Shape evolution of parallelogrammic magnesium oxalate controlled by phosphate species

Abstract

Phosphate species are capable of playing a crucial role in manipulating the shapes and properties of inorganic materials. Herein, we examined the shape evolution of magnesium oxalate dihydrate (MgC₂O₄·2H₂O) to parallelogrammic microparticles with sheet-like structure in a precipitation process under the influence of phosphate species in the form of sodium tripolyphosphate (Na₅P₃O₁₀). Supported by a series of time-dependent experiments, the shape evolution of MgC₂O₄·2H₂O was evidenced by the complexation and blocking effect of Na₅P₃O₁₀ via a later adsorption at the surface of the formed MgC₂O₄·2H₂O product. The results from scanning electron microscopy (SEM) images, energy dispersive spectroscopy (EDS) and X-ray photoelectron spectrometry (XPS) measurements show that the shape evolution of parallelogrammic particles is accompanied by a modification in the chemical composition via a later adsorption. Our results demonstrated that the amount of Na₅P₃O₁₀ participating into the self-assembly of the layer-like parallelogram varied with reaction time. In the initial reaction stage, little amount of Na₅P₃O₁₀ is involved in the particle formation, and the participating amount increased with extent of reaction. This investigation presents the effect of Na₅P₃O₁₀ on the self-assembly of MgC₂O₄·2H₂O particles with layer-like structure and highlights its role in the controllable synthesis of microparticles.

---

1. Introduction

Over the past few decades, considerable attention has been attracted to the synthesis of inorganic materials, especially novel self-assembly of meso-, micro- and nano-structured building blocks with specific structures or interesting morphologies, due to their unusual properties and potential applications. As is well-known, magnesium oxalate dihydrate (MgC₂O₄·2H₂O) undergoes thermal decomposition at a relatively low temperature to yield magnesium oxide (MgO), as an exceptionally important material for use in catalysts, toxic waste remediation, and as additives in refractory, paint and superconductor products.^1–5 Because the physicochemical properties of MgO are significantly influenced by its morphology and structure,^3,6–8 different MgC₂O₄·2H₂O samples with a variety of morphologies, such as hollow nanospheres and nanorod-assembled hollow nanospheres,⁹ distorted rods,¹⁰ spherical and cubic agglomerates,^11,12 stacks of plates,¹³ polyhedron-like structure,¹⁴ wormhole-like porous structure,¹⁵ trapezoid-like and parallelogram-like structures,¹⁶ have been synthesized through different techniques including solid phase, sol–gel and precipitation methods. The formation basis of the above complex structures is achieved through a high degree of control over the nucleation, growth mechanisms and synthetic parameters. A useful approach to tune the shapes of the resulting products is to adjust the types of morphology regulators or ligands and their concentration so that certain crystallographic facets are favored during growth,^17,18 however, this strategy has rarely been used in adjusting MgC₂O₄·2H₂O morphology for synthesis of MgO.¹⁶

Phosphate species have demonstrated their unique capacities in tuning the shapes of various inorganic materials and even polymers.^19–21 For example, Ozaki et al. reported different spindle-type colloidal hematite particles with narrow size distribution in the presence of small amounts of phosphate or hypophosphite ions, which had a significant effect on the resulting particle shape.¹⁹ Zhang et al. synthesized spherical MgO from basic magnesium carbonate Mg₅(CO₃)₄(OH)₂·4H₂O in the presence of trace amount of phosphate species, and their influence on the shape of the resultant particles increased with increase in the polymerization of phosphate.²⁰ Mansourpanah and Amiri presented polyethersulfone membranes with more dense and compressed skin layer in the presence of Na₅P₃O₁₀ relative to that without, in which the former membranes demonstrated a favorable capability to separate bivalent anions from monovalent anions, as well as color from the feed.²¹ Inspired by these successful examples, more recently we obtained parallelogram-like MgC₂O₄·2H₂O microparticles with uniform size distribution by using Na₅P₃O₁₀ to tune the morphology. The resultant MgO, obtained by calcination of MgC₂O₄·2H₂O has demonstrated the characteristics of mesocrystals^22–27 and a superior catalytic performance in the Meerwein–Ponndorf–Verley reaction between benzaldehyde and ethanol than that obtained in the reaction system in the absence of Na₅P₃O₁₀.¹⁶ For this reason, it is interesting to systematically examine the influence of Na₅P₃O₁₀ on the shape evolution of MgC₂O₄·2H₂O.

In a liquid-phase synthesis, the resultant particles undergo different stages from nucleation over growth to ripening, which are directly bound up with the supersaturation of the reaction solution and its depletion. Herein, we present a deep insight into the formation of trapezoid-like MgC₂O₄·2H₂O and their shape evolution to parallelogrammic particles with sheet-like structure directed by Na₅P₃O₁₀. The resulting morphology of MgC₂O₄·2H₂O is highly dependent on the reaction time, the addition period of Na₅P₃O₁₀ and its involved amount in the resulting particles. From the current investigation, we can rationalize that the added Na₅P₃O₁₀ plays a major role in the shape evolution of parallelogrammic MgC₂O₄·2H₂O. SEM images and XPS and EDS measurements are presented to explain why trapezoid-like MgC₂O₄·2H₂O evolved into parallelogrammic ones upon the influence of Na₅P₃O₁₀. This study contributes to a better understanding of the shape evolution of MgC₂O₄·2H₂O directed by Na₅P₃O₁₀, and expand the knowledge on manipulation of morphology regulators such as phosphate species in designing the shape of inorganic materials for different applications.

2. Experimental

2.1 Synthesis of trapezoid-like MgC₂O₄·2H₂O

In a typical procedure, 7.69 g of Mg(NO₃)₂·6H₂O was dissolved into 50 mL of double-deionized water. Then, the Mg(NO₃)₂ solution was transferred to a 250 mL three-necked flask and heated to 323 K. Subsequently, 0.03 mol Na₂C₂O₄ was dissolved into 100 mL of double-deionized water, and then the Na₂C₂O₄ solution was also heated to 323 K. Under a vigorous stirring (ca. 800 rpm), the Na₂C₂O₄ solution was poured into the Mg(NO₃)₂ solution in 4–5 s. The mixture was further stirred for 1 min and then maintained at the temperature of 323 K under static conditions for different times. Then the white precipitate was collected at different reaction times, filtered off, and washed with double deionized water and absolute ethanol several times, respectively. Finally, the obtained product was dried at 343 K for several hours.

2.2 Synthesis of parallelogram-like MgC₂O₄·2H₂O

The major synthesis procedures for preparing the samples in the presence of Na₅P₃O₁₀ were similar to those described above except before reaction between Mg(NO₃)₂ and Na₂C₂O₄, 0.05 g of Na₅P₃O₁₀ was added into 100 mL of 0.03 mol Na₂C₂O₄ solution.

2.3 Synthesis of the products by adding 0.05 g of Na₅P₃O₁₀ into the reaction solution with different delay times

The preparation procedure was similar to the described above, and the only difference between them was the addition time of Na₅P₃O₁₀. Namely, after pouring Na₂C₂O₄ solution into Mg(NO₃)₂ solution, 2 mL aqueous solution containing 0.05 g Na₅P₃O₁₀ was added with various delay times ranging from 0 to 10 min into the above reaction mixture.

2.4 Synthesis of the products by adding 0.05 g of Na₅P₃O₁₀ into the reaction solution after a delay time of 2.5 min with different total reaction periods

Similar to the above preparation procedure, the product for the reaction time of 2.4 min was obtained by directly reacting the mixture of Na₂C₂O₄ and Mg(NO₃)₂ for 2.4 min. After that, a 2 mL aqueous solution containing 0.05 g Na₅P₃O₁₀ was added into the above reaction solution under a vigorous stirring (ca. 800 rpm) for 10 s followed by collecting the product with a total reaction time of 2.5 min. By a similar procedure, the products with a total reaction time of 3.5, 5, 10, 20 and 60 min were also prepared. Before each collection, the reaction solution was stirred for 10 s in order to obtain a uniform solution.

2.5 Characterization

The morphology, size and energy dispersive spectra (EDS) of the obtained particles were examined by a JEOL JSM-6390A scanning electron microscope (SEM). Chemical bonding states were investigated by X-ray photoelectron spectrometry (XPS, Kratos, UK). XPS spectra were obtained with monochromatic Al-Kα (1486.71 eV) radiation at a power of 150 W (10 mA, 15 kV). The charge neutralizer was used to compensate for surface charge effects, and binding energies were calibrated using the C 1s hydrocarbon peak at 284.8 eV. The crystal structures of as-synthesized products were characterized with X-ray diffraction (XRD) on a XRD-6000 diffractometer using Cu-Kα radiation. The operation voltage was 40 kV, and the current was 30 mA.

3. Results and discussion

3.1 Shape evolution of MgC₂O₄·2H₂O by Na₅P₃O₁₀

In an earlier study, we observed that the addition of trace amount of Na₅P₃O₁₀ can elaborately smooth a suture in the middle of spherical Mg₅(CO₃)₄(OH)₂·4H₂O by a “burst” mixture of Mg(NO₃)₂ and K₂CO₃ during precipitation.²⁰ The tuning effect of Na₅P₃O₁₀ over the shape of MgC₂O₄·2H₂O was also obtained more recently in the reaction process of Mg(NO₃)₂ and Na₂C₂O₄.¹⁶ To get a deep insight into the occurring shape evolution, we investigated the reaction process via a series of time-dependent experiments with involvement of equal amount of Na₅P₃O₁₀. In a typical synthesis, 0.05 g Na₅P₃O₁₀ was added into the reaction solution of Mg(NO₃)₂ and Na₂C₂O₄ followed by aging at 323 K for various periods.

In an experiment without Na₅P₃O₁₀, the shape evolution follows the reported mechanism where kinetic growth control leads to the resulting product with a trapezoid-like structure.¹⁴ Due to the slow growth rate after pouring Na₂C₂O₄ into Mg(NO₃)₂ solution, the reaction solution is almost clear and it is hard to obtain any product in the initial reaction stage. Therefore the product was collected from a reaction time of 2.5 min. As shown in Fig. S-1a,† the obtained product demonstrates a trapezoid-like structure, and further increasing the reaction time has little influence on the morphologies of the products (Fig. S-1b–f†). The only change with variation of reaction time is the increase in the particle sizes of the trapezoid-like particles. It is obvious in Fig. S-1g† that the sizes of the upper and lower base edges quickly increased in the first 10 min and then maintained almost constant with values of about 6 and 10 μm, respectively. These results reveal that over the course of the reaction without Na₅P₃O₁₀, the basic morphology of the resulting MgC₂O₄·2H₂O keeps constant from the initially collected particles to the final product aside from the increase in particle size.

As a comparison, in the presence of Na₅P₃O₁₀, the surface structure and particle size of MgC₂O₄·2H₂O obtained varied with the reaction time. Fig. 1 shows the shape evolution of the resultant product over the course of reaction. When the reaction time was 2 min, the collected product demonstrates a parallelogram-like structure, and there are a few sheet-like architectures piled up layer-by-layer along both sides of the diagonal line of as-synthesized particle (Fig. 1a). With the extension in the reaction time, the number of layers in both sides gradually increased (Fig. 1b–f). Finally, the multilayered structure facilitated the formation of a canyon along the diagonal line (Fig. 1g). From the whole reaction process, it can also be seen that the self-assembly of the layer-like structure mainly occurred in the first 10 min, and after that the number of piled layers upon the parallelogram-like structure increased slowly, which has a similar trend as the size of the diagonal line with reaction time (Fig. 1h). These results suggest that with the addition of Na₅P₃O₁₀ into the reaction system of Na₂C₂O₄ and Mg(NO₃)₂, the growth orientation of the resulting MgC₂O₄·2H₂O has been significantly altered probably due to complexation and blocking effects between Mg²⁺ and P₃O₁₀⁵⁻,^20,28–33 and the product favored to growth from its front view rather than side view (Fig. S-2†). More importantly, the obtained products are prone to be assembled into rough layer-like structures in the presence of Na₅P₃O₁₀, similar to the architectures of other magnesium salts.^20,34–38 After a careful examination of the surface structure of the obtained MgC₂O₄·2H₂O in the absence of Na₅P₃O₁₀ (Fig. S-2c†), it is obvious that its composition unit is also a fine sheet-like architecture. This information to some extent shows that the introduction of Na₅P₃O₁₀ would not make the crystal nuclei of MgC₂O₄·2H₂O form layer-like structure, but could prevent the self-assembly of the formed sheet-like particles into a fine surface structure, thus leading to the formation of rough layer-like particles.

Fig. 1 SEM images of the particles from various reaction times at 323 K by pouring Na₂C₂O₄ into Mg(NO₃)₂ solution in the presence of 0.05 g Na₅P₃O₁₀: (a) 2 min; (b) 3 min; (c) 4 min; (d) 5 min; (e) 10 min; (f) 20 min; (g) 60 min; (h) plot of the diagonal size of the obtained particles with variation of the reaction time, and inset is a typical particle obtained.

To investigate changes in the shape evolution of the resultant parallelogram-like structure produced by the introduced phosphate species, synchrotron XPS and EDS data of the obtained products prepared from different reaction stages in the presence of Na₅P₃O₁₀ were compared in Fig. 2. Apparently, there is a clear P 2p peak centered at 133.5 eV in the XPS data of the resulting product (Fig. 2a), with a shifted binding energy from the position for a pure Na₅P₃O₁₀ sample (132.9 eV). This shift could be ascribed to the interaction between the added Na₅P₃O₁₀ and the surface of MgC₂O₄. Quantitative analysis of the surface P, Mg, C and O elements shows that the content of P increased with the reaction time (Fig. 2b and c). When the reaction time was 2 min, the amount of P₃O₁₀⁵⁻ per MgC₂O₄ molecule is about 9.25 × 10⁻³. Further extending the reaction time led to a sharp increase in the number of P₃O₁₀⁵⁻ molecules, and this value became as large as 1.72 × 10⁻² for a reaction time of 5 min. After that the number of P₃O₁₀⁵⁻ steadily increased up to about 2.28 × 10⁻² per MgC₂O₄ molecule as the reaction proceeded to 60 min. The increase in the number of P₃O₁₀⁵⁻ molecules obtained from XPS data has a similar trend as that from the EDS analysis performed on single micrometer-sized particles obtained in different reaction stages (Fig. 2c). These results demonstrate that in the self-assembly of the sheet-like MgC₂O₄·2H₂O particles, the introduced Na₅P₃O₁₀ molecules is not uniformly distributed in the structure, but gradually increased with the reaction time. Namely, a small amount of Na₅P₃O₁₀ molecule participated into the reaction between Na₂C₂O₄ into Mg(NO₃)₂ in the initial reaction stage, while as the reaction proceeded, the content of Na₅P₃O₁₀ per MgC₂O₄ molecule increased. From Fig. 2c, careful observation shows that the amount of P₃O₁₀⁵⁻ molecules obtained from XPS data is much higher than that from the EDS measurements in the same reaction period. As is well known, the analysis depth with XPS is in the level of nanometers (around 2–10 nm), whereas using the EDS measurement the quantitative analysis is performed as deep as micrometers. These results suggest that in the growth of parallelogram-like MgC₂O₄·2H₂O, the introduced Na₅P₃O₁₀ more favorably adsorbs at the surface of the resultant particles, rather than being incorporated into the inner structure of the generated product.

Fig. 2 (a) XPS data of the region of P 2p from as-synthesized parallelogram-like MgC₂O₄·2H₂O (peak position: 133.5 eV) and commercially available Na₃P₃O₁₀ sample (peak position: 132.9 eV); (b) XPS data of the region of P 2p with variation of the reaction time on pouring Na₂C₂O₄ into Mg(NO₃)₂ solution in the presence of 0.05 g Na₅P₃O₁₀; (c) plots of the amount of P₃O₁₀⁵⁻ per MgC₂O₄ molecule from XPS and EDS measurements with reaction time. For the XPS analysis, the areas of P signals in (c) are normalized to the respective Mg 2p signals to derive the atomic ratios. For the EDS analysis, the atomic composition was collected from a single particle, and each analysis was carried out on three individual particles. The average of the obtained data was used to derive the respective atomic ratios. The amount of P₃O₁₀⁵⁻ per MgC₂O₄ molecule increased with increasing the reaction time for both XPS and EDS analyses.

Based on the changes in the amount of Na₅P₃O₁₀ molecules and the shape evolution of the obtained particles with reaction time (Fig. 1), a possible process for the growth of the parallelogram particles is speculated as below. In the initial reaction stage, the crystal nuclei of MgC₂O₄·2H₂O tended to quickly form parallelogram-like structure with few layers at the surface probably owing to the low amount of Na₅P₃O₁₀ involved. With the extension of the reaction, Na₅P₃O₁₀ favored to interact with the surface of the formed MgC₂O₄·2H₂O particles and disturbed their normal growth due to a complexation or blocking effect as discussed above. Owing to the effect of Na₅P₃O₁₀, the basic component unit of the resultant particle, namely the sheet-like nanoparticles assembled by the MgC₂O₄·2H₂O nuclei, could not produce trapezoid-like micro-sized particles with smooth surface (Fig. S-2a–c†), whereas they do facilitate growth itself. Along with the process, the amount of the surface adsorbed Na₅P₃O₁₀ would increase. On the one hand, Na₅P₃O₁₀ at the surface would block the active growth sites of the resultant particles,^28,29,31,33 and lead to growth orientation change. As a result, the layer-like structure grew from the front view of the resulting product rather than from the side view (Fig. S-2†). On the other hand, the adsorbed Na₅P₃O₁₀ would interact with the remaining Mg²⁺ in reaction solution due to its strong complexation ability,^20,30 hence causing the growth of another layer. As a result, particles were observed to pile up layer-by-layer with rough surfaces (Fig. 1).

3.2 Incorporation of Na₅P₃O₁₀ into the resultant particles via later adsorption

In order to obtain more convincing evidence on the shape evolution of the parallelogram-like MgC₂O₄·2H₂O upon treatment with Na₅P₃O₁₀, a series of time-dependent experiments by introducing Na₅P₃O₁₀ into the reaction system at various times were carried out. Fig. 3 shows the typical SEM images of the products from the reactions by adding 2 mL aqueous solution containing 0.05 g Na₅P₃O₁₀ into the reaction mixture of Na₂C₂O₄ and Mg(NO₃)₂ at different reaction times, with a total reaction time of 60 min. Surprisingly, when the Na₅P₃O₁₀ solution was added into the reaction solution in the period ranging from 0 to 2.5 min (Fig. 3a–e), the resultant products also demonstrated a parallelogram-like structure assembled by layer-like architectures from the front view. However, with delay in adding the Na₅P₃O₁₀ solution, the number of assembled layers gradually decreased. For Na₅P₃O₁₀ introduced into the reaction system after a reaction time of 5 min (Fig. 3f and g), the product was trapezoid-like, similar to the collected particles in the absence of Na₅P₃O₁₀ (Fig. S-1†). With the shape evolution of the resulting products, their particle size varied as well. As can be seen in Fig. 4a, the obtained particle size was around 13 μm when the delay time was in the range from 0 to 2.5 min. With the increase in the delay time to 5 min, the particle size demonstrated a decreasing trend with a value of ca. 8 μm, but further increasing the time up to 10 min resulted in an increase in the particle size to about 12 μm. These results illustrate that by adding the Na₅P₃O₁₀ solution into the reaction mixture of C₂O₄²⁻ and Mg²⁺ at the initial reaction stage, the resultant product can also be altered into a parallelogram structure assembled by layer-like architecture even with a delay time of 2.5 min. Due to the low solubility product constant of MgC₂O₄ (K_sp = 8.57 × 10⁻⁵), C₂O₄²⁻ and Mg²⁺ tend to nucleate quickly after mixing, suggesting that the nucleation of MgC₂O₄·2H₂O was not affected by the later addition of Na₅P₃O₁₀. In another words, the alteration in the shape of the resulting product could be attributed to the change in the growth orientation of MgC₂O₄·2H₂O nuclei upon the influence of Na₅P₃O₁₀. Because MgC₂O₄·2H₂O can also form a fine sheet-like structure without Na₅P₃O₁₀ (Fig. S-2†), the generation of parallelogram-like particles with a rough surface could be ascribed to the complexation or blocking effect of Na₅P₃O₁₀ via a later adsorption at the surface of the formed MgC₂O₄·2H₂O product. The earlier the added Na₅P₃O₁₀ into the reaction system, the more influence of the morphology was observed due to the smaller MgC₂O₄·2H₂O nuclei. Thus when Na₅P₃O₁₀ was added into the reaction solution in a shorter delay time, the adsorbed Na₅P₃O₁₀ would have a more pronounced effect on the morphology of the resulting product. With the growth of the formed particles, the effect of Na₅P₃O₁₀ gradually disappeared because of the larger sizes of the formed MgC₂O₄·2H₂O particles, and the product even returned back to the trapezoid-like structure like that without Na₅P₃O₁₀ (Fig. S-1†). To get a direct relationship between the shape evolution of the parallelogram-like particles and the content of Na₅P₃O₁₀, the products were examined with XPS and EDS. As shown in Fig. 4b, the amount of P₃O₁₀⁵⁻ per MgC₂O₄ molecule sharply decreased with the delay of introducing the Na₅P₃O₁₀ solution into the reaction mixture of Na₂C₂O₄ and Mg(NO₃)₂ in the range of 0–0.67 min, and further increasing the delay time (1.5–10 min) resulted in a gradual decreasing of P₃O₁₀⁵⁻ content, which is opposite with the trend for the products obtained in various aging periods (Fig. 2c). These results further indicate that the shape evolution of the resulting particles is closely related to the level of P₃O₁₀⁵⁻ participating in the reaction. Besides that, some obvious differences between XPS and EDS were observed. Namely, the number of P₃O₁₀⁵⁻ per MgC₂O₄ molecule is somewhat higher with EDS than that with XPS when Na₅P₃O₁₀ was introduced into the reaction system with a shorter delay time (0–1.5 min). With the further extension in the delay time (2.5–10 min), the number of P₃O₁₀⁵⁻ kept almost constant with a value of about 5.5 × 10⁻³ per MgC₂O₄ molecule determined by XPS, whereas its content still maintained a decreasing trend with the delay time for the EDS measurements. These results suggest that with a shorter delay time of adding Na₅P₃O₁₀ into the reaction, P₃O₁₀⁵⁻ more favorably interacts with MgC₂O₄·2H₂O and participates in the formation of the resulting particles. Therefore, the content of P₃O₁₀⁵⁻ in the structure of the resultant particles is higher than that at their surfaces, probably attributable to the smaller particles of MgC₂O₄·2H₂O formed in a shorter delay time as stated above. With extension in the delay time, the formed MgC₂O₄·2H₂O became larger, and the influence of P₃O₁₀⁵⁻ on the resultant particles became weaker, hence leading to the gradual formation of trapezoid-like rather than parallelogram-like product. On the other hand, P₃O₁₀⁵⁻ still could adsorb at the surface of the resulting particles due to its strong complexation effect with Mg²⁺.^20,28–33 Although a small amount of P₃O₁₀⁵⁻ involved in the particles has little effect on the morphology of MgC₂O₄·2H₂O, the crystal phase of the products is still impacted. As shown in Fig. 5, the XRD patterns of the particles obtained from different delay times show little difference, unlike the products obtained in the presence and absence of Na₅P₃O₁₀.¹⁶ These results reveal that by introducing Na₅P₃O₁₀ into the reaction between Na₂C₂O₄ into Mg(NO₃)₂ with a shorter delay time, P₃O₁₀⁵⁻ will have a great influence not only on the morphology of the obtained particles, but also on the crystal phase. By contrast, with the increase in delay time, the impact of P₃O₁₀⁵⁻ on the product is only demonstrated on the crystal phase, and little effect was observed on the resulting morphologies.

Fig. 3 Typical SEM images of the particles from the reaction systems by adding 2 mL aqueous solution containing 0.05 g Na₅P₃O₁₀ into the reaction solution of Na₂C₂O₄ and Mg(NO₃)₂ at various reaction periods followed by aging for a total reaction time of 60 min: (a) 0 min; (b) 0.33 min; (c) 0.67 min; (d) 1.5 min; (e) 2.5 min; (f) 5 min; (g) 10 min.

Fig. 4 (a) Change in the diagonal size of single particles from the reaction systems by adding 2 mL aqueous solution containing 0.05 g Na₅P₃O₁₀ into the reaction solution of Na₂C₂O₄ and Mg(NO₃)₂ at various reaction periods followed by aging for a total reaction time of 60 min; (b) plots of the amount of P₃O₁₀⁵⁻ per MgC₂O₄ molecule from XPS and EDS measurements with varying the addition periods of 2 mL aqueous solution containing 0.05 g Na₅P₃O₁₀ into the reaction solution of Na₂C₂O₄ and Mg(NO₃)₂. The calculations of the atomic ratios for XPS and EDS analyses were same as those in Fig. 2.

Fig. 5 Typical XRD patterns of the products from the reaction systems by adding 2 mL aqueous solution containing 0.05 g Na₅P₃O₁₀ into the reaction solution of Na₂C₂O₄ and Mg(NO₃)₂ at various reaction periods (0–10 min) followed by aging for a total reaction time of 60 min.

3.3 Shape evolution via later adsorption

To study the influence of P₃O₁₀⁵⁻ on the shape evolution of MgC₂O₄·2H₂O particles via later adsorption, the products were collected from different reaction stages by adding 2 mL aqueous solution containing 0.05 g Na₅P₃O₁₀ into the reaction between Na₂C₂O₄ into Mg(NO₃)₂ with a delay time of 2.5 min. For each collection, the reaction solution was stirred for 10 s in order to obtain a uniform solution. It should be pointed out here that adding Na₅P₃O₁₀ with 2.5 min delay time was based on the following reasons. First, with such a delay time parallelogram-like product still could be observed (Fig. 3e). Second, 2.5 min delay time just lies in the margin between parallelogram-like particles and trapezoid-like structures (Fig. 3), which may be more likely to observe intermediates from trapezoid-like to parallelogram-like structures. Fig. 6 shows the shape evolution of the products obtained with the above procedure. Before adding Na₅P₃O₁₀ into the reaction system, the product with a reaction time of 2.4 min was collected, which showed a typical trapezoid-like structure with a diagonal size of about 4.3 μm (Fig. 5a). When the reaction time was around 2.5 min, Na₅P₃O₁₀ was introduced into the reaction system under a vigorous stirring for 10 s. After that, the product was collected, which is demonstrated in Fig. 6b. As can be seen, the morphology is much different from that in the absence of Na₅P₃O₁₀. Specifically, the size from the side view became narrower, and the edges and corners from the front view became less sharp relative to that without Na₅P₃O₁₀. Also, the diagonal size diminished to around 3.7 μm, which revealed that upon the addition of Na₅P₃O₁₀, the reaction solution underwent a significant change. This phenomenon could be assigned to the strong complexation ability between P₃O₁₀⁵⁻ and Mg²⁺.^20,28–33 Owing to this effect, the introduced P₃O₁₀⁵⁻ would interact with Mg²⁺ in the reaction solution and at the surface of the formed MgC₂O₄·2H₂O. For the latter, P₃O₁₀⁵⁻ will not only adsorb at the surface of the produced MgC₂O₄·2H₂O, but also replace the C₂O₄²⁻ in the formed particles with production of soluble Mg²⁺–P₃O₁₀⁵⁻ complexes. Thereby, changes in the surface structure and particle size of the collected product were observed for the particles before and after adding Na₅P₃O₁₀. With extension in the reaction from 2.5 min to 5 min, the size of the obtained particles from the side view gradually increased as shown in the upper insets of Fig. 6b–d; and the surface from the front view steadily became coarse, via assembly of sheet-like structures, and the number of layers increased with reaction time (below insets in Fig. 6b–d). Further increasing the reaction time up to 10 min led to the generation of parallelogram-like structures with particle sizes in the range from 7.3 to 7.7 μm similar to that in Fig. 3, and there are more sheets at the surface of the obtained particles (insets in Fig. 6e). Apart from that, there are some sheet-like structures involved in the products as indicated with green arrows (Fig. 6e). A typical detailed surface structure of the sheet-like product is shown in Fig. 6f, from which it is clear that the sheet-like structures were self-assembled via an interlacing manner. More interesting is that some sheet-like structures can incorporate into the parallelogram-like structures (Fig. 6g), which might give us some hints on the formation of the parallelogram-like MgC₂O₄·2H₂O. To our knowledge, the generated sheet-like structures might arise from the self-assembly of the dissolved remaining Mg²⁺ and C₂O₄²⁻ upon the effect of the introduced P₃O₁₀⁵⁻. However, these particles were not observed for the product by introducing Na₅P₃O₁₀ into the reaction system between Na₂C₂O₄ and Mg(NO₃)₂ with a delay time of 2.5 min as shown in Fig. 3e. The difference might result from the vigorous stirring (ca. 800 rpm) before each collection, which disturbed their self-assembly orientation on the formed MgC₂O₄·2H₂O particles. To exhaust the remaining Mg²⁺ and C₂O₄²⁻ in solution, on the one hand a second nucleation event occurred with production of the interlaced sheet-like structures upon the effect of P₃O₁₀⁵⁻. On the other hand, the produced MgC₂O₄·2H₂O nuclei would deposit at the surface of bigger particles, thus leading to the generation of parallelogram-like structures with more sheets at their surface as shown in Fig. 6g. To get a direct relationship between the added Na₅P₃O₁₀ and the above structures, an EDS analysis was carried out on some typical single particles. The results demonstrated that a higher content of P₃O₁₀⁵⁻ per MgC₂O₄ was present in the sheet-like structure (0.0074 ± 0.0020) than that in the parallelogram-like particles (0.0056 ± 0.0006). These data to some extent suggest that Na₅P₃O₁₀ is the major formative agent for these sheet-like products (Fig. 6f) and the different layers at the surface of the parallelogram-like structures. When the reaction time was further extended to more than 20 min, the number of sheets at the surface of the generated particles gradually increased as shown in the insets of Fig. 6h and i. The quantitative analysis with XPS data (Fig. S-3†) reveals that the introduced Na₅P₃O₁₀ was responsible for such structures. During the shape evolution, careful observation also revealed that the edges and corners became more and more coarse with reaction time (insets in Fig. 6), and the sheet-like structures were gradually exposed. These results illustrate that with the extension of reaction, the later introduced Na₅P₃O₁₀ would be favourable to not only the self-assembly of sheet-like structures, but also dissolution of the surface structure of formed particles with generation of a coarser architecture.

Fig. 6 Typical SEM images of the collected products from the reaction systems (a) by direct reaction of Na₂C₂O₄ and Mg(NO₃)₂ for 2.4 min, and (b)–(i) by adding 2 mL aqueous solution containing 0.05 g Na₅P₃O₁₀ into the above reaction solution at different stages of (b) 2.5 min; (c) 3.5 min; (d) 5 min; (e)–(g) 10 min; (h) 20 min and (i) 60 min, followed by aging for a total reaction time of 60 min. Insets in (a)–(e), (h) and (i) are the typical particles from the side (upper) and front (below) views].

3.4 Formation mechanism of parallelogram MgC₂O₄·2H₂O

Based on the above shape evolution and structural analysis of the resultant products from the reaction between Na₂C₂O₄ and Mg(NO₃)₂ upon the effect of Na₅P₃O₁₀, the formation process of parallelogrammic MgC₂O₄·2H₂O structures is illustrated in Fig. 7. First, the precipitation of Mg²⁺ and C₂O₄²⁻ favored the nucleation and growth of micro-sized MgC₂O₄·2H₂O polyhedra. During such a process, the introduced small amount of Na₅P₃O₁₀ had little influence on the resulting product probably due to the high amount of Mg²⁺ and C₂O₄²⁻ in the initial reaction solution. Another reason could be attributed to the formation of soluble Mg²⁺–tripolyphosphate complex.³² As a result, little Na₅P₃O₁₀ participated into the nucleation and growth of MgC₂O₄·2H₂O (Fig. 2b and c), and the product demonstrated a trapezoid-like shape from the side view and a parallelogram-like structure with few sheet-like architectures piled up layer-by-layer along both sides of the diagonal line from the front view (Fig. 1a). With the exhaustion of Mg²⁺ and C₂O₄²⁻ in solution by production of MgC₂O₄·2H₂O, the soluble Mg²⁺–P₃O₁₀⁵⁻ complex would adsorb at the surface of the resulting product, followed by interaction with the remaining C₂O₄²⁻. Owing to the complexation and blocking effects between Mg²⁺ and P₃O₁₀⁵⁻,^20,28–33 the involved P₃O₁₀⁵⁻ would disturb the normal self-assembly and growth of MgC₂O₄·2H₂O. Specifically, in the reaction system in the absence of Na₅P₃O₁₀ the surface structure of the resultant product was composed of fine plate-like structures (Fig. S-2c†). When Na₅P₃O₁₀ was added into the reaction system, the self-assembly process of the fine plate-like units would be blocked, and their growth by themselves became dominant. Subsequently, the primary plate-like nanoparticles assembled into larger particles. The above process was repeated until Mg²⁺ and C₂O₄²⁻ in the reaction solution were depleted, and the collected product demonstrated sheet-like architectures piled up layer-by-layer along both sides of the diagonal line (Fig. 1a–g). According to the above scheme, it is easy to understand that with the extension of the reaction, the concentration of P₃O₁₀⁵⁻ gradually increased with the layers piled up along the sides of the diagonal line (Fig. 2b and c).

Fig. 7 Schematic illustration detailing all major steps and changes involved in the formation of parallelogrammic MgC₂O₄·2H₂O particles in the presence of trace amount of Na₅P₃O₁₀.

4. Conclusions

In this contribution we have shown that phosphate species such as Na₅P₃O₁₀ have a great impact on the shape evolution in the precipitation synthesis of parallelogram-like MgC₂O₄·2H₂O. In the presented preparation method, Na₅P₃O₁₀ influenced both the self-assembly and crystal phase of the reaction thus leading to the product with sheet-like structure, which experienced a gradual piling process of sheets at the surface of the generated particles. We found that such an additive was effective through a later adsorption to influence the morphological evolution of MgC₂O₄·2H₂O. Time-dependent experiments reveals that the later adsorption led to a complex process for generation of the final structures owing to the effect of Na₅P₃O₁₀, including dissolution of the formed MgC₂O₄·2H₂O and self-assembly of sheet-like units at the surface of formed particles. XPS and EDS data showed that the introduced Na₅P₃O₁₀ was more favourably adsorbed at the surface of the resulting particles than incorporated into the inner structures, and was also responsible for the oriented self-assembly of sheet-like units at the surface of formed product. We believe that the knowledge would give us an insight into better understanding the surface evolution of MgC₂O₄·2H₂O parallelograms and improved understanding of the properties of phosphate species in designing novel materials for catalysis and other fields.^16,20

Acknowledgements

This work was supported by the National Natural Science Foundation of China (No. 21205093), Key Science and Technology Program of Shaanxi Province of China (No. 2014K13-16), Special Research Project of Shaanxi Provincial Department of Education in China (2013JK0674), Xi’an Science and Technology Project in China (No. CXY1434(6)) and Graduate Innovation Foundation of Xi’an Shiyou University (No. 2014cx130737).

Notes and references

1. B. Eckhardt, E. Ortel, J. Polte, D. Bernsmeier, O. Gorke, P. Strasser and R. Kraehnert, Adv. Mater., 2012, 24, 3115.
2. Y. Lv, Z. Zhang, Y. Lai, J. Li and Y. Liu, CrystEngComm, 2011, 13, 3848.
3. K. Zhu, J. Hu, C. Kübel and R. Richards, Angew. Chem., Int. Ed., 2006, 45, 7277.
4. S. Makhluf, R. Dror, Y. Nitzan, Y. Abramovich, R. Jelinek and A. Gedanken, Adv. Funct. Mater., 2005, 15, 1708.
5. R. Richards, W. Li, S. Decker, C. Davidson, O. Koper, V. Zaikovski, A. Volodin, T. Rieker and K. J. Klabunde, J. Am. Chem. Soc., 2000, 122, 4921–4925.
6. N. Sutradhar, A. Sinhamahapatra, S. K. Pahari, P. Pal, H. C. Bajaj, I. Mukhopadhyay and A. B. Panda, J. Phys. Chem. C, 2011, 115, 12308.
7. J. V. Stark and K. J. Klabunde, Chem. Mater., 1996, 8, 1913.
8. K. J. Klabunde, J. Stark, O. Koper, C. Mohs, D. G. Park, S. Decker, Y. Jiang, I. Lagadic and D. Zhang, J. Phys. Chem., 1996, 100, 12142.
9. D. Jia, L. Zhang, Y. Cao, L. Liu and D. Xiao, J. Electroceram., 2008, 21, 218.
10. A. Kumar and J. Kumar, J. Phys. Chem. Solids, 2008, 69, 2764.
11. M. S. Mastuli, R. Rusdi, A. M. Mahat, N. Saat and N. Kamarulzaman, Adv. Mater. Res., 2012, 545, 137.
12. F. Mohandes, F. Davar and M. Salavati-Niasari, J. Phys. Chem. Solids, 2010, 71, 1623.
13. M. Sharma and P. Jeevanandam, J. Alloys Compd., 2011, 509, 7881.
14. W. Zhu, L. Zhang, G.-L. Tian, R. Wang, H. Zhang, X. Piao and Q. Zhang, CrystEngComm, 2014, 16, 308.
15. H.-B. Guan, P. Wang, H. Wang, B.-Y. Zhao, Y.-X. Zhu and Y.-C. Xie, Acta Phys.-Chim. Sin., 2006, 22, 804.
16. X. Zhang, Y. Zheng, H. Yang, Q. Wang and Z. Zhang, CrystEngComm, 2015, 17, 2642.
17. C. R. Bealing, W. J. Baumgardner, J. J. Choi, T. Hanrath and R. G. Hennig, ACS Nano, 2012, 6, 2118.
18. M. Meyns, F. Iacono, C. Palencia, J. Geweke, M. D. Coderch, U. E. A. Fittschen, J. M. Gallego, R. Otero, B. H. Juárez and C. Klinke, Chem. Mater., 2014, 26, 1813.
19. M. Ozaki, S. Kratohvil and E. Matijevic, J. Colloid Interface Sci., 1984, 102, 146.
20. Z. Zhang, Y. Zheng, J. Chen, Q. Zhang, Y. Ni and X. Liang, Adv. Funct. Mater., 2007, 17, 2447.
21. Y. Mansourpanah and Z. Amiri, Desalination, 2014, 335, 33.
22. R. Q. Song and H. Colfen, Adv. Mater., 2010, 22, 1301.
23. J. S. Chen, T. Zhu, C. M. Li and X. W. Lou, Angew. Chem., Int. Ed., 2011, 50, 650.
24. J. Fang, B. Ding and H. Gleiter, Chem. Soc. Rev., 2011, 40, 5347.
25. M. Xu, F. Wang, B. Ding, X. Song and J. Fang, RSC Adv., 2012, 2, 2240.
26. Z. Bian, T. Tachikawa, P. Zhang, M. Fujitsuka and T. Majima, Nat. Commun., 2014, 5, 3038.
27. X. Fei, W. Li, Z. Shao, S. Seeger, D. Zhao and X. Chen, J. Am. Chem. Soc., 2014, 136, 15781.
28. C. Li and Y. Lü, Int. J. Miner. Process., 1983, 10, 219.
29. A. Breeuwsma and J. Lyklema, J. Colloid Interface Sci., 1973, 43, 437.
30. F. Rashchi and J. A. Finch, Miner. Eng., 2000, 13, 1019.
31. P. Parsonage, D. Melven, A. F. Healy and D. Watson, Depressant function in flotation of calcite, apatite and dolomite, in Reagents in the minerals industry, ed. M. J. Jones and R. Oblatt, Institution of Mining and Metallurgy, London, 1984, ch. 294, pp. 33–45.
32. S. T. Liu, A. Hurwitz and G. H. Nancollas, J. Urol., 1982, 127, 351.
33. M. X. Zhao, H. Y. Yu, X. M. Wu and J. M. Ouyang, J. Jinan Univ., 2006, 27, 705.
34. Z. P. Zhang, Y. J. Zheng, Y. W. Ni, Z. M. Liu, J. P. Chen and X. M. Liang, J. Phys. Chem. B, 2006, 110, 12969.
35. Z. Zhang, Y. Zheng, J. Zhang, Q. Zhang, J. Chen, Z. Liu and X. Liang, Cryst. Growth Des., 2007, 7, 337.
36. C. Yan and D. Xue, J. Phys. Chem. B, 2005, 109, 12358.
37. K. Mitsuhashi, N. Tagami, K. Tanabe, T. Ohkubo, H. Sakai, M. Koishi and M. Abe, Langmuir, 2005, 21, 3659.
38. Q. Li, Y. Ding, G. Yu, C. Li, F. Li and Y. Qian, Solid State Commun., 2003, 125, 117.

---

Footnote

† Electronic supplementary information (ESI) available: SEM images of the particles from various reaction times at 323 K by adding Na₂C₂O₄ into Mg(NO₃)₂ solution, and SEM images of the particles obtained from the reaction between Na₂C₂O₄ and Mg(NO₃)₂ solutions in the presence or absence of 0.05 g Na₅P₃O₁₀ at 323 K, and plots of the amount of P₃O₁₀⁵⁻ per MgC₂O₄ molecule from XPS measurements by adding 2 mL aqueous solution containing 0.05 g Na₅P₃O₁₀ into the reaction solution of Na₂C₂O₄ and Mg(NO₃)₂ with a delay time of 2.5 min followed by various total reaction periods. See DOI: 10.1039/c5ra10807a

---