Prolonged Electrochemical Cycling Characteristics of ZnSiP₂ Prepared with Mixed Crystalline-Amorphous Domains

Abstract

This paper presents an innovative approach wherein mechanical alloying and mechanical cation-disordering techniques are combined to synthesize peculiar zinc silicide phosphide (ZnSiP₂) anode materials under controlled atmosphere. In this method, Zn atoms and P atoms are simultaneously incorporated into the parent Si crystal structure, resulting in A(II)_xB(IV)_yP_x+y solid solutions with precise control over nanodomain structures of mixed crystalline-amorphous phases. This distinctive nanoarchitecture of the ZnSiP₂ anode, featuring an amorphous ionic-conduction phase network, facilitates the smooth transport of Li⁺ ions, thereby enabling an exceptionally prolonged electrochemical cycling performance, surpassing 200 cycles. In this study, we attempted to unravel the microstructure of ZnSiP₂ using transmission electron microscopy (TEM). It was observed that when synthesized under an inert Argon atmosphere, the material formed a polycrystalline structure consisting of numerous nanocrystals (5–10 nm) assembled. Additionally, when attempts were made to reduce synthesis costs by conducting the synthesis under ambient atmospheric (Air) conditions, amorphous regions were generated. This amorphous region within the polycrystalline ZnSiP₂ microstructure represents a novel finding. The electrochemical impedance measurements and galvanostatic intermittent titration technique (GITT) analysis conducted in this study not only revealed but also characterized the enhanced cycling performance of this unique ZnSiP₂ anode structure.

1. Introduction

Mechanical milling, also known as mechanochemical methods, represents an innovative approach that harnesses mechanical forces, such as impact and shear, to activate chemical species, thereby inducing chemical transformations. In this study, we focus on the utility of mechanical milling, particularly the ball milling technique, to facilitate chemical reactions, commonly referred to as mechanical milling.^1–4 Mechanical alloying stands as a technique that enables the creation of solid solutions through the cyclic interplay between cold welding and fracturing processes, thus facilitating atomic-level mixing. In the domain of metals, cold welding manifests as the phenomenon where metallic bonds form even at temperatures below the melting point when two metallic surfaces undergo mechanical compression. Notably, similar phenomena have been observed in ion-crystalline materials. Consequently, particles undergoing this process establish atomic-level bonds as they agglomerate. Upon reaching a critical size, mechanical forces lead to their fragmentation, aptly termed fracturing. Smaller particle sizes yield larger surface areas, increasing the likelihood of cold welding—a surface-bonding phenomenon. Conversely, larger particles tend to experience more pronounced fracturing. Therefore, as mechanical alloying progresses, equilibrium ensues when the rates of cold welding and fracturing align. Several reports have detailed the preparation of various A(III)B(IV)_xP solid solutions, including AlGeP, via mechanical alloying.^5,6 In this process, Al atoms and P atoms are simultaneously introduced into the parent Si or Ge crystal structure, resulting in a single-phase A(III)B(IV)_xP solid solution.^5,6

While mechanical alloying primarily aims to solidify binary and ternary systems, mechanical milling in single-component systems serves a distinct purpose: disrupting crystalline order and inducing phenomena such as amorphization—a technique aptly named mechanical disordering. When applying mechanical disordering to ionic crystals, the objective is to disrupt the orderly arrangement of cations while preserving the anionic framework.^7–9 This process, known as “mechanical cation-disordering,” necessitates the presence of two or more cationic species and a single anionic species, capable of maintaining the anionic framework through mechanical milling.

ZnSiP₂, classified as an A(II)_xB(IV)_yP_x+y compound and categorized among ternary silicon phosphide materials,^10–12 has garnered significant attention as a promising anode material since its initial report in 2019.¹⁰ This interest stems from its comparatively high theoretical capacity (1966 mAh g⁻¹, equivalent to accommodating 11.4 Li⁺) compared to conventional graphite anode (372 mAh g⁻¹) across a broad potential range between 0.005 and 1.6 V vs. Li/Li⁺. ZnSiP₂ also holds the potential for superior cycling performance when compared to other high-capatity anodes such as silicon (Si), owing to more modest substantial volumetric change (∼250 %) than Si (280–400 %)^13–16 during complete lithiation/delithiation. The introduction of caton-disorder between Zn and Si atoms in ZnSiP₂ crystal structure may offer an additional enhancement in cycling performance compared to the thin layer of cation-ordered ZnSiP₂ prepared via chemical vapor deposition,^10,11 a material that has been extensively investigated due to its wide band gap properties.^17–19 It should be also noted that, although binary transion metal phosphide (M-P_x: M = Fe,²⁰ Co,^21,22 Ni,^23,24 Sn,²⁵ Cu,²⁶ Ge²⁷ etc.) materials has been rigorously investigated, there have been yet relatively few reports on ternary silicon phosphide materials as anode materials.^10–12

This paper introduces an innovative approach where mechanical alloying and mechanical cation-disordering techniques are combined to synthesize ZnSiP₂ anode materials. In this method, Zn atoms and P atoms are simultaneously incorporated into the parent Si crystal structure, resulting in A(II)_xB(IV)_yP_x+y solid solutions with precise control over nanodomain structures of crystalline-amorphous phases. It is imperative to emphasize that the straightforward synthesis of these cation-disordered compounds solely from a mixture of starting materials holds practical and substantial significance in the realm of materials science. Previous reports on the crystal structure of zinc silicide phosphide (ZnSiP₂) have revealed a disordered arrangement of cation species (Zn(II) and Si(IV)).^10,11 However, the surrounding environment of the crystal phase and the overall structural organization of the crystalline material remained unclear.

In this study, we attempted to unravel the microstructure of ZnSiP₂ using transmission electron microscopy (TEM). It was observed that when synthesized under an inert Argon atmosphere, the material formed a polycrystalline structure consisting of numerous nanocrystals (5–10 nm) assembled. Additionally, when attempts were made to reduce synthesis costs by conducting the synthesis under ambient atmospheric (Air) conditions, amorphous regions were generated. This amorphous region within the polycrystalline ZnSiP₂ microstructure represents a novel finding.

This structure demonstrates remarkably prolonged electrochemical cycling, exceeding 200 cycles. Notably, achieving such extended cycling characteristics with a specific electrolyte configuration is a groundbreaking achievement in the field. The TEM analysis presented here makes a significant contribution to the understanding of the outstanding electrochemical cycling performance of this material, marking it as a novel discovery.

2. Experimental

2.1 Synthesis of sphalerite-type ZnSiP₂

As starting materials, zinc (Zn) powder [purity ≥ 99.9 %, Fuji Film Wako Pure Chemical] was used as the Zn source, silicon (Si) powder [purity ≥ 99.99 %, Kanto Chemical] as the Si source, and red phosphorus (P) powder [purity ≥ 99.999 %, Nakalai Tesque] as the P source. Sphalerite-type ZnSiP₂ was obtained by ball milling 1.2622 g (1.0 eq) of zinc powder, 0.5422 g (1.0 eq) of silicon powder, and 1.1957 g (2.0 eq) of red phosphorus powder using a ball mill [Premium line PL-7, Fritsch] at 600 rpm for 12 h. The ball mill container was maintained either under an argon atmosphere or in ambient Air atmosphere at room temperature. Conducting the synthesis under an argon atmosphere helps suppress impurity formation and is expected to yield the theoretical capacity. On the other hand, synthesis in ambient air offers the potential for high-efficiency and cost-effective material synthesis. Additionally, there have been no prior reports of synthesizing ZnSiP₂ in ambient air, making it the focus of this investigation.

2.2 X-ray diffraction (XRD) and X-ray photoelectron spectroscopy (XPS)

A desktop MiniFlex (Rigaku Corporation) was employed for XRD measurements for the prepared samples. The measurement conditions included a range of 10–80°, a step size of 0.01°, and a scan speed of 2.0° per minute. XPS analysis was performed using a scanning X-ray photoelectron spectrometer (PHI X-tool, Ulvac-Phi). The X-ray source used was AlKα, with conditions set at 15 kV and 50 W. The analysis was conducted for C_1s, N_1s, O_1s, Zn_2p3/2, Si_2p, and P_2p orbitals. Narrow scans for specified elements were conducted at any three points on the powdered samples. Depth profiling analysis for the targeted elements was also performed. Sputtering treatment at 5 kV for 8 s and narrow scans were repeated 10 times. The obtained spectra were subjected to peak deconvolution based on the previous reports.^10,12,28,29 Gaussian functions were assumed during this process.

2.3 SEM and TEM observation

Morphological observations were carried out using a scanning electron microscope (SEM) [S-5500, Hitachi High-Technologies]. A small amount of powdered sample was fixed onto double-sided carbon tape. Crystalline microstructure observations were conducted using a transmission electron microscope (TEM) [H-9500, Hitachi High-Technologies]. A small amount of sample was dispersed in approximately 2 mL of dehydrated ethanol [CH₃CH₂OH, purity 99.5 vol%, Wako Pure Chemicals]. Ca. 10 µL of the dispersion was dropped onto a microgrid and allowed to air-dry for sample preparation.

2.4 Electrode fabrication of ZnSiP₂/AB composite

To ensure the electronic conductivity throughout electrode materials, the synthesized sphalerite-type ZnSiP₂ (7.0 eq) was composited with the conductive agent acetylene black (AB) [HS-100, Denka] (2.0 eq). The compositing (ZnSiP₂ : AB = 70 : 20 wt%) was achieved by mixing the materials at 300 rpm for 12 h using a planetary ball mill. Lithium polyacrylic acid (PAALi, 10 wt%), dissolved in an aqueous solution, was used as a binding agent. The mixed slurry was coated onto a copper foil collector with thickness of 8 µm. Subsequently, it was dried for more than 12 h under vacuum at 80 °C. The obtained electrode was subjected to two pressing cycles at 3-ton air hydro press. Following this, it was dried for overnight under vacuum at 80 °C. The loading electrode mass was 0.5–1.0 mg cm⁻².

2.5 Coin-cell assembly and charge-discharge characterization

Coin cell assembly was conducted using CR2032 coin cell components. A single-layer polypropylene separator was used. The electrolyte used was 1 M LiPF₆/(EC : DEC)(1 : 1 in vol) [Kishida Chemical]. Lithium metal [Honjo Metal] was used as the counter electrode, and ca. 200 µL of the electrolyte was added. Testing was performed using a battery charge-discharge unit [HJ1001SD8, Hokuto Denko] at a current density of 197 mA g⁻¹ (0.1 C, 1 C = 1966 mA g⁻¹) within the voltage range of 0.005–3.0 V vs. Li/Li⁺. All the cells were assembled in a dry room (dew-point temperature < −40 °C).

2.6 Electrochemical characterization

EIS measurements were conducted using a potentiostat/galvanostat for electrochemical measurements [ModuLab XM ECS, AMETEK Solartron Analytical]. Measurements were taken under conditions of 10 mV AC perturbation amplitude and a frequency range from 1 MHz to 0.01 Hz. To calculate the Li⁺ diffusion coefficient within the solid, GITT (galvanostatic intermittent titration technique) measurements were performed. The potential range was set to 0.005–3.0 V vs. Li/Li⁺. Constant current was applied for 15 min at a current density of 197 mA g⁻¹ (0.1 C-rate) followed by 4 h of relaxation, and this cycle was repeated until the voltage reached its cut-off value. The Li⁺ diffusion coefficient (D_Li+) was calculated from the voltage drop observed during the relaxation periods.

3. Results and Discussion

Figure 1 displays XRD patterns of ZnSiP₂ samples obtained through high-energy ball milling in both Argon and Air atmospheres, along with those for the initial/raw materials and a reference for chalcopyrite-type (cation-ordered) ZnSiP₂. In both ZnSiP₂ samples (Argon and Air), discernibly absent are the peaks associated with the starting materials (Zn, Si, and P). Instead, their diffraction patterns strikingly resemble that of pristine Si with a diamond-type structure, signifying the formation of a single-phase cubic compound. This transformation is attributed to the effective integration of Zn (atomic radius: 1.24 Å) and P (atomic radius: 0.98 Å) into the diamond-type Si structure (atomic radius: 1.11 Å), resulting in a diminished lattice constant (a = b = c = 5.34 Å) when compared to pristine Si (a = b = c = 5.43 Å). The obtained XRD patterns are well attributed to the cation-disordered ZnSiP₂,^10,11 and a comparison with the JCPDS card for the chalcopyrite-type ZnSiP₂ (cation-ordered) and prsior XRD data for cation-disordered sphalerite-type ZnSiP₂ further validates our samples as conforming to the target structure. This precision in our synthesis process points towards an inherently disordered arrangement of cation species, specifically Zn(II) and Si(IV).⁷ Notably, the indices for samples synthesized under Argon and Air atmospheres (111, 200, 220, 311, 400, 331) are precisely half of those for the cation-ordered structure (112, 200, 220, 312, 400, 332), serving as a compelling indicator of the cation arrangement’s disorder. A telling revelation emerges when we compare the lattice constants (a = b = c = 5.34 Å) of cation-disordered ZnSiP₂ with those of cation-ordered ZnSiP₂ (a = b = 5.34 Å, c = 10.435 Å). It becomes evident that the unit cell of each sample closely resembles a fractional division of the c-axis of the cation-ordered ZnSiP₂’s tetragonal structure (space group: I-42d). Furthermore, the XRD pattern fitting (Fig. S1) consistently yields the space group F-43m, firmly confirms the sphalerite (ZnS) type. Within this structure, phosphorus (P) occupies sulfur (S) sites in a cubic close-packed arrangement, while zinc (Zn) and silicon (Si) are randomly distributed within the Zn sites.

Figure 1.

Energy Ball-Milling preparation process of ZnSiP₂ under varied atmospheres (Argon and Air) with corresponding XRD patterns revealing cation-disordered sphalerite structures. All the raw material powder samples were enclosed in a hermetically sealed ball-milling container designed for PL-7 to minimize contamination during the ball-milling process. For Argon, the mixing and enclosure of raw material powder samples were carried out within the Ar-filled glove box.

3.1 Charge-discharge characteristics

A half-cell was constructed by pairing a ZnSiP₂/AB composite with Li metal as the counter electrode to evaluate the fundamental electrochemistry. We conducted charge-discharge tests at a constant current of 0.1 C-rate (197 mAh g⁻¹). In Fig. 2, typical charge-discharge curves are shown for the fifth cycle, which is considered to have achieved a steady state. In both atmospheric conditions (Argon/Air), a gradual sloping behavior is evident, starting at around 1.2 V and descending to a terminal voltage of 0.005 V. This behavior is attributed to a multistage Li deintercalation reaction, in accordance with the findings reported by W. Li et al.¹⁰ For ZnSiP₂(Argon), the reversible capacity reaches 1819 mAh g⁻¹, equivalent to 10.6 electron reactions, achieving 93 % of the theoretical capacity. Conversely, in the case of ZnSiP₂(Air), the reversible capacity amounts to 1574 mAh g⁻¹, corresponding to 7.8 electron reactions, achieving 80 % of the theoretical capacity. Inlet of Fig. 2, we present potential charge-discharge reactions for ZnSiP₂.

Figure 2.

Comparative charge-discharge profiles of ZnSiP₂ anodes fabricated in Air and Argon atmospheres, evaluated as half-cells with Lithium metal in a 1 M LiPF₆/(EC/DEC) electrolyte at a constant current of 0.1 C-rate rate at 5th cycle.

3.2 Impedance analysis of electrochemical resistance components

EIS measurement was executed at the potential of 0.4 V. The fitting of the impedance data was performed using the equivalent circuit (inset), where R1 represents the solution resistance. R2 signifies the resistance at the Li metal interface, R3 denotes the resistance at the ZnSiP₂ electrode interface, and R4 corresponds to the charge transfer resistance of the ZnSiP₂ electrode. Since the same electrolyte was utilized throughout the study, the value of R1 remained constant.

From the Nyquist plot shown in Fig. 3, it became evident that the total magnitude of the arc associated with R2 + R3 + R4 revealed a significantly lower value for ZnSiP₂(Argon) at 46 Ω cm² compared to 69 Ω cm² for ZnSiP₂ (Air). Clearly, ZnSiP₂(Argon) exhibited approximately two-thirds less resistance. In both atmospheric conditions, it was confirmed that the component R4, which corresponds to the charge transfer resistance of the ZnSiP₂ electrode, comprised approximately 75 % of the total value of R2 + R3 + R4. When comparing the charge transfer resistance values of the ZnSiP₂ electrode, ZnSiP₂(Argon) demonstrated approximately three-quarters less resistance, which is likely connected to its ability to achieve the theoretical capacity in charge-discharge tests.

Figure 3.

Nyquist plots and corresponding fitted curves for ZnSiP₂(Argon) and ZnSiP₂(Air). Electrochemical Impedance Spectroscopy (EIS) was conducted at a constant potential of 0.4 V vs. Li/Li⁺ during charge-discharge operations at the 5th cycle. The EIS measurements were conducted using a half-cell configuration with Lithium metal in a 1 M LiPF₆/(EC/DEC) electrolyte.

3.3 Calculation of apparent Li⁺ diffusion coefficients using GITT

We conducted the calculation of apparent Li⁺ diffusion coefficients within the crystal lattice using the galvanostatic intermittent titration technique (GITT). An enlarged view of the GITT test step and an overview of the measurement are shown in Fig. 4. Li⁺ diffusion coefficients was determined from equation inset. The IR region primarily encompasses ion conduction resistance in the electrolyte, electronic resistance, and charge transfer resistance at the electrolyte-electrode interface. The ΔE_τ region includes Li⁺ diffusion. It should be noted that we assumed that there is no significant difference in average Li⁺ diffusion length within ZnSiP₂, as the particle sizes are more or less the same between ZnSiP₂(Argon) and ZnSiP₂(Air), as observed in their SEM images (Fig. S2). Both samples possess irregulay shaped primary particles with various sizes ranging from 50 to 100 nm. These primary particles aggregate to form secondary particles on the order of several micrometers, displaying variety of shapes. During the lithiation process, the apparent Li⁺ diffusion coefficients in the ZnSiP₂(Argon) were approximately one-third lower compared to those for ZnSiP₂(Air). This trend was consistent during the Li extraction reaction process as well. From these findings, it is suggested that the ZnSiP₂(Argon) exhibits seemingly lower solid-state Li⁺ diffusion properties compared to the ZnSiP₂(Air).

Figure 4.

Galvanostatic Intermittent Titration Technique (GITT) analysis illustrating the Li⁺ insertion process across ZnSiP₂ anodes synthesized in Air and Argon environments, employing a 1 M LiPF₆ EC/DEC electrolyte. The Diffusion Coefficient evaluated for each case is shown against SOC.

3.4 XPS analysis on the bonding state of ZnSiP₂

A comprehensive XPS analysis was employed to investigate the precise bonding state of ZnSiP₂ (Fig. 5) XPS spectra were acquired at 5 kV following the sputtering of samples synthesized under both Argon and Air atmospheres, with measurements taken at 24-s intervals.

Figure 5.

Depth profiles of XPS spectra during 0–72 s Ar etching on the surface of ZnSiP₂ anode prepared under air and Argon environment showing that the surface of ZnSiP₂ (Air) is more oxidized than ZnSiP₂(Argon), implying the existence of possibly ZnO species.

Regarding Zn_2p3/2, for ZnSiP₂(Argon), the peak position remained constant for up to 72 s of sputtering, indicating a consistent chemical state from the surface to the interior. In contrast, for ZnSiP₂(Air), the peak position notably shifted from the energy corresponding to Zn-P bonding to that of Zn-O bonding, strongly suggesting the presence of Zn(II)-O bonding extending from the surface to the interior. Concerning Si_2p, the peak position transitioned to the vicinity of the energy value associated with SiO during sputtering, signifying silicon in the IV oxidation state on the surface and silicon in the II oxidation state within the interior. In the case of P_2p, the peak associated with P-O bonding vanished immediately after sputtering, implying the exclusive presence of P(IV)-O bonding on the surface. Accordingly, the existence of ZnO and SiO oxides in Air is thought to reduce the electronic conductivity of active material, leading to an increase in charge transfer resistance (R4 in Fig. 3) associated with lithiation of ZnSiP₂ or Zn, P, Si. Such increase in R4 for ZnSiP₂ (Air) leads to an augmentation in overpotential of its charge discharge curves characterized by a gradual slope (see Fig. 2). This renders charge discharge curves for ZnSiP₂ (Air) more susceptible to reach the cut-off voltage, consequently decreasing the capacity at 0.1 C. Nonetheless, unsolved question remains: What is the factor contributing to the threefold increase in the apparent Li⁺ diffusion coefficient for ZnSiP₂ (Air). To address this inquiry, we have delved deeper into the difference in nanostructure of ZnSiP₂ crystals between ZnSiP₂ (Air) and ZnSiP₂ (Argon) through high-resolution transmission electron microscopy (HR-TEM) observations.

3.5 Nanostructure characterization by HR-TEM

Figure 6 displays typical HR-TEM images of ZnSiP₂ samples for Argon and Air comparatively. Under Argon atmosphere, the uniform generation of nanocrystals is observed, while, under Air condition, one can see the formation of a heterogeneous, crystalline/amorphous nature. In Fig. S3, more detailed depiction including selected-area electron diffraction (ED) patterns are shown. The ED patterns clearly reveal diffraction rings corresponding to the (111), (220) and (311) planes, aligning well with the previously discussed XRD results. For ZnSiP₂(Argon), it was observed to be a polycrystalline assembly of crystallites with sizes of up to approximately 10 nm. Clear lattice fringes corresponding to the (111) plane with an interplanar spacing of d₁₁₁ = 3.09 Å, as determined by XRD measurements, were prominently observed. Additionally, minor gaps between crystallites and amorphous regions at the surface were detected. For ZnSiP₂(Air), it was determined to consist of a polycrystalline structure composed of crystallites with sizes of approximately 5 nm and clear lattice fringes corresponding to the (111) plane with an interplanar spacing of d₁₁₁ = 3.09 Å. Unlike the Argon, for the ZnSiP₂(Air), one can observe the presence of individual aggregates composed of solely 1–2 nanocrystals each. Within these nanocrystals, lattice fringes attributed to ZnSiP₂ were not discernible, further suggesting the presence of amorphous regions. Accordingly, the uniform generation of nanocrystals in Argon facilitated electron transfer, leading to the attainment of a capacity exceeding 90 % of the theoretical value. Conversely, under Air condition, the formation of a heterogeneous, crystalline/amorphous nature, have made local Li⁺ diffusion more accessible through amorphous region. However, this advantage is accompanied by an overall impeded electron transfer, a capacity of approximately 80 % was achieved to the theoretical one during initial cycling, even at a slow C-rate of 0.1 C.

Figure 6.

High-resolution transmission electron microscopy (HR-TEM) images (700,000× at 300 kV) of ZnSiP₂ anodes prepared in Air and Argon environments, highlighting distinct arrangements of crystalline and amorphous domains, as depicted.

3.6 Proposed mechanism for two distinct structures of ZnSiP₂

The proposed mechanism underlying the formation of two distinct ZnSiP₂ structures in different environments, namely Argon and Air, is illustrated in Fig. 7. In the initial stages of the mechanochemical process under an Argon atmosphere, it is postulated that Zn and P exhibit higher reactivity compared to Si, primarily due to their respective “Mohs hardness” values (Zn: 2, P: 0.5, Si: 7). This hypothesis gains support from XRD patterns obtained after 1-h milling, where discernible ZnP₂ peaks are evident while Si peaks remain unaltered (as shown in Fig. S4), aligning with previous findings.^10,11 Subsequent intense collisions between ZnP₂ and Si particles lead to the formation of minute quantities of ZnSiP₂ on the surface of Si particles. Given the high-energy ball milling process with a rotation speed of 600 rpm, the synthesized ZnSiP₂ is delaminated from the Si particle surfaces. Subsequent re-collisions between ZnP₂ and freshly exposed Si surfaces facilitate further ZnSiP₂ formation. The repetitive nature of such vigorous collisions ultimately results in the creation of nanocrystalline ZnSiP₂ particles, as observed in the HRTEM image (Fig. 5 and S3). In contrast, when the mechanochemical reaction occurs in the presence of Air, simultaneous oxidation of raw materials takes place. This dual process yields ZnSiP₂ as well as oxidized compounds, including amorphous ZnO. The incorporation of this amorphous ZnO contributes to the development of a heterogeneous, polycrystalline/amorphous microstructure of ZnSiP₂ under air conditions.

Figure 7.

Speculated mechanism for two distinct structures of ZnSiP₂ materials under two different atmospheres.

3.7 Electrochemical cycling performance for ZnSiP₂

To assess the impact of the variation in polycrystalline microstructure of ZnSiP₂, cycling performances were conducted at different current densities, namely 0.1 (low C-rate) and 1.0 C (high C-rate). Two distinct electrolyte compositions were employed: electrolyte (A), a conventional 1 M LiPF₆/(EC/DEC) (50 : 50 in volume), and electrolyte (B), specially formulated for Si-based anodes,^30,31 containing 1.2 M LiPF₆ + 0.1 LiBF₄/(EC/FEC/DMC) (25 : 5 : 70 in volume) with the addition of 1 wt% VC.

As shown in Fig. 8, all samples tested with electrolyte (B) exhibited significantly extended cycling performance compared to those using electrolyte (A). This enhancement in capacity retention is attributed to the formation of FEC- and VC-derived polymeric solid electrolyte interphase (SEI), where the former enhances the Li⁺ ion conductivity, and the latter provides flexibility and electrochemical reversibility.^32,33 It is considered that the combination of these attributes from FEC and VC plays a critical role in mitigating the approximately twofold volume expansion during the lithiation of ZnSiP₂.^10,11

Figure 8.

Electrochemical cycling performances of ZnSiP₂ anodes synthesized under Argon and Air atmospheres at low (0.1 C-rate) and high (1 C-rate) rates. Charge-discharge tests were conducted over 100 cycles in two distinct electrolytes: (A) 1 M LiPF₆/(EC/DEC) and (B) 1.2 M LiPF₆ + 0.1 M LiBF₄/(EC/FEC/DMC)(25 : 5 : 70 in vol%) containing 1 wt% VC as an additive. Prior to any 1 C-rate cycling tests, cells were pre-cycled for 5 times at 0.1 C-rate.

At a low C-rate of 0.1 C-rate, when employing electrolyte (B), both ZnSiP₂ samples (Argon and Air) maintained a high capacity of over 1500 mAh g⁻¹ throughout 200 cycles, corresponding to an impressive capacity retention over 90 %. During initial cyclings, ZnSiP₂(Air) exhibited lower capacity compared to ZnSiP₂(Argon), attributable to its higher charge transfer resistance, reflecting the differences in their polycrystalline microstructures as well as oxides (ZnO and SiO). After 100 cycles, however, the capacity for ZnSiP₂(Air) gradually increases, narrowing the gap with ZnSiP₂(Argon). By the 200th cycle, the capacity reaches 1523 mAh g⁻¹, equivalent to 108 % of its initial capacity, whereas the capacity of ZnSiP₂(Argon) decrease to 1572 mAh g⁻¹, corresponding to 90 % of its initial capacity. Such capacity rise after long cycling has also been previously reported for other metal oxide anodes exhibiting conversion/alloying reaction such as CuO, CoO, Fe₂O₃, Fe₃O₄, and SnO₂.^34–40 For ZnSiP₂(Air), the capacity increase is thought to be a consequence of the ongoing micronization of ZnSiP₂ nanoparticles during repeated conversion and alloying process, leading to the rearrangement of nanocrystalline and amorphous domains. This rearrangement gradually removes inhibitroy factor for charge transfer during lithiation/delithiation and mitigates capacity degradation, approaching to the theoretical capacity. However, the ratio of capacity increase slowly decreases around the 200th cycle, indicating that the increased capacity is not expected to surpass the theoretical capacity. Figure S5 illustrates that up to the 10th cycle, ZnSiP₂(Air) exhibits slightly higher coulombic efficiency compared to ZnSiP₂(Argon). After stabilizing over the 10th cycle, the coulombic efficiency for ZnSiP₂(Air) [98.6–99.0 %] remains nearly equivalent to those for ZnSiP₂(Argon) [96.6–99.0 %]. These results demonstrate that neither oxides nor amorphous domains in ZnSiP₂(Air) significantly deteriorate the electrochemical reversibility of active materials. Instead, they contribute to the stabilization of charge discharge process.

At a high C-rate of 1.0 C, where apparent Li⁺ diffusion within the ZnSiP₂ host material played a dominant role, the overall cycling performance of ZnSiP₂ (Air) surpassed that of ZnSiP₂(Argon) over 200 cycles, indicating the more favorable Li⁺ diffusion in ZnSiP₂(Air), which exhibited a threefold increase. Taking into account for these electrochemical characteristics and the convenience of synthesizing under ambient conditions without the need for atmosphere control equipment, such as a globe box, this paper concludes that ZnSiP₂(Air) is the superior choice over ZnSiP₂(Argon).

4. Conclusions

In this study, the authors conducted with a special focus on an analysis of the synthesis atmosphere conditions and their impact on the fundamental electrochemical characteristics of sphalerite-type ZnSiP₂ materials. Through crystallographic, spectroscopic measurements, and electron microscopy observations, we confirmed that ZnSiP₂ synthesized under both inert argon and ambient (Air) atmospheres consisted of polycrystalline structures composed of 10 nm nanocrystals, exhibiting a cation-disordered arrangement. Notably, the synthesis of nanoscale primary particles (5–10 nm) with nanocrystalline structure was achieved in a single-step mechanical milling process using elemental components (Zn, Si, P) as raw materials, a particularly intriguing outcome. Magnified HRTEM observation revealed that ZnSiP₂ (Air) was comprised nanodomain structures with a mixture of crystalline and amorphous phases, along with the presence of oxides such as ZnO and SiO.

In initial electrochemical evaluations, ZnSiP₂ (Ar) atmosphere demonstrated consistent values matching the theoretical capacity (1966 mAh g⁻¹). In contrast, ZnSiP₂ (Air) exhibited a capacity of 1574 mAh g⁻¹, which corresponds to 80 % of the theoretical value, primaly due to its relatively high R_ct. Over 200 cycles, however, ZnSiP₂ (Air) displayed exceptional cycling stability, achieving a capacity of over 1500 mAh g⁻¹, thereby narrowing the gap with ZnSiP₂(Argon). This result signifies the achievement of sustaining a discharge capacity equivalent to three times the theoretical capacity of conventional graphite anode over an extended period. Furthermore, at a high C-rate of 1.0 C, ZnSiP₂ (Air) outperformed ZnSiP₂(Argon) in terms of cyclability over 200 cycles. This observation suggests that ZnSiP₂ (Air) benefits from more favorable Li⁺ diffusion in ZnSiP₂(Air), facilitated by the presence of amorphous ionic-conduction phase.

Importantly, this accomplishment is not limited to ZnSiP₂ alone but holds promise for other sphalerite-type compounds. Particularly noteworthy is the outstanding electrochemical performance achieved under the relatively cost-effective synthesis conditions of ambient atmosphere, showcasing the broader applicability of this research.

Acknowledgments

This study was supported by JSPS Grant-in Aid for Scientific Research (KAKENHI) C under Grant No. JP 21K05241.

Data Availability Statement

The data that support the findings of this study are openly available under the terms of the designated Creative Commons License in J-STAGE Data at https://doi.org/10.50892/data.electrochemistry.24484144.

• 1) This paper presents an innovative approach wherein mechanical alloying and mechanical cation-disordering techniques are combined to synthesize peculiar zinc silicide phosphide (ZnSiP₂) anode materials under controlled atmosphere. In this method, Zn atoms and P atoms are simultaneously incorporated into the parent Si crystal structure, resulting in A(II)_xB(IV)_yP_x+y solid solutions with precise control over nanodomain structures of mixed crystalline-amorphous phases. This distinctive nanoarchitecture of the ZnSiP₂ anode, featuring an amorphous ionic-conduction phase network, facilitates the smooth transport of Li⁺ ions, thereby enabling an exceptionally prolonged electrochemical cycling performance, surpassing 200 cycles. In this study, we attempted to unravel the microstructure of ZnSiP₂ using transmission electron microscopy (TEM). It was observed that when synthesized under an inert Argon atmosphere, the material formed a polycrystalline structure consisting of numerous nanocrystals (5–10 nm) assembled. Additionally, when attempts were made to reduce synthesis costs by conducting the synthesis under ambient atmospheric (Air) conditions, amorphous regions were generated. This amorphous region within the polycrystalline ZnSiP₂ microstructure represents a novel finding. The electrochemical impedance measurements and galvanostatic intermittent titration technique (GITT) analysis conducted in this study not only revealed but also characterized the enhanced cycling performance of this unique ZnSiP₂ anode structure.

CRediT Authorship Contribution Statement

Etsuro Iwama: Conceptualization (Lead), Data curation (Lead), Funding acquisition (Lead), Supervision (Lead), Writing – original draft (Lead), Writing – review & editing (Lead)

Toyomi Takazawa: Data curation (Lead), Investigation (Lead), Methodology (Lead)

Koji Matsuyama: Investigation (Equal), Methodology (Lead)

Daisuke Yamaguchi: Investigation (Lead), Methodology (Equal)

Conflict of Interest

The authors declare no conflict of interest in the manuscript.

Funding

Japan Society for the Promotion of Science: 21K05241

Footnotes

E. Iwama: ECSJ Active Member

References

• 1)  J. S. Benjamin and T. E. Volin, Metall. Trans., 5, 1929 (1974).
• 2)  J. S. Benjamin, Mater. Sci. Forum, 88–90, 1 (1992).
• 3)  M. H. Enayati and F. A. Mohamed, Int. Mater. Rev., 59, 394 (2014).
• 4)  A. G. Lehmann, M. Bionducci, and F. Buffa, Phys. Rev. B, 58, 5275 (1998).
• 5)  W. Li, J. Wen, A. Chen, J.-H. Wang, M. Liu, and H. S. Park, J. Mater. Chem. A, 10, 25329 (2022).
• 6)  J. Wen, X. Liu, Z. Li, and W. Li, J. Alloys Compd., 934, 167622 (2023).
• 7)  C. Baur, J. Chable, F. Klein, K. Chakravadhanula Venkata Sai, and M. Fichtner, ChemElectroChem, 5, 1484 (2018).
• 8)  M. N. Obrovac, O. Mao, and J. R. Dahn, Solid State Ionics, 112, 9 (1998).
• 9)  Q.-J. Li, H. Sheng, and E. Ma, Nat. Commun., 10, 3563 (2019).
• 10)  W. Li, X. Li, J. Liao, B. Zhao, L. Zhang, L. Huang, G. Liu, Z. Guo, and M. Liu, Energy Environ. Sci., 12, 2286 (2019).
• 11)  W. Li, J. Liao, X. Li, L. Zhang, B. Zhao, Y. Chen, Y. Zhou, Z. Guo, and M. Liu, Adv. Funct. Mater., 29, 1903638 (2019).
• 12)  W. Li, Q. Ma, P. Shen, Y. Zhou, L. Soule, Y. Li, Y. Wu, H. Zhang, and M. Liu, Nano Energy, 80, 105506 (2021).
• 13)  X. Zhao and V.-P. Lehto, Nanotechnology, 32, 042002 (2021).
• 14)  X. H. Liu, L. Zhong, S. Huang, S. X. Mao, T. Zhu, and J. Y. Huang, ACS Nano, 6, 1522 (2012).
• 15)  D. Uxa, B. Jerliu, E. Hüger, L. Dörrer, M. Horisberger, J. Stahn, and H. Schmidt, J. Phys. Chem. C, 123, 22027 (2019).
• 16)  B. Jerliu, E. Hüger, L. Dörrer, B. K. Seidlhofer, R. Steitz, V. Oberst, U. Geckle, M. Bruns, and H. Schmidt, J. Phys. Chem. C, 118, 9395 (2014).
• 17)  E. Buehler and J. H. Wernick, J. Cryst. Growth, 8, 324 (1971).
• 18)  A. D. Martinez, E. L. Warren, P. Gorai, K. A. Borup, D. Kuciauskas, P. C. Dippo, B. R. Ortiz, R. T. Macaluso, S. D. Nguyen, A. L. Greenaway, S. W. Boettcher, A. G. Norman, V. Stevanović, E. S. Toberer, and A. C. Tamboli, Energy Environ. Sci., 9, 1031 (2016).
• 19)  A. D. Martinez, E. M. Miller, A. G. Norman, R. R. Schnepf, N. Leick, C. Perkins, P. Stradins, E. S. Toberer, and A. C. Tamboli, J. Mater. Chem. C, 6, 2696 (2018).
• 20)  S. Boyanov, K. Annou, C. Villevieille, M. Pelosi, D. Zitoun, and L. Monconduit, Ionics, 14, 183 (2008).
• 21)  D. Yang, J. Zhu, X. Rui, H. Tan, R. Cai, H. E. Hoster, D. Y. W. Yu, H. H. Hng, and Q. Yan, ACS Appl. Mater. Interfaces, 5, 1093 (2013).
• 22)  J. Yang, Y. Zhang, C. Sun, H. Liu, L. Li, W. Si, W. Huang, Q. Yan, and X. Dong, Nano Res., 9, 612 (2016).
• 23)  Y. Lu, J. P. Tu, J. Y. Xiang, X. L. Wang, J. Zhang, Y. J. Mai, and S. X. Mao, J. Phys. Chem. C, 115, 23760 (2011).
• 24)  K. Yamaguchi, Y. Domi, H. Usui, A. Ueno, T. Komura, T. Nokami, T. Itoh, and H. Sakaguchi, Chem. Lett., 47, 1416 (2018).
• 25)  Q. Liu, C. Liu, Z. Li, Q. Liang, B. Zhu, J. Chai, X. Cheng, P. Zheng, Y. Zheng, and Z. Liu, ACS Appl. Energy Mater., 4, 11306 (2021).
• 26)  G.-A. Li, C.-Y. Wang, W.-C. Chang, and H.-Y. Tuan, ACS Nano, 10, 8632 (2016).
• 27)  K.-W. Tseng, S.-B. Huang, W.-C. Chang, and H.-Y. Tuan, Chem. Mater., 30, 4440 (2018).
• 28)  W. Li, P. Shen, L. Yang, A. Chen, J.-H. Wang, Y. Li, H. Chen, and M. Liu, Energy Environ. Sci., 14, 2394 (2021).
• 29)  G. Liu, L. Zhang, Y. Zhou, L. Soule, Y. Mu, W. Li, and Z. Shi, J. Mater. Chem. A, 9, 9124 (2021).
• 30)  T. Hirose, M. Morishita, H. Yoshitake, and T. Sakai, Solid State Ionics, 303, 154 (2017).
• 31)  T. Hirose, M. Morishita, H. Yoshitake, and T. Sakai, Solid State Ionics, 304, 1 (2017).
• 32)  T. Jaumann, J. Balach, U. Langklotz, V. Sauchuk, M. Fritsch, A. Michaelis, V. Teltevskij, D. Mikhailova, S. Oswald, M. Klose, G. Stephani, R. Hauser, J. Eckert, and L. Giebeler, Energy Storage Mater., 6, 26 (2017).
• 33)  Y. Li, Z. Cao, Y. Wang, L. Lv, J. Sun, W. Xiong, Q. Qu, and H. Zheng, ACS Energy Letters, 8, 4193 (2023).
• 34)  M. F. Hassan, Z. Guo, Z. Chen, and H. Liu, Mater. Res. Bull., 46, 858 (2011).
• 35)  C. Ban, Z. Wu, D. T. Gillaspie, L. Chen, Y. Yan, J. L. Blackburn, and A. C. Dillon, Adv. Mater., 22, E145 (2010).
• 36)  H. Wang, Q. Pan, J. Zhao, and W. Chen, J. Alloys Compd., 476, 408 (2009).
• 37)  C. Chen, N. Ding, L. Wang, Y. Yu, and I. Lieberwirth, J. Power Sources, 189, 552 (2009).
• 38)  Y. Yu, C.-H. Chen, J.-L. Shui, and S. Xie, Angew. Chem., Int. Ed., 44, 7085 (2005).
• 39)  J.-M. Tarascon, S. Grugeon, M. Morcrette, S. Laruelle, P. Rozier, and P. Poizot, C. R. Chim., 8, 9 (2005).
• 40)  S. Grugeon, S. Laruelle, L. Dupont, and J. M. Tarascon, Solid State Sci., 5, 895 (2003).