Effect of Transition Metal Cations on Stability ... - ACS Publications

May 8, 2017 - Department of Physics and Astronomy, Uppsala University, 751 20 Uppsala, Sweden. ‡. School of Engineering and Information Technology, ...
0 downloads 0 Views 7MB Size
Subscriber access provided by CORNELL UNIVERSITY LIBRARY

Article

Effect of transition metal cations on stability enhancement for molybdate based hybrid supercapacitor Teeraphat Watcharatharapong, Manickam Minakshi Sundaram, Sudip Chakraborty, Dan Li, GM Shafiullah, Robert D. Aughterson, and Rajeev Ahuja ACS Appl. Mater. Interfaces, Just Accepted Manuscript • Publication Date (Web): 08 May 2017 Downloaded from http://pubs.acs.org on May 8, 2017

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

ACS Applied Materials & Interfaces is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 43

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Applied Materials & Interfaces

Effect of transition metal cations on stability enhancement for molybdate based hybrid supercapacitor Teeraphat Watcharatharapong1, Manickam Minakshi Sundaram2,*, Sudip Chakraborty1,*, Dan Li2, GM Shafiullah2, Robert D. Aughterson3 and Rajeev Ahuja1 1

Department of Physics and Astronomy, Uppsala University, Sweden School of Engineering and Information Technology, Murdoch University, Murdoch, WA 6150, Australia 3 Australian Nuclear Science and Technology Organization, Locked Bag 2001, Kirrawee DC, NSW 2232, Australia 2

Abstract The race for better electrochemical energy storage systems has prompted to examine the stability in the molybdate framework (MMoO4; M = Mn, Co or Ni) based on a range of transition metal cations from both computational and experimental approaches. Molybdate materials synthesized with controlled nanoscale morphologies (such as nanorods, agglomerated nanostructures, and nanoneedles for Mn, Co and Ni elements, respectively) have been used as a cathode in hybrid energy storage systems. The computational and experimental data confirms that the MnMoO4 crystallised in β form with α-MnMoO4 type whereas Co and Ni cations crystallised in α form with α-CoMoO4 type structure. Among the various transition metal cations studied, hybrid device comprising NiMoO4 vs. activated carbon exhibited excellent electrochemical performance having the specific capacitance 82 F g-1 at a current density of 0.1 A g-1 but the cycling stability need to be significantly improved. The specific capacitance of the NiMoO4 electrode material is shown to be directly related to the surface area of the electrode / electrolyte interface but the CoMoO4 and MnMoO4 favoured a bulk formation that could be suitable for structural stability. The useful insights from the electronic structure analysis and effective mass have been provided to demonstrate the role of cations in the molybdate structure and its influence in electrochemical energy storage. With improved cycling stability, NiMoO4 can be suitable for renewable energy

1 ACS Paragon Plus Environment

ACS Applied Materials & Interfaces

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

storage. Overall, this study will enable the development of next generation molybdate materials with multiple cation substitution resulting in better cycling stability and higher specific capacitance. * M. Minakshi Sundaram (

) E-mail: [email protected]; *Sudip

Chakraborty ( ) E-mail: [email protected]

Keywords: Capacitor; molybdate; stability; structural; computation; theoretical

2 ACS Paragon Plus Environment

Page 2 of 43

Page 3 of 43

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Applied Materials & Interfaces

1. Introduction Renewable energy resources such as solar, wind and hydro have been considered as promising alternatives to the finite fossil fuels (such as petroleum, coal, oil and natural gas) and also have the additional desirable property of reducing greenhouse gas emissions. However, the available renewable energy sources are intermittent due to the variability of these sources and this has led to serious concerns regarding their reliability for supply to an electric grid.

1

The challenges associated with meeting the variation in demand while

providing consistent output has led to the development of energy storage such as rechargeable batteries. Hence, to make the renewable energy sources more viable, integrating energy storage systems with photovoltaic solar panels (for an instance) is vital. Rechargeable batteries and supercapacitors are types of energy storage devices that can store and release energy when required.

2

They convert chemical into electrical energy and have been

commercialised for efficient energy utilization until now. To enable the renewable integration into a large scale deployment, the amount of storage (which is the energy density of the system) depends on the chosen electrode material, and its electrochemistry. 2 The state-of-art lithium-ion battery technology possesses the highest energy density of the current known energy storage devices and has found immediate applications in electric vehicles. However, scaling up this technology for stationary energy storage applications is challenging and expensive, this is due to the use of the non-aqueous electrolytes. Compared with non-aqueous electrolytes, the aqueous medium provides a much higher conductivity leading to higher power density 3 with no formation of solid electrolyte interphase (SEI) layer. Until recently, the most commonly used aqueous electrolytes were 1M H2SO4, 6M KOH and 2M LiOH. 4 In recent years, there has also been a renewed interest in rechargeable sodium-ion batteries in which the hypothesis is quite similar but with Na+ rather Li+ because of concerns regarding 3 ACS Paragon Plus Environment

ACS Applied Materials & Interfaces

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

the safety issues and scarcity of lithium in the earth’s crust.

Page 4 of 43

3

The race for better

electrochemical energy storage systems relies on two typical characteristics which are high power and high energy density. On the other hand, electrochemical capacitors possessing fast charge-discharge performance and constant cycling stability are of primary interest in applications ranging from portable devices to large-scale technologies.

4-5

To improve the

energy density, the current generation capacitors have modified the conventional double layered capacitor electrodes with redox type battery electrodes. In this paper, the hybrid (asymmetric) energy storage system 4 using battery type electrode as cathode coupled with a capacitor type electrode as anode has been reported. The thermodynamic and kinetic limitations have left a large performance gap in hybrid (battery coupled with capacitor) systems. Hence, in this work, computational and electrochemical studies have been carried out to explore and advance the crystal configuration of chosen molybdate cathode material.

Chemical aspects of renewable energy storage The typical transition metal oxide cathodes (MO2) such as cobalt, nickel, and manganese have been extensively researched. 5-7 During the discharge/charge mechanism, the insertion / removal of each Na+ in the cathode is associated by the reduction/oxidation of a transition metal ion to accommodate the compensating intercalation/de-intercalation of an electron.4

The

performance

of

a

hybrid

system4

depends

crucially

on

the

adsorption/desorption of ions from/into the electrolyte along with the insertion/extraction of alkali ions in the host cathode, in the aqueous 2M NaOH electrolyte, and in the electrodeelectrolyte interface. Though the performance characteristics of the hybrid system based on a single electrode (three electrode configuration) have been reported 4-8 but it has not been well substantiated with the support of a computational approach. Also, the electrochemical performance available for the hybrid device for the molybdates is scant. generation oxide materials

5-7

9-10

The last

suffer from low rate capability due to its relatively high 4

ACS Paragon Plus Environment

Page 5 of 43

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Applied Materials & Interfaces

resistance against lithium/sodium diffusion. On the other hand, the current generation phosphate framework (MPO4)

11-12

generally offers high structural stability as well as high

cathode redox potential and they are found to be attractive for battery applications. In particular, phospho-olivine type LiFePO4 11, has excellent thermal stability and rate capability but its low discharge voltage, low ionic and electrical conductivity greatly limited its electrochemical performance. Other transition metal cations 12-14 have also been investigated in olivine compounds but their combined properties including power density, energy density, and structural stability are not comparable for practical applications. To overcome the drawbacks, in recent years, metal molybdates have attracted great interest in both lithium-ion batteries and supercapacitors.

15-16

Molybdate materials are well

known for excellent catalytic activity exhibiting good electrochemical properties and they are also environmentally friendly. Manganese (molybdate) is one of the most cost-effective and abundantly available electrode materials with high cycling stability. However, its poor electrical conductivity and low specific capacitance have limited its application.

17

Cobalt

molybdate has attracted increasing interests due to its fast reversible redox reaction 9 which is a pre-requisite for capacitor applications. Nevertheless, the poor conductivity and cycle stability of CoMoO4 limits its application. Nickel molybdates gained great interest due to their high redox potential, low cost, low toxicity and environmentally friendliness.

8, 16

Although the electrochemical studies for molybdates as cathodes have been reported for supercapacitor and battery applications 15-18, to the best of our knowledge, no work has been reported for a range of transition metal cations (MMoO4; where M = Mn, Ni or Co) outlining the advantages and limitations for each of the systems from chemical aspects that is suitable for hybrid devices. Moreover, in this work, we have shown the computational studies on individual transition metal molybdates for studying faster ion conduction, and how these methods have provided useful insights into hybrid device design. The computational results 5 ACS Paragon Plus Environment

ACS Applied Materials & Interfaces

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 6 of 43

are supported by both physicochemical and electrochemical properties of the synthesized molybdate materials. However, the focus of this work is not to compare and draw a benchmark of the reported capacitance values available in the literature which could be a very routine study. 2. Methodology 2.1

Computational

In this work, the ground state properties of the MMoO4 systems (where M = Mn, Co or Ni) are studied through electronic structure calculations based on density functional theory (DFT) formalism.

19

The projector-augmented wave (PAW)

20

framework has been used as

implemented in the Vienna Ab-initio Simulation Package (VASP).

21

The energy cutoff of

500 eV for the plane wave basis set and the Monkhorst-Pack k-point mesh 22 of 2 x 2 x 4 are consciously chosen through convergence test so as to approximate Brillouin zone integrals. We have started our DFT calculation using PBE (Perdew Burke Erzerhof) type GGA (Generalized Gradient Approximation) exchange correlation functional

23

and found that the

density of states (DOS) for CoMoO4 manifests metallic behavior while the 3d-electrons in three compounds undergo the delocalization. Owing to the self-interaction error, the DFT +U functional is requisite to be used so that the correlation effect for the d-electrons is properly described. The effective interaction parameters (Ueff = U - J) for Mn, Co, and Ni are 3.9, 3.4, and 6.0 eV, respectively within the Dudarev framework

24

as chosen from the reported

results. 25-26 All configurations are fully relaxed until the total energy is less than 10-5 eV and the Hellmann-Feynman forces are below 0.005 eV/Å, in order to find the optimized geometries of each case. It is important to mention that all calculations are spin-polarized in order to obtain the magnetic ground states of the molybdate systems. While performing the spin polarized calculations for both FM (Ferromagnetic) and AFM (Anti-ferromagnetic) configurations, it was found that the AFM analogue of MnMoO4 and CoMoO4 are relatively 6 ACS Paragon Plus Environment

Page 7 of 43

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Applied Materials & Interfaces

more stable than the FM with the difference in total energy of 0.005 eV/f.u., while for NiMoO4, the energy difference is 0.001 eV/f.u. This is in agreement with the experimental observation of α-NiMoO4 exhibiting AFM ordering below 19K.

27

Therefore, only AFM

configuration will be considered for all molybdate systems in this work. To have a profound understanding about charge transfer, we use the Bader charge approach 28, where an amount of charge on each atom is considered in the region separated by the minimum of electronic charge density between each atom. 2.2 Experimental

(a) Synthesis of MMoO4 (M = Mn, Co, or Ni) materials Three transition metal cation molybdate (MMoO4; M = Mn, Co, or Ni) samples were combustion synthesised using analytically pure M(NO3)2.6H2O; (NH4)6Mo7O24.4H2O and urea (as a fuel) in stoichiometric ratios, with all chemicals supplied by Sigma Aldrich. The metal to fuel ratio was kept constant (1:1) for all the samples. The metal and molybdate reactants were dissolved in 20 mL of de-ionised water at 80 °C with an effective stirring to obtain a homogenous solution. The pH of the solution was adjusted by drop wise addition of ammonia solution. Subsequently, the mixture was dried at 150 °C in a hot air oven for 12 hours. The MMoO4 material was synthesised at 300 °C for 3 hours. All the single metal molybdates were synthesized under identical conditions.

(b) Physical characterisation The materials synthesised were characterized extensively by physical and electrochemical techniques. X-ray diffraction (XRD) was used to identify the crystal structure of synthesized

MMoO4 product using a PANalytical X’pert Pro powder diffractometer housed at ANSTO with an X’celerator high-speed detector. Samples were placed on zero-background aluminium sample holders, scanned using a Cu-kα radiation source (λ= 1.5418 Å) with a step size of 0.02o. The voltage and current were 45 kV and 40 mA, respectively. A high magnification Zeiss Neon 40ESB Field Emission Scanning Electron Microscope (FESEM) 7 ACS Paragon Plus Environment

ACS Applied Materials & Interfaces

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

instrument was also used to acquire morphological and microstructure information of samples. The nanostructure and lattice imaging of the MMoO4 materials was characterised by transmission electron microscopy (TEM). TEM specimens were prepared by grinding a small amount of molybdate powder material under ethanol in an agate mortar and pestle and dispensing onto a holey carbon film. High-resolution images were collected using a JEOL 2200FS TEM operated at 200 keV. Pore structures and surface area of the samples were characterized by nitrogen adsorption/ desorption Micromeritics Tristar II Surface area and porosity analyser. Before analysing, the molybdate samples were degassed at 100 °C overnight.

(c) Electrochemical characterisation For electrochemical measurement, the electrode was prepared by mixing MMoO4 (85 wt. %), carbon black (10 wt. %) and PVDF (5 wt. %) with 0.4 mL of NMP to make slurry. This was coated on a small piece of carbon sheet (area of coating, 1 cm2). The remainder of the carbon strip was masked using an insulation film to obtain a coated surface area of 1 cm2 exposed to the aqueous (NaOH) electrolyte. The loaded active material was 11.25 mg in each case. To prepare the active negative electrode material, activated carbon (AC) (90 wt. %) and PVDF (10 wt. %) were used. A full cell termed as “hybrid device” (AC||MMoO4) was fabricated by using polypropylene separator. The cyclic voltammetry and galvanostatic charge-discharge studies of the composites were carried out using SP-150, Bio-Logic Science instruments in 2M NaOH electrolyte at room temperature. For the three-electrode tests, a platinum wire of 10 cm length and 1 mm diameter in dimension and mercury–mercuric oxide (Hg/HgO) served as the counter and reference electrodes, respectively. MMoO4 served as the working electrode. Galvanostatic chargedischarge cycles (two electrode systems) were performed using an 8-channel battery analyser from MTI Corp. USA. The cyclic voltammetric profiles of MMoO4 were obtained at various

8 ACS Paragon Plus Environment

Page 8 of 43

Page 9 of 43

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Applied Materials & Interfaces

scan rates between 1 mV s-1 and 20 mV s-1 within the voltage range between 0 and 0.75 V. Charge–discharge studies of the hybrid device were carried out at various current densities between 0.1 A g-1 and 5 A g-1. The cut-off charge and discharge voltages were 1.6 (for Mn and Co cations) and 1.8 V (for Ni cation) and 0 V, respectively. Specific capacitance and energy density of the device were calculated at the end of each discharge. For activated carbon (AC), the working electrode was cycled between 0 and -1.0 V at scan rates between 1 mV s-1 and 5 mV s-1. For a hybrid device, in order to maintain the charge conservation between the two electrodes, the mass ratio was calculated using the equation (1) m+ /m ̶ = (C ̶ * ∆E ̶ ) / (C+ * ∆E+)

Eq. (1)

where m represents the mass in g, C- and C+ represents the specific capacitance for the AC and MMoO4, respectively; ∆E- and ∆E+ the discharge/charge potential range for the AC and

MMoO4 electrodes, respectively. The specific capacitance of the single electrode or hybrid device (asymmetric capacitor) were calculated from the charge-discharge curves using the equation (2) C = I t / m ∆V

Eq. (2)

where I (A) represents the discharge current, t (s) is the discharge time, m is the mass of the individual electrode or the sum of the AC and MMoO4 electrode corresponding to the hybrid device. The energy density of the hybrid device can be calculated using the equation (3) E = 0.5 CV2 / 3.6

Eq. (3)

where E is the energy density (Wh kg-1), C represents the specific capacitance and V is the voltage window during the discharge process. The specific capacitance of the AC and MMoO4 after subtracting from contributions of carbon black (acetylene black) and binder were calculated to be 210 F g-1 (for activated 9 ACS Paragon Plus Environment

ACS Applied Materials & Interfaces

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

carbon) and 120 F g-1 (for Mn), 240 F g-1 (for Co), and 412 F g-1 (for Ni) respectively. Based on the single electrode characteristics, from the above equation (1), the optimal mass ratio between AC and MMoO4 was determined to be 3.24 (for Mn), 1.37 (for Co), and 0.86 (for Ni) for the fabricated hybrid capacitor. Therefore, for a mass of 15 mg MMoO4 material the mass of the AC was varied between 4.65 mg (for Mn), 11 mg (for Co), and 17.5 (for Ni) respectively. This corresponds to different mass ratios of positive and negative electrodes (Mmolybdate:MAC = 1:0.3 (Mn), 1:0.75 (Co); 1:1.15 (Ni). 3. Results and Discussion 3.1. Theoretical Investigation 3.1.1. Crystal structures and corresponding energetics In general, there are two main polymorphs belonging to centered monoclinic C2/m space group for the MMoO4 framework, one is isotypic with α-CoMoO4 type structure known as “α-phase”, while the other is with α-MnMoO4 type homologue known as “β-phase”. The GGA calculations reported by S.F. Matar and co-workers

27

showed NiMoO4 and MgMoO4

were more energetically stable in α- and β-phases, respectively. Likewise, experimental validations also showed Mg- and Mn-based compounds crystallize in the β-phase while the Co- and Ni-based crystallize in α-phase.

29

This is owing to the larger size and lower

electronegativity of M cations (XM), where the latter could be ordered in the series of Mg (1.31) < Mn (1.55) < Co (1.88) < Ni (1.91). However, the abrupt phase transition from α- to β- can take place around ~ 400 °C and ~ 600 °C in the case of Co and Ni, respectively, and those phases might coexist at room temperature.

29- 30

Thus, all the compounds considered in

this work have been computationally investigated for both the phases in order to perceive their correlations between energy trend and structural information. With the help of fully structural optimization, the crystallographic data for all three cations in the molybdate systems are obtained and shown in Table 1. Clearly, it can be seen that α- and β-phase structures possess the β angles of 114° and 106°, respectively, as one of 10 ACS Paragon Plus Environment

Page 10 of 43

Page 11 of 43

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Applied Materials & Interfaces

determining parameters that can differentiate both isomorphs. The experimental lattice parameters evaluated from the X-ray diffraction patterns (XRD) (discussed in section (3.2.1) supports the β angle values confirming that Ni- and Co-based compounds crystallize in αphase whereas Mn-based analogue crystallize in β-phase. Although the calculated lattice parameters, angle, and volume are slightly overestimated as a consequence of GGA+U approximation, however, they are still in good agreement with our experimental findings and the reported values in ref. 31. Table 1 also shows calculated Eg of three systems in the presence of Hubbard Ueff parameters along with reported experimental values.32-33 It is important to note that the Hubbard Ueff parameters are not introduced on the purpose to manipulate the Eg but actually to obtain an enhanced DOS, which is detailed in section (3.1.2). In table 1, our calculated Eg for each compound is well aligned with the corresponding experimental values, notably for MnMoO4, the value is very close to α-phase. Moreover, the cell volume and Eg for α-phase are likely to be smaller than those for β-phase. This can be attributed to the anisotropic coordination of Mo which is octahedral MoO6 for αphase but becomes tetrahedral MoO4 for β-phase. The magnitude of Mn2+, Co2+, and Ni2+ spin moments are approximately 4.61, 2.73, 1.80 µB, respectively, as shown in Table 1, and they all prefer to possess high-spin states. Figure 1(a) and (b) illustrates the optimized configurations of MnMoO4 in β-phase and NiMoO4 in α-phase, respectively. Only NiMoO4 is shown in the figure due to the structural similarity between Co- and Ni-based compounds. From the side-view perspective in Figure 1, four octahedral MO6 are connected by edge-sharing and can be regarded as a structural unit of 4MO6 in both isomorphs. Each of this unit is linked and separated by MoOx units in different ways as it can be clearly noticed from their anisotropic co-ordinations of Mo. In this unit, Mo ions are tetrahedrally coordinated (MoO4) by oxygen ions in α-phase and octahedrally coordinated MoO6 in β-phase. Four octahedral MoO6 are closely packed and

11 ACS Paragon Plus Environment

ACS Applied Materials & Interfaces

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

connected in a similar way to 4MO6 in α-phase, whereas each tetrahedral MoO4 is preferably at distances sharing the corners with MO6 in β-phase. These structural demarcations in

MMoO4 frameworks are the prime factors to describe the variation of their M2+/3+ redox potentials that will be discussed subsequently. Moreover, one can see from the top-view perspective that MoO4 tetrahedron in MnMoO4 is originated from two Mo-O broken bonds in MoO6 resulting in an increased volume and a smaller β angle. Interestingly, the cooperative distortion in MnMoO4 leads to the creation of an additional channel for the intercalation process and, thus, enriches the electrochemical activities through diffusion mechanism. Figure 2(a) demonstrates the decreasing tendency of M-O and M-M bond distances from Mnto Ni-based compounds (the left vertical axis), and the presence of octahedral Jahn-Teller distortion coefficient of M-O unit (∆(M-O)), for all compounds in both phases. As similar to the trend of bond distances, the ∆(M-O) is also getting lower with the increase of atomic number. The Jahn-Teller distortion has occurred more distinctly in Mn- and Co-based compounds while both isomorphs of Ni-based are having smaller value of ∆(M-O). We assume that this distortion coefficient could relate to the redox processes. The low value of ∆(M-O) in NiMoO4 reflects the fact that its redox reaction is unlikely to occur through intercalation in bulk structure 34 while the high value leads to a spacious structure with larger cavity aiding the diffusion process for Mn molybdate. Thus, Mn molybdate enhances the possibility to store charges via the redox intercalation reaction. More details on the electrochemical activity of molybdates and its effect to cation are discussed in section (3.1.4). Regarding their energetics, the total energy difference between MMoO4 in α- and βphases is here defined as ∆Etot = Etot (α) − Etot (β) . In Figure 2(b), the ∆Etot calculated with different exchange correlation functions are relatively plotted in order to check the influence of Hubbard Ueff parameter on Mo (UMo=3.5eV) and to perceive the energy trend of their favorable phases. The resulting ∆Etot within PBE functional indicates that CoMoO4 and 12 ACS Paragon Plus Environment

Page 12 of 43

Page 13 of 43

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Applied Materials & Interfaces

NiMoO4 are more energetically favorable in α-phase whereas MnMoO4 in β-phase, which is consistent with our experimental findings. However, it can be seen that the absence of UMo gives a contradicting result showing Mn-, Co-, and Ni-based compounds prefer forming in βphase. Fortunately, the inclusion of UMo is found to resolve this contradiction by providing the actual energy trend. This is, therefore, an evidence of the effect of UMo parameter that essentially influences the phase stability of molybdate systems and also signifies the more pronounced 4d-electron delocalization of Mo in α-phase compounds. In order to maintain the actual trend of phase stability, the appropriate UMo parameter should be necessarily applied. Nevertheless, UMo does not make any prominent changes in DOS besides the marginal increase of intensity of Mo in PDOS. Furthermore, we have computed the cohesive energy (Ecoh) to assess and compare the relative structural stability of MMoO4. As compared to the other two compounds, MnMoO4 emerges with highest stability having lowest Ecoh, following by Co- and Ni-based molybdates. Possibly, it implies that MnMoO4 is likely to preserve its bulk characteristic and tends to merely undergo surface relaxation. Conversely, because of the distinctly higher Ecoh for Co and Ni cations, it is anticipated that the surface reconstruction possibly occurs to help minimizing the total energies and, thus, leads to a larger surface area. Moreover, the small difference seen in Ecoh between the two phases confirms the possibility of phase transition and a co-existence of two-phases in Co- and Ni-based molybdates as previously found in [29]. The morphology and surface area of electrode materials have significant effects on their specific capacitances. The decrease in particle size can increase the number of active sites leading to the initiation of ball milling process that will help enhancing both electronic conductivity and specific capacitance of materials. This is because its electrochemical storage, namely electrical double layer (EDL) or pseudocapacitance, takes place dominantly on the electrode surface. Referring to the calculated cell volume shown in Table 1, we find

13 ACS Paragon Plus Environment

ACS Applied Materials & Interfaces

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

that NiMoO4 and MnMoO4 are the smallest and largest crystal lattices. In addition to the cohesive energy, this result further reinforces that the Ni based molybdate is regarded as an extrinsic pseudocapacitor in which its pseudocapacitive behavior apparently depending on the amount of surface area. The implication of Ecoh and cell volume is well corresponding to the tendency of high specific surface area and specific capacitance obtained from our experimental measurement that is discussed in section (3.1.4). 3.1.2 Projected Density of States Analysis In order to envisage the hybridization in these three molybdate compounds, site projected density of states (PDOS) are determined and shown in Figure 3. Due to the similar features of PDOS, M and Mo species residing in two non-equivalent crystal sites, namely iand h-sites, will be grouped. The introduction of different Ueff values in this material primarily affects the energy shift of the valence band maximum (VBM) and conduction band minimum (CBM), and therefore the band gap value. For all systems, the valence band within the region between -6 and -0.5 eV is dominated mostly by O-2p contribution, whereas the conduction band is originated from the hybridization between Mo-3d and O-2p orbitals in CB. In the case of MnMoO4, the electron contribution in the energy interval between -6 to 3.9 eV shows a strong interaction between Mo-3d and O-2p orbitals and is found to originate from the tetrahedral MoO4 unit. Since this contribution of the MoO4 unit becomes less and less in Co- and Ni-cased compounds, we believe that it certainly plays a crucial role in stabilizing MnMoO4 structure and can be correlated with its phase stability suggested by the cohesive energy calculations. Moreover, a very small electron contribution of Mo near the VBM indicates that it is not involved in redox reactions of MMoO4. We also observed that the PDOS intensity of O-2p increases while the contribution of M-3d decreases near the VBM region. This result in the weaker hybridization between these orbitals in case of NiMoO4 where Ni-3d contribution is not substantial near the VBM and leads to the less electronic conductivity compared to the other cations. Despite the lower intensity of O-2p 14 ACS Paragon Plus Environment

Page 14 of 43

Page 15 of 43

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Applied Materials & Interfaces

near the VBM, the overlap of M-3d and O-2p orbitals reveals their strong hybridizations and uniform electron localization around the VBM in Mn- and Co-based compounds. 3.1.3. Electron effective mass The shape of the band structure, particularly, at the VBM and CBM determines the effective mass of hole and electron (m*), and subsequently the charge carrier mobility. In this work, we have evaluated the relative conductivity of bulk-MMoO4 on a basis of the effective mass entity as it can be used to interpret the electron acceleration or carrier mobility under a periodic potential of the crystal in more convenient way. 35 In principle, the effective mass is , where d 2 E / dk 2 is the

determined using the dispersion relation (E-k) of

second derivative of the energy with respect to the distance in reciprocal space and

is the

reduced Planck constant. Here, we employed the parabolic approximation to evaluate the second derivative of E-K in x, y, and z direction. 36 Figure 4 shows the individual electronic band structure of MMoO4 (M = Mn, Co, Ni) along the corresponding high-symmetry points. One can see all three compounds have indirect band gap where their VBM and CBM are located at different symmetry points. As for the CBM, it appears at Γ(0 ,0, 0), Y(0 ,0, ½), and A(-½, ½, 0) symmetry points, for Mn-, Co, Ni-based molybdates, respectively. We find that the symmetry points representing the VBM and CBM, and the band structure profile for CoMoO4 turn to resemble those for NiMoO4 when UCo is increased to 5.7 eV. Based on the electron effective mass of these three bulk systems, it shows that NiMoO4 has lowest electronic conductivity due to its relatively higher m* as well as larger band gap, which required more external energy to thermally excite an electron to the conduction band. As the electronic conductivity influences the electrochemical properties of electrode materials

8, 9, 17

, NiMoO4 should have exhibited

inferior electrochemical properties among the other cation studied. Nevertheless, this is not in agreement with our experimental counterpart that reveals NiMoO4 as a superior 15 ACS Paragon Plus Environment

ACS Applied Materials & Interfaces

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

supercapacitor in terms of higher redox potential and specific capacitance. This implies that the existing electrical conductivity is enhanced through the contribution of surface active sites more significantly in Ni cation than that in Co- and Mn-based compounds. Furthermore, the anisotropy of electron conduction is observed for all three systems in the order of m*(z)