English

• Manganese is one of the transition metals and it has attached much attention due to the rich resources, low cost, environment-friendly and diversified nature. It usually exists in the valence states of +2 to +7, while the manganese oxides mainly appear in Mn state of +2, +3 and +4, such as MnO₂, Mn₂O₃, Mn₃O₄. Besides, it is common that different Mn states may exist in the oxide phase simultaneously, which provides the oxide with unique characteristics [1].

  Manganese dioxide (MnO₂), the relatively common oxide among them, displays outstanding performance in adsorption, electrochemistry, oxidation, and so on, while the properties of MnO₂ behave diversely attributing to the polymorphs in practical terms [2]. Attentively, there are numerous crystalline structures of MnO₂ and the crystallinity such as α-, β-, γ-, δ- and λ- seems to be explored frequently. As for the crystal diversity, it is originated from the stacking or linking variety of [MnO₆] units, and the structures could be divided into three types, including one-dimensional (1D) tunnel or chain framework (α-, β-, γ-), two-dimensional (2D) layer or sheet-like (δ-) and three-dimensional (3D) pores structure (λ-), thus the natural fabrication difference determines the various performance in all aspects [3, 4]. Musil et al. found that α-, β-, γ-, and δ-MnO₂ differed in surface appearance, especially the layered δ-MnO₂ [5]. Shafi et al. employed β-, γ-, δ- and λ-MnO₂ for the cyclic voltammetry determination, and the results showed that δ- and λ-MnO₂ displayed the superior electro catalysis behavior [6]. Li et al. applied α-, β- and δ-MnO₂ for the 1, 2, 4-trichlorobenzene removal, and the results manifested that δ-MnO₂ with the highest Mn³⁺/Mn⁴⁺ gained the superior efficiency [7]. Hence, there have been numerous researches on MnO₂ polymorphs, that is, synthesizing MnO₂ with different crystallinity based on the targeted properties and realizing the desired result.

  To synthesis of MnO₂ with different crystallography, several methods including thermal, reflux, hydrothermal, and sol-gel methods have been reported before [8]. With regard to the hydrothermal synthesis, some previous works pointed out that the K⁺ ions played the role of framework stabilizing and the crystallinity changed with the K⁺ content; meanwhile, K⁺ was closely related to the formation of oxygen vacancies (OVs) that further impacted on the production of reactive oxygen species (ROS) such as O₂^•− [9-12]. Wang et al. claimed that K⁺ amount affected the crystal formation, and layer structure would form under the higher K⁺ content while tunnel skeleton was fabricated with the lower K⁺ [13]. Cao et al. revealed that α- and δ-MnO₂ exhibited better electrocatalysis activities than λ-MnO₂ since higher K⁺ content was presented during the preparation [14]. Zhu et al. reported that the higher K⁺ content would slash the formation energy of oxygen vacancies, thus bringing better ozone elimination [10]. Therefore, it seems that K⁺ content is the critical factor for the characteristics of MnO₂ polymorphs in all aspects, such as morphology, electrochemistry, oxidizability. However, the intrinsic relationships among them remain ambiguous and further clarification is demanding to optimize the oxidation efficiency. To shed light on this, MnO₂ polymorphs including α-, β-, and δ-MnO₂ were prepared through the hydrothermal synthesis and investigations including morphology, electrochemistry and oxidizability were conducted. Furthermore, malachite green (MG), a kind of organic compound, was commonly utilized in dyestuff and pigment industries, however, it was proved to be harmful to human beings and other organisms [15, 16]. Therefore, MG was chosen as the target pollutant, which could not only evaluate the oxidation performance of x-MnO₂ but also propose efficient methods for the degradation of similar pollutants. Details of experiments were exhibited in Supporting information.

  It has been recognized that synthetic MnO₂ with different crystal structures was actually composed of basic [MnO₆] octahedral units, and the bonding motifs diversity resulted in the polymorphs of MnO₂, that is, the linking way was varied among the [MnO₆] units [17]. Obvious difference can be observed from the X-ray diffraction (XRD) peaks of different x-MnO₂ crystalline in Fig. 1a. The characteristic crystal facets of α-MnO₂ ((110), (200), (310), (211), (301), (411) and (521)), β-MnO₂ ((110), (101), (111), (211), (220), (310) and (112)) and δ-MnO₂ ((001), (002), (200) and (020)) were coincided with XRD patterns of JCPDS No. 44–0141 (cryptomelane), JCPDS No. 24–0735 (pyrolusite) and JCPDS No. 80–1098 (birnessite), respectively [13, 18]. The crystalline difference was the reflection of K⁺ concentration discrepancy in the preparation, since K⁺ played the important role of cation template and stabilizer, which could not only broaden the crystal cells but also impact on the balance of charge [10, 18, 19]. In short, δ-MnO₂ with 2D layer structure would be fabricated in the process of chemical reaction and precipitation and acted as the framework precursor, and the sequent hydrothermal was the critical procedure for the crystallinity transformation, since MnO₂ crystal structure would curl under the elevated temperature and pressure, thus MnO₂ with 1D structure was formed in this plication [2, 13]. Consequently, as shown in Fig. S1 (Supporting information), δ-MnO₂ structure with abundant K⁺ in the interlayer kept the skeleton stable and would be hardly affected by the hydrothermal, thus gaining the most expansive space (0.70 nm) for ions or molecules traversing, such as H₂O or OH⁻ species [5, 7]. As for α-MnO₂, crystal structure with limited K⁺ content could not completely resist the force brought by high temperature and pressure. Therefore, it curled to the structure of [2 × 2] 1D tunnels and occupied the size of 0.46 nm × 0.46 nm, which possessed the edge-sharing [MnO₆] octahedra through the double-chain connection. However, β-MnO₂ framework without extra K⁺ addition would collapse and transform to the cramped 1D [1 × 1] tunnels during the hydrothermal process, and the tunnel size was further narrowed to 0.23 nm × 0.23 nm as a result of the dense oxygen atom array in the forms of twisted hexagon [20, 21].

  Figure 1

  

  Figure 1.  Physiochemical properties analysis of x-MnO₂ and the application for MG degradation: (a) XRD patterns; (b–d) SEM images; (e) LSV and (f) EIS measurement; (g, h) batch experiments for MG degradation comparison. (MnO₂ dosage (g): 1–5 mg (pH 4.9 ± 0.1); pH (h): 3–7 (dosage = 3 mg)).

  DownLoad: Full-Size Img PowerPoint

  Based on the discrepancy of crystallinity, characterization instruments such as scanning electron microscope (SEM), transmission electron microscopy (TEM) and Brunauer-Emmett-Teller (BET) were applied for the observation of morphology difference of x-MnO₂. Apart from this, it was accepted that the K⁺ could modulate the ions distribution in crystal cells, which may further have a significant impact on the oxidant's properties such as electrochemistry and oxidizability. Hence, linear sweep voltammetry (LSV) and electrochemical impedance spectroscopy (EIS) were employed for physiochemical property analysis, while MG removal efficiency was utilized for the oxidation capacity comparison of x-MnO₂. All the experiments above were conducted for expounding the effect of crystallinity discrepancy on the expression of morphology, electrochemistry and oxidizability, thus grasping the mutual influence factors among the above characteristics overall.

  As shown in Figs. 1b–d, it was obvious that x-MnO₂ exhibited differently in surface appearance. It could be observed that β-MnO₂ was rod-shaped (Fig. 1c) while α-MnO₂ was more fluffy and filamentous-like, bringing the larger surface area (Fig. 1b). TEM images (Fig. S2 in Supporting information) showed that β-MnO₂ displayed higher aspect ratio at the same multiple than α-MnO₂, which was in line with SEM images. Completely different from α- and β-MnO₂, morphology of δ-MnO₂ was similar to the broccoli and it seemed to be composed of numerous micro spheres (Fig. 1d), and TEM images showed that it was lamellar microscopically. Correspondingly, it was shown in Table 1 and Fig. S3 (Supporting information) that δ-MnO₂ possessed the highest specific surface area (SSA) while β-MnO₂ performed the least, as well as the pore volume (Pv). Although the average pore diameter (Pd) of x-MnO₂ followed the order of β- > α- > δ-MnO₂, mesopore (2–50 nm) was the dominant structure in all of them (Fig. S3b). It could be acquainted that K₂CO₃ was involved in both α- and δ-MnO₂, which may be the core of discrepancy above, since it had been reported that K₂CO₃ would be transferred to KHCO₃ through the reaction with H₂O and CO₂ in the autoclave, and then it would decompose quickly between 100 ℃ and 200 ℃ [22]. Therefore, CO₂ gas would be generated via KHCO₃ decomposition in the hydrothermal preparation of α- and δ-MnO₂ under the temperature of 180 ℃ and super-atmospheric pressure, which facilitated the formation of abundant pores and made the surface more uneven. Besides, the additional metal cations of K⁺ of α- and δ-MnO₂ could incorporate into the structure and improve the surface area remarkably, resulting in the better morphology of them [23]. In addition, the extra addition of KOH in δ-MnO₂ preparation made the solution more alkaline (5 g KOH/60 mL H₂O), thus the alkalinous-hydrothermal process could stimulate the structure splitting and enlarge the surface area, explaining why δ-MnO₂ was the superlative in surface and pores structure [24]. Consequently, δ- and α-MnO₂ performed much better in SSA and Pv than β-MnO₂.

  Table 1

  

  Table 1.  Morphology characterization and elements analysis of x-MnO₂.

  DownLoad: CSV

  Since electrochemistry properties could also reflect the oxidation–reduction reactivity (ORR) to a certain extend, which may provide proper methods for the improvement of organic pollutants degradation, thus LSV and EIS measurements were conducted based on this. Fig. 1e exhibited the LSV measurement of x-MnO₂, and it could be accepted that onset potential of α-, β- and δ-MnO₂ was 0.98, 0.80 and 0.99 V vs. SCE while the limiting current density was 2.69, 2.08 and 3.11 mA/cm², respectively, suggesting that δ-MnO₂ exhibited the highest ORR kinetics while β-MnO₂ did the worst [25]. On the basis of the phenomena above, Tafel slope was drawn into the further analysis, and the results were shown in Fig. S4 (Supporting information). It was apparent that the slopes of α-, β- and δ-MnO₂ were 206, 224 and 194 mV/dec, respectively, thus δ-MnO₂ held the optimum catalytic activity due to the lowest slope [26]. Integrating all the substances above, it could make the conclusion that the ORR activity obeyed the order of δ- > α- > β-MnO₂, which may be due to the special layered structure of δ-MnO₂, thus promoting the capacity of ions traversing and reaction activity [27]. In addition, charge transfer numbers (n), which could describe the ORR pathway and reflect the activity, were calculated according to the equation below (Eq. 1) [28, 29].

  (1)

  where I_d represented the current at the disk while I_r was that at the ring in rotating ring-disk electrode (RRDE) measurement, N was the collection efficiency of ring and disk electrodes (0.37). As exhibited in Fig. 1e, charge transfer numbers of α-, β- and δ-MnO₂ were 3.93, 3.79 and 3.91, respectively, insinuating that ORR pathway of x-MnO₂ adhered to the 4e transferring, however, the electrocatalytic activity of β-MnO₂ was less prominent than that of α- and δ-MnO₂, which was also verified in LSV and Tafel slop results above. For the fact that K⁺ was conducive for the charge balance with relatively higher electron transfer mobility and favorable for fast ions diffusion, therefore, it was not surprising that δ-MnO₂ with the highest K⁺ content displayed the most excellent electrochemical performance exhibited above [30].

  In the light of the results above, EIS was adopted for the subsequent inquiry, and the Nyquist spectrums were exhibited in Fig. 1f. It could be observed that δ-MnO₂ got the lowest semi-circle diameter while that of β-MnO₂ was the highest, suggesting that the charge transfer resistance (R_CT) between the electrode and electrolyte interface complied with the order of δ- > α- > β-MnO₂, and the results were accordant with that shown in LSV determination. For the more intuitive comparison, ZsimpWin software was applied for the experimental fitting, and the results showed that the electrolyte resistance in solution (R_S) of α-, β- and δ-MnO₂ were similar, locating in 6.6, 7.1 and 6.9 Ω around, respectively. However, R_CT values were quite different among them according to the fitting results in Fig. S5 (Supporting information). δ-MnO₂ held the lowest R_CT with 626 Ω and that of α-MnO₂ increased obviously (5986 Ω), yet β-MnO₂ possessed the highest R_CT, much higher than that of both δ- and α-MnO₂ (1.21 × 10⁴ Ω). Therefore, it was much more propitious for δ-MnO₂ to electron transfer, which was possibly benefited from the superior 2D-sheet structure with K⁺ filling in the interlayer, thus providing the higher surface area as well as allowing the electrolyte species to get through quickly [6].

  For the direct comparison of x-MnO₂ performance diversity, MG removal experiments were proceeded. It could be observed that δ-MnO₂ performed the most satisfied MG removal under the different dosage, and the dosing of δ-MnO₂ in the experimental range affected the performance slightly because MG removal efficiency was always over 95.00% (from 95.53% to 99.75%) (Fig. 1g). As for α-MnO₂, MG removal efficacy increased from 74.27% to 96.42% with the dosage increasing. However, it seemed that β-MnO₂ had inferior ability to degraded MG, since MG removal efficiency maintained subtle changes with the dosage increasing (from 38.76% to 46.58%). For the fact that MG was removed through the MnO₂ solids in the aqueous environment, thus the reaction was actually supposed to be the heterogeneous process. Therefore, pH was an essential factor for the above system since it was not only related to surface charge but also may affect the oxidation of MnO₂, since it had been confirmed that MnO₂ displayed better oxidation capacity in the acid environment [31, 32]. In order to verify whether MG removal performance of x-MnO₂ would be restrained by the improper pH, batch experiments were conducted under the pH range of 3~7 for further investigation. It should be noticed that initial pH of the original MG solution was 4.9 ± 0.1, thus pH 5 represented the stock solution. Furthermore, pHpzc of x-MnO₂ was determined for the verification of the hypothesis that whether the surface repulsion would hold back the heterogeneous reaction, since MG was cationic and existed in the forms of positive charge [33]. As shown in Fig. 1h and Fig. S6 (Supporting information), it could be concluded that MG removal of β-MnO₂ was always poorer than that of α- and δ-MnO₂ although the reaction was conducted under the pH that was favorable for β-MnO₂ but detrimental to α- and δ-MnO₂, implying that surface contacting interaction was not the resistance for the lower MG removal of β-MnO₂. Therefore, the conjecture could be proposed that oxidation capacity or some oxygen species related to MG removal of β-MnO₂ was much lower than that of α- and δ-MnO₂, thus β-MnO₂ always displayed the inferior MG removal capacity under the favorable conditions even.

  Though MnO₂ was able to remove organic contaminants via the common oxidation, it had been also reported that MnO₂ could degrade contaminants through the radical or non-radical pathway, such as ^•OH, ¹O₂, and O₂^•− [34, 35]. Therefore, quenching experiments were conducted based on the hypothesis that some oxidative radicals and non-radicals were generated for boosting the degradation of MG. Hence, ethanol (EtOH), furfuryl alcohol (FFA) and p-benzoquinone (BQ) were used to capture the ^•OH, ¹O₂ and O₂^•−, respectively, with the reaction constants of k(^•OH) = 1.9 × 10⁹, k(¹O₂) = 1.2 × 10⁸ and k(O₂^•−) = 9.8 × 10⁸ mol L⁻¹ s⁻¹, respectively [36, 37]. All the molar ratio of x-MnO₂ to scavengers above was set as 1:20, and the results were shown in Fig. 2. It could be noted that MG removal was significantly impeded with the addition of BQ, implying that O₂^•− was momentous for all of x-MnO₂ and played a notable role. However, MG removal was hardly affected when the adequate EtOH existed, and the addition of FFA decreased the performance of α- and δ-MnO₂ to some extent while it had a negligible impact on β-MnO₂. In order to examine whether the results above were caused by the insufficiency of EtOH and FFA, the ratio of x-MnO₂ to EtOH or FFA was reset to 1:50, and the results were furnished in Fig. S7 (Supporting information). With the increase of EtOH, MG removal efficacy of all x-MnO₂ was still in line with the control groups, showing that there may not have any ^•OH existing in the systems, since ^•OH was the nonselective radical species and it would react with most organic contaminants. As for FFA, the elevated dosage further slashed MG removal in α- and δ-MnO₂ while the effect was not obvious in β-MnO₂ yet, demonstrating that ¹O₂ concentration was so scarce. Hence, it could be comprehended why β-MnO₂ did worst in MG removal, since ROS in β- was not affluent enough compared with α- and δ-MnO₂. Besides, it should be noted that ROS could not obliterate the MG removal performance thoroughly, and it had been also reported that MnO₂ was capable of removing some organic contaminants through the general redox reaction, thus the sequent experiments were conducted based on the hypothesis above.

  Figure 2

  

  Figure 2.  Reactive oxygen species detection. (dosage of MnO₂ = 3 mg, pH 4.9 ± 0.1; mole ratio of MnO₂ to scavengers = 1:20).

  DownLoad: Full-Size Img PowerPoint

  It had been mentioned that several Mn states may simultaneously exist in MnO₂ while Mn³⁺ and Mn⁴⁺ were usually considered to be the important states species [38, 39]). Therefore, X-ray photoelectron spectroscopy (XPS) was employed for the subsequent research. Attentively, peaks at 641.3 eV and 643.0 eV represented Mn³⁺ and Mn⁴⁺ in Mn 2p, respectively; as for O 1s images, peaks of lattice oxygen (O_latt), adsorbed oxygen (O_ads) and H₂O were determined at 529.8 eV, 530.4 eV, and 531.9 eV, respectively [40-44]. It could be cognized in Figs. 3a–c that the content of Mn³⁺ in α-, β- and δ-MnO₂ was 53.3%, 47.6%, and 54.5%, respectively. It had been reported that Mn³⁺ could be generated at a certain molar ratio of Mn²⁺/KMnO₄ in the preparation of MnO₂, and the oxygen vacancies formation was companied with the simultaneous process of Mn⁴⁺ reduction, suggesting that there would be some trivalent Mn existing in MnO₂ [45, 46]. Hence, it could be observed that there was a large amount of Mn³⁺ appearing in all of x-MnO₂. In addition, it had been mentioned in the previous paragraphs that there were some K⁺ ions filling in the tunnels or interlayer of α- and δ-MnO₂, therefore, more Mn³⁺ would be formed for the charge balance of the foreign ions [2]. Therefore, the discrepancy of K/Mn ratio was another reason why Mn³⁺ content adhered the order of δ- > α- > β-MnO₂. In addition, it could be noticed that the order of Mn³⁺ content was in line with MG removal efficiency, suggesting that the role of Mn³⁺ was of great concern. It had been reported that O₂ could be adsorbed on MnO₂ surface, and then e_g¹ electron of Mn³⁺ could be transferred to the adsorptive O₂, and then O₂^•− was generated [7, 17]. As a result, the higher Mn³⁺ species content existed, the more prosperous O₂^•− would be produced, thus accelerating MG degradation. Hence, δ-MnO₂ with the most affluent Mn³⁺ performed best. Besides, O_ads was another important factor since it could be utilized for O₂^•− generation. Therefore, it could be discovered in Figs. 3d–f that the ratio of O_ads to O_latt followed the order of δ- > α- > β-MnO₂, manifesting that δ-MnO₂ possessed the most outstanding capacity for surface oxygen sorption, which was also in accordance with both Mn³⁺ content and MG removing. As for ¹O₂, it was probably formed through the way that O₂^•− reacted with itself under the existence of H⁺ [47, 48]. In consequence, α- and δ-MnO₂ with relatively ample O₂^•− could produce ¹O₂ via the reaction above while it may be limited in β-MnO₂, since the inadequate O₂^•− would be consumed by MG and there was too little O₂^•− left to generate ¹O₂ further, thus ¹O₂ mass in β-MnO₂ was unsatisfied.

  Figure 3

  

  Figure 3.  XPS analysis of MnO₂: Mn 2p (a–c) and O 1s (d–f).

  DownLoad: Full-Size Img PowerPoint

  In fact, the generation of O₂^•− was also controlled by the oxygen vacancies (OVs) since OVs could capture the adsorbed O₂ and motivated it to be converted to O₂^•− when it got close to the subsurface oxygen defect [49]. Therefore, the content of OVs was another essential factor for oxidation capacity of MnO₂. With regard to the formation of OVs, it was companied with the loss of O_latt, that is, O atom of Mn(Ⅳ)-O⁻²-Mn(Ⅳ) escaped from the framework and then the new structure of Mn(Ⅲ)-□-Mn(Ⅲ) formed for the charge balance, consequently, the oxygen vacancy (□) generated while the state of Mn decreased [50-52]. Hence, the higher Mn³⁺ content suggested the more presentation of OVs, which further resulted in a higher contaminants removal capacity. Moreover, the ratio of Mn/O for x-MnO₂, indicating the potential of O absconding and OVs emergence, was displayed in Table 1. It could be realized that the Mn/O of α-, β- and δ-MnO₂ was 0.484, 0.434 and 0.492, respectively, which was abided by the OVs production order of δ- > α- > β-MnO₂ and coincided with the MG removal efficacy. Moreover, it should be paid significant attention that K⁺ had a potential for cutting down OVs formation energy (E_OV) since K⁺ had the ability to lead to the electrostatic interaction between K⁺ and O_latt, thus boosting the generation of OVs [10, 53]). Therefore, OVs were easier to be produced in δ-MnO₂ with the most abundant K⁺ mass and MG could be degraded more excellently in this paper. Furthermore, it had been reported that OVs on MnO₂ surface were pivotal in ¹O₂ formation, since it favored capturing O₂ and promoting the mobility, thus facilitating the generation of O₂^•− and the following conversion of ¹O₂ [54]. Hence, α- and δ-MnO₂ with more OVs had abundant sites for ¹O₂ fabrication, which was in accordance with the experiments of both MG removal and ROS quenching.

  Otherwise, TGA determination, which reflected the surface oxygen species and manganese oxides transformation, was applied for the thermal analysis of x-MnO₂, conducting under the determined temperature of 30–1000 ℃ and with the temperature rising rate of 10 K/min. It was revealed in Fig. 4 that the decomposition of x-MnO₂ could be generally divided into four steps as the temperature increased. In step 1 (< 250 ℃), the mass loss of α-, β- and δ-MnO₂ was 4.59%, 1.10% and 9.93%, respectively, which was due to the deprivation of physically adsorbed or interlayered water, and the least loss of β-MnO₂ may be attributed to less water entrapped in the parochial tunnel ([1 × 1]) [55]. As the temperature rose (step 2, 250–550 ℃), mass loss of α- and δ-MnO₂ were 1.63% and 3.37%, respectively, while that of β-MnO₂ was only 0.53%, signifying that surface active-oxygen disappeared. The total loss of α-, β- and δ-MnO₂ below 550 ℃ followed the order of δ- > α- > β-MnO₂, implying that δ-MnO₂ owned the amplest surface oxygen species, which may be conducive to the generation of O₂^•−. In step 3 (550–800 ℃), the conspicuous loss of β-MnO₂ (4.16%) demonstrated that MnO₂ crystal missed O atom and then transformed into Mn₂O₃, however, α- (0.35%) and δ-MnO₂ (0.00% nearly) were relatively stable and maintained the original components. When the temperature was higher than 800 ℃ (step 4), the mass loss was due to the further decomposition from Mn₂O₃ to Mn₃O₄. It could be discovered that the loss of δ-MnO₂ was the lowest, since abundant K⁺ ions in the layers held back the transformation of Mn₂O₃ [56]. In general, both α- and δ-MnO₂ possessed affluent surface oxygen species while there was little in β-MnO₂, which may be another explanation for MG removal behavior, since more O₂^•− could be generated on α- and δ-MnO₂, thus improving the higher oxidative capability.

  Figure 4

  

  Figure 4.  Thermogravimetric analysis of MnO₂.

  DownLoad: Full-Size Img PowerPoint

  Except for the reactive factors above, Mn³⁺ itself could be also taken into consideration for MG degradation, since Mn³⁺ could receive an electron through the empty б orbit coordination with MG without changing the Mn spin state [57]. As for Mn⁴⁺, the process above could be realized through the outer sphere electron transferring and the spin state was required to be transformed, thus it was more difficult compared with Mn³⁺; meanwhile, Mn³⁺ possessed four 3d orbital electrons and the anti-bonding electron (e_g¹) would bring about the longer (Jahn-Teller distortion), leading to the weaker Mn−O bond and the higher ORR activity compared with Mn⁴⁺ [58, 59]). Therefore, it could be better understood that superior MG removal was realized under the higher Mn³⁺ rather than Mn⁴⁺. Furthermore, it had been mentioned above that Mn³⁺ and OVs were generated simultaneously and the higher K⁺ content was conducive to the richer OVs production, thus generating more Mn³⁺ content. Hence, δ-MnO₂ with the amplest K⁺ reached the best MG removal owing to the highest Mn³⁺ content while β-MnO₂ did worst. Moreover, MnO₂ would react with H₂O and then turned into the active site (MnOOH) through accepting an electron from H₂O, thus MnOOH may be another trigger for MG removal and the ORR kinetics was probably controlled by the MnOOH oxidation [14, 60].

  According to all the results above, it could be considered that the performance diversity was related to K⁺ content and it seemed that there was some connection among the properties. Therefore, this part was conducted for the thorough expounding of the discrepancy and intrinsic relationship (Fig. 5). Firstly, it could be accepted that crystallinity diversity of α-, β-, and δ-MnO₂ was attributed to the K⁺ content, since K⁺ ions would act as the template and stabilizer, which was useful for tightening up the 2D layer structure of δ-MnO₂ and allowed more ions to pass through the interlayer. However, the limited or the absent K⁺ in α- and β-MnO₂ could not prevent the skeleton from curling and collapsing to the 1D tunnel structure with [2 × 2] or [1 × 1], respectively, and the capacity of ions containing in the tunnel decreased compared with δ-MnO₂. Consequently, electrochemical properties interrelating with ORR kinetics analyzed via LSV and EIS exhibited that δ-MnO₂ gained the most prominent performance, coinciding with the tunnel size or layer distance characteristic above. Secondly, it could be observed that the overall MG removal performance adhered to the order of δ- > α- > β-MnO₂, which kept pace with that of ORR activity behavior suggested in LSV and EIS examination. According to the incisive analysis of MG removal in previous paragraphs, it could be acquainted that MG removal was mainly profited from the oxidative attacking via ROS as well as the electrons transferring between Mn³⁺ and MG, thus δ-MnO₂ with the highest ORR activity favored the interaction with MG and realized the more tremendous removal. Attentively, it had been mentioned that K⁺ was conducive to the E_OV decreasing, thus δ-MnO₂ with the highest K⁺ content possessed the most fertile Mn³⁺ and OVs and degraded MG in the most conspicuous capacity, since Mn³⁺ and OVs were important sources for the generation of O₂^•− and ¹O₂, directly oxidizing MG and achieving the rapid degradation. Meanwhile, it had also been mentioned that OVs could facilitate the electrolytic ions diffusion and charges transferring, thus δ-MnO₂ with the richest OVs displayed the best electrocatalysis and the lowest resistance in LSV and EIS [61]. Hence, it could be proposed that the higher K⁺ stimulated the more OVs production, which optimized both electrochemistry and oxidizability of MnO₂, thus δ-MnO₂ gained the best performance. Moreover, it should be admitted that adsorptive O₂ was another essence for MG removal. Therefore, α- and δ-MnO₂ with better morphology could provide more active sites for the more excellent O₂ adsorption capability and MG degradation, which was as a result of the ample K⁺ addition and exhibited in SEM images and S_BET determination.

  Figure 5

  

  Figure 5.  Intrinsic relationship among the characteristics of MnO₂ polymorphs.

  DownLoad: Full-Size Img PowerPoint

  Furthermore, it had been mentioned that crystallinity was connected with ORR activity, and it could be discovered that facets of x-MnO₂ slackened after reaction (Fig. 1a), confirming that they were participated in MG removal process [62]. The facets of (001) and (002) in δ-MnO₂ decreased remarkably since they had a bearing on the formation of OVs, and the more consumption of (001) was a result of the higher activity [27]. As for α-MnO₂, it could be caught sight of the slight abating of the (110), (200) and (211) facets, signifying that there were utilized in the reaction due to the reactive activity. However, the lessening of facets in β-MnO₂ was less visible than α- and δ- even though the relatively active facet of (110) decreased slightly, which may be due to the finite activity [62]. Generally, the crystal facets cutting variation above not only reflected the MG removal performance in essence, but also manifested the crystallinity activity diversification resulted from K⁺ content, which directionally provided the pathway of ORR activity promoting.

  In conclusion, MG degradation capacity adhered to the order of δ- > α- > β-MnO₂, according with K⁺ content diversity among them, since it could whittle down the formation energy of OVs and enhance the production of both Mn³⁺ and OVs, thus facilitating the sequent ROS generation and MG degradation. In addition, electrochemistry and oxidizability were turned out to be positively correlated due to the reflection of OVs content and ORR activity; meanwhile, the brilliant morphology provided more active sites for O₂ adsorption, indirectly promoting the oxidation capacity. Therefore, δ-MnO₂ with the highest K⁺ content was the most ORR active and gained the richest active-sites as well as OVs, thus performing the optimum in electrochemistry and oxidizability.

  Declaration of competing interest

  The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

  Acknowledgments

  This work was financially supported by the Hubei Provincial Key Lab of Water System Science for Sponge City Construction (No. 2019–06) and the Fundamental Research Funds for the Central Universities (No. 215206002).

  Supplementary materials

  Supplementary material associated with this article can be found, in the online version, at doi:10.1016/j.cclet.2022.01.082.

  
  
• 1. [1]

     S.G. Ma, X. Ye, X. Jiang, et al., J. Alloy. Compd. 852 (2021) 157007 doi: 10.1016/j.jallcom.2020.157007

  2. [2]

     Q. Feng, H. Kanoh, K. Ooi, J. Mater. Chem. 9 (1999) 319–333 doi: 10.1039/a805369c

  3. [3]

     Y. Yuan, C. Liu, B.W. Byles, et al., Joule 3 (2019) 471–484 doi: 10.1016/j.joule.2018.10.026

  4. [4]

     J. Shin, J.K. Seo, R. Yaylian, et al., Int. Mater. Rev. 65 (2020) 356–387 doi: 10.1080/09506608.2019.1653520

  5. [5]

     M. Musil, B. Choi, A. Tsutsumi, et al., J. Electrochem. Soc. 162 (2015) 2058–2065 doi: 10.1149/2.0201510jes

  6. [6]

     P.M. Shafi, C. Johnson, A.C. Bose, AIP Conf. Proc. 1942 (2018) 050069 doi: 10.1063/1.5028700

  7. [7]

     Q.Q. Li, X.C. Huang, G.J. Su, et al., Environ. Sci. Technol. 52 (2018) 13351–13360 doi: 10.1021/acs.est.8b03789

  8. [8]

     S.N. Guan, W.Z. Li, J.R. Ma, et al., J. Ind. Eng. Chem. 66 (2018) 126–140 doi: 10.1016/j.jiec.2018.05.023

  9. [9]

     T.D. Xiao, E.R. Strutt, M. Benaissa, et al., Nanostruct. Mater. 10 (1998) 1051–1061 doi: 10.1016/S0965-9773(98)00137-8

  10. [10]

      G.X. Zhu, J.G. Zhu, W.L. Li, et al., Environ. Sci. Technol. 52 (2018) 8684–8692 doi: 10.1021/acs.est.8b01594

  11. [11]

      M.D. Ma, Q. Zhu, Z.Y. Jiang, et al., J. Colloid Interface Sci. 598 (2021) 238–249 doi: 10.1016/j.jcis.2021.04.043

  12. [12]

      Y. Li, D.D. Li, S.S. Fan, et al., Catal. Sci. Technol. 10 (2020) 864–875 doi: 10.1039/c9cy01849b

  13. [13]

      X. Wang, Y.D. Li, Chem. Eur. J. 9 (2003) 300–306 doi: 10.1002/chem.200390024

  14. [14]

      Y.L. Cao, H.X. Yang, X.P. Ai, et al., J. Electroanal. Chem. 557 (2003) 127–134 doi: 10.1016/S0022-0728(03)00355-3

  15. [15]

      Y.F. Chen, C.H. Xue, J. Wang, et al., Chin. Chem. Lett. 2021. 10.1016/j. cclet. 2021.09.088 doi: 10.1016/j.cclet.2021.09.088

  16. [16]

      J. Li, W.L. Li, Y. Rao, et al., Chin. Chem. Lett. 32 (2021) 150–153 doi: 10.1016/j.cclet.2020.10.043

  17. [17]

      Z.Q. Wang, H.Z. Jia, T. Zheng, et al., Appl. Catal. B Environ. 272 (2020) 119030 doi: 10.1016/j.apcatb.2020.119030

  18. [18]

      W.H. Yang, Z.A. Su, Z.H. Xu, et al., Appl. Catal. B Environ. 260 (2020) 118150 doi: 10.1016/j.apcatb.2019.118150

  19. [19]

      J. Liu, V. Makwana, J. Cai, et al., J. Phys. Chem. B107 (2003) 9185–9194 doi: 10.1021/jp0300593

  20. [20]

      Y. Dai, J.H. Li, Y. Peng, et al., Acta Phys. Chim. Sin. 28 (2012) 1771–1776 doi: 10.3866/PKU.WHXB201204175

  21. [21]

      X.K. Huang, D.P. Lv, Q.S. Zhang, et al., Electrochim. Acta 55 (2010) 4915–4920 doi: 10.1016/j.electacta.2010.03.090

  22. [22]

      R.L. Lehman, J.S. Gentry, N.G. Glumac, Thermochim. Acta 316 (1998) 1–9 doi: 10.1016/S0040-6031(98)00289-5

  23. [23]

      C. Dong, H. Wang, Y.W. Ren, et al., J. Environ. Sci. 104 (2021) 102–112 doi: 10.1016/j.jes.2020.11.003

  24. [24]

      S.H. Wang, Q. Liu, Z.Q. Zhao, et al., Eng. Chem. Res. 59 (2020) 6556–6564 doi: 10.1021/acs.iecr.0c00373

  25. [25]

      Z.Q. Chen, X.C. Zhong, Y.M. Xie, et al., J. Alloy. Compd. 884 (2021) 160980 doi: 10.1016/j.jallcom.2021.160980

  26. [26]

      J. Lee, N. Son, J. Shin, et al., J. Alloy. Compd. 869 (2021) 159265 doi: 10.1016/j.jallcom.2021.159265

  27. [27]

      X.H. Zheng, J.M. Cai, Y.N. Cao, et al., Appl. Catal. B Environ. 297 (2021) 120402 doi: 10.1016/j.apcatb.2021.120402

  28. [28]

      G.X. Zhang, Q.L. Wei, X.H. Yang, et al., Appl. Catal. B Environ. 206 (2017) 115–126 doi: 10.1016/j.apcatb.2017.01.001

  29. [29]

      B. Lan, X.Y. Zheng, G. Cheng, et al., Electrochim. Acta 283 (2018) 459–466 doi: 10.1016/j.electacta.2018.06.195

  30. [30]

      S. Islam, M.H. Alfaruqi, D.Y. Putro, et al., J. Mater. Chem. A 7 (2019) 20335–20347 doi: 10.1039/c9ta05767f

  31. [31]

      W.Q. Huang, G.W. Wu, H. Xiao, et al., Environ. Pollut. 256 (2020) 113408 doi: 10.1016/j.envpol.2019.113408

  32. [32]

      L. Gan, X.Y. Fang, L.J. Xu, et al., Mater. Des. 203 (2021) 109596 doi: 10.1016/j.matdes.2021.109596

  33. [33]

      X.J. Fan, L.G. Deng, K. Li, et al., Colloid Interface Sci. Commun. 44 (2021) 100485 doi: 10.1016/j.colcom.2021.100485

  34. [34]

      Y. He, L.J. Wang, Z. Chen, et al., Sci. Total Environ. 785 (2021) 147328 doi: 10.1016/j.scitotenv.2021.147328

  35. [35]

      W.J. Zong, Z.M. Guo, M.H. Wu, et al., Sci. Total Environ. 761 (2021) 143201 doi: 10.1016/j.scitotenv.2020.143201

  36. [36]

      D. Ouyang, Y. Chen, J.C. Yan, et al., Chem. Eng. J. 370 (2019) 614–624 doi: 10.1016/j.cej.2019.03.235

  37. [37]

      O. Fónagy, E. Szabó-Bárdos, O. Horváth, J. Photochem. Photobiol. A Chem. 407 (2021) 113057 doi: 10.1016/j.jphotochem.2020.113057

  38. [38]

      W.X. Hou, S.H. Wang, X.R. Bi, et al., Chin. Chem. Lett. 32 (2021) 2513–2518 doi: 10.1016/j.cclet.2021.01.023

  39. [39]

      S.P. Mo, Q. Zhang, J.Q. Li, et al., Appl. Catal. B Environ. 264 (2020) 118464 doi: 10.1016/j.apcatb.2019.118464

  40. [40]

      S.S. Zhu, X.J. Li, J. Kang, X.G. Duan, S.B. Wang, Environ. Sci. Technol. 53 (2019) 307–315 doi: 10.1021/acs.est.8b04669

  41. [41]

      Y. Yang, J. Huang, S.W. Wang, et al., Appl. Catal. B Environ. 142-143 (2013) 568–578 doi: 10.1016/j.apcatb.2013.05.048

  42. [42]

      R.J. Yang, Z.J. Guo, L.X. Cai, et al., Small (2021) 2103052 doi: 10.1002/smll.202103052

  43. [43]

      Y. Chen, X. Zhang, C. Xu, H. Xu, Electrochim. Acta 309 (2019) 424–431 doi: 10.1016/j.electacta.2019.04.072

  44. [44]

      N.N. Xu, J.W. Liu, J.L. Qiao, et al., J. Power Sources 455 (2020) 227992 doi: 10.1016/j.jpowsour.2020.227992

  45. [45]

      D.N. Rao, B. Sun, J.L. Qiao, X.H. Guan, Prog. Chem. 29 (2017) 1142

  46. [46]

      J.A. Dawson, H. Chen, I. Tanaka, ACS Appl. Mater. Interfaces 7 (2015) 1726–1734 doi: 10.1021/am507273c

  47. [47]

      Z.H. Diao, W.X. Zhang, J.Y. Liang, et al., Chem. Eng. J. 409 (2021) 127684 doi: 10.1016/j.cej.2020.127684

  48. [48]

      T.C. Zang, H. Wang, Y.H. Liu, et al., Chemosphere 261 (2020) 127616 doi: 10.1016/j.chemosphere.2020.127616

  49. [49]

      Y.D. Liu, X.J. Zhang, J. Bian, et al., Appl. Catal. B Environ. 254 (2019) 260–269 doi: 10.3390/e21030260

  50. [50]

      J. Jia, P. Zhang, L. Chen, Catal. Sci. Technol. 6 (2016) 5841–5847 doi: 10.1039/C6CY00301J

  51. [51]

      Z.Q. Wang, H.Z. Jia, Z.W. Liu, et al., J. Hazard. Mater. 413 (2021) 125285 doi: 10.1016/j.jhazmat.2021.125285

  52. [52]

      C. He, Y.C. Wang, Z.Y. Li, et al., Environ. Sci. Technol. 54 (2020) 12771–12783 doi: 10.1021/acs.est.0c05235

  53. [53]

      S.P. Rong, P.Y. Zhang, F. Liu, Y.J. Yang, ACS Catal. 8 (2018) 3435–3446 doi: 10.1021/acscatal.8b00456

  54. [54]

      G.F. Xiao, T.T. Xu, M. Faheem, et al., Int. J. Environ. Res. Public Health 18 (2021) 3344 doi: 10.3390/ijerph18073344

  55. [55]

      M.Q. Liu, Q.H. Zhao, H. Liu, et al., Nano Energy 64 (2019) 103942 doi: 10.1016/j.nanoen.2019.103942

  56. [56]

      S. Devaraj, N. Munichandraiah, J. Phys. Chem. C 112 (2008) 4406–4417 doi: 10.1021/jp7108785

  57. [57]

      J.Z. Huang, S.F. Zhong, Y.F. Dai, C.C. Liu, H.C. Zhang, Environ. Sci. Technol. 52 (2018) 11309–11318 doi: 10.1021/acs.est.8b03383

  58. [58]

      D.M. Robinson, Y.B. Go, M. Mui, et al., J. Am. Chem. Soc. 135 (9) 20133494–3501 doi: 10.1021/ja310286h

  59. [59]

      L. Ukrainczyk, M.B. McBride, Clays Clay Miner. 40 (1992) 157–166 doi: 10.1346/CCMN.1992.0400204

  60. [60]

      X. Shi, H. Zheng, A.M. Kannan, K. Pérez-Salcedo, B. Escobar, Inorg. Chem. 58 (2019) 5335–5344 doi: 10.1021/acs.inorgchem.9b00492

  61. [61]

      M.X. Lin, F.Q. Shao, S.T. Weng, et al., Electrochim. Acta378 (2021) 138147 doi: 10.1016/j.electacta.2021.138147

  62. [62]

      M. Rittiruam, S. Somdee, P. Buapin, N. Aumnongpho, N. Kerdprasit, et al., J. Alloy. Compd. 869 (2021) 159280 doi: 10.1016/j.jallcom.2021.159280

• 

  Figure 1  Physiochemical properties analysis of x-MnO₂ and the application for MG degradation: (a) XRD patterns; (b–d) SEM images; (e) LSV and (f) EIS measurement; (g, h) batch experiments for MG degradation comparison. (MnO₂ dosage (g): 1–5 mg (pH 4.9 ± 0.1); pH (h): 3–7 (dosage = 3 mg)).

  下载: 全尺寸图片 幻灯片
  

  Figure 2  Reactive oxygen species detection. (dosage of MnO₂ = 3 mg, pH 4.9 ± 0.1; mole ratio of MnO₂ to scavengers = 1:20).

  下载: 全尺寸图片 幻灯片
  

  Figure 3  XPS analysis of MnO₂: Mn 2p (a–c) and O 1s (d–f).

  下载: 全尺寸图片 幻灯片
  

  Figure 4  Thermogravimetric analysis of MnO₂.

  下载: 全尺寸图片 幻灯片
  

  Figure 5  Intrinsic relationship among the characteristics of MnO₂ polymorphs.

  下载: 全尺寸图片 幻灯片

  Table 1.  Morphology characterization and elements analysis of x-MnO₂.

  下载: 导出CSV
•