Synthesis of Nitride Zeolites in a Hot Isostatic Press

Abstract

The recently introduced nitridophosphate synthesis in a hot isostatic press (HIP) enabled simple access to large‐scale product quantities starting from exclusively commercially available starting materials. Herein, we show that this method is suitable for the synthesis of highly condensed functional nitridophosphates, as well. Hence, the syntheses of the nitridophosphate zeolites Ba₃P₅N₁₀ X (X=Cl, Br) are presented as proof of concept for this innovative access. Furthermore, samples of unprecedented Sr₃P₅N₁₀ X (X=Cl, Br) were prepared and characterized to demonstrate the advantages of this synthetic approach over commonly used methods. Luminescence investigations on Eu²⁺‐doped samples of AE ₃P₅N₁₀ X (AE=Sr, Ba; X=Cl, Br) were carried out and characteristics of observed emission bands are discussed.

Nitridophosphate zeolites AE ₃P₅N₁₀ X (AE=Sr, Ba; X=Cl, Br) were synthesized under medium‐pressure conditions (150 MPa N₂, 1200 °C), extending the range of the degree of condensation for nitridophosphates accessible by hot isostatic pressing from κ=1/3 to 1/2. AE ₃P₅N₁₀ X:Eu²⁺ (AE=Sr, Ba; X=Cl, Br) reveals natural‐white to deep‐red emission.

The versatile compound class of nitridophosphates has been researched in detail for more than 30 years. ^[1] The structural diversity of nitridophosphates can be derived from their close relationship to oxosilicates, as the element combination Si/O is isoelectronic with P/N, and similar with oxosilicates which are built up from SiO₄ tetrahedra, the PN₄ tetrahedron is the fundamental building unit in nitridophosphates. However, nitridophosphates can even feature edge‐sharing tetrahedra due to the higher covalence (and lower polarity) of P−N bonds compared to Si−O in oxosilicates.[ ¹ , ² ] Additionally, the higher valence of nitrogen allows for higher degrees of condensation (i.e. atomic ratio κ of tetrahedra centers to tetrahedra vertices).[ ¹ , ³ ] But despite numerous investigations, nitridophosphate synthesis has ever been challenging, as the most common starting material phosphorus(V) nitride (P₃N₅) is prone to thermal decomposition above 850 °C. ^[4] Therefore, high‐pressure high‐temperature methods (e.g. multianvil technique) have been employed for their synthesis, since elevated nitrogen pressure suppresses the thermally induced elimination of N₂ from P₃N₅. ^[1] This high‐pressure strategy led to a large number of nitridophosphates with various incorporated electropositive elements.[ ⁵ , ⁶ , ⁷ , ⁸ , ⁹ ] However, sample quantities have intrinsically been limited by high‐pressure techniques. Investigations on the associated optical and physical properties of nitridophosphates, such as ion conductivity or luminescence, have however quickly revealed the potential of this functional materials class.[ ⁵ , ¹⁰ , ¹¹ , ¹² ] In particular, the intriguing luminescence properties of nitridophosphates like Ba₃P₅N₁₀ X:Eu²⁺ (X=Cl, Br, I) and AEP₈N₁₄:Eu²⁺ (AE=Ca, Sr, Ba) clearly underline the quest for a synthetic approach that can be transformed to a large batch scale.[ ³ , ¹¹ , ¹³ ] Here, ambient and medium‐pressure methods that have been used for the early syntheses of phosphorus nitrides and nitridophosphates come to mind.[ ¹⁴ , ¹⁵ , ¹⁶ , ¹⁷ , ¹⁸ , ¹⁹ , ²⁰ , ²¹ , ²² , ²³ ] However, only a limited number of such nitride compounds could be synthesized applying these techniques, as gentle conditions or tailored starting materials had to be used. This limitation has changed only recently, as improved high‐temperature ammonothermal techniques and the use of red phosphorus as starting material enabled synthesis of diverse nitridophosphates.[ ²⁴ , ²⁵ , ²⁶ , ²⁷ ] However, ammonothermal syntheses of nitridophosphates may not be performed industrially due to its demanding handling and the stated goal for a large‐scale access remained. Recently, we have reported on the successful synthesis of Ca₂PN₃ in a hot isostatic press (HIP) under nitrogen pressure, which thus appears as a promising innovative approach for nitridophosphate synthesis. ^[28] HIPs do not only provide large sample quantities, but also shorten reaction times and facilitate crystal growth under comparatively gentle reaction conditions. Moreover, we have demonstrated that red phosphorus can serve as a starting material in HIPs as well, simplifying nitridophosphate syntheses even more. ^[28]

Within the scope of this work, this approach is further developed to grant access to highly condensed nitridophosphates with a degree of condensation κ=1/2. For this purpose, Ba₃P₅N₁₀ X (X=Cl, Br) were chosen as model compounds, because of their above‐mentioned luminescence properties.[ ¹¹ , ¹³ ] Furthermore, hitherto unknown compounds Sr₃P₅N₁₀ X (X=Cl, Br) were synthesized and investigated with regard to luminescence properties.

All alkaline earth nitridophosphate zeolites AE ₃P₅N₁₀ X (AE=Ba, Sr; X=Cl, Br) were synthesized under nitrogen atmosphere in a HIP applying hot isostatic conditions (150 MPa N₂, 1000 °C, Figures S1 and S2). In an initial attempt, the title compounds were synthesized from specially prepared P₃N₅ and the respective alkaline earth azides and halides following Equation 1 (Table S1).

$${\displaystyle 15AE\left(\mathrm{N}{}_3\right){}_2+3AEX{}_2+10\mathrm{P}{}_3\mathrm{N}{}_5\to 6AE{}_3\mathrm{P}{}_5\mathrm{N}{}_{10}X+40\mathrm{N}{}_2}$$ (1)

Subsequently, all variants of AE ₃P₅N₁₀ X (AE=Ba, Sr; X=Cl, Br) were prepared using commercial red phosphorus (P_red), replacing P₃N₅ as phosphorus source [Eq. 2, Table S1]. The reaction equation is balanced with P₄, since P_red is considered to transform into molecular/gaseous phosphorus at reaction conditions of 150 MPa and 1000 °C. ^[28]

$${\displaystyle 10AE\left(\mathrm{N}{}_3\right){}_2+2AEX{}_2+5\mathrm{P}{}_4\to 4AE{}_3\mathrm{P}{}_5\mathrm{N}{}_{10}X+10\mathrm{N}{}_2}$$ (2)

Subsequent oxidation into P^V is coupled to the disproportionation of azide ions into network‐forming nitride anions and elemental nitrogen. Thereby, it is conceivable that activated P₄ is gradually oxidized and present as P^III in form of molecular “PN” after initial reaction with N₂. In a second oxidization step, halide‐containing intermediates, such as (P^VNCl₂)₃, might be formed. (PNCl₂)₃ in turn serves as starting material for laboratory synthesis of P₃N₅ and may facilitate the in situ formation of the latter, leading to the desired zeolites AE ₃P₅N₁₀ X. An experimental evidence for this hypothesis by in situ measurements is still pending.

To investigate luminescence properties of AE ₃P₅N₁₀ X, Eu²⁺‐doped samples were prepared by adding 3 mol % EuCl₂ (with regard to alkaline earth ions) to the mixture of starting materials.

The undoped samples are yielded as colorless cube‐like crystals, while sinter cakes of Eu²⁺‐doped products exhibit yellow (Ba compounds) to orange body colors (Sr compounds). All products have been washed with de‐ionized water and are not sensitive towards air or moisture. SEM imaging of the products reveals the microcrystalline character of Ba compounds (edge length up to ≈3 μm, Figure 1). Sr compounds form slightly larger crystals with edge lengths up to 15–20 μm (Figure 1). Phase purity of Ba₃P₅N₁₀ X (X=Cl, Br) was confirmed by Rietveld refinements, using literature known structures as starting models.[ ¹¹ , ¹³ ] Detailed information on the refinements is provided in the Supporting Information (Figures S3 and S4, Table S2).

Figure 1.

SEM images of obtained AE ₃P₅N₁₀ X samples (AE=Sr, Ba; X=Cl, Br) containing single crystals.

The structures of Sr₃P₅N₁₀ X (X=Cl, Br) were elucidated from a single‐crystal XRD measurement of Sr₃P₅N₁₀Cl. Sr₃P₅N₁₀Br was refined using the Rietveld method and using the structure of Sr₃P₅N₁₀Cl as starting model (Figure S5, Table S7). Both compounds crystallize homeotypically to the Ba compounds in the JOZ zeolite structure type (Pnma; Z=8; Sr₃P₅N₁₀Cl: a=12.240(3); b=12.953(3); c=13.427(3) Å; Sr₃P₅N₁₀Br: a=12.297(1); b=12.990(1); c=13.458(1) Å.[ ²⁹ , ³⁰ , ³¹ ] The crystal structure of Sr₃P₅N₁₀ X is shown exemplarily for X=Cl in Figure 2. The as‐refined crystallographic data is summarized in the Supporting Information (Tables S3–S5) and a more detailed description of the crystal structure is provided in literature.[ ¹¹ , ¹³ ] Phase purity of Sr₃P₅N₁₀Cl has been confirmed by Rietveld refinement (Figure S6, Table S7). In contrast to the Ba compounds, Sr₃P₅N₁₀ X (X=Cl, Br) show split positions of the Sr5 site. Although the split position is not present in the Ba compounds the corresponding Ba5 site in Ba₃P₅N₁₀ X, however, shows the most elongated ellipsoids and Ba−X distances (Figure S7).[ ¹¹ , ¹³ ] This observation can be explained by the significantly smaller space filling of Sr²⁺ compared to Ba²⁺ and the associated increasing displacement of ions.[ ³² , ³³ ] Moreover, owing to the larger radius of Br⁻, the distance between the split position is reduced in Sr₃P₅N₁₀Br (Figure S8).

Figure 2.

Projection of the crystal structure of Sr₃P₅N₁₀Cl along [100]: PN₄ tetrahedra blue, Cl atoms pink, Sr1–Sr4 atoms gray, split position of Sr5 black. All atoms are displayed with 95 % probability. ^[29]

These observations may explain the reason for Sr₃P₅N₁₀ X (X=Cl, Br) being exclusively accessible at medium‐pressure conditions, as the pressure has to be well balanced. Although increased pressure is necessary to prevent thermal decomposition, it must not be chosen too high for Sr₃P₅N₁₀ X syntheses to prevent collapsing of the large cages, considering the reduced space filling of the Sr²⁺ ions. In line with this hypothesis, the Ba compounds are rather difficult to access at ambient pressure, as well, and phase pure samples have only been obtained at 1–5 GPa, thus far.[ ¹¹ , ¹³ ]

The interatomic P–N distances and N‐P‐N angles of Sr₃P₅N₁₀ X (X=Cl, Br) are in very good agreement with values reported for other nitridophosphates (Table S9).[ ³ , ¹¹ , ¹² , ¹³ ] In line with previous refinements, the observed AE–N/X distances differ significantly depending on the coordination number of the alkaline earth metal ion (Figure S8). A detailed list of the interatomic AE–N/X distances is provided in literature for Ba₃P₅N₁₀ X and in the Supporting Information for Sr₃P₅N₁₀ X (Table S8).[ ¹¹ , ¹³ ]

The chemical compositions of the title compounds were confirmed by energy dispersive X‐ray spectroscopy (EDX), with details provided in the Supporting Information (Tables S10).

Eu²⁺‐doped samples of Ba₃P₅N₁₀ X (X=Cl, Br) have already been discussed as promising phosphor materials, but any industrial application had been ruled out, owing to limited sample volumes.[ ¹¹ , ¹³ ] With the innovative approach that is presented herein, large‐volume samples become accessible. Additionally, hitherto unknown Sr₃P₅N₁₀ X:Eu²⁺ (X=Cl, Br) is introduced as new luminescent material.

Excitation with UV to blue light (λ _exc=420 nm) induces natural‐white (Ba₃P₅N₁₀Br:Eu²⁺), orange (Ba₃P₅N₁₀Cl:Eu²⁺), and deep‐red emission (Sr₃P₅N₁₀ X:Eu²⁺), which features two emission maxima for each compound (Figures 3 and S9–13). The two emission maxima likely correspond to the different Eu²⁺ coordination spheres that are provided by the host lattice. The sites AE1, AE4, and AE5 are coordinated by eight N and two X ions (CN=10) and thus, feature rather elongated AE–X and AE–N distances, causing a weak crystal field (Figure S8, Table S8). Therefore, the higher energetic emission bands can be assigned to Eu²⁺ ions occupying these sites according to the parity‐allowed transition 4f⁷ → 4f⁶5d¹. Consequently, the second emission band is assigned to Eu²⁺ ions located on AE sites with lower coordination number (AE2, AE3: CN=8, six N and two X ions), which feature shorter AE–X and AE–N distances and thus a strong crystal field (Figure S8, Table S8). A detailed illustration of the emission and excitation spectra for each element combination AE‐X is provided in the Supporting Information (Figures S10–S13).

Figure 3.

Measured emission spectra of AE ₃P₅N₁₀ X:Eu²⁺ (AE=Sr, Ba; X=Cl, Br) at room temperature with a nominal doping level of 3 mol % Eu²⁺ referred to AE ²⁺.

When comparing the emission spectra, two trends are particularly striking. First, emission bands of AE ₃P₅N₁₀Cl and AE ₃P₅N₁₀Br are shifted comparing AE=Ba and AE=Sr. This effect is attributable to the different AE radii and the associated AE–N/X and Eu²⁺–N/X distances, leading to an increased crystal field splitting for Eu²⁺ on the Sr site and a red‐shift in emission. The second trend describes the influence of the halide ions on the position of the emission maxima. While the higher energetic bands are shifted red with increasing size of X, lower energetic bands are shifted blue. A detailed discussion of these observations is provided in the Supporting Information.

Furthermore, measurements of the internal quantum efficiency have been carried out for non‐optimized powder samples of the title compounds. The IQE of Sr₃P₅N₁₀ X:Eu²⁺ was determined to 29 % (X=Cl) and 32 % (X=Br). Measurements on Ba₃P₅N₁₀ X:Eu²⁺ yield quantum efficiency of 12 % each, which shows potential for improvement of the investigated samples, as values of >60 % have been reported for Ba₃P₅N₁₀ X:Eu²⁺ (X=Cl, Br; 2 mol % Eu²⁺ referred to Ba²⁺), previously.[ ¹¹ , ¹³ ] The luminescence characteristics of the title compounds are summarized in Table S11.

Recapping, we have extended the possibilities of nitridophosphate syntheses by hot isostatic pressing. Prior to this work, this approach has been limited to preparation of lowly condensed Ca₂PN₃:Eu²⁺. In this work, we have succeeded in synthesizing highly condensed nitridophosphates AE ₃P₅N₁₀ X (AE=Sr, Ba; X=Cl, Br), increasing the maximum degree of condensation reachable for nitridophosphates by HIP synthesis to 1/2. It is particularly noteworthy that the used conditions allow for the synthesis of Sr₃P₅N₁₀ X, which was not accessible by conventional methods, so far. Presumably, in this case only medium‐pressure methods, such as hot isostatic pressing, enable pressure balancing in a way that the synthesis is possible at all. The pressure should be high enough to prevent thermal decomposition, but not too high, otherwise large cages collapse due to lower space filling of Sr²⁺.

The presented results suggest that the pressure range generally applied under HP/HT conditions may exceed the actually required synthesis pressure for nitridophosphates by far. Since the minimum pressure to suppress thermal decomposition is not known, numerous published and some novel nitridophosphates are already accessible under medium‐pressure conditions. Consequently, future investigations may focus on synthesis of Ca₃P₅N₁₀ X, other P/N based zeolites (e.g. NPO or NPT), and even higher condensed nitridophosphates (e.g. AEP₈N₁₄). In particular, the successful activation of red phosphorus as starting material could contribute to a considerable acceleration of these investigations. It should also be examined whether other starting materials, such as nitrides, can be replaced by precursors, like metals or alloys. The fact that N₂ can serve as necessary redox partner has already been shown in the synthesis of Ca₂PN₃. ^[28] These significant simplifications in synthesis, the wide range of achievable degrees of condensation and the large sample volumes may allow that nitridophosphates could find their way into industrial application as phosphor materials.

Conflict of interest

The authors declare no conflict of interest.

Supporting information

As a service to our authors and readers, this journal provides supporting information supplied by the authors. Such materials are peer reviewed and may be re‐organized for online delivery, but are not copy‐edited or typeset. Technical support issues arising from supporting information (other than missing files) should be addressed to the authors.

Supplementary

S. Wendl, M. Zipkat, P. Strobel, P. J. Schmidt, W. Schnick, Angew. Chem. Int. Ed. 2021, 60, 4470.