Understanding hydrothermal transformation from Mn2O3 particles to Na0.55Mn2O4·1.5H2O nanosheets, nanobelts, and single crystalline ultra-long Na4Mn9O18 nanowires

Manganese oxides are one of the most valuable materials for batteries, fuel cells and catalysis. Herein, we report the change in morphology and phase of as-synthesized Mn2O3 by inserting Na+ ions. In particular, Mn2O3 nanoparticles were first transformed to 2 nm thin Na0.55Mn2O4·1.5H2O nanosheets and nanobelts via hydrothermal exfoliation and Na cation intercalation, and finally to sub-mm ultra-long single crystalline Na4Mn9O18 nanowires. This paper reports the morphology and phase-dependent magnetic and catalytic (CO oxidation) properties of the as-synthesized nanostructured Na intercalated Mn-based materials.

for Na-ion rechargeable batteries 32,[38][39][40][41][42][43][44][45][46][47][48][49][50][51][52][53][54][55][56] . Several methods have been used to synthesize the material, including sol-gel/high-temperature calcinations 32,42,43,52 , solid-state reaction 27,39 , thermal-conversion of a precursor 41 ,polymer-pyrolysis 45 , and hydrothermal method 52 . Hosono et al. used a hydrothermal method (Teflon-lined autoclave at 205 °C for 2 days) using Mn 3 O 4 powder in a 5.0 M NaOH solution and obtained single-crystalline Na 0.44 MnO 2 nanowires with superior capacity of 120 mAh/g and high charge-discharge cyclability 52 . In these cases, the efficiency of the material was shown to be dependent on the surface area and morphology; hence, an understanding of the change in morphology during Na (or Li and K) ion-insertion is very important. Liu et al. prepared Na 0.44 MnO 2 nanorods with recipes of MnSO 4 , KMnO 4 and NaOH solutions by a hydothermal method 56 . Le et al. reported a change in morphology (from nanosheets to nanowires) and crystal structure (from Mn 2 O 3 to birnessite and Na 0. 44 MnO 2 ) after the hydrothermal reaction of Mn 2 O 3 powder in a 5.0 M NaOH solution 48 . Although many studies have reported the electrochemical properties of Na-inserted MnO x materials 32,[38][39][40][41][42][43][44][45][46][47][48][49][50][51][52][53][54][55][56] , this study examined the undiscovered Na-insertion and morphological behaviors of Mn 2 O 3 nanoparticles during a hydrothermal reaction process. This paper reports a facile process to control the morphology and phase of alkali metal intercalated Mn oxides using a simple hydrothermal technique. Three different alkali metals (Li, Na, and K) were intercalated into the Mn 2 O 3 powder (particles) to nanosheets, nanobelts and nanowires. In particular, quantum-thick Na 0.55 Mn 2 O 4 ·1.5H 2 O nanosheets, nanobelts and single crystalline ultra-long Na 4 Mn 9 O 18 nanowires were produced by inserting Na with different concentrations and reaction durations. The magnetic and catalytic (CO oxidation) properties of the as-synthesized Mn oxides are reported in detail. In addition to the new findings of the morphological behaviors (by Na-insertion)/detailed characterization and magnetic properties, the laser-induced Na-deinsertion behavior was also examined by Raman spectroscopy. The present study provides several new insights into the development of alkali metal ion intercalated Mn materials.  Fig. 1  NPs hydrothermally in a 1.0 M NaOH solution at 200 °C, the intensity of these new peaks (Δ ) did not increase significantly, even though the reaction was performed for 3 weeks, which was attributed to the lack of Na + ions. On the other hand, in the 10 M NaOH solution, these two diffraction peaks (Δ ) for Na 0.55 Mn 2 O 4 ·1.5H 2 O showed significant intensities upon a reaction for less than 3 days. Upon the reaction for 1 week, the XRD peaks corresponding to the cubic phase Mn 2 O 3 were disappeared completely. Interestingly, several new diffraction peaks (ο ) appeared. With further increases in the reaction time to 1~3 weeks, the two peaks (Δ ) for Na 0.55 Mn 2 O 4 ·1.5H 2 O at 12.5° and 25.1° 2θ decreased in intensity. After a reaction for 3 weeks, the newly appeared peaks (ο ) were mainly present, which matched orthorhombic (Pbam) Na 4 Mn 9 O 18 (JCPDS 27-0750) (Supporting Information Fig. S1 and S2) 32,42,46 . This suggests a complete change in the crystal phase of Na 0. 55 Fig. 1 shows the observed and Rietveld refinement XRD patterns (see Supporting Information, Fig. S3). The crystal structures were fully refined, and the detailed structural parameters are provided in the Supporting Information Fig. S3, Tables S1 and S2.

Results and Discussion
The SEM and TEM/HRTEM images of the corresponding samples were examined to further understand the recrystallization mechanism of Mn 2 O 3 NPs in a NaOH solution under hydrothermal conditions at 200 °C for the specified duration. Figure 2 shows SEM images of the starting materials (Mn 3 O 4 and Mn 2 O 3 ) and the synthesized materials prepared by a hydrothermal method in 1.0 M NaOH, LiOH and KOH solutions for 24 hrs. The starting Mn 3 O 4 and Mn 2 O 3 showed particle morphologies with different sizes. On the other hand, after a hydrothermal reaction (1.0 M NaOH) at 200 °C, the surface morphology had changed entirely to ultrathin nanosheets. Under LiOH and KOH solution conditions, the surface morphologies were also changed to nanosheets, but were thicker than those prepared in the NaOH solution. Supporting Information, Fig. S4 provides additional SEM images of the nanosheets obtained by Na, Li and K intercalation. The SEM images and the XRD patterns ( Fig. 1) indicate that the sheet morphology originated from the monoclinic Na 0.55 Mn 2 O 4 ·1.5H 2 O phase, which was formed by the exfoliation of Mn 2 O 3 upon Na and H 2 O concomitant intercalation. On the other hand, the presence of a Mn 2 O 3 phase for the samples prepared in a short duration (< 3 weeks in 1 M NaOH or < 3 days in 10 M NaOH) was attributed to the incomplete conversion of Mn 2 O 3 present primarily in the core part of the powder, whereas the surface consisted mainly of ultra-thin nanosheets (Fig. S5,SI). TEM, HRTEM images and electron diffraction patterns were also obtained for the ultrathin nanosheets, as shown in Fig 57 . On the other hand, they reported Na + -ion free birnessite-related layered MnO 2 nanobelts (5-15 nm width), which is inconsistent with the present study.
To measure the accurate thickness of the ultrathin nanosheets discussed above, a more skillful technique was employed, as described in Fig. 3. The nanosheets were first sandwiched between epoxy supported by disks, as illustrated in the Figure. Various treatment steps such as bonding, slicing, disk cutting, and ion milling, were then performed to make a suitable TEM specimen. The thickness of the TEM specimen was finally less than 5 μ m. TEM, HRTEM and high-angle annular dark field (HAADF) images were taken, which clearly showed the edge of the nanosheets. Mn in the nanosheets edge was also confirmed by an EDX profile ( Supporting Information, Fig. S6). The HRTEM image showed lattice fringes with neighboring distances of 0.25 nm, corresponding to the (200) plane of monoclinic Na 0.55 Mn 2 O 4 ·1.5H 2 O as mentioned above. The thickness of the nanosheet edge was measured to be 2 nm, which is close to the unit cell thickness (also see Supporting Information, Fig. S7).
Because the crystal phase of Mn 2 O 3 was not completely changed using 1.0 M NaOH, the NaOH concentration was increased to 10.0 M and a hydrothermal reaction was performed for various reaction durations. The morphologies and microstructures of the samples obtained by the hydrothermal treatment of Mn 2 O 3 in 10 M NaOH for 20 h at 200 °C were examined further by SEM and TEM/HRTEM, as shown in Figs 4 and 5. The Mn 2 O 3 particles initially changed to nanosheets and nanobelts with a few nanowires (or nanothreads) for a reaction duration of less than 1 week (Supporting Information, Fig. S8), whereas the Mn 2 O 3 nanoparticles were still present in the synthesized samples. This was supported by the corresponding XRD patterns (Fig. 1). As the reaction time increased, the nanobelts evolved slowly to ultra-long nanowires. Mixed morphologies were observed in the SEM images (Supporting Information, Fig. S9). For the corresponding XRD results (Fig. 1), the XRD patterns (∆) of Na 0.55 Mn 2 O 4 ·1.5H 2 O were diminished slowly and those (ο ) of Na 4 Mn 9 O 18 were remarkable. Upon the reaction for 3 weeks, the SEM image in Fig. 5 showed mostly ultra-long (sub-mm) nanowires (also see Supporting Information, Fig. S10). The corresponding optical microscopy images showed that the black color of the Mn 2 O 3 (with particle morphology) changed to a brown color as the crystal phase changed to Na 0. 55 Fig. S11). The morphology appeared like nanofibers for the final Na-intercalated Mn product. HRTEM images of the nanobelts showed a clear lattice spacing of 0.23 nm, corresponding to the (200) plane of monoclinic Na 0.55 Mn 2 O 4 ·1.5H 2 O (Fig. 4). This was also observed for the ultrathin nanosheets (Figs 2 and 3), suggesting a similar growth direction of nanosheets and nanobelts. The SAED pattern confirmed the single crystal nature of the Na 0.55 Mn 2 O 4 ·1.5H 2 O nanobelts. Supporting Information, Fig. S12 shows the corresponding simulated diffraction patterns. A structure projection model in Fig. 4 Figure 5 shows representative SEM, TEM, and HRTEM images of the Na 4 Mn 9 O 18 nanowires obtained using 10 M NaOH at 200 °C for 3 weeks. The HRTEM image shows a lattice spacing of 0.442 nm for the nanowires, which is in accordance with the (200) plane of orthorhombic Na 4 Mn 9 O 18 32 . The spot SAED pattern confirms the single crystal structure of these nanowires. The wire grew along the [100] direction. Figure 6 shows the corresponding structure projections and crystal models of the Na-intercalated samples. In the case of the Na 0.55 Mn 2 O 4 ·1.5H 2 O nanobelts, H 2 O and Na cations were concomitantly intercalated between the skeletons of Mn-O sheets. For the ab plane structure of the Na 4 Mn 9 O 18 nanowires, Na was embedded in the Mn-O tunnel frame, which is consistent with the MnO 5 square pyramids and MnO 6 octahedra 58 . The Na cations are situated in two different sites (with a unique tunnel structure) and the c-axis is the charge-discharge paths of Na cation diffusion 27,32,44 . The SAED and simulated patterns of the starting material, i.e. Mn 2 O 3 , are provided in the Supporting Information, Fig. S13.
The change in crystal phase was further confirmed by FT-IR spectroscopy ( Supporting Information, Fig. S14). The characteristics of the Mn-O vibrational peaks were observed between 500 and 800 cm −1 for all samples 13 . No OH stretching bands at approximately 3400 cm −1 was observed for the starting material, i.e. Mn 2 O 3 powder. Upon the formation of Na 0.55 Mn 2 O 4 ·1.5H 2 O, strong OH stretching bands were observed at 3430 and 3350 cm −1 . On the other hand, the FTIR peaks became weaker and broader for the Na 4 Mn 9 O 18 nanowires ( Fig. 5 and Fig. S10). The much weaker broad band at 3400 cm −1 for Na 4 Mn 9 O 18 was attributed to the adsorbed H 2 O (and OH) species. Figure 7 shows the Raman spectra of the Na 4 Mn 9 O 18 nanowires measured with different laser powers (0.004 mW to 2.7 mW). At a low laser power (< 0.012 mW), no obvious signal was observed. With increasing laser power to 0.19 mW, the Raman signals became clear at 637.9 cm −1 and a shoulder was observed at 561.8 cm −1 (see Supporting Information, Fig. S15). Upon further increases in the laser power to 2.7 mW, a strong fluorescence signal was observed (also see Supporting Information, Fig. S16) and the peak at 637.9 cm −1 was decreased significantly. Upon reducing the laser power to 0.19 mW, critically different Raman signals were obtained (Supporting Information, Fig. S15). This suggests that the crystal phase of Na 4 Mn 9 O 18 had changed irreversibly to Mn 2 O 3 by the high power laser irradiation. The laser light induces the de-insertion of Na cations in the structure, which requires further study. The newly obtained Raman spectrum shows peaks at 312.7, 374. 3 Fig. S15, S16 and S17).
X-ray photoelectron spectroscopy (XPS) was used to examine the chemical states of Na 4 Mn 9 O 18 nanowires and compared with those of the starting material, i.e., hydrothermally synthesized Mn 2 O 3 powders, as displayed in Fig. 8. A typical survey XPS scan of Mn 2 O 3 showed Mn, O and impurity carbon signals, whereas that of Na 4 Mn 9 O 18 showed additional Na as well as Mn, O and C ( Supporting Information, Fig. S18). The distinct peaks at ~653.8 and ~642.1 eV (Fig. 8, top left) were assigned to the Mn 2p 1/2 and Mn 2p 3/2 XPS peaks, respectively, with a spin-orbit energy splitting of 11.7 eV 49 . The Mn 2p XPS peaks for Na 4 Mn 9 O 18 were shifted slightly to a lower binding energy, confirming the Na insertion and reduction of the oxidation state of Mn 59,60 . The O 1s XPS spectra showed two broad peaks at 532.0 and 529.7 eV (Fig. 8, top right) due to the absorbed surface oxygen (e.g., OH, H 2 O, and O 2 ) species and lattice oxygen atoms of the Mn oxides, respectively 13 . The Na 1s XPS and Na KLL Auger peaks for Na 4 Mn 9 O 18 (Fig. 8, bottom panel) were observed at 1070.7 and 494.2 eV, respectively 49 .
The magnetic properties of the Na 4 Mn 9 O 18 nanowires were examined by SQUID. Figure 9 presents zero-field-cooling (ZFC) and field-cooling (FC) magnetization curves measured at an applied field of H = 100 Oe (0.1 kOe) over the temperature range of 5− 300 K. The top inset in Fig. 9 shows the magnetization (M− H) curves measured at various temperatures from 5 K to 300 K and magnetic fields from − 50 to 50 kOe. An ideal linear plot (with no hysteresis loop) of magnetization was obtained with an applied magnetic field at temperatures between 300 K and 50 K, indicating the paramagnetic and antiferromagnetic properties of the Na 4 Mn 9 O 18 nanowires. The M− H curves showed no saturation magnetism in the external fields up to 50 kOe. A magnetization of 2.19 emu g −1 was measured at 50 kOe and 300 K. The mass magnetic susceptibility of the nanowires at 300 K was 4.39 × 10 −5 emu·g −1 ·Oe −1 . This increased with decreasing temperature and was determined to be 5.58 × 10 −5 emu·g −1 ·Oe −1 at 50 K. Interestingly, a magnetic hysteresis loop was clearly observed at 5 K ( Supporting Information, Fig. S19), suggesting typical ferromagnetic behavior. On the other hand, the M− H curve showed no saturation, indicating antiferromagnetic property. The residual magnetism (or remanence) and coercive force were measured to be 0.136 emu·g −1 and 475 Oe, respectively. A coercivity of 10.7 kOe at 5 K was reported for the Mn 3 O 4 nanowires 22 . For single unit cell thickness Mn 3 O 4 nanosheets, Huang et al. observed paramagnetic and ferromagnetic (with a coercivity of 5.8 kOe) behaviors at room temperature and 5K, respectively 21 . The FC magnetization curve increased with decreasing temperature. On the other hand, the ZFC curve was increased slowly with decreasing temperature to 25 K, and decreased below that temperature. The ZFC curve showed a maximum at 25 K. This suggests a clear transition from paramagnetic to ferromagnetic at a temperature below 25 K. The FC and ZFC curves showed no overlap at all temperatures up to 300 K.
The surface resistance of Na 4 Mn 9 O 8 nanowires was measured as a function of temperature ( Supporting  Information, Fig. S20). The resistance of 12.5 MΩ at room temperature decreased linearly to 1.0 MΩ with increasing temperature to 200 °C. For the Mn 3 O 4 (in Fig. 2) and Mn 2 O 3 powder samples, the surface resistance could not be measured because of the high resistance.
The CO oxidation activities (Supporting Information, Fig. S21) of Mn 3 O 4 (in Fig. 2), Mn 2 O 3 (in Fig. 2), and Na 0.55 Mn 2 O 4 ·1.5H 2 O nanosheets (or Mn 2 O 3 @Na 0.55 Mn 2 O 4 ·1.5H 2 O core-shell structures; sample prepared with NaOH solution in Fig. 2) was tested for catalytic applications, such as CO oxidation using low cost materials 13 . In the first CO oxidation runs, the CO oxidation onsets were observed in the order of  nanowires (T 10% ≈ 180 °C) showed much catalytic activity than the others (T 10% ≈ 220 °C) and Na 2 Mn 8 O 16 did not relate to their high catalytic activity. Their conclusions are in good agreement with the present study.

Conclusion
Na-ion intercalation into Mn 2 O 3 was initially transformed into ultra-thin monoclinic Na 0.55 Mn 2 O 4 ·1.5H 2 O nanosheets and nanobelts. The nanobelts were then evolved to single crystalline ultra-long orthorhombic Na 4 Mn 9 O 18 nanowires with a (Na-ion mobile) tunnel structure. This synthesis process was extended further to other alkali metals (Li and K) using a simple hydrothermal method in a Mn 2 O 3 -dispersed alkali hydroxide (LiOH, NaOH and KOH) solution. SEM and TEM confirm the transformation of the morphology. XRD and HRTEM were used to examine the crystal phase change and microstructure. Detailed crystal structural parameters were obtained  by Rietveld refinement analysis. XPS confirmed the presence of inserted Na cation. Moreover, high power laser irradiation readily induces the irreversible Na-deinsertion behavior from Na 4 Mn 9 O 18 to Mn 2 O 3 , as confirmed by Raman spectroscopy. The Na 4 Mn 9 O 8 nanowires exhibited ferromagnetic behavior at temperatures below 25 K and paramagnetic behavior at above that temperature. The surface resistance of Na 4 Mn 9 O 8 nanowires was 12.5 MΩ at room temperature and decreased linearly to 1.0 MΩ with increasing temperature to 200 °C. The CO oxidation activity (T 10% = 230 °C) of the Mn 2 O 3 nanoparticles was substantially decreased after Na-intercalation. The very detailed transformation mechanism and the new fundamental characterization provide new insights into the development of alkali metal cation intercalated Mn oxides.

Methods
Material synthesis. Mn 3 O 4 was synthesized by a hydrothermal method, as described below. Briefly, 10 mL of 0.1 M Mn(II) nitrate tetrahydrate (Sigma-Aldrich. > 97.0%) was mixed with 10 mL of deionized water (18.2 MΩ cm resistivity) in a Teflon jar (120 mL capacity), and 1.0 mL of an ammonia solution was then added to obtain the precipitates. The reaction jar was capped tightly and placed in an oven (120 °C) for 12 hours, after which the oven was cooled naturally to room temperature. The brown precipitate was collected after washing with deionized water followed by ethanol, and then dried in an air convection oven (80 °C). Bulk Mn 2 O 3 was obtained by the post-annealing of Mn 3 O 4 at 750 °C for 4 hrs. To synthesize the Na(or Li and K)-intercalated Mn materials, the Mn 2 O 3 (~25 mg) was dispersed in a 20.0 mL 1.0 M (or 10 M) NaOH (or LiOH and KOH) solution. The solution in a Teflon-lined stainless autoclave was placed at 200 °C for a reaction time, which was varied from 12 hrs to 3 weeks. After a specified time (12 hrs, 1 day, 3 days, 1, 2 and 3 weeks were selected to show in the present article), the oven was stopped and cooled naturally to room temperature and the powder product was collected by centrifuging. The powder was finally washed and dried for further characterization. Although the slow reaction process took time and patience (and somewhat industrially impractical) we employed the slow process to disclose new findings and to carefully examine change in morphology which has never been reported for Mn oxide material.
Material characterization. The surface morphology of the synthesized powder samples was examined by field emission scanning electron microscopy (FE-SEM, Hitachi SE-4800). High resolution transmission electron microscopy (HRTEM) and the electron diffraction patterns were obtained using a FEI Tecnai G2 F20 at an operating voltage of 200 kV. The powder X-ray diffraction (XRD) patterns were obtained using a PANalytical X'Pert Pro MPD diffractometer operated at 40 kV and 30 mA using Cu Kα radiation. The Rietveld refinement was performed using the TOPAS software program (ver. 4.2, Bruker 2005). Further details are described elsewhere 61 . The Fourier-transform infrared (FT-IR) spectroscopy was performed using a Thermo Scientific Nicolet iS10 spectrometer in ATR (attenuated total reflectance) mode. The X-ray photoelectron spectra were obtained using a Thermoscientific K-alpha X-ray photoelectron spectrometer with a monochromated Al Kα X-ray source, a pass energy of 20.0 eV, and an analyzed spot size of 400 μ m. Confocal Raman microscopy (PRISM, NOST Co., South Korea) was conducted to take the Raman spectra for the powder samples at a laser wavelength of 532 nm and a 100 × , 0.9NA microscope objective. The laser intensity was varied from 0.004 mW to 2.7 mW. All the Raman spectra were referenced to the Raman spectrum of cyclohexane. The magnetic properties of the Na 4 Mn 9 O 18 nanowires were examined using a MPM5-XL-7 superconducting quantum interference device (SQUID) magnetometer (Quantum Design, Inc.) at various temperatures. CO oxidation and surface resistance tests. The CO oxidation experiments were performed on a continuous flow quartz U-tube reactor with a 10 mg sample. A mixed gas (1% CO and 2.5% O 2 in N 2 balance) was introduced into the reactor at a flow rate of 40 mL/min. The temperature heating rate was fixed to 20 °C/min. The reaction gas products were analyzed using a SRS RGA200 quadrupole mass spectrometer. The surface resistance of the pelletized sample was measured using a home-built four-probe resistance measurement instrument.