobicity increased as the carbon chain of silane became longer. However, the experiment of Zhou et al. had higher feasibility because it was simple to operate and was carried out at room temperature, and it also caused the phosphor to have better moisture resistance. Arunkumar et al. [38] used oleic acid (OA) as a hydrophobic layer, and the hydrophobic tail of OA formed a thin layer to protect phosphors from atmospheric moisture and prevent the hydrolysis of [MnF6] 2− . K2SiF6-OA and DenKa-based K2SiF6 phosphors were aged at HT and HH for 450 h and then packaged into WLEDs. The emission intensity of K2SiF6-OA-WLED decreased by 15%, while the commercial DenKa-based K2SiF6-WLED decreased by 23% under the device current of 120 mA. The water stability and efficient red emission of the encapsulated K2SiF6@OA phosphors indicated that employing OA to treat fluorine phosphors was a feasible method and could be apply in WLED. However, the synthesis temperature of this method was high, and there was a risk of toxic by-products. Luo et al. [40] used pyruvate to prepare KFSM-98PA with a bilayer shell, and its internal quantum efficiency (IQE) value was as high as 99.71%. However, due to the influence of the shell, the PL intensity was 90% of the original KSFM. After soaking in water for 360 h, the PL intensity was maintained at 88.5% of the original intensity, while the shell-free KSFM was only 51.6%. Moreover, the color of KSFM-98PA did not change, even after boiling in water for 1 h. In addition, PA could recover the fluorescence of hydrolyzed fluoride, and the IQE of hydrolyzed h-KSFM could recover up to 95.24%.  Compared with K2TiF6, the luminescence intensity of the K2TiF6@CaF2 phosphor prepared by Dong et al. [41] increased slightly. The change of luminescence intensity after surface modification was mainly due to the removal of surface hydroxyl groups by surface coating modification, which enhanced the emission. As shown in Figure 10, with the change of the CaF2 coating amount, the KTF@CaF2 had a blue shift of 2-5 nm. The CaF2 coating changed the crystal field environment around Mn 4+ , resulting in a blue shift in its Kim et al. [37], both of them formed hydrophobic organic layers on the surface of phosphors, which significantly reduced the -OH or H2O molecules on the surface of the phosphors and improved the water-resistance of the phosphor, and the hydrophobicity increased as the carbon chain of silane became longer. However, the experiment of Zhou et al. had higher feasibility because it was simple to operate and was carried out at room temperature, and it also caused the phosphor to have better moisture resistance. Arunkumar et al. [38] used oleic acid (OA) as a hydrophobic layer, and the hydrophobic tail of OA formed a thin layer to protect phosphors from atmospheric moisture and prevent the hydrolysis of [MnF6] 2− . K2SiF6-OA and DenKa-based K2SiF6 phosphors were aged at HT and HH for 450 h and then packaged into WLEDs. The emission intensity of K2SiF6-OA-WLED decreased by 15%, while the commercial DenKa-based K2SiF6-WLED decreased by 23% under the device current of 120 mA. The water stability and efficient red emission of the encapsulated K2SiF6@OA phosphors indicated that employing OA to treat fluorine phosphors was a feasible method and could be apply in WLED. However, the synthesis temperature of this method was high, and there was a risk of toxic by-products. Luo et al. [40] used pyruvate to prepare KFSM-98PA with a bilayer shell, and its internal quantum efficiency (IQE) value was as high as 99.71%. However, due to the influence of the shell, the PL intensity was 90% of the original KSFM. After soaking in water for 360 h, the PL intensity was maintained at 88.5% of the original intensity, while the shell-free KSFM was only 51.6%. Moreover, the color of KSFM-98PA did not change, even after boiling in water for 1 h. In addition, PA could recover the fluorescence of hydrolyzed fluoride, and the IQE of hydrolyzed h-KSFM could recover up to 95.24%.  Compared with K2TiF6, the luminescence intensity of the K2TiF6@CaF2 phosphor prepared by Dong et al. [41] increased slightly. The change of luminescence intensity after surface modification was mainly due to the removal of surface hydroxyl groups by surface coating modification, which enhanced the emission. As shown in Figure 10, with the change of the CaF2 coating amount, the KTF@CaF2 had a blue shift of 2-5 nm. The CaF2 coating changed the crystal field environment around Mn 4+ , resulting in a blue shift in its Compared with K 2 TiF 6 , the luminescence intensity of the K 2 TiF 6 @CaF 2 phosphor prepared by Dong et al. [41] increased slightly. The change of luminescence intensity after surface modification was mainly due to the removal of surface hydroxyl groups by surface coating modification, which enhanced the emission. As shown in Figure 10, with the change of the CaF 2 coating amount, the KTF@CaF 2 had a blue shift of 2-5 nm. The CaF 2 coating changed the crystal field environment around Mn 4+ , resulting in a blue shift in its luminescence. After 2 h of immersion, the luminescence intensity of KTF@CaF 2 decreased by only 13.6%, while KTF decreased by 93.2%. Yu et al. [63] prepared K 2 SiF 6 : Mn 4+ @CaF 2 with high water resistance and thermal stability by the hydrothermal method and surface coating process. Compared with KSF: Mn 4+ , the CaF 2 shell could effectively prevent the hydrolysis of surface [MnF 6 ] 2− groups to MnO 2 . After soaking in water for 6 h, the luminescence intensity of the uncoated product decreased to 41.68% of the initial product, while the coated material was 88.24% of the initial value. Compared with Dong's KTF: Mn 4+ @CaF 2 , KSF: Mn 4+ @CaF 2 had a negative thermal quenching effect (NTQ), and the mechanism can be considered to be a thermal-light energy conversion mechanism. Fang et al. [43] modified the SrF 2 shell on the K 2 TiF 6 :Mn 4+ surface, but the PL intensity slightly decreased with the amount of SrF 2 due to increased light scattering and absorption at the core-shell interface. The PL intensity of K 2 TiF 6 : Mn 4+ @SrF 2 phosphor soaked in distilled water for 2 h maintained more than 90% of the initial value, while the PL of K 2 TiF 6 : Mn 4+ was only 21.2% of the initial value. The results showed that the SrF 2 shell could effectively cut off the hydrolysis of the inner [MnF 6 ] 2− groups. Quan et al. [46] synthesized K 2 SiF 6 @SiO 2 , which had good thermal stability in the range of 300 • C. The PL intensity was 100% of the initial value at 250 • C and 75% at 300 • C, which was better than the thermal stability of K 2 SiF 6 (82% at 250 • C and 27% at 300 • C). After soaking in water for 1 h, the luminescence intensity values of K 2 SiF 6 @SiO 2 and K 2 SiF 6 were 43% and 7% of the initial values, respectively. SiO 2 could significantly improve the water-resistance of K 2 SiF 6 , but K 2 SiF 6 @SiO 2 would hydrolyze in a short time, and its external quantum efficiency was as low as 0.494%.  Yu et al. [47] synthesized (i) K2SiF6: Mn 4+ , (ii) K2SiF6: Mn 4+ , Na + , (iii) K2SiF6: Mn 4+ , Na + @GQDS, and (iv) K2SiF6: Mn 4+ , Na + @GQDs@K2SiF6 by the hydrothermal method and the room temperature coating method. The PL intensity values of (ii), (iii), and (iv) were 1.21, 1.47, and 1.71 times that of (i), respectively, which means that Na + co-doping, GQDs, and KSF shells could further enhance the emission intensity of the samples. At 180 °C, (iii) and (iv) had obvious NTQ effects because their PL intensity values at 180 °C were 217% and 298% of that at 30 °C, respectively. The mechanism of NTQ was further regarded as thermo-optic energy conversion. After being immersed in deionized water for 360 min, the emission intensities of (iii) and (iv) decreased from 100% to 70.57% and 91.63%, respectively. The water resistance of KSFM with a double coating structure improved significantly. As shown in Figure 11, the K2SiF6: Mn 4+ @C phosphor synthesized by Liu et al. [48] could maintain 73% of the initial luminescence intensity after soaking in aqueous solution for 8 h at room temperature, while K2SiF6: Mn 4+ was only 0.7% of the initial value under the same conditions. Although the C deposition layer could improve the waterresistance of KSF, it also decreased the luminescence intensity of KSF. Yu et al. [47] synthesized (i) K 2 SiF 6 : Mn 4+ , (ii) K 2 SiF 6 : Mn 4+ , Na + , (iii) K 2 SiF 6 : Mn 4+ , Na + @GQDS, and (iv) K 2 SiF 6 : Mn 4+ , Na + @GQDs@K 2 SiF 6 by the hydrothermal method and the room temperature coating method. The PL intensity values of (ii), (iii), and (iv) were 1.21, 1.47, and 1.71 times that of (i), respectively, which means that Na + co-doping, GQDs, and KSF shells could further enhance the emission intensity of the samples. At 180 • C, (iii) and (iv) had obvious NTQ effects because their PL intensity values at 180 • C were 217% and 298% of that at 30 • C, respectively. The mechanism of NTQ was further regarded as thermo-optic energy conversion. After being immersed in deionized water for 360 min, the emission intensities of (iii) and (iv) decreased from 100% to 70.57% and 91.63%, respectively. The water resistance of KSFM with a double coating structure improved significantly. As shown in Figure 11, the K 2 SiF 6 : Mn 4+ @C phosphor synthesized by Liu et al. [48] could maintain 73% of the initial luminescence intensity after soaking in aqueous solution for 8 h at room temperature, while K 2 SiF 6 : Mn 4+ was only 0.7% of the initial value under the same conditions. Although the C deposition layer could improve the water-resistance of KSF, it also decreased the luminescence intensity of KSF. Among the coating materials, alkyl coupling agents, oleic acid, pyruvate, CaF2, and nano-carbon can improve the water resistance of the material. Alkyl phosphate, SiO2, and GQDs have certain effects on the enhancement of the water resistance of the materials, but they are not ideal enough. Among them, GQD shells can also improve the thermal stability of Mn 4+ -doped fluoride, which possess the obvious NTQ effect. The heterogeneous shell isolates the luminescent center [MnF6] 2− from the external moisture, which decreases the water sensitivity of the Mn 4+ -doped fluoride. However, some areas of phosphor particles cannot be coated entirely due to the uneven and incomplete deposition of the coating material.

A2MF6: Mn 4+ with Homogeneous Shell Layer
Huang et al. [49] decomposed the oxides and hydroxide of Mn 4+ in situ in an aqueous environment, as shown in Figure 12a,b. The aqueous solution produced by the hydrolysis of KGFM was brown. The aqueous solution of KGFM improved in transparency and clearness after the addition of DL-mandelic acid (MA), and the optical properties of KGFM could also be recovered. Figure 12c,d shows that KGFM@MA had the same high luminescence performance as KGFM. After 168 h of immersion in water, the luminescence intensity of KGFM@MA was almost unchanged (98%), while the luminescence intensity of KGFM was only 33% of the original. As a mild reducing acid material, MA could decompose brown hydrolases in situ to recover the optical properties of fluoride. Furthermore, it was also suitable for other commercial materials to improve the water-resistance, such as K2SiF6: Mn 4+ and K2TiF6: Mn 4+ , which provided a new approach for synthesizing waterresistant Mn 4+ -doped fluoride phosphors with narrow bands. After 168 h of a water erosion experiment, the PL intensities of the original KSFM and KTFM remained at 38.7% and 12.8%, respectively, which was significantly lower than 100% of KSFM@MA and 108% of KTFM@MA. Huang et al. [50] synthesized red phosphors by coating and surface passivation methods, respectively. After soaking in water for 6 h, the WR-KSFM-8 synthesized by the surface passivation method retained 76% of its initial emission intensity, which was much higher than 11% that of IE-KSFM synthesized by the coating method. Zhou et al. [51] reestablished the luminescence properties of hydrolyzed phosphor by adding H2O2. As shown in Figure 13a, the hydrolysis of Mn 4+ quickly reacted with H2O2 until the surface oxidation-reduction was complete and the internal [MnF6] 2− groups were separated from the external moisture. As shown in Figure 13b,c, the same humidity test was performed on P-KSF and the commercial water-resistance phosphor powder C-KSF. Their wet strength values were 97.63% and 80.84% of their initial values, respectively. The passivation surface of the H2O2 aqueous solution provided an effective method for constructing core-shell structures, and the fluoride phosphor prepared by this method had high water-resistance and did not need to consume additional HF solution. The surface redox method could be extended to other doping systems, opening up a new perspective Figure 11. (a) The photographs of K 2 SiF 6 : Mn 4+ and K 2 SiF 6 : Mn 4+ @C after immersion for different times; (b) emission intensity curves of K 2 SiF 6 : Mn 4+ and K 2 SiF 6 : Mn 4+ @C phosphors after hydrolysis for different times. (Reprinted from Ref [48]. Copyright from Elsevier, 2020).
Among the coating materials, alkyl coupling agents, oleic acid, pyruvate, CaF 2 , and nano-carbon can improve the water resistance of the material. Alkyl phosphate, SiO 2 , and GQDs have certain effects on the enhancement of the water resistance of the materials, but they are not ideal enough. Among them, GQD shells can also improve the thermal stability of Mn 4+ -doped fluoride, which possess the obvious NTQ effect. The heterogeneous shell isolates the luminescent center [MnF 6 ] 2− from the external moisture, which decreases the water sensitivity of the Mn 4+ -doped fluoride. However, some areas of phosphor particles cannot be coated entirely due to the uneven and incomplete deposition of the coating material.

A 2 MF 6 : Mn 4+ with Homogeneous Shell Layer
Huang et al. [49] decomposed the oxides and hydroxide of Mn 4+ in situ in an aqueous environment, as shown in Figure 12a,b. The aqueous solution produced by the hydrolysis of KGFM was brown. The aqueous solution of KGFM improved in transparency and clearness after the addition of DL-mandelic acid (MA), and the optical properties of KGFM could also be recovered. Figure 12c,d shows that KGFM@MA had the same high luminescence performance as KGFM. After 168 h of immersion in water, the luminescence intensity of KGFM@MA was almost unchanged (98%), while the luminescence intensity of KGFM was only 33% of the original. As a mild reducing acid material, MA could decompose brown hydrolases in situ to recover the optical properties of fluoride. Furthermore, it was also suitable for other commercial materials to improve the water-resistance, such as K 2 SiF 6 : Mn 4+ and K 2 TiF 6 : Mn 4+ , which provided a new approach for synthesizing water-resistant Mn 4+ -doped fluoride phosphors with narrow bands. After 168 h of a water erosion experiment, the PL intensities of the original KSFM and KTFM remained at 38.7% and 12.8%, respectively, which was significantly lower than 100% of KSFM@MA and 108% of KTFM@MA. Huang et al. [50] synthesized red phosphors by coating and surface passivation methods, respectively. After soaking in water for 6 h, the WR-KSFM-8 synthesized by the surface passivation method retained 76% of its initial emission intensity, which was much higher than 11% that of IE-KSFM synthesized by the coating method. Zhou et al. [51] reestablished the luminescence properties of hydrolyzed phosphor by adding H 2 O 2 . As shown in Figure 13a, the hydrolysis of Mn 4+ quickly reacted with H 2 O 2 until the surface oxidation-reduction was complete and the internal [MnF 6 ] 2− groups were separated from the external moisture. As shown in Figure 13b,c, the same humidity test was performed on P-KSF and the commercial water-resistance phosphor powder C-KSF. Their wet strength values were 97.63% and 80.84% of their initial values, respectively. The passivation surface of the H 2 O 2 aqueous solution provided an effective method for constructing core-shell structures, and the fluoride phosphor prepared by this method had high water-resistance and did not need to consume additional HF solution. The surface redox method could be extended to other doping systems, opening up a new perspective for the development of luminescent materials that enhance the stability of the surface and the duration of the device. Liu et al. [52] synthesized CSFM-P by passivating Cs 2 SiF 6 : Mn 4+ with H 2 O 2 by low-temperature coprecipitation. Compared with the original phosphor CSFM, the emission intensity of the original phosphor decreased to 13.6% after immersion in water for 168 h, while CSFM-P still maintained 74.0% of the initial value. Under excitation at 460 nm, the IQE, EQE, and absorption efficiency of CSFM-P phosphor was 98%, 85%, and 86.8%, whereas the corresponding values of the commercial KSFM were 92%, 67.08%, and 72.92%, respectively. for the development of luminescent materials that enhance the stability of the surface and the duration of the device. Liu et al. [52] synthesized CSFM-P by passivating Cs2SiF6: Mn 4+ with H2O2 by low-temperature coprecipitation. Compared with the original phosphor CSFM, the emission intensity of the original phosphor decreased to 13.6% after immersion in water for 168 h, while CSFM-P still maintained 74.0% of the initial value. Under excitation at 460 nm, the IQE, EQE, and absorption efficiency of CSFM-P phosphor was 98%, 85%, and 86.8%, whereas the corresponding values of the commercial KSFM were 92%, 67.08%, and 72.92%, respectively.  Yu et al. [53] constructed K2GeF6: Mn 4+ @K2GEF6 (T-KGF) with a uniform Mn 4+ -free surface layer. As shown in Figure 14a, the position and shape of the characteristic peaks did not change significantly. H2C2O4 solution treatment did not affect the luminescence characteristics of Mn 4+ ions in the K2GeF6 host, because the PL intensity, fluorescence lifetime, and thermal quenching behavior were nearly unchanged. Figure 14c,d show that the emission intensity of T-KGF kept 95.8% of the initial value while that of KGF was only 36.2% of the original value after 5 h of immersion in water. The disadvantage was that the internal Mn 4+ ions were difficult to reduce. Jiang et al. [54] passivated Rb2SnF6: Mn 4+ with oxalic acid to construct a protective layer of Mn 4+ ion passivation on the surface. The PL intensity of Rb2SnF6: Mn 4+ decreased to a certain extent. After soaking in boiling water for 3 h, the PL intensity of shell Rb2SnF6: Mn 4+ was 95% of that at room temperature (RT) after (c) normalized excitation emission spectra for K 2 GeF 6 : Mn 4+ and K 2 GeF 6 : Mn 4+ @MA; (d) normalized emission spectra of K 2 GeF 6 : Mn 4+ and K 2 GeF 6 : Mn 4+ @MA before and after immersion in water for 168 h. (Reprinted from Ref [49]. Copyright from Royal Society of Chemistry, 2018). for the development of luminescent materials that enhance the stability of the surface and the duration of the device. Liu et al. [52] synthesized CSFM-P by passivating Cs2SiF6: Mn 4+ with H2O2 by low-temperature coprecipitation. Compared with the original phosphor CSFM, the emission intensity of the original phosphor decreased to 13.6% after immersion in water for 168 h, while CSFM-P still maintained 74.0% of the initial value. Under excitation at 460 nm, the IQE, EQE, and absorption efficiency of CSFM-P phosphor was 98%, 85%, and 86.8%, whereas the corresponding values of the commercial KSFM were 92%, 67.08%, and 72.92%, respectively.  Yu et al. [53] constructed K2GeF6: Mn 4+ @K2GEF6 (T-KGF) with a uniform Mn 4+ -free surface layer. As shown in Figure 14a, the position and shape of the characteristic peaks did not change significantly. H2C2O4 solution treatment did not affect the luminescence characteristics of Mn 4+ ions in the K2GeF6 host, because the PL intensity, fluorescence lifetime, and thermal quenching behavior were nearly unchanged. Figure 14c,d show that the emission intensity of T-KGF kept 95.8% of the initial value while that of KGF was only 36.2% of the original value after 5 h of immersion in water. The disadvantage was that the internal Mn 4+ ions were difficult to reduce. Jiang et al. [54] passivated Rb2SnF6: Mn 4+ with oxalic acid to construct a protective layer of Mn 4+ ion passivation on the surface. The PL intensity of Rb2SnF6: Mn 4+ decreased to a certain extent. After soaking in boiling water for 3 h, the PL intensity of shell Rb2SnF6: Mn 4+ was 95% of that at room temperature (RT) after Yu et al. [53] constructed K 2 GeF 6 : Mn 4+ @K 2 GEF 6 (T-KGF) with a uniform Mn 4+ -free surface layer. As shown in Figure 14a, the position and shape of the characteristic peaks did not change significantly. H 2 C 2 O 4 solution treatment did not affect the luminescence characteristics of Mn 4+ ions in the K 2 GeF 6 host, because the PL intensity, fluorescence lifetime, and thermal quenching behavior were nearly unchanged. Figure 14c,d show that the emission intensity of T-KGF kept 95.8% of the initial value while that of KGF was only 36.2% of the original value after 5 h of immersion in water. The disadvantage was that the internal Mn 4+ ions were difficult to reduce. Jiang et al. [54] passivated Rb 2 SnF 6 : Mn 4+ with oxalic acid to construct a protective layer of Mn 4+ ion passivation on the surface. The PL intensity of Rb 2 SnF 6 : Mn 4+ decreased to a certain extent. After soaking in boiling water for 3 h, the PL intensity of shell Rb 2 SnF 6 : Mn 4+ was 95% of that at room temperature (RT) after soaking in boiling water for 3 h. Liu et al. [55] successfully restored the luminescence characteristics of hydrolyzed K 2 SiF 6 : Mn 4+ with oxalic acid and significantly improved the water resistance. Figure 15 showed a schematic diagram of the reverse strategy to restore luminescence and enhance water resistance. During the repair process, oxalic acid reacted with hydrolyzed dark brown material to reduce Mn 4+ ions to soluble low-state Mn ions, releasing K, Si, and F elements into the supernatant. The solubility of K 2 SiF 6 in solution was particularly low, and it would precipitate out and form a K 2 SiF 6 shell on the surface of K 2 SiF 6 : Mn 4+ particles. The emission intensity of K 2 SiF 6 : Mn 4+ phosphors repaired by oxalic could reestablish to 103.68% of the original K 2 SiF 6 : Mn 4+ red phosphors (O-KSFM). The recouped K 2 SiF 6 : Mn 4+ (R-KSFM), kept around 62.3% of the beginning relative emission, escalated after 5 h of submersion in deionized water. Moreover, the luminescence intensity of the degraded K 2 TiF 6 : Mn 4+ (D-KTFM) fluorophores could recover 162.59% of the original K 2 TiF 6 : Mn 4+ (O-KTFM). Zhong et al. [57] used surface passivation to synthesize LiNaSiF 6 : Mn 4+ -CA, Li + co-doping, and surface passivation improved the luminescence intensity of LiNaSiF 6 : Mn 4+ . As shown in Figure 16, after 3 and 6 h of immersion, the PL intensity of NSF: Mn 4+ decreased to 31.26 and 17.54%, LNSF: Mn 4+ decreased to 60.11% and 42.73%, and LNSF: Mn 4+ -CA maintained 96.84% and 92.33% of the initial value, respectively. Even after immersion in water for 30 days, the luminescence intensity of LNSF: Mn 4+ -CA was still 76.16% of the original. LNSF: Mn 4+ -CA had 154% and 118% fluorescence intensities at 120 • C and 150 • C compared with that at 30 • C, respectively. The NTQ effect could be explained by the large number of electron traps formed by LNSF: Mn 4+ -CA after surface passivation. As the temperature increased, the electrons obtained additional energy compensation from the electron traps generated by the matrix defects and then transferred the energy to Mn 4+ ions, inducing the NTQ effect. The luminescence intensity of CSNage 0.5 Sn 0.5 F 6 : Mn 4+ synthesized by Li et al. [56] decreased sharply after immersion in water for more than 5 min, which may be caused by the fact that the matrix CNGSF with double-central ions is more easily hydrolyzed due to larger crystal distortion. The luminescence intensity of the phosphors C1-CNGSFM and C2-CNGSFM modified by citric acid or oxalic acid slightly increased in the first 2 h, which may have been due to the diffusion of the phosphor in water. The reflection and refraction of the excitation light were increased, while the PL intensity remained unchanged for 16'h. When the quenched CNGSFM was stirred in the modifier for 30 min, the luminescence of CNGSFM could be restored to the initial brightness. Cai et al. [58] treated K 2 SiF 6 with FeCl 2 to form a protective shell on the surface and obtained T-KSF, whose PL intensity was significantly higher than KSF, effectively reducing the non-radiative transition probability of Mn 4+ ions. The Mn 4+ -free shell protected the T-KSF particles and effectively reduced the luminescence center ions in the particles that were in direct contact. After immersion for 320 min, the relative luminescence intensity of the KSF phosphors decreased sharply to only 63.4% of the initial intensity value, while the T-KSF samples still maintained 80.3% of the initial intensity.      (c) temperature-dependent relative PL intensities KGF and T-KGF; (d) relative PL intensities of immersion in water for different times of KGF and T-KGF. (Reprinted from Ref [53]. Copyright from Elsevier, 2018).   The K 2 SiF 6 : Mn 4+ @K 2 SiF 6 synthesized by Li et al. [59] kept the initial PL intensity of 88% after soaking in water for 300 min, while the strength of the uncoated sample decreased to 1%. Moreover, the PL intensities of K 2 SiF 6 : Mn 4+ @K 2 SiF 6 at 120 • C, 150 • C, 180 • C, and 210 • C were 176%, 198%, 214%, and 213% of the initial PL intensities at 30 • C, respectively. The K 2 SiF 6 coating had multiple effects on the luminescence properties of K 2 SiF 6 : Mn 4+ red phosphors. In addition to preventing the hydrolysis of Mn 4+ , the energy transfer to surface defects was also prevented. The probability of radiative transition increased with temperature more quickly than that of non-radiative transition. Huang et al. [28] synthesized K 2 TiF 6 : Mn 4+ @K 2 TiF 6 phosphors by the reverse ion strategy. As shown in Figure 17a,b, the yellow KTF: Mn 4+ sample quickly changed to brown after 5 min, while the yellow KTF: Mn 4+ @KTF remained yellow even after 5 h in water. The relative fluorescence intensity can be seen in Figure 17c. The relative PL intensities of KTF: Mn 4+ @KTF and KTF: Mn 4+ after aging 480 h in the HT and HH environment are shown in Figure 17f, which retained 89% and 45% of the initial value, respectively. the yellow KTF: Mn @KTF remained yellow even after 5 h in water. The cence intensity can be seen in Figure 17c. The relative PL intensities of and KTF: Mn 4+ after aging 480 h in the HT and HH environment are show which retained 89% and 45% of the initial value, respectively. Wan et al. [29] proposed a new idea of reduction-assisted surface (RSRC) to reconstruct the Mn 4+ -free shell of fluoride. Using α-hydroxy a process could improve the water-resistance of the fluoride. The PL inte RSRC fluoride containing LA, MA, CA, and AA was maintained at 90% 97% of the initial PL intensity after the soaking for 360 h in water, respec in Figure 18a. They also prepared KSFM-SP and KSFM-CE phosphors sivation and cation exchange, respectively. The PL intensities of KSFM ph with RSRC, SP, and CE maintained 97%, 100%, and 97% of the origina tively. The PL intensity of KSFM did not decrease by different treatmen in Figure 18b, after boiling in water for 20 min, the PL intensities of KFSM Wan et al. [29] proposed a new idea of reduction-assisted surface recrystallization (RSRC) to reconstruct the Mn 4+ -free shell of fluoride. Using α-hydroxy acid in the RSRC process could improve the water-resistance of the fluoride. The PL intensity of KSFM-RSRC fluoride containing LA, MA, CA, and AA was maintained at 90%, 96%, 94%, and 97% of the initial PL intensity after the soaking for 360 h in water, respectively, as shown in Figure 18a. They also prepared KSFM-SP and KSFM-CE phosphors by surface passivation and cation exchange, respectively. The PL intensities of KSFM phosphors treated with RSRC, SP, and CE maintained 97%, 100%, and 97% of the original KSFM, respectively. The PL intensity of KSFM did not decrease by different treatments. As presented in Figure 18b, after boiling in water for 20 min, the PL intensities of KFSM-RSRC, KFSM-SP, KFSM-CE, and the original KSFM were maintained at 96%, 65%, 31%, and 25% of that before boiling, respectively. In contrast, the core-shell fluoride constructed by the reduction-assisted surface recrystallization method had better water resistance. Compared with the commercial K 2 SiF 6 : Mn 4+ -CP, K 2 SiF 6 : Mn 4+ prepared by Zhou et al. [30] had better thermal stability and water-resistance. There was almost no quenching at 200 • C, the fluorescence intensity was 100.8% of the initial value, which was almost unchanged, while the luminescence intensity of KFSM-CP was just 87.1%. After soaking in deionized water for 12 h, the initial values were 97.6% and 80.8%, respectively. KSFM single crystal fluorophores formed an almost Mn 4+ -free surface and possessed excellent water resistance. LiNaSiF 6 : Mn 4+ with a core-shell structure prepared by Zhong et al. [50] had an NTQ effect, while the Na 2 SiF 6 : Mn 4+ did not. The PL intensity values at 150 • C were 148% and 43% of those at 30 • C, respectively. The thermal stability of Mn 4+ was even greater than that of K 2 SiF 6 : Mn 4+ -C (commercial K 2 SiF 6 : Mn 4+ ). As shown in Figure 19, after soaking in water, the fluorescence intensities of the LiNaSiF 6 : Mn 4+ , Na 2 SiF 6 : Mn 4+ , and K 2 SiF 6 : Mn 4+ -C samples maintained 87.16%, 66.16%, and 30.15% of their initial values, respectively. Co-doping Li + not only led to carrier transfer (CT), which induced the NTQ effect, but it also led to the surface formation, the CT produced the NTQ effect, and the surface prevented the hydrolysis of Mn 4+ on the surface of the sample. Jiang et al. [61] used anti-solvent to induce the epitaxial growth of deposited fluoride to prepare K 2 SiF 6 : Mn 4+ @K 2 SiF 6 , which retained 82% of the initial emission intensity after submerging in water for 4 h and retained 90% of the initial emission intensity after 10 days under HT and HH conditions, while K 2 SiF 6 : Mn 4+ only retained 38% after immersion for 4 h. The epitaxial development of the deposition caused by ethanol improved the water-resistance and luminescence of K 2 SiF 6 : Mn 4+ .
fluorescence intensity was 100.8% of the initial value, which was almost unchanged, while the luminescence intensity of KFSM-CP was just 87.1%. After soaking in deionized water for 12 h, the initial values were 97.6% and 80.8%, respectively. KSFM single crystal fluorophores formed an almost Mn 4+ -free surface and possessed excellent water resistance. LiNaSiF6: Mn 4+ with a core-shell structure prepared by Zhong et al. [50] had an NTQ effect, while the Na2SiF6: Mn 4+ did not. The PL intensity values at 150 °C were 148% and 43% of those at 30 °C, respectively. The thermal stability of Mn 4+ was even greater than that of K2SiF6: Mn 4+ -C (commercial K2SiF6: Mn 4+ ). As shown in Figure 19, after soaking in water, the fluorescence intensities of the LiNaSiF6: Mn 4+ , Na2SiF6: Mn 4+ , and K2SiF6: Mn 4+ -C samples maintained 87.16%, 66.16%, and 30.15% of their initial values, respectively. Co-doping Li + not only led to carrier transfer (CT), which induced the NTQ effect, but it also led to the surface formation, the CT produced the NTQ effect, and the surface prevented the hydrolysis of Mn 4+ on the surface of the sample. Jiang et al. [61] used anti-solvent to induce the epitaxial growth of deposited fluoride to prepare K2SiF6: Mn 4+ @K2SiF6, which retained 82% of the initial emission intensity after submerging in water for 4 h and retained 90% of the initial emission intensity after 10 days under HT and HH conditions, while K2SiF6: Mn 4+ only retained 38% after immersion for 4 h. The epitaxial development of the deposition caused by ethanol improved the water-resistance and luminescence of K2SiF6: Mn 4+ .  The core-shell structure of K2SiF6: Mn 4+ coated with an alkyl phosphate layer im- the luminescence intensity of KFSM-CP was just 87.1%. After soaking in deionized water for 12 h, the initial values were 97.6% and 80.8%, respectively. KSFM single crystal fluorophores formed an almost Mn 4+ -free surface and possessed excellent water resistance. LiNaSiF6: Mn 4+ with a core-shell structure prepared by Zhong et al. [50] had an NTQ effect, while the Na2SiF6: Mn 4+ did not. The PL intensity values at 150 °C were 148% and 43% of those at 30 °C, respectively. The thermal stability of Mn 4+ was even greater than that of K2SiF6: Mn 4+ -C (commercial K2SiF6: Mn 4+ ). As shown in Figure 19, after soaking in water, the fluorescence intensities of the LiNaSiF6: Mn 4+ , Na2SiF6: Mn 4+ , and K2SiF6: Mn 4+ -C samples maintained 87.16%, 66.16%, and 30.15% of their initial values, respectively. Co-doping Li + not only led to carrier transfer (CT), which induced the NTQ effect, but it also led to the surface formation, the CT produced the NTQ effect, and the surface prevented the hydrolysis of Mn 4+ on the surface of the sample. Jiang et al. [61] used anti-solvent to induce the epitaxial growth of deposited fluoride to prepare K2SiF6: Mn 4+ @K2SiF6, which retained 82% of the initial emission intensity after submerging in water for 4 h and retained 90% of the initial emission intensity after 10 days under HT and HH conditions, while K2SiF6: Mn 4+ only retained 38% after immersion for 4 h. The epitaxial development of the deposition caused by ethanol improved the water-resistance and luminescence of K2SiF6: Mn 4+ .  The core-shell structure of K2SiF6: Mn 4+ coated with an alkyl phosphate layer improved the water resistance of the phosphor. Subsequently, a series of Mn 4+ -doped fluo- The core-shell structure of K 2 SiF 6 : Mn 4+ coated with an alkyl phosphate layer improved the water resistance of the phosphor. Subsequently, a series of Mn 4+ -doped fluorides based on a core-shell structure was successfully synthesized. Table 1 shows the comparison of water resistance and thermal stability of A 2 MF 6 : Mn 4+ fluoride with and without the core-shell structure reported recently. It is clear from the table that the relative PL intensity of A 2 MF 6 : Mn 4+ fluoride with the core-shell structure was higher than that of Mn 4+ -doped fluorides without the shell, whether in the HT and HH condition or soaking in water for a long time, which showed that the water-resistance was significantly improved. Moreover, the thermal stability of some A 2 MF 6 : Mn 4+ fluorides was also improved. Thus, the core-shell structure could markedly increase the water-resistance and stability of A 2 MF 6 : Mn 4+ fluorides. As summarized above, the homogeneous shell layer included an organic shell, such as alkyl phosphates, octadecyl trimethoxy silanes, silane coupling agents, oleic acid, pyruvate, and the inorganic shell, such as SiO 2 , Al 2 O 3 , TiO 2 , CaF 2 , nano-carbon, GQD, SrF 2 , and the homogeneous shell layer coated by the same material as itself. These shell materials can improve the water resistance of A 2 MF 6 : Mn 4+ , but the luminescence intensities of A 2 MF 6 : Mn 4+ may be affected by the transparency and thickness of the shell layer material, especially the heterogeneous shell layer. The homogeneous shell is constructed by the coating construction method, which is the earliest method to construct the core-shell structure of A 2 MF 6 : Mn 4+ . The synthetic process of this method is mature. However, there are certain requirements for the synthesis conditions and equipment, which may involve high temperature and high-pressure conditions. At the same time, the existence of a shell interface may affect the optical properties of the phosphors, resulting in the shell falling off [44]. Therefore, when using this method, the shell layer material that is more matched with the lattice of the matrix can be selected to reduce the influence of the interface. The homogeneous shell layer is mainly constructed by the surface passivation method and the saturated crystallization method. The synthesis process of the surface passivation method is simple. Homogeneous shell layer phosphors can be formed by stirring with the reducing agent solution, such as DL-mandelic acid, oxalic acid, citric acid, L-tartaric acid, DL-malic acid, ascorbic acid, DL-lactic acid, H 3 PO 4 /H 2 O 2 , and FeCl 2 , for a period of time. This method requires the matrix itself to be slightly soluble or insoluble in water to obtain excellent water resistance. The saturated crystallization method can synthesize single crystals with high crystallinity and few defects. However, this method also requires the solubility of the matrix material. In addition, the saturated crystallization method needs more time to ensure the formation of a shell free of Mn 4+ ions. The homogeneous core-shell structure crystallized with the saturated solution has better water-resistance, which may be due to the presence of fewer Mn 4+ ions on the shell surface, and the core is coated completely, while the surface passivation method may leave more Mn 4+ ions on the surface due to an insufficient redox reaction. It is hoped that these summaries will be helpful for the design, synthesis, and optical performance optimization of A 2 MF 6 : Mn 4+ with improved water resistance in the future.

Application of A 2 MF 6 : Mn 4+ in WLED
A 2 MF 6 : Mn 4+ exhibits a series of narrow sharp emissions within the wavelength range of 600-650 nm, showing unique excitation and emission characteristics, making it a candidate for WLED, and especially K 2 SiF 6 : Mn 4+ has been commercialized. Huang et al. [50] packaged ternary (WR-K 2 SiF 6 : Mn 4+ -8) and binary (without red phosphor) WLEDs using blue LED chips. The correlation color temperature (CCT), color rendering index (Ra), and the luminous efficacy (LE) of binary WLED were 5661 K, and 69.8 and 168 lm/W, respectively, while the CCT, Ra, and LE values of ternary WLED were 5398 K, and 80.5 and 96 lm/W, respectively. Obviously, with the addition of A 2 MF 6 : Mn 4+ red phosphor, ternary WLED did have lower CCT and higher color rendering index than binary WLED. However, the LE of ternary WLED greatly decreased. Therefore, more attention should be paid to the trade-offs between Ra, R9, CCT, and LE of WLED according to the field of practical application. Table 2 lists the basic optoelectronic parameters of recently reported WLED devices based on A 2 MF 6 : Mn 4+ red phosphors with or without the core-shell structure. As listed in it, the color rendering index (such as Ra and R9) and LE of A 2 MF 6 : Mn 4+ were not distinctive, regardless of whether it possessed the shell layer, indicating that the core-shell structure had little effect on the luminescent properties of the WLED. However, it had a significant effect on the service life of the device. Zhou et al. [51] packaged WLEDs by using YAG&K 2 TiF 6 : Mn 4+ (LED-1), and YAG&P-K 2 TiF 6 : Mn 4+ (LED-2). When LED-1 was aged in HT and HH conditions for 60 days, its LE began to decrease, and it finally failed after aging for about 77 days. The LE of LED-2 also gradually decreased during the aging process, but it still remained at 39.5% of the initial LE after being aged in HT and HH conditions for 100 days. Obviously, the A 2 MF 6 : Mn 4+ deteriorates much less than the shell-free fluoride under HT and HH conditions. Thus, the improved water resistance of A 2 MF 6 : Mn 4+ by constructing a core-shell can significantly prolong the lifetime of WLED devices, which is beneficial to environmental protection and resource saving.

Conclusions
The market demand for WLEDs of high quality is still growing rapidly. Compared with rare earth doped red phosphors, A 2 MF 6 : Mn 4+ red phosphors have strong broadband excitation at nearly 460 nm and stable narrow-band red emission at about 630 nm, as well as the advantages of high thermal stability, low cost, and room temperature synthesis, which are beneficial to the development of WLED. Many research results summarized in this review show that constructing core-shell structures is an effective approach to improve the moisture resistance of A 2 MF 6 : Mn 4+ red phosphors and avoid the degradation of device performance caused by the deliquescence of the phosphors. There are still some challenges to be further studied to meet the commercial application requirements. Firstly, high concentrations of HF can decompose KMnO 4 to produce manganese (VI), provide F − ions in an aqueous solution to obtain [MnF 6 ] 2− complexes instead of MnO 2 , and facilitate the incorporation of Mn 4+ into the host lattice, thereby enhancing the luminescent efficiency of phosphors. However, as a corrosive and volatile acid, HF introduces hazards in terms of safety and the environment. Although low toxicity H 3 PO 4 /KHF 2 have been used to replace highly toxic HF and good luminescent efficiency has been achieved, Mn 4+ is prone to be reduced to Mn 2+ under high temperature and high humidity conditions, which decreases the luminescent properties. Hence, it is urgent to develop new green synthesis routes with low HF or without HF without sacrificing the luminescent performance and water resistance. Secondly, there is no uniform standard for measuring the water resistance of A 2 MF 6 : Mn 4+ red phosphors. Different dosages and different test methods always show different results of water resistance. Therefore, relevant test standards should be developed, which is very important in terms of commercial applications. Thirdly, some optoelectronic parameters of WLED can be improved by adding red powder. However, some other parameters (such as the luminous efficacy) may be decreased at the same time. Thus, from the device perspective, more attention needs to be paid to the trade-offs between Ra, R9, CCT, and light efficiency, as well as the costs and lifespans of WLEDs. Moreover, though many Mn 4+ -doped fluoride red phosphors have been discovered in the past few decades, it is still valuable to explore new Mn 4+ -doped fluoride red phosphors with more excellent comprehensive performances to meet the demand of commercialization.