Annealing induced a well-ordered single crystal δ-MnO2 and its electrochemical performance in zinc-ion battery

Herein, the formation and electrochemical performance of a novel binder-free turbostratic stacked/ well-ordered stacked δ-MnO2-carbon fiber composite cathodes in deep eutectic solvent (DES) based zinc-ion battery (ZIB) is reported. Results of morphological, elemental, and structural analyses revealed directly grown and interconnected δ-MnO2 crumpled nanosheets on a carbon fiber substrate. Moreover, an improvement via a simple annealing strategy in the stacking, surface area and conductivity of the δ-MnO2 sheets was observed. Annealing induces the rearrangement of δ-MnO2 sheets resulting in the transformation from turbostratic stacking to a well-ordered stacking of \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${\delta }$$\end{document}δ-MnO2 sheets, as indicated by the selected area electron diffraction (SAED) hexagonal single crystal pattern. Besides, the formation of the well-ordered stacking of \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${\delta }$$\end{document}δ-MnO2 sheets exhibited improved electrochemical performance and cyclability, as cathode material for ZIB. The novel strategy described in this study is an essential step for the development of binder-free δ-MnO2-C fiber composite with a well-ordered stacking of δ-MnO2 sheets. This study also demonstrated comparable electrochemical performance between the turbostratic \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${\delta }$$\end{document}δ-MnO2 sheets and the well-ordered stacked δ-MnO2 sheets.

a n o ur c hi n, n a n or o d s, n a n o s p h e r e, n a n o s h e et s, n a n ofl a k e s, n a n o c or al s, n a n ofl o w e r s, a n d n a n o wi r e s, i nt e r e sti n g 1 D e p art m e nt of C h e mi c al E n gi n e eri n g, F a c ult y of E n gi n e eri n g, C h ul al o n g k or n U ni v er sit y, B a n g k o k, 1 0 3 3 0, T h ail a n d. 2 D e p art m e nt of P h y si c s, S c h o ol of S ci e n c e a n d E n gi n e eri n g, At e n e o d e M a nil a U ni v er sit y, Q u e z o n Cit y, 1 1 0 8, P hili p pi n e s. 3 D e p art m e nt of C h e mi c al E n gi n e eri n g, F a c ult y of E n gi n e eri n g, U ni v er sit y of S a nt o T o m a s, M a nil a, 1 0 1 5, P hili p pi n e s. 4 Hi g h-p erf or m a n c e c o m p uti n g u nit ( C E C C-H C U), C e nt er of E x c ell e n c e o n C at al y si s a n d C at al yti c R e a cti o n E n gi n e eri n g ( C E C C), C h ul al o n g k or n U ni v er sit y, B a n g k o k, 1 0 3 3 3, T h ail a n d. 5 D e p art m e nt of M at eri al s S ci e n c e, F a c ult y of S ci e n c e, C h ul al o n g k or n U ni v er sit y, B a n g k o k, 1 0 3 3 0, T h ail a n d. 6 R e s e ar c h U nit of A d v a n c e d M at eri al s f or E n er g y St or a g e, C h ul al o n g k or n U ni v er sit y, B a n g k o k, 1 0 3 3 0, T h ail a n d. 7 C e nt er of E x c ell e n c e i n P etr o c h e mi c al a n d M at eri als T e c h n ol o g y, C h ul al o n g k or n U ni v ersit y, B a n g k o k, 1 0 3 3 0, T h ail a n d. 8 Di visi o n of M at eri als S ci e n c e a n d E n gi n e eri n g, F a c ult y of E n gi n e eri n g, H o k k ai d o U ni v er sit y, Kit a 1 3 Ni s hi 8, S a p p or o, H o k k ai d o, 0 6 0-8 6 2 8, J a p a n. * e m ail: s o or at h e p. k @ c h ul a. a c.t h o p e n nanostructures: 1D, 2D, and 3D self-assembled into varieties of forms such as tunnels, sheets etc. which could serve as intercalation type electrode for ZIBs. Further, MnO 2 has ease of synthesis via common scalable routes: electrodeposition, chemical reduction, thermal decomposition, and hydrothermal methods 14,15 .
Although Zn-MnO 2 based batteries are well-documented, most of the literature published in this field has been typically reported under aqueous [16][17][18][19] and non-aqueous 20,21 electrolyte systems. Accordingly, a previous paper had investigated the deep eutectic solvent (DES) based electrolyte system and applied it to Zn-MnO 2 secondary battery 22 . DES 23,24 , which is typically formed from the eutectic mixture of Lewis or Bronsted acids and bases, is interesting for the fact that H + ion intercalation and electrolysis of water (hydrogen evolution) can be avoided. Besides, anode passivation and formation of ZnO and dendrites commonly observed in aqueous systems could be prevented. This development will eventually make Zn 2+ intercalation as the governing route to understand the complex interaction mechanism of divalent cations with MnO 2 cathode. An understanding of the said mechanism is not only important experimentally but also important for simulation and theoretically researches whereby improvement can be undertaken to boost the electrochemical performance of ZIBs.
At present, a common obstacle encountered in using MnO 2 cathode is its poor conductivity 25 which could increase the internal resistance of the ZIBs resulting in poor electrochemical performance. A remedy frequently used by researchers is the addition of conductive materials such as carbon black 26 , reduced graphene oxide (RGO) 27 , conducting polymer 28 and carbon nanotubes (CNTs) 29 usually with the aid of polymeric binder to enhance the adhesion of the additives and active material on the current collector. However, this not only complicates the system but also makes the cost of producing the ZIBs expensive. In this study, a novel strategy was developed to address the poor conductivity of MnO 2 as well as the undesirable corrosion problem which is widely encountered in aqueous based ZIBs. A first approach, based on direct deposition and growth of active material MnO 2 on a conductive carbon fiber, was to synthesize a binder-free MnO 2 -C composite cathode via an economical hydrothermal process. This method not only eliminated the need for expensive polymer binders, but most importantly improved the adhesion and connectivity of the active materials to the substrate which is necessary for efficient charge transfer. Secondly, the introduction of an annealing technique was undertaken to enhance δ-MnO 2 sheets stacking order, surface area, and conductivity of the active materials. This was done to generate more redox and intercalation sites for charge storage. Further, it was found that annealing at 300 °C for 1 h can induce the transformation from turbostratic stacking to a well-ordered stacking of δ-MnO 2 sheets. Hence, in a review of the literature, it would appear that this study is the first to analyze the transformation from turbostratic stacking to a well-ordered stacking of δ-MnO 2 sheets via annealing, and its effect on the electrochemical performance of δ-MnO 2 , as cathode material for ZIB utilizing DES as electrolyte.

Experimental
Materials. All chemicals were used as received without further purification: potassium permanganate Substrate preparation. The carbon cloth was surface treated with 1.0 M sulfuric acid for 1 h, washed with DI water several times, and vacuum dried at 60 °C for 2 h before usage.
Cathode synthesis and fabrication. In a typical experiment, 0.1264 g KMnO 4 and 0.0428 g (NH 4 ) 2 SO 4 were dissolved and mixed in 40 mL deionized water. The resulting solution was then sonicated for 1 h and hydrothermally synthesized at 110 °C for 18 h using Teflon lined autoclave decorated with carbon cloth on its inner wall. Next, the carbon cloth, deposited with MnO 2 particles, was washed with deionized water several times and rinsed with IPA. Finally, the fabricated cathode was vacuum dried for 2 h and annealed for 1 h at 300 °C with a heating rate of 10 °C/min. DES synthesis and electrolyte preparation. In a typical experiment, 50 g choline chloride and 43 g urea were mixed together inside a glove box until a homogenous solution was achieved. Then, 4.1 g ZnCl 2 was added to the mixture and thoroughly mixed overnight.
ZIB assembly. CR2032 cell was used to assemble the ZIB. The assembled battery typically consisted of an anode, cathode, and electrolyte made up of Zn electrodeposited on Ni foam. Further, MnO 2 hydrothermally was grown on carbon cloth, and ZnCl 2 electrolyte was dissolved in DES, respectively. The electrodes were separated with microfiber glass and were enclosed within circular metal cases. To prevent the electrode from moving around, a metal spacer and spring were placed adjacent to the anode and the negative case, respectively, as shown in Fig. S1.

Results and Discussion
The aim of this study is to synthesize a binder free MnO 2 -C fiber composite with improved conductivity without the addition of a conductive agent e.g. conductive carbon, graphene or carbon nanotube (CNT). A further objective is to investigate its electrochemical performance as a cathode material for ZIBs. To realize this, it is imperative to produce particles with uniform morphology i.e. size, shape, and structure. In so doing, a relationship among morphology, structure, and property can be established. This is important not only for basic or fundamental science but especially important for enhancing the performance of ZIBs. Hence, this study focuses on thorough optimization of the synthesis parameters viz. time and temperature which can significantly affect the resulting morphology, structure, and the performance of the battery [30][31][32][33] . Figure 1 shows the morphology of samples synthesized at 110 °C for 13 h in un-annealed and annealed conditions. It can be observed that the particles hydrothermally synthesized have a flower-like morphology (Fig. 1a) which was retained even after annealing at 300 °C (Fig. 1b). EDS analysis of the samples, as shown in Fig. S2a,b, revealed the presence of elements Mn, O and K in the produced microsphere, for both un-annealed and annealed conditions which strengthen the likelihood of the existence of MnO 2 . To investigate the phase and structure of the synthesized powder, XRD analysis was done for the as-synthesized powder in its un-annealed and annealed conditions, as shown in Fig. Sα. XRD analyses revealed that the as-synthesized powder (un-annealed) had a δ phase that can be judged from the broad prominent peaks around 2θ = 12.636°, 25.43°, 36.551°, 54.031° and 65.432° corresponding to (200), (400), (020), (613), and (033) planes respectively of the orthorhombic δ-MnO 2 polymorph [34][35][36][37][38][39][40] . On the other hand, the annealed powder showed similar peaks with that of the un-annealed sample with more intense and sharper peaks at 2θ of ~36° and ~65°. These peaks can be attributed to the change in the www.nature.com/scientificreports www.nature.com/scientificreports/ stacking of δ-MnO 2 and not to the formation of another compound since the annealing temperature employed is not enough to cause phase or structural transformation 41,42 . This phenomenon will be discussed in the following paragraph.
δ-MnO 2 has a 2-D layered structure (Fig. S3b) wherein the layers typically consist of edge-sharing MnO 6 octahedra with space in between layers which can be occupied by cations such as K + , Li + , Na + , other alkaline metals ions, and water molecules to stabilize the randomly distributed deficiency of charges in the structure 36 . The orthorhombic δ-MnO 2 (K 0.33 MnO 2 ·0.66H 2 O), was found to be in good agreement with that of the EDS elemental analysis (Fig. S2c), wherein the amount of K + per Mn atom in both un-annealed and annealed samples are ~14-20% which is one of the polymorph of K 0.33 MnO 2 ·0.66H 2 O 34 . The broadening of peaks that can be seen are usually attributed to small nanoparticles, amorphous, and disordered structures. These are commonly observed phenomenon in hydrothermally synthesized δ-MnO 2 . However, in some instances, this occurrence can also be associated with the variation in the interlayer spacing or interlayer order 34,43 . Figure 2a,b show the TEM images of the particles formed for both un-annealed and annealed samples. It was observed at micro-level that the flower-like morphology under SEM were actually crumpled nanosheets that have a flower-like morphology. The formation of sheet-like morphology was found to be in good agreement with the Wulff-constructed equilibrium shape of δ-MnO 2 (Fig. S4) wherein the surface of the nanosheets contains the low surface energy plane {001}. According to Chen et al. 34 , this structure had an average surface energy of 0.11 Jm −2 and an atomic volume of 32.9 Å 3 /Mn atom. The sheet-like morphology was examined under SAED, as shown in Fig. 2c,d. It was observed that the un-annealed sample had a polycrystalline structure which is indexed to the planes of the orthorhombic δ-MnO 2 . The annealed sample exhibited single crystal diffraction which is indexed to {020} and {510} for the inner hexagonal pattern and {530} and {10 0 0} for the outer hexagonal pattern. These  34,43 . Hence, after annealing, the hexagonal single crystal pattern SAED profile indicated that the well-ordered MnO 2 structure was obtained. Figure 3 illustrates the formed particles before and after annealing. Hydrothermal synthesis at 110 °C for 13 h utilizing KMnO 4 and (NH 4 ) 2 SO 4 , produced crumpled sheet-like morphology of orthorhombic δ-MnO 2 , which were directly grown on the carbon current collector in the absence of any polymeric binder. These particles have a high degree of disordered stacking i.e. turbostratic stacking, as indicated by the circular rings (polycrystalline) in the SAED analysis and the broad XRD peaks. However, these disordered structures realigned to form a well-ordered structure after annealing at 300 °C for 1 h. This phenomenon was clearly observed through the transition from the polycrystalline SAED pattern of un-annealed to the hexagonal single-crystal SAED pattern of the annealed sample.
As shown in Fig. 4 and Table S1, BET analysis and pore size distribution of the un-annealed and annealed MnO 2 were measured. This was carried out to determine the change in the surface area upon the transition from turbostratic stacking to well-organized stacking. The adsorption-desorption of both the un-annealed (turbostratic stacking) and annealed (well-ordered stacking) sample exhibits Type II isotherm, as shown by the large measured C value (BET) in Table S1. Both samples also follow an H 3 hysteresis loop in the range (0.5 to 1.0) P/P o and (0.4 to 1.0) P/P o for un-annealed (turbostratic stacking) and annealed (well-ordered stacking) samples, respectively. According to IUPAC classification, this indicates the presence of sheet-like particles forming slit-shaped pores 44 . Computation of its corresponding BET surface area revealed that the annealed δ-MnO 2  www.nature.com/scientificreports www.nature.com/scientificreports/ (well-ordered structure) had a surface area equal to 246.18 m 2 g −1 greater than the surface area of the un-annealed δ-MnO 2 (turbostratic stacking) which was around 106.05 m 2 g −1 . Moreover, the pore volume (1.3188 cm 3 g −1 ) of the annealed sample increased in comparison to the pore volume of the un-annealed sample (0.7776 cm 3 g −1 ). The increase in the BET surface area and BJH pore volume after annealing can be attributed to the removal of adsorbed and interlayer water molecules, as illustrated in Fig. 5. Hereafter, the term annealed and un-annealed is used to represent well-ordered stacking and turbostratic stacking, respectively, for simplicity. Figures 6 and S5, representing the galvanostatic charge-discharge capacity profile, show the galvanostatic charge-discharge capacity at different current densities. It was observed that the un-annealed sample exhibited a higher initial discharge capacity of about 208 mAhg −1 in comparison to the annealed sample having an initial discharge capacity of 158 mAhg −1 . However, during cycling, in the 2 nd and 10 th cycle, the discharge capacity of the un-annealed sample drastically dropped to 130 mAhg −1 and then dropped down to 88 mAhg −1 , respectively. As for the annealed sample, the discharge capacity did not drastically change. However, in the 2 nd and 10 th cycle, the capacity slowly dropped to 140 and 105 mAhg −1 , respectively.
Subsequently, when the current density increased to 100 mAg −1 , it was found that the un-annealed sample reached a plateau of around 75 and 66 mAhg −1 . Nevertheless, the annealed sample attained a plateau at 84 and 74 mAhg −1 . At 150 mAg −1 , the un-annealed sample approached a plateau at around 56 mAhg −1 and the annealed sample at 60 mAhg −1 . Furthermore, at 200 mAg −1 , the un-annealed sample displayed a plateau around 45 mAhg −1 and the annealed one displayed a plateau around 50 mAhg −1 . Finally, when the applied current density was  www.nature.com/scientificreports www.nature.com/scientificreports/ returned to 50 mAhg −1 , the un-annealed sample showed a plateau around 70 mAhg −1 and the annealed sample showed a higher discharge capacity plateau of 80 mAhg −1 .
After the 50 th cycle (Fig. S6), both un-annealed and annealed samples showed a stable discharge capacity of around 70 mAhg −1 at an applied current density of 100 mAg −1 . During the 1 st and 2 nd cycle, especially for the www.nature.com/scientificreports www.nature.com/scientificreports/ un-annealed sample, discharge capacity abruptly faded. This most probably happened due to irreversible structural/volumetric change 45 during the insertion of Zn 2+ and the dissolution of the active material MnO 2 . This is a common phenomenon since MnO 2 is somewhat soluble in DES 46 . Besides, CV information after 50 th cycle is shown in Fig. S9. The CV result revealed that the structure did not significantly change due to insignificant change in position of the redox peaks. Figure 7 shows the cyclic voltammogram of un-annealed and annealed samples within the voltage window: (0.4-1.9) V measured at a scan rate of 0.5 mV/s (a, c). Prominent peaks around 1.05 V and around 1.5 V were observed which proved to be in good agreement with the results obtained by other researchers 22 . These peaks are attributed to the insertion (discharge)/extraction of Zn 2+ (charge) on the MnO 2 cathode during the discharging and charging of the ZIB. It can also be inferred that overlapping curves of the first 3 cycles in both un-annealed and annealed conditions (a, c) imply good cycling stability of both samples. At low scan rate (b, d), obvious peaks were observed implying redox reaction on the electrodes. However, at high scan rate, the redox peaks are no longer visible, and both samples were showing surface capacitive-like behavior.
To understand the overall governing charge storage mechanism, Eq. (1) was employed:  (Table S2) obtained from the relationship given in Eq. (1), it can be assumed that the governing charge storage mechanism, for both un-annealed and annealed samples in the considered scan rates, is dominated by diffusion. This result is consistent with the b values obtained which are around 0.6 ( Fig. 7e and Table S3) for both conditions and is therefore predominantly diffusion controlled, as in Eq. (2): where a and b are adjustable parameters. If b is approximately equal to 1, the charge storage mechanism is predominantly surface capacitive in nature, whereas if b is approximately equal to 0.5, then it is diffusion controlled [30][31][32][33] . However, of the two considered conditions, the un-annealed sample is more capacitive than the annealed one and to understand this behavior further, the electrochemical impedance of the samples was investigated via EIS. As shown in Fig. 8, the Bode magnitude and phase plots of the samples can be designated into high, mid, and low-frequency regions. Thus, the high-frequency region corresponds to the anodic reaction (>10 3 Hz). The mid-frequency region corresponds to the faradaic reaction (10 1 -10 3 ) Hz. The low-frequency region corresponds to the cathodic reaction (10 −2 -10 1 ) Hz. It can be observed that of the three regions, the cathodic region displayed an obvious difference in magnitude and phase which could be the consequence of annealing the MnO 2 cathode.
Herein, the annealed sample demonstrated lower values of magnitude and phase in comparison to the un-annealed one. On the other hand, fitting of the Nyquist plot using the equivalent circuit model (Fig. S7 and Table S4) showed that the annealed sample (139.8 Ω) had lower resistance in the cathodic region than the un-annealed one (171.8 Ω). This implied therefore that indeed the conductivity of the cathode was enhanced after annealing treatment through the reorientation of the δ-MnO 2 sheets to form a well-ordered stacking.