ZnH₂P₄N₈: Case Study on Topochemical Imidonitridophosphate High‐Pressure Synthesis

Abstract

Nitridophosphates are subject of current research, as they have a broad spectrum of properties and potential applications, such as ion conductors or luminescent materials. Yet, the subclass of imidonitridophosphates has been studied less extensively. The primary reason is that the controlled N−H functionalization of nitridophosphates is not straight forward, making targeted synthesis more challenging. Inspired by the high‐pressure (HP) post‐synthetic modification of nitridophosphates, we present the topochemical HP deprotonation of phosphorus nitride imides using the high‐pressure polymorph β‐PN(NH) as an example. Additional incorporation of Zn²⁺ results in the first quaternary transition metal imidonitridophosphate ZnH₂P₄N₈. The crystal structure was elucidated by single‐crystal X‐ray diffraction (SCXRD), energy‐dispersive X‐ray spectroscopy (EDX), powder X‐ray diffraction (PXRD) and solid‐state magic angle spinning nuclear magnetic resonance spectroscopy (MAS NMR). In addition, the presence of H as part of an imide group was confirmed by IR spectroscopy. The potential of this defunctionalization approach for controlling the N−H content is demonstrated by the preparation of partially deprotonated intermediates Zn _x H_4−2x P₄N₈ (x≈0.5, 0.85). This topochemical high‐pressure reaction represents a promising way to prepare, control and manipulate new imide‐based materials without altering their overall anionic framework.

ZnH₂P₄N₈ was synthesized under elevated pressures and characterized by SCXRD, PXRD, SEM‐EDX and NMR. The novel high‐pressure deprotonation of phosphorus nitride imide polymorphs is a promising strategy for the specific preparation of future imidonitridophosphates and offers a significant increase in synthetic control. NMR studies on partially defunctionalized intermediates give insights on the local structure and the assumed reaction mechanism.

Introduction

Silicate analogous materials have become increasingly important in materials chemistry. Nitride compounds, along with oxides and oxonitrides, are already used in high‐performance materials for applications such as high‐temperature, thermal and ion conduction, and semiconductor materials.[ ¹ , ² , ³ , ⁴ , ⁵ ] The classes of tetrahedron‐based structures also include the (imido)‐nitridophosphates. Their structural similarities to silicates are frequently discussed due to the isoelectronic element combinations of Si/O and P/N. ^[5] The ternary PN(NH) illustrates the similarities to silicates directly. As a consequence of the isoelectronicity to SiO₂, this nitride imide crystallizes in two modifications, structurally related to high and low cristobalite.[ ⁶ , ⁷ ] In contrast to ternary P/N/H compounds, quaternary ones have been sparsely studied. The major reason for this is the low thermal stability of the nitride precursors required for synthesis. P₃N₅ is one of the most common starting materials for the preparation of (imido)nitridophosphates. It has a decomposition temperature of 850 °C, α‐PN(NH) even only 570 °C.[ ⁷ , ⁸ ] According to Le Chatelier, this restriction can be circumvented by application of elevated pressure.

For example the lowly condensed representatives Na₁₀[P₄(NH)₆N₄](NH₂)₆(NH₃)_0.5 and Rb₈[P₄N₆(NH)₄](NH₂)₂ were synthesized by Jacobs et al. under ammonothermal conditions in autoclaves.[ ⁹ , ¹⁰ ] The synthesis of highly condensed networks, such as the imidonitridophosphate BaP₆N₁₀NH, requires not only high pressure but also high temperatures to facilitate bond cleavage and reformation. ^[11] Targeting nitridophosphates a number of synthesis routes under medium‐/high‐pressure have been established in recent years. ^[5]

However, these are only restrictedly transferable to imidonitridophosphates. For example, the mineralizer‐assisted route can be used for certain imidonitridophosphates, as illustrated by SrP₃N₅NH. The latter was synthesized under high‐pressure and high‐temperature (HP/HT) conditions using stoichiometric amounts of Sr(N₃)₂, P₃N₅, and NH₄Cl. ^[12] However, the consistent incorporation of N−H groups into the target compound is not observed, as shown by the synthesis of many pure nitridophosphates crystallized by the addition of ammonium halides.[ ¹³ , ¹⁴ , ¹⁵ , ¹⁶ ] More efficient could be the direct addition of phosphorus nitride imides (e. g. amorphous or crystalline α‐PN(NH)) to the reaction mixture, as exemplified in the syntheses of MH₄P₆N₁₂ (M=Mg, Ca). ^[17] Here, the tendency is evident that providing pre‐organized anionic [PN₃(NH)]⁶⁻ motifs leads to their integration in the desired product. The aforementioned methods are employed to attain a specific molecular formula with a desired degree of condensation, with the stoichiometry selected accordingly. However, the directed influence on structural details is typically limited. Nevertheless, this control is essential for the development of new compounds with desired properties, as it ensures that the outcome is not left to chance. In contrast to the direct synthesis routes mentioned above, post‐synthetic modification is a way of preserving the anionic main motif to a certain extent. Wendl et al. showed that highly condensed 3D frameworks in particular allow for topotactic control over the cation environments. ^[18] Starting from a nitridophosphate‐based precursor, ion exchange or metathesis reaction has been carried out using metal halides. In this process, cations with larger ionic radii are typically replaced by smaller ones under high/medium pressure conditions. More stable by‐products drive this process. This reaction strategy uses the stronger interaction between smaller cations and the anionic backbone, resulting in their incorporation into a more energetically favorable product. That such exchanges are not limited to this reaction direction is shown by the deprotonation of the nitride imide sodalite Zn₅H₄[P₁₂N₂₄]Cl₂ with ZnCl₂ resulting in Zn₇[P₁₂N₂₄]Cl₂.[ ¹⁹ , ²⁰ ] Transferred to the high‐pressure conditions required for the stabilization of many imidonitridophosphates, this reaction provides a possibility for selective deprotonation of existing H/P/N compounds. Due to the 3D structure with channels in which amide groups are localized, the ternary high‐pressure polymorph β‐PN(NH) was considered as a model compound for this purpose. In this contribution, we report on the synthesis of the first quaternary transition metal imidonitridophosphate, namely ZnH₂P₄N₈, by controlled de‐functionalization of a phosphorus nitride imide high‐pressure polymorph, retaining its anionic backbone. The structure was elucidated by single‐crystal X‐ray diffraction data (SCXRD), and confirmed using MAS NMR, EDX, FTIR and powder X‐ray diffraction data (PXRD). Furthermore, a series of partially deprotonated representatives with the general stoichiometric formula Zn _x H_4−2x P₄N₈ (x≈0, 0.5, 0.85, 1) were prepared and analyzed by PXRD, MAS NMR and EDX spectroscopy. Additionally, a direct synthesis approach of the stoichiometric ZnH₂P₄N₈ is presented.

Results and Discussion

Synthesis

Zn _x H_4−2x P₄N₈ (x≈0.5, 0.85, 1) was synthesized according to two different routes (Equations (1) and (2)) after empirical optimization.

Both, topochemical deprotonation and nitride/azide reaction lead to the stoichiometric imidonitridophosphate ZnH₂P₄N₈. According to Equation (1) an excess (10 mol‐%) of the respective halide stoichiometry is necessary to prepare also partially de‐functionalized representatives.

Topochemical deprotonation approach starting from ternary β‐PN(NH):

(1)

Nitride/azide approach:

(2)

The generated HCl favors the formation and growth of single crystals of ZnH₂P₄N₈ with a length of up to 20 μm (Figure 1). However, according to Equations (1) and (2) different amounts of the title compound were obtained together with Zn₈P₁₂N₂₄O₂, β‐HP₄N₇ and unknown side phases. Due to the better crystal quality after synthesis according to Equation (2), single‐crystals of this approach were used for structure analysis of ZnH₂P₄N₈. Detailed information on both high‐pressure approaches can be found in the Experimental Section. Since the title compound is stable to moisture and air, potentially occurring water‐soluble by‐products such as the starting material ZnCl₂, were removed by washing with de‐ionized water. Zn _x H_4−2x P₄N₈ (x≈0.5, 0.85, 1) were isolated as dark grayish crystalline solids. The respective elemental compositions were analyzed via Rietveld refinements on PXRD data (with free refinement of the Zn occupations and subsequent fixation within the errors) and EDX spectroscopy on individual crystals (Tables S1–S4, Figure S1). As the latter method is very surface‐sensitive, the low oxygen contents detected can be attributed to surface hydrolysis caused by the washing step. However, the results obtained are within the usual error range of the analysis method. Synthesis experiments based on Equation (1) but using the low‐pressure polymorph α‐PN(NH) instead of β‐PN(NH) resulted in an unidentifiable product with a high amorphous content.

Figure 1.

SEM images of ZnH₂P₄N₈ crystals with a maximum diameter of ca. 10–20 μm; left: obtained from nitride/azide direct synthesis, right: obtained from topochemical deprotonation reaction.

Structure Determination

Structure elucidation was performed by single‐crystal X‐ray diffraction (SCXRD). Deposition Numbers 2300728 and 2379986 contain the supplementary crystallographic data for this paper. These data are provided free of charge by the joint Cambridge Crystallographic Data Centre and Fachinformationszentrum Karlsruhe Access Structures service. ZnH₂P₄N₈ was solved and refined in the monoclinic space group P2/c (no.13) with Z=2, with a unit cell dimension a=7.9230(4), b=4.8990(2), c=7.9708(3) Å, and β=107.725°, comparable to the initial ternary imide β‐PN(NH) (a=7.89365(5), b=4.81867(2), c=8.11718(4) Å, β=108.0548°). ^[6] Refined crystallographic data are given in Table 1, more detailed information on the single‐crystal refinement is provided in the Supporting Information (Table S5–S8).

Table 1.

Crystallographic data from single‐crystal refinement of ZnH₂P₄N₈.

Formula	ZnH2P4N8
Crystal system Space group Lattice parameters [Å, °] Cell volume [Å3] Formula units [cell] Density [g cm−3] μ [mm−1] Diffractometer Radiation Temperature [K] F(000) Θ range [°] Total no. of reflections Independent reflections (>2σ) Refined parameters Restraints R int; R σ R1 (all data); R1 (F 2>2σ(F 2)) wR2 (all data); wR2 (F 2>2σ(F 2)) Goodness of fit Δρ max; Δρ min [e Å−3]	monoclinic P2/c (No.13) a=7.9230(4) b=4.8990(2) c=7.9708(3) β=107.725(2) 294.70(2) 2 3.419 5.197 Bruker D8 Venture Mo‐K α (λ=0.71073 Å) 299(2) 296 4.16–32.02 4131 850 69 1 0.0296; 0.0327 0.0400; 0.0307 0.0784; 0.0750 1.064 0.615; −0.575

^[a] Estimated standard deviations are given in parentheses.

During structure refinement, valence electrons of N−H bonds could be determined from difference Fourier maps. The corresponding interatomic N−H distance was restrained for common distances in imidonitridophosphates (0.90(2) Å) and the H atoms were refined isotropically, while all non‐hydrogen atoms were refined anisotropically without any further restraints.[ ¹² , ¹⁷ ] The single‐crystal structure of ZnH₂P₄N₈ was confirmed by Rietveld refinement of a representative powder X‐ray diffraction pattern (Figure 2). The chemical composition was verified by energy‐dispersive X‐ray spectroscopy (EDX) on individual crystals. No other elements than Zn, P, N and O were detected, having an atomic ratio of Zn : P : N≈1 : 4 : 9. The nitrogen content is slightly lower than the expected value, but can be explained by shadowing due to the unfavorable orientation of the crystals to the EDX detector during the analysis. ^[21]

Figure 2.

Rietveld refinement of ZnH₂P₄N₈; observed (black) and simulated (red) powder X‐ray diffraction pattern with difference profile (gray). Vertical bars indicate the positions of the Bragg reflections of the desired product (blue) and oxonitridic by‐product (green).

In contrast to Zn, lighter elements such as N are affected more strongly, therefore their content is often underestimated. However, the average atomic ratio of Zn : P agrees with the values corresponding to the empirical sum formula. The results of the performed CHNS analysis are consistent with the expected nitrogen content (Table S9). However, the hydrogen content is also elevated and significantly higher than the theoretical values. This can be attributed to the already mentioned surface hydrolysis, as well as small amounts of the sodalite‐type Zn_7−x H_2x [P₁₂N₂₄]Cl₂ which is homeotypic with Zn₈P₁₂N₂₄O₂.[ ¹⁹ , ²² ] FTIR spectra of the sample (Figure S2) show an N−H band in the range 2700–3400 cm⁻¹ and confirm imide functionality in the sample. Additional absorption bands in the fingerprint region are due to P−N−P framework vibration modes.

The analysis of the phase width of partially defunctionalized representatives of the series Zn _x H_4−2x P₄N₈ (x≈0.5, 0.85, 1) was carried out analogously by performing PXRD, EDX and MAS NMR. Rietveld refinements on PXRD data were based on the structural model obtained from SCXRD data as described above (see Supporting Information).

Structure Description

As expected for a product obtained by a topochemical reaction, ZnH₂P₄N₈ shows the same network topology of the anionic P/N framework as the precursor β‐PN(NH), which crystallizes in a distorted α‐crystobalite type variant. It can be classified as a highly condensed transition metal imidonitridophosphate with a degree of condensation of κ=n(P):n(N)=1/2, consisting of a three‐dimensional network of all‐side vertex‐sharing PN₄ tetrahedra. According to the silicate nomenclature of Liebau the PN₄ tetrahedra form three different sechser‐ring types, which assume a distorted armchair conformation. ^[23] Two of these sechser‐ring types can be seen along [010] (see Figure 3). A stacking sequence A B A is created by alternating between larger and smaller channels. The larger channels contain zinc atoms, while the smaller channels contain only hydrogen atoms, which are covalently bound to one quarter of the N atoms. In the larger channels, zinc occupies a split position (half occupation of Wyckoff position 4 g) without any indication of zinc alignment (see Figure 4). The pore size of the smaller channels of β‐PN(NH) is apparently insufficient for Zn²⁺ incorporation, which supports the assumption that the P/N framework is maintained during the reaction.

Figure 3.

Schematic reaction for β‐PN(NH) with ZnCl₂. Zn atoms in yellow, protons in rose, PN₄ tetrahedra in green.

Figure 4.

Crystal structure of ZnH₂P₄N₈ along [010]. The half occupation of Wyckoff sites 4 g by zinc was highlighted, resulting in a tetrahedral coordination of zinc by nitrogen.

The zinc ions show a tetrahedral coordination by nitrogen with Zn−N bond lengths in a range of 1.961(3)–2.158(2) Å, which is in good agreement with the sum of the corresponding ionic radii reported by Shannon (ionic radii Zn²⁺= 0.6 Å, N³⁻= 1.46 Å). ^[24] Furthermore, these interatomic distances are comparable to those of related compounds such as Zn₂PN₃ or Zn₆[P₁₂N₂₄].[ ²¹ , ²⁵ ] The P−N bond lengths of ZnH₂P₄N₈ are between 1.608(3) Å and 1.667(2) Å and are in good agreement with the bond lengths of other known nitridophosphates.[ ²⁶ , ²⁷ , ²⁸ ] The observed PN₄ tetrahedra are distorted, as evidenced by the N−P−N bond angles (104.22(11)–117.88(12)°) which diverge from the regular tetrahedral angle 109.5°. Furthermore, the P−N−P bond angles show values between 125.8(2)° and 141.03(12)°. These are plausible for the existing sechser‐ring types and show comparable angles to those of the host structure β‐PN(NH) (P−N−P bonding angles: 126.0(2)–147.85(4)°). ^[6] More detailed information on interatomic distances and angles is given in Tables S7 and S8.

Solid‐State NMR Studies

In order to confirm the obtained structure model from X‐ray diffraction data the prepared samples were investigated by solid‐state NMR spectroscopy. NMR spectroscopy is a highly effective technique for investigating the local structure of selected nuclei within a given compound.[ ²⁹ , ³⁰ ] The focus within these investigations was on the H atoms, as their scattering factor in X‐ray diffraction experiments is low. During the Rietveld refinements (Table S10, Figure S3), the H content was just deduced from the degree of occupation of Zn²⁺ with respect to charge neutrality (premise of the topochemical reaction when starting from β‐PN(NH)). For this purpose, ¹H, ³¹P and ³¹P{¹H} cross‐polarization experiments were performed on the title compound ZnH₂P₄N₈ first. For interpretation of the obtained spectra, as well as to gain structural insights from topochemical HP partially de‐protonated intermediates, these experiments were also performed on Zn _x H_4−2x P₄N₈ (x≈0, 0.5, 0.85). All respective spectra are shown in Figures S4–S7 in the Supporting Information. As the unit cell of ZnH₂P₄N₈ features one crystallographic H site (Wyck. 4 g, Table S5), only one resonance line is expected in the ¹H MAS spectrum. The spectrum shows one main signal at 5.9 ppm (Figure 5), which can be assigned to ZnH₂P₄N₈, and a small narrow signal at 1.2 ppm resulting from surface hydrolysis, which is in line with the EDX results (spinning side bands are marked with asterisks). ^[6] Figure 6a shows the ³¹P spectrum (black), which reveals four signals with a chemical shift of δ=6.4, −4.7, −10.3, and −22.1 ppm. The signal at δ = 6.4 ppm can be assigned to the minor side phase of Zn₈P₁₂N₂₄O₂, observed in the PXRD as well, which is in line with the absence of this signal in the ³¹P[¹H} cross‐polarization experiment (Figure 6a, gray). The three major signals at δ=−4.7, −10.3, and −22.1 ppm are preserved in the cross‐polarization spectrum and can be assigned to the title compound. Upon closer examination of the three signals, it becomes evident that the signals are not perfectly symmetric but exhibit additional intensity besides their maxima. Especially the one at −22.1 ppm shows a significant shoulder on the left. For this reason, a deconvolution of these ³¹P signals was initially carried out with four Voigt functions. Nevertheless, the signal shoulders and areas of low intensity could only be satisfactorily fitted with a minimum of four additional Voigt functions (total=eight Voigt functions, Figure 6b). The four major signals (green) are superimposed on four minor signals (blue). The integral ratio of the four major signals is about 1.3 : 1.1 : 1 : 1.9 (A : B : C : D).

Figure 5.

The ¹H MAS spectrum shows one signal at 5.9 ppm, corresponding to one crystallographic H position in the structure model of ZnH₂P₄N₈; weak signal at 1.2 ppm can be attributed to small amounts of hydrolysis product; observed sidebands are marked with asterisks.

Figure 6.

a) The ³¹P spectrum (black) shows three major signals at −4.7, −10.3, and −22.1 ppm, which can be assigned to ZnH₂P₄N₈ (they are preserved in the ³¹P{¹H} cross‐polarization spectrum (gray)). The fourth signal at 6.4 ppm can be assigned to Zn₈P₁₂N₂₄O₂. The section box shows a possible local arrangement of the partial real structure in ZnH₂P₄N₈, ZnN₄ tetrahedra in yellow, PN₄ tetrahedra in green, N−H functionality in blue. b) The ³¹P{¹H} cross‐polarization NMR spectrum of ZnH₂P₄N₈ (gray); Deconvolution by eight Voigt functions indicates a lower local symmetry, since significantly more than two signals expected for two P sites with equal multiplicity in the structure model (obtained by SCXRD) are observed.

However, the structure model obtained from X‐ray diffraction data shows only two crystallographic P sites with equal site multiplicity (Wyck. 4 g), which does not seem to be consistent with the observed signals. Nevertheless, the structural model obtained from the X‐ray diffraction data represents an average over long distances, rather than over specific areas. In contrast, the local surroundings of the investigated nuclei are the cause of the respective resonance positions in the NMR spectra. The observed discrepancy can be attributed to the split positions of the Zn²⁺ ions, which in the real structure lead to numerous local environments for the ³¹P atoms. These environments are differentially shielded/unshielded by proximity to the zinc (high electron density) or to the imide groups (lower electron density). This indicates that the SCXRD structure model assumes a higher symmetry than is actually present locally.

Returning to the observed signals in Figure 6, a possible explanation for the integral ratios of the major signals is that a symmetry reduction to Pc (no. 7) has occurred locally, comparable to the mineral γ‐eucryptite (LiAlSiO₄).[ ³¹ , ³² ] This would result in four independent crystallographic P sites with the same site multiplicity (Wyck. 2a) and two independent Zn sites with a site occupancy of 1/2 (Figure 6a, section box; Table S11). Regardless the occupation Zn1=1 and Zn2=0 or vice versa, the resulting Zn−P distances can be divided into three categories: short (occur.: 1×; Zn−P: 2.88 Å), medium (occur.: 1×; Zn−P: 3.03 Å), and long (occur.: 2×; Zn−P: 3.06–3.07 Å). These findings align with the observed integrals of approx. 1 : 1 : 2 (signals A : B : D). However, it is not possible to ascertain unequivocally which of the two occupation variants is present in the title compound. In order to gain a more comprehensive understanding of the data, it is necessary to include the ³¹P{¹H} cross‐polarization spectra from β‐PN(NH) and Zn _x H_4−2x P₄N₈ (x≈0.5, 0.85). This will allow us to interpret the signals more accurately. Figure 7 illustrates the strong dependence of the local environment on the Zn content, which is highlighted by color and small arrows. The lowest Zn content (x≈0.5) indicates the presence of numerous potential local variants, with a tendency towards a real symmetry in space group Pc for the title compound ZnH₂P₄N₈. The shoulder of the main signal (Figure 6, signal C, integral=1) still indicates regions in ZnH₂P₄N₈ that correspond to those of β‐PN(NH) and in which almost no Zn has been incorporated (compare Figure 7, spectra x=0 with x≈0.5 and x=1), which is in line with the Zn split positions. This observation is also consistent with assumption of a topochemical reaction according Equation (1), from which we can now even deduce a possible topotactic reaction‐mechanism with complete preservation of the β‐PN(NH) framework during the reaction. However, this conclusion should be confirmed by in situ experiments.

Figure 7.

³¹P{¹H} cross‐polarization experiments of Zn _x H_4−2x P₄N₈ (x≈0, 0.5, 0.85, 1). Signal ranges that can be identified in the individual experiments are highlighted in color.

Conclusions

In this contribution, we present the synthesis and structural elucidation of the quaternary imidonitridophosphate ZnH₂P₄N₈ via high‐pressure/high‐temperature syntheses. The crystal structure refined from single‐crystal X‐ray diffraction data (SCXRD) is supported by powder X‐ray diffraction data (PXRD), energy‐dispersive X‐ray spectroscopy (EDX) and solid‐state NMR experiments. Two reaction approaches are presented, with the main focus on the novel synthetic route of (partial) high‐pressure defunctionalization of a pre‐synthesized phosphorus nitride imide. By partial defunctionalization and incorporation of Zn²⁺ ions into the P/N‐framework, this synthetic approach is a promising extension to the known synthesis routes of (imido)nitridophosphates, preserving the anionic P/N‐structure and gaining enhanced structural control during synthesis. Solid‐state MAS NMR experiments on different stoichiometric compositions of Zn _x H_4−2x P₄N₈ provide insight into the real structure of ZnH₂P₄N₈ (local environment) and allow conclusions to be drawn about the reaction mechanism of a topochemical reaction. The crystal structure of ZnH₂P₄N₈ can be described as P/N based cation filling variant of distorted low‐cristobalite. Zn²⁺ is located in the larger channels, while hydrogen is covalently bound to N in the smaller ones. Further studies could investigate the complete deprotonation of β‐PN(NH) using smaller cations like lithium. A comparative analysis on the topologically similar alumosilicate γ‐eucryptite (a polymorph of LiAlSiO₄, with filled quartz structure) demonstrates that lithium can occupy both larger and smaller channels in this related substance class. ^[32] In summary, this route extends the synthetic possibilities for the synthesis of (imido)nitridophosphates and shows great potential for synthetic control with respect to classical direct syntheses. Furthermore, it seems to be a suitable method for the preparation of transition metal (imido)nitridophosphates, since the competitive reaction for the formation of quite stable phosphides is prevented by the choice of relatively moderate high pressure/high temperature conditions. Future research should concentrate on the expansion of this synthetic route to encompass other phosphorus nitride imide polymorphs, such as β‐P₄N₆(NH), as well as the incorporation of additional cations to provide a more comprehensive assessment of the synthetic potential of this route. Furthermore, these initial conclusions regarding the reaction mechanism should be subjected to further investigation through in situ experiments. ^[33]

Experimental Section

Synthesis of P₃N₅

P₃N₅ was synthesized according to Stock et al. by an ammonolysis reaction of P₄S₁₀ (Sigma Aldrich, 99 %) at 850 °C for 4 h (heating rate: 5 °C min⁻¹). ^[34] For this purpose, the silica reaction tube was heated out in advance and then saturated with NH₃ for 4 h. After cooling down to room temperature (5 °C min⁻¹), the received orange product is used without further processing. Phase purity was confirmed by PXRD, Elemental analysis (CHNS) and FTIR measurements.

Synthesis of NH₄N₃

NH₄N₃ was synthesized according to Frierson et al. by sublimation of NaN₃ (Acros Organics, 99 %) and NH₄NO₃ (Grüssing, 99 %). ^[35] The starting materials were ground and transferred to a Schlenk tube. The lower part of the Schlenk tube was placed into a glass oven and its valve was opened before starting the heating step at 200 °C for 12 h (heating rate: 5 °C min⁻¹). The product was obtained as colorless crystals in the top of the Schlenk tube. Phase purity was confirmed by PXRD and FTIR measurements.

Synthesis of ZnH₂P₄N₈

The title compound was prepared by high‐pressure/high‐temperature synthesis according two distinct approaches (1) and (2). According (1) different equivalents of pre‐synthesized β‐PN(NH) and ZnCl₂ (10 % excess, Merck, >97.0 %) were used. According Equation (2) stoichiometric amounts of Zn₃N₂ (AlfaAesar, 99.99 %), P₃N₅ and NH₄N₃ were ground together under inert gas conditions. The reaction conditions of 800 °C and 5 GPa were achieved by combining a hydraulic 1000 t press (Voggenreiter, Mainleus, Germany) with the multianvil technique using a modified Walker module. The starting materials were mixed in an argon‐filled glove box (<1 ppm O₂, <0.1 ppm H₂O; Unilab, MBraun, Garching) and placed in an h‐BN crucible of size 18/11 (cavity diameter=1.6 mm, cavity depth=2.3 mm; HeBoSint® S100, Henze, Kempten, Germany). The crucible was sealed with a cap of h‐BN, inserted into two 18/11‐assembly sized graphite furnaces and centered using MgO spacers. For thermal insulation this construct was transferred into a zirconia sleeve; electrical contact was achieved by two Mo discs. The as‐prepared assembly was placed into a center drilled octahedron consisting of Cr₂O₃ (6 %) substituted MgO (Ceramic Substrates & Components, Isle of Wight, U.K.). Eight tungsten carbide (substituted with 7 % Co) cubes (Hawedia, Marklkofen, Germany) with truncated edges (edge length=11 mm) were used as anvils. Initially the sample was pressed to 5 GPa within 200 min before being heated to the target temperature of 800 °C (heating rate: 100 °C min⁻¹). After 15 min dwell time the sample was cooled down to ambient temperature within 15 min. The assembly was decompressed within 600 min and the obtained product was recovered from the crucible and washed with deionized water for purification. Schematic drawings of the octahedral pressure cell and the Walker‐type module are illustrated in Figure S8. Additional information regarding high‐pressure/high‐temperature synthesis can be found in literature. ^[36]

Single‐Crystal X‐ray Diffraction

For single‐crystal XRD measurements ZnH₂P₄N₈ single crystals were isolated by a MicroMount^TM (Bruker). The single‐crystal X‐ray diffraction data were collected on a Bruker D8 Venture TXS diffractometer (rotating anode, Mo‐Kα radiation, λ=0.71073 Å, multilayer monochromator). Indexing and integration as well as determination of the space group was performed by the APEX3 software package.[ ³⁷ , ³⁸ , ³⁹ ] The crystal structure was solved using the SHELXS‐97 algorithm and refined by full matrix least‐squares methods using WinGX.[ ⁴⁰ , ⁴¹ ]

Powder X‐ray Diffraction

The ground product was placed and sealed in a glass capillary (d=0.3 mm, Hilgenberg GmbH) for PXRD measurement. The measurement was performed using a Stoe STADI P diffractometer with Mo‐Kα ₁ (λ=0.71073  Å) radiation, Ge(111) monochromator and Mythen 1 K detector in modified Debye‐Scherrer geometry. Rietveld refinement of the measured data was performed using TOPAS software. ^[42]

Scanning Electron Microscopy

SEM imaging and EDX measurements were made by a Dualbeam Helios Nanolab G3 UC (FEI, Hillsboro) equipped with a X–Max 80 SDD detector (Oxford Instruments, Abingdon). For this purpose single crystallites were fixed on an adhesive carbon pad and coated by a high‐vacuum sputter coater (BAL‐TEC MED 020, Bal Tec A) to ensure electrical conductivity.

Fourier Transform Infrared Spectroscopy

The FTIR spectrum was collected on a Spectrum BX II spectrometer with DuraSampler ATR‐device (Perkin Elmer) at ambient conditions.

Solid‐state magic angle spinning (MAS) NMR Spectroscopy

³¹P, ¹H, and ³¹P{¹H} cross‐polarization NMR experiments were performed with a AVANCE DSX 500 MHz NMR spectrometer (Bruker) with a magnetic field of 11.7 T. The samples were filled and compacted into a 2.5 mm rotor, which was mounted on a commercial MAS probe (Bruker). The sample rotation frequency was about 20 kHz. The obtained data were analysed using ORIGIN Pro 2022b. All spectra were indirectly referenced to ¹H in 100 % TMS at −0.1240 ppm.

Conflict of Interests

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.

Supporting Information

Pritzl R. M., Steinadler J., Buda A. T., Wendl S., Schnick W., Chem. Eur. J. 2024, 30, e202402741. 10.1002/chem.202402741

Data Availability Statement

The data that support the findings of this study are available in the supplementary material of this article.