Cobalt Doping as a Pathway To Stabilize the Solid-State ConversionChemistry of Manganese Oxide Anodes in Li-Ion BatteriesAlessandro Palmieri,‡ Sajad Yazdani,§ Raana Kashfi-Sadabad,∥ Stavros G. Karakalos,⊥

Michael T. Pettes,§ and William E. Mustain*,†

†Department of Chemical Engineering and ⊥College of Engineering and Computing, Swearingen Engineering Center, University ofSouth Carolina, Columbia, South Carolina 29208, United States‡Department of Chemical & Biomolecular Engineering, University of Connecticut, Storrs, Connecticut 06269-3222, United States§Department of Mechanical Engineering, University of Connecticut, Storrs, Connecticut 06269-3139, United States∥Institute of Material Science, University of Connecticut, Storrs, Connecticut 06269-3136, United States

*S Supporting Information

ABSTRACT: Metal oxides have been widely studied in recent years to replace commercial graphite anodes in lithium ionbatteries. Among the metal oxides, manganese oxide has a high theoretical capacity, low cost, and is environmentally friendly.However, many MnO materials have shown limited reaction reversibility and poor conversion kinetics. To understand why, inthis paper we investigate the mechanism, kinetics, and reversibility for the solid-state conversion reaction of MnO with Li+. Wedefinitively show, for the first time, that during repeated reaction cycles, multiple reaction pathways occur that lead not only tothe reformation of MnO but also higher oxidation-state Mn3O4which when combined with the poor intrinsic electronicconductivity of both manganese oxide species results in a rapid loss in the amount of charge that can be stored in these materials.Learning this, the approach in this study was to use cobalt doping to concomitantly stabilize the redox behavior of manganese(allowing for the gradual transformation of MnO to Mn3O4 over time) and to increase the intraparticle electronic conductivity ofthe active layer. The result is an active material, Mn0.9Co0.1O, that exhibits excellent charge stability and conversion kinetics (near600 mAh/g at a rate of 400 mA/g), even over hundreds of reaction cycles.

INTRODUCTION

Li-ion batteries (LIBs) have been the dominant energy sourcein the portable electronic devices market since they werecommercialized by Sony in the early 1990s,1 due to their highspecific energy and high specific power densities.2 CommercialLi-ion batteries utilize a graphite anode, lithium metal oxidecathode, and LiPF6 salt dissolved in a mixture of organiccarbonates as the electrolyte. These batteries can supply up to150 Wh/kg, which is 5 times lower than the requirement forlong-range fully electric vehicles and far away from otheremerging applications.3,4 Therefore, there is the need to findalternative materials to increase Li-ion battery energy andpower density without sacrificing longevity and compactness.The cathode side of the cell presents very few choices of

materials that are suitable to be used in LIBs (typically LiCoO2,LiMnO2, LiFeO4), all with a theoretical capacity lower than 300

mAh/g. This is due to the fact that the active compound mustinclude lithium, be resistant to corrosion, and have a highreversible Nernst potential. In contrast, the anode side of thecell presents a wide variety of chemistries with theoreticalcapacities significantly higher than that of graphite that could besuitable for Li-ion battery operation. In fact, several classes ofmaterials have been proposed in recent years including pureelements/metals (Si, Ge, Sn, Sb)5,6 as well as metal fluorides,7

nitrides,8 phosphides,9 hydrides,10 and oxides.11−13

Among the list above, one of the most promising categoriesof compounds to replace graphite in the LIB anode is metaloxides, and manganese oxides in particular, due to the multiple

Received: January 12, 2018Revised: February 27, 2018Published: March 16, 2018

Article

pubs.acs.org/JPCCCite This: J. Phys. Chem. C 2018, 122, 7120−7127

© 2018 American Chemical Society 7120 DOI: 10.1021/acs.jpcc.8b00403J. Phys. Chem. C 2018, 122, 7120−7127

Dow

nloa

ded

via

UN

IV O

F C

ON

NE

CT

ICU

T o

n O

ctob

er 2

, 201

8 at

02:

54:4

1 (U

TC

).

See

http

s://p

ubs.

acs.

org/

shar

ingg

uide

lines

for

opt

ions

on

how

to le

gitim

atel

y sh

are

publ

ishe

d ar

ticle

s.

advantages that they offer: they are relatively inexpensive,highly abundant in nature, and environmentally friendly.However, not all metal oxides behave the same, and Li storagecan occur through three different mechanisms: intercalation,alloying, and conversion.14 Conversion metal oxides undergo abond-breaking reaction with Li+, which causes a chemicaltransformation from the oxide to the metallic state, followingthe general formula given in eq 1

+ + ↔ +−y y x yM O 2 Li 2 e M Li Ox y 2 (1)

Although this is a solid-state reaction that leverages chemicaltransformations (where bonds are broken and formed) to storeand deliver energy, the overall anode electronic conductivitycan be reaction-limiting for two reasons. First, metal oxides aretypically semiconductors; therefore, their intrinsic electronicconductivity is low (10−8−10−3 Ω−1cm−1). Second, theformation of the Li2O phase is problematic. Above a few nmin size, Li2O is not electrically conductive; the result of which isthe trapping of the material in the metallic (charged) state,resulting in very poor reaction reversibility that would meanrapid fade in battery capacity with cycling and limited cycle life.For these reasons, raw manganese oxides have shown very poorcapacity retention with deep capacity fade even after only 10cycles of operation.15

In order to improve MnO electrochemical kinetics andreaction reversibility, various approaches have been tried, suchas creating complex high surface area nanostructures. Li et al.synthesized interconnected porous MnO nanoflakes on nickelfoam, retaining 400 mAh/g over 100 cycles at 1230 mA/g16

(approximately a 2C rate). Cui et al. have developed a radio-sputtering technique to deposit a MnO thin film on top ofcopper electrodes, showing 400 mAh/g at a C/20 currentrate.17 Although these nanostructures may help reactionreversibility by shortening the Li-ion diffusion pathway,showing good results at low current rates, they still were notable to achieve high capacities at fast rates due to the limitedelectronic conductivity. To overcome the limited electrodeconductivity, one popular strategy consists of pairing the MnOactive material with a highly conductive matrix such asgraphene oxide or carbon nanotubes (Figure S1 in theSupporting Information), which dramatically increases theinterparticle electronic conductivity,18 resulting in higherreaction reversibility and therefore higher capacity. Increasingthe electrode conductivity has also been suggested to helpdecrease the domain size of the Li2O + Metal phases andtherefore improve the metal oxide conversion reactionreversibility19,20 as well as provide a buffer for the volumetricexpansion that occurs during lithiation/delithiation.21−23

Although synthesizing metal oxide/carbon composites maybe one of the best strategies to achieve stable electrochemicalconversion in some metal oxide anodes, this is not sufficient forMnO because of its very poor intrinsic electronic conductivity(10−8 cm−1Ω−1, one of the lowest among all metal oxides).Therefore, a more advanced strategy to increase the reactionreversibility of MnO, and metal oxides in general, is codopingthe active material in order to manipulate its intraparticleelectronic conductivity.24,25 One element with proven electro-chemical synergy with Mn is Co. A relevant example hasrecently been reported for pseudocapacitors,26,27 where Co canstabilize the redox behavior of manganese oxide, inhibitinganodic dissolution and improving stability over many cycles.Also, the electronic conductivity of Co-derived oxide species ishigher than that of Mn-based oxides. Adding just 10 atomic %

Co to MnO increases its intrinsic conductivity by more than anorder of magnitude, as confirmed by Van Der Pauwmeasurements28 (Table S1 in the Supporting Information).However, to the best of our knowledge, no studies have been

conducted either on the influence of doping on the kinetics ofthe MnO conversion reaction or on its reaction reversibility,leaving these as important open questions to address. In orderto do so, five MnxCo1−xO/CNT materials with differentpercentages of Co inclusion (x = 0, 0.05, 0.1, 0.15, and 0.2)were synthesized and characterized both physically andelectrochemically.

EXPERIMENTAL DETAILS

The samples were prepared in two steps.18,29 First, multiwallcarbon nanotubes (MWCNTs, Sigma-Aldrich cataloguenumber 724769) were oxidized in a manner similar to amodified Hummers’ method with a lower concentration ofoxidizing agent30,31 (details in the Supporting Information). Inthe second step, 90 mg of the oxidized MWCNTs were welldispersed in a 122.5 mL ethanol/DI water (50:1 volume ratio)solution for 1 h, in order to achieve a final CNT loading of 10wt % for each sample. Next, 3 mL of manganese(II) acetatetetrahydrate (0.6 M in DI water) solution was added. Toachieve the desired level of Co incorporation in Mn,stoichiometric moles of Mn precursor were replaced withcobalt(II) acetate tetrahydrate. The mixture was refluxed in anoil bath at 90 °C for 24 h after adding 2.5 mL of ammoniumhydroxide solution. The mixture was centrifuged, and the solidswere dried in vacuum at room temperature for 80 h. Lastly, theMnxCo1−xO/CNT anode material was obtained by annealingthe dried sample at 600 °C for 3 h in an argon atmosphere.

RESULTS AND DISCUSSION

Figure 1a shows the XRD patterns for the MnxCo1−xO/CNTmaterials with various cobalt inclusion. The peak positions forthe entire material set are reported in Table S2 of theSupporting Information. The 0, 5, and 10% samples show anaccurate coincidence with the reported peak position in theliterature32 for a MnO face centered cubic (FCC) crystalstructure, showing all of the dominant peaks at 2θ = 35.7, 40.3,58.7, 70.2, and 73.8°. The XRD pattern for the 15% Coimpregnated material shows a slight displacement to higherangles, which indicates the presence of a second material(cobalt) that alters the d-spacing and therefore the crystalstructure of the MnO active material. There is also peakwidening and what appears to be a slight disruption of the(111) peak at the higher 2θ edge, which suggests the possibilityfor incomplete phase separation. Regarding high Co content(20%), a dramatic change is observed; indeed, two sets of peaksare clearly present. The first set can still be ascribed to MnO,although shifted due to the introduction of Co into the FCCMnO crystal structure. The second set of peaks at 2θ = 35.7,41.4 and 60.2°, corresponds to the (111), (200), and (220)Miller planes of cubic CoO.33

Figure 1b shows a typical XPS survey for Mn0.9Co0.1O/CNT.The results confirm the presence of Mn and Co on the activematerial surface in the expected atomic ratio (10%, Figure S2 inthe Supporting Information). Moreover, the Mn highresolution scan contained only two peaks that correspond tothe 2p(1/2) and 2p(3/2) orbital energies of MnO.34 Peakdeconvolution and the peak displacement (11.8 eV) confirmsno metallic or higher oxidation states of manganese (3+, 4+)

The Journal of Physical Chemistry C Article

DOI: 10.1021/acs.jpcc.8b00403J. Phys. Chem. C 2018, 122, 7120−7127

7121

were present on the active material surface before electro-chemical experiments were done.Figure 2a shows a TEM image of Mn0.9Co0.1O nanoparticles

impregnated onto the CNT matrix. The image confirms a lowloading of carbon nanotubes present in the active material andshows a pseudospherical shape of the active material nano-particles. The average particle size is between 25 and 40 nm,which is in good agreement with Scherrer equation calculationsfrom the XRD pattern (Table S3 in the Supporting

Information). Moreover, there was not a significant increasein the lattice parameter or in the particle size for Mn0.85Co0.15Oor Mn0.8Co0.2O when compared to Mn0.9Co0.1O. In fact, theapparent “MnO” domain size decreased as Co was added athigh levels, which confirmed that for higher Co content, cobaltwas not doped into the FCC MnO structure but instead phase-separated.Figure 2b presents a high magnification image of the

Mn0.9Co0.1O nanoparticles where the lattice structure of theactive material is clearly observed. The particle shows a d-spacing of 1.66 A° (in good agreement with XRD results) andan almost spherical shape, which has been shown to helpdispersion and adhesion of metal oxide nanoparticles on carbonsubstrates.35 Figure 2c−f shows the STEM mapping for carbon,manganese, oxygen, and cobalt, respectively, using the exactTEM grid position as Figure 2a. The analysis shows thehomogeneous presence of cobalt throughout the activematerial, which is greatly helped by the fact that CoO issoluble on MnO over the entire range of MnxCo(1−x)O

36

compositions (0 < x < 1), with a complete solubility at roomtemperature and a maximum miscibility gap in the calculatedphase diagram occurring at 242 K, due to a similar Co2+ andMn2+ cation size and rock salt structure.37 However, somephase separation can occur on the surface under the presentconditions, as shown in this and other studies.36 Moreover, theoxygen mapping matches all of the nanoparticles, supportingthe XPS result that metallic manganese is very rarely present, ifat all.The electrochemical characterization of the chemistry,

kinetics, and reversibility of the MnxCo1−xO/CNT conversionreaction is shown in Figure 3. Figure 3a−c shows the cyclicvoltammograms for MnxCo1−xO/CNT with 0, 10, and 20% Coinclusion, respectively. These were selected because they arerepresentative cases of no doping (3a), inclusion in the materialwithout showing any phase separation (3b), and clear phaseseparation (3c). There are some common features that can beobserved regardless of the cobalt concentration. In the firstcathodic scan, a large peak at 0.05 V is observed, correspondingto the formation of the solid electrolyte interphase (SEI). In theMn0.8Co0.2O voltammogram, two peaks are present in the firstpolarization. The smaller peak at 0.2 V is characteristic in shapeand position to SEI formation on CoO,38 further confirmingphase separation at high cobalt content; the larger peak at 0.05V can still be ascribed to SEI formation on “MnO”. Insubsequent cathodic scans, two peaks were always observed forall materials, the first at 0.45 V and the second at 0.3 V. In thereverse (anodic) scan, two oxidative peaks were alwaysobserved, a large peak at 1.3 V and a smaller peak at 2.0 V.With regards to the cathodic peaks, there is much evidence in

the literature that suggests the peak at 0.45 V can be ascribed tothe conversion of MnO to metallic Mn and Li2O

39 (light brownarrow in Figure 4). The secondary peak at 0.3 V is notcorrelated to CoO conversion,35 since it was detected in allcases, even when no cobalt was present in the active material,suggesting the possibility of a higher oxidation state of Mnbeing present after only one redox cycle.The most likely higher oxidation-state Mn species to be

formed are Mn3O4 and MnO2. There are two compelling piecesof evidence to support Mn3O4 as the secondary oxide phase.First, Figure S3 in the Supporting Information shows the cyclicvoltammograms for Mn3O4 supported on CNTs. All five scansshow a clear single cathodic peak at 0.3 V and two anodic peaksat 1.3 and 2.2 V, which is consistent with what was observed in

Figure 1. (a) XRD pattern for MnxCo1−xO materials with 0, 5, 10, 15,and 20% Co addition. (b) XPS general survey spectrum and Mn highresolution scan for Mn0.9Co0.1O.

Figure 2. (a) TEM image of Mn0.9Co0.1O/CNT. (b) Highmagnification image of one single Mn0.9Co0.1O/CNT nanoparticle.STEM mapping images for the particles in (a) showing the elementscarbon (c), manganese (d), oxygen (e), and cobalt (f).

The Journal of Physical Chemistry C Article

DOI: 10.1021/acs.jpcc.8b00403J. Phys. Chem. C 2018, 122, 7120−7127

7122

Figure 3a−c. Second, to better understand the MnxCo1−xOreaction behavior, Mn0.9Co0.1O was cycled 300 times between 0and 3 V vs Li/Li+ to allow a large amount of the secondaryphase to accumulate on the surface, and the resulting specieswere probed by XPS (Figure S4 in the SupportingInformation). It was possible to deconvolute the Mn 2p regionin two doublets, one that is ascribed to MnO and the othercorresponding to Mn3O4. Moreover, the peak sets show almostthe same intensity, which suggests that the two different MnOphases were present in a similar concentration (Table S4,Supporting Information), which is in good agreement with thecyclic voltammetry analysis. This is the first time in theliterature that the conversion of MnO to higher oxidation statesduring the conversion reaction lithiation/delithiation has beendefinitively found, providing new insight into the evolvingunderstanding of the chemistry of metal oxide anode materials.In order to account for the higher oxidation-state Mn (red

pathway in Figure 4) after repeated oxidation and reductioncycles, eq 2, there are three main possibilities.

+ + ↔ ++ −Mn O 8Li 8e 3Mn 4Li O3 4 2 (2)

First, the Mn3O4 that is formed contains oxygen vacancies,which has been found previously40 in a nonbattery environ-ment. The second possibility is that MnO is converted to ahigher oxidation state by reacting with oxygen species (i.e.,carbonates) that are created during the formation of the SEIthe products of electrolyte decomposition. Third, it is alsopossible that local stoichiometry post reduction (after charging)kinetically favors the oxidation of Mn3O4 (during discharge).This would also result in domains of metallic Mn post cycling,which were detected by XPS (Table S4), although not in asignificant quantity as one would have expected if this was thepredominant pathway.The kinetics and reversibility of the conversion reaction was

probed by carrying out charge/discharge tests at different ratesfor all MnxCo1−xO materials, and the results are shown inFigure 3d. First, the results generally showed a decrease in theextent of reaction at higher rates, which is expected because ofthe increasing kinetic overpotentials the battery is experiencingwith increasing reaction rate. In order to evaluate theoverpotentials in situ, the charge/discharge curves for all ofthe MnxCo1−xO chemistries were deconvoluted at four reactionrates (Figure S5 in the Supporting Information). Among the

Figure 3. Cyclic voltammograms for MnO (a), Mn0.9Co0.1O (b), and Mn0.8Co0.2O (c), showing their redox reaction peaks. (d) Rate capability forMnxCo1−xO investigating kinetic overpotentials as a function of Co inclusion. (e) Initial charge storage ability of MnxCo1−xO at a current rate of 400mA/g. (f) 300 charge/discharge cycles for MnxCo1−xO at a current rate of 400 mA/g, showing their long-term charge storage ability as a function ofCo inclusion in MnO.

Figure 4. Proposed mechanism for MnO solid-state conversion reaction.

The Journal of Physical Chemistry C Article

DOI: 10.1021/acs.jpcc.8b00403J. Phys. Chem. C 2018, 122, 7120−7127

7123

MnxCo1−xO materials, increasing the Co content led to loweroverpotentials, with the most notable differences at highreaction rates (Table S5 in the Supporting Information).Because of the lower overpotentials, the cells were able tooperate over a larger effective voltage window and thus wereable to achieve a higher extent of reaction (higher capacity). Infact, MnO (0% Co doped), shows an extent of reaction of only23.1% (175 mAh/g) at 1600 mA/g. The Mn0.95Co0.05O,Mn0.9Co0.1O, Mn0.85Co0.15O, and Mn0.8Co0.2O active materialswere able to achieve significantly higher extents of reaction andcapacities of 284, 355, 390, and 554 mAh/g (at a current rate of1600 mA/g), respectively. The Mn0.8Co0.2O in this work wasable to achieve one of the highest capacities reported to date forany Mn anode in lithium ion batteries to date at a rate of 1600mA/g.13,16,32,39,41−45

Knowing the overpotentials for all of the active materials as afunction of the current, a Tafel slope analysis was carried outfor both the reduction (lithiation) and oxidation (delithiation)reactions in order to identify the rate limiting step in theelectrochemical conversion reaction. Three assumptions weremade: the Mn3O4 reaction pathway is negligible initiallybecause of the very small number of charge/discharge cyclesthat occurred during the testing; the first data point at 40 mA/gwas neglected, because the overpotential was too small (∼10mV) to justify its use given the inherent assumptions in theTafel equation derivation (eq 3); the Mn0.8Co0.2O was notconsidered because of the phase separation and consequentincreasing number of reactions occurring simultaneously. Theoverpotentials were evaluated for both charge and discharge bytaking the midpoint value between the cutoff voltage at the endof each cycle and the following data point, and the Tafel slopewas evaluated for each MnxCo1−xO chemistry

η = +a b ilog (3)

Where η is the overpotential at a specific current density i, andb is the Tafel slope, which is defined as

α=

·b

RTF

2.303

eff (4)

Where F = 96485.3 C mol−1 is Faraday constant, R = 8.314 Jmol−1 K−1 is the ideal gas constant, T = 298 K is thetemperature, and αef f is the effective transfer coefficient. FigureS6 in the Supporting Information shows an example of theevaluation of the Tafel slope for both forward and reverse scansfor MnO. From the Tafel slope, αeff was determined by eq 4and averaged over the four MnxCo1−xO materials in order todetermine the nature of the rate-determining step (RDS) usingeq 5.

α γν

ρβ= +eff (5)

Where γ is the number of electron transfer steps preceding therate-determining step, ν is the RDS stoichiometric coefficient, βis the transfer coefficient for a reversible reaction (which is 0.5for most systems of interest), and ρ is a coefficient equal to 1 ifthe RDS is an electron transfer step or 0 if it is a chemical step.The results showed an αeff of 0.54 (∼0.5) for the reductionreaction and an αeff of 1.54 (∼1.5) for the oxidation reaction,from which a reaction scheme proposed in eqs 6−8 can beinferred. When αeff ≈ 0.5 (Tafel slope ≈ 120 mV/decade), ρ =1, and γ = 0, the RDS is an electrochemical step, and morespecifically, the first electrochemical reaction (eq 6), namely,the formation of the MnOLi complex. During the oxidation

(from eq 8 to eq 6), αeff ≈ 1.5 (Tafel slope ≈ 40 mV/decade),ρ = ν = γ = 1, which identifies the second electrochemical stepas the RDS, strongly supporting the reduction data pointing toeq 6 as the RDS. The consistent result also suggests that theactive sites for the reaction do not change whether oxidation orreduction is occurring, and the primary activation hurdle lies inthe MnOLi intermediate. Lastly, the individual values for theTafel slope, and hence αeff, with varying Co content were quiteconsistent, show that the underlying mechanism for theconversion of MnO is not affected by the presence of Co.

+ + ↔+ −MnO Li e MnOLi (6)

+ + ↔+ −MnOLi Li e MnOLi2 (7)

↔ +MnOLi Mn Li O2 2 (8)

To begin probing the reaction reversibility, initial charge/discharge cycles and long-term stability tests for all MnxCo1−xOactive materials at 400 mA/g are reported in Figure 3e,f. For alimited number of redox cycles (Figure 3e), all of theMnxCo1−xO materials show good reversibility, with theexception of the phase-separated x = 0.15 and x = 0.2, whichexperienced slight decline even within 50 cycles. When lookingat extensive long-term cycling (Figure 3f), it is clearlydemonstrated that the phase-separated materials are veryunstable. In fact, Mn0.85Co0.15O and Mn0.8Co0.2O show rapidloss in reversibility (capacity fade), only realizing 347 and 133mAh/g, respectively, after 300 cycles at a rate of 400 mA/g.This might be due to the fact that the phase separation in theactive material limits the number of surface active sites availableand lengthens the path for Li-ion solid-state diffusion. At theother extreme, low Co inclusion had minimal impact on thereaction reversibility, showing equally poor reversibility to rawMnO. Indeed, MnO and Mn0.05Co0.15O show a final extent ofreaction after 300 cycles of only 299 and 286 mAh/g,respectively.Interestingly, Mn0.9Co0.1O showed the highest charge

reversibility over the 300 charge/discharge cycles and is oneof the best results for an MnO-based anode to date at a rate of400 mA/g (Table S6 in the Supporting Information). In fact,10% Co inclusion in MnO is able to achieve excellent capacityretention, ca. 550 mAh/g for 300 cycles. Comparing charge/discharge curves at cycles 3 and 300 for the various materials(Figure S7 in the Supporting Information), only Mn0.9Co0.1Oshowed low hysteresis, confirming the high reversibility of theconversion reaction in this case. This behavior appears to resultfrom a balance between the intrinsic capacity fade of MnO withcycling and the gradual emergence of Mn3O4 with cyclenumber, which is supported below.Features characteristic of the more gradual emergence of

metallic manganese to Mn3O4 for Mn0.9Co0.1O compared tothe other MnxCo1−xO species can be detected from cyclicvoltammograms and post cycling XPS. In fact, Figure 3a,cshows clear anodic peaks at 2.2 V (a fingerprint for thepresence of Mn3O4, Figure S3), but Figure 3b only shows avery low intensity peak. Moreover, though the height and areaof the two cathodic peaks during the second redox scan iscomparable for the three analyzed materials, in the followingredox cycles, the peak ratio changes. Only for Figure 3b is thearea under the cathodic peak at 0.45 V substantively larger thanthe area under the peak at 0.35 V, confirming that thepredominant species after five cycles is MnO for Mn0.9Co0.1O.

The Journal of Physical Chemistry C Article

DOI: 10.1021/acs.jpcc.8b00403J. Phys. Chem. C 2018, 122, 7120−7127

7124

As further proof, the higher Co-content MnxCo1−xOmaterials (Mn0.85Co0.15O and Mn0.8Co0.2O) had much highercapacities in the first 30 cycles (Figure 3e). This behaviorsuggests rapid transformation of MnO to higher capacityMn3O4 upon cycling, which was followed by the characteristiccapacity fade of Mn3O4, shown in Figure S8 in the SupportingInformation. It should be noted that Figure S8 itself suggeststhat Mn3O4 in isolation may not be a desirable starting oxidefrom which to create an Li-ion battery anode.The higher immediate capacity of the higher Co-content

oxides in Figure 3c−e also gives insight into which of the threemechanisms discussed previously is primarily responsible forthe formation of Mn3O4reaction with the SEI. This can beconcluded because of the limited oxygen that is available in theform of Li2O after the conversion of MnO to Li2O + Mn. SinceMn3O4 requires additional oxygen to form per Mn atom (eq 2)than MnO (eq 1), the reaction of Li2O + Mn to Mn3O4 wouldnot result in an increased capacity but only orphaned metallicMn (which was not readily observed). Also, if oxygen-vacancyladen Mn3O4 were formed, it could only come from thereaction of three Li2O molecules, which would only yield sixelectrons, not eight, and maintain the same 2e−/Mn ratio asMnOagain not increasing capacity. However, it does appearthat the amount of oxygen that is available to react from the SEIis limited, which explains why we observe the extensive (andrapid) formation of Mn3O4 but not higher oxidation states, likeMnO2, which we know are possible based on post cycling XPSof anodes consisting of Mn3O4 initially, where almost half ofthe Mn becomes MnO2 over time (Figure S9 and Table S7 inthe Supporting Information). It is interesting that MnO2 wasnot observed, because the transformation of higher oxidation-state Mn oxides to higher oxidation states during cycling isthermodynamically favored. The standard reduction potentialsfor the conversion reactions of MnO, Mn3O4, and MnO2 werecalculated to be 1.03, 1.25, and 1.70 V, respectively. Thus, thefact that MnO2 is not detected supports the hypothesis ofreactant deficiency (oxygen) in the system.The final piece of data supporting the slow transformation of

MnO to Mn3O4 in Mn0.9Co0.1O is post cycling XPS, which wascarried out after 300 cycles. With pure MnO (Figure S10 in theSupporting Information), 100% of the active material isreconverted to Mn3O4 at the end of the redox cycles. Recall(Table S4) that for Mn0.9Co0.1O, around half of the manganeseremained as MnO, being a potential indication that theinclusion of Co limits the amount of available oxygen for MnOby forming different cobalt oxide species (CoO and Co3O4 inequal amounts, supported by Figure S11 and Table S8 in theSupporting Information). Therefore, it is the competition foroxygen between Mn and Co that controls the materialschemistry and gradual transformation of MnO to Mn3O4, a newand very interesting result. This was not observed for the higherCo-containing materials due to phase separation of CoO; thus,MnO acted as a pure, not doped, compoundrapidly formingMn3O4 upon cycling. Also exciting is that by occurring overmany cycles, though all of the individual oxide phasesthemselves have limited stability, in operating batteries, theseanodes showed very good capacity retention. This shows that itis possible to decouple material stability (reaction reversibility)and capacity retention or to even engineer materials withprescribed failure mechanisms that do not sacrifice perform-ance.

CONCLUSIONIn conclusion, a detailed study on the MnO solid-stateconversion reaction in Li-ion batteries has been reported. Allactive materials showed the presence of secondary phaseMn3O4, which was not detected before. The degree of Coinclusion in MnO has been shown to be a critical variable tomaintain the balance between MnO intrinsic capacity fade dueto particle agglomeration and Mn3O4 formation, by allowinggradual transformation of MnO to Mn3O4 during cycling. Co-content optimization produced an active material able to retainstable capacity over hundreds of cycles, being one of the bestMnO materials reported in the literature to date. This work alsoopens up new pathways for investigation, including the study ofother conversion reactions for low oxidation-state metal oxidecompounds, which could help explain some of the works in theliterature where the capacity values are above the theoretical.Moreover, the presented Co inclusion approach can beextended to many other materials; the possibility of findingother synergistic combinations of chemistries might be usefulfor the development of new anodes for advanced, high powerdemanding Li-ion battery applications.

ASSOCIATED CONTENT*S Supporting InformationThe Supporting Information is available free of charge on theACS Publications website at DOI: 10.1021/acs.jpcc.8b00403.

Experimental details, charge/discharge curves at differentrates and cycles, XPS details, conversion reversibility forMnO vs MnO/Co, conductivity data, Sherrer equationdata, Tafel slope data (PDF)

AUTHOR INFORMATIONCorresponding Author*Phone: 803-777- 4181; E-mail: mustainw@mailbox.sc.edu.ORCIDAlessandro Palmieri: 0000-0002-1375-8625Sajad Yazdani: 0000-0003-1667-8249Stavros G. Karakalos: 0000-0002-3428-5433Michael T. Pettes: 0000-0001-6862-6841William E. Mustain: 0000-0001-7804-6410NotesThe authors declare no competing financial interest.

ACKNOWLEDGMENTSThis work was funded by the Ford Motor Company throughthe Ford University Research Program, which supported theefforts of A.P. and W.E.M. This work was also partiallysupported by the National Science Foundation under GrantNo. CAREER-1553987 (M.T.P., S.Y.), the UConn ResearchFoundation, award number PD17-0137 (R.K.-S., S.Y.), and aGE Graduate Fellowship for Innovation (S.Y.). We are verygrateful to Prof. Elena Silva, Prof. Ali Gokirmak, and LhaceneAdnane for the use of the Van Der Pauw setup.

REFERENCES(1) Nagaura, T.; Tozawa, K. Lithium Ion Rechargeable Battery. Prog.Batteries Sol. Cells 1990, 9, 209−217.(2) Tarascon, J. M.; Armand, M. Issues and Challenges FacingRechargeable Lithium Batteries. Nature 2001, 414, 359−367.(3) Thackeray, M. M.; Wolverton, C.; Isaacs, E. D. Electrical EnergyStorage for TransportationApproaching the Limits of, and Going

The Journal of Physical Chemistry C Article

DOI: 10.1021/acs.jpcc.8b00403J. Phys. Chem. C 2018, 122, 7120−7127

7125

Beyond, Lithium-Ion Batteries. Energy Environ. Sci. 2012, 5, 7854−7863.(4) Zackrisson, M.; Avellan, L.; Orlenius, J. Life Cycle Assessment ofLithium-Ion Batteries for Plug-In Hybrid Electric Vehicles − CriticalIssues. J. Cleaner Prod. 2010, 18, 1519−1529.(5) Derrien, G.; Hassoun, J.; Panero, S.; Scrosati, B. NanostructuredSn−C Composite as an Advanced Anode Material in High-Performance Lithium-Ion Batteries. Adv. Mater. 2007, 19, 2336−2340.(6) Liu, N.; Lu, Z.; Zhao, J.; McDowell, T. M.; Lee, H. W.; Zhao, W.;Cui, Y. A Pomegranate-Inspired Nanoscale Design for Large-Volume-Change Lithium Battery Anodes. Nat. Nanotechnol. 2014, 9, 187−192.(7) Li, H.; Richter, G.; Maier, J. Reversible Formation andDecomposition of LiF Clusters Using Transition Metal Fluorides asPrecursors and Their Application in Rechargeable Li Batteries. Adv.Mater. 2003, 15, 736−739.(8) Snyder, Q. M.; Trebukhova, A. S.; Ravdel, B.; Clayton Wheeler,M.; DiCarlo, J.; Tripp, P. C.; DeSisto, J. W. Synthesis andCharacterization of Atomic Layer Deposited Titanium Nitride ThinFilms on Lithium Titanate Spinel Powder as a Lithium-Ion BatteryAnode. J. Power Sources 2007, 165, 379−385.(9) Kim, E.; Son, D.; Kim, T. G.; Cho, J.; Park, B.; Ryu, K. S.; Chang,S. H. A Mesoporous/Crystalline Composite Material Containing TinPhosphate for Use as the Anode in Lithium-Ion Batteries. Angew.Chem., Int. Ed. 2004, 43, 5987−5990.(10) Oumellal, Y.; Rougier, A.; Nazri, G. A.; Tarascon, J. M.; Aymard,L. Metal Hydrides for Lithium-Ion Batteries. Nat. Mater. 2008, 7,916−921.(11) Reddy, M. V.; Subba Rao, G. V.; Chowdari, B. V. R. MetalOxides and Oxysalts as Anode Materials for Li Ion Batteries. Chem.Rev. 2013, 113, 5364−5457.(12) Poizot, P.; Laruelle, S.; Grugeon, S.; Dupont, L.; Tarascon, J. M.Nano-Sized Transition-Metal Oxides as Negative-Electrode Materialsfor Lithium-Ion Batteries. Nature 2000, 407, 496−499.(13) Wei, W.; Wang, Z.; Liu, Z.; Liu, Y.; He, L.; Chen, D.; Umar, A.;Guo, L.; Li, J. Metal Oxide Hollow Nanostructures: Fabrication and LiStorage Performance. J. Power Sources 2013, 238, 376−387.(14) Palmieri, A.; Liu, Y.; He, J.; Meng, Y.; Suib, S. L.; Mustain, W. E.Metal Oxide/Reduced Graphene Oxide Anodes for Lithium-IonBatteries. ECS Trans. 2015, 66, 47−55.(15) Wu, M. S.; Chiang, P. C. J.; Lee, J. T.; Lin, J. C. Synthesis ofManganese Oxide Electrodes with Interconnected Nanowire Structureas an Anode Material for Rechargeable Lithium Ion Batteries. J. Phys.Chem. B 2005, 109, 23279−23284.(16) Li, X.; Li, D.; Qiao, L.; Wang, X.; Sun, X.; Wang, P.; He, D.Interconnected Porous MnO Nanoflakes for High-PerformanceLithium Ion Battery Anodes. J. Mater. Chem. 2012, 22, 9189−9194.(17) Cui, Z.; Guo, X.; Li, H. High Performance MnO Thin-FilmAnodes Grown by Radio-Frequency Sputtering for Lithium IonBatteries. J. Power Sources 2013, 244, 731−735.(18) Yazdani, S.; Kashfi-Sadabad, R.; Palmieri, A.; Mustain, W. E.;Thompson Pettes, M. Effect of Cobalt Alloying on the ElectrochemicalPerformance of Manganese Oxide Nanoparticles Nucleated onMultiwalled Carbon Nanotubes. Nanotechnology 2017, 28, 155403.(19) Spinner, N.; Zhang, L.; Mustain, W. E. Investigation of MetalOxide Anode Degradation in Lithium-Ion Batteries via Identical-Location TEM. J. Mater. Chem. A 2014, 2, 1627−1630.(20) Spinner, N.; Palmieri, A.; Beauregard, N.; Zhang, L.;Campanella, J.; Mustain, W. E. Influence of Conductivity on theCapacity Retention of NiO anodes in Li-ion batteries. J. Power Sources2015, 276, 46−53.(21) Zhao, Y.; Li, X.; Yan, B.; Li, D.; Lawes, S.; Sun, X. SignificantImpact of 2D Graphene Nanosheets on Large Volume Change Tin-Based Anodes in Lithium-Ion Batteries: A Review. J. Power Sources2015, 274, 869−884.(22) Reddy, A. L. M.; Shaijumon, M.; Gowda, S. R.; Ajayan, P. M.Coaxial MnO2/Carbon Nanotube Array Electrodes for High-Perform-ance Lithium Batteries. Nano Lett. 2009, 9, 1002−1006.

(23) Zhang, J.; Jiang, J.; Zhao, X. S. Synthesis and CapacitiveProperties of Manganese Oxide Nanosheets Dispersed on Function-alized Graphene Sheets. J. Phys. Chem. C 2011, 115, 6448−6454.(24) Liu, Y.; Palmieri, A.; He, J.; Meng, Y.; Beauregard, N.; Suib, S.L.; Mustain, W. E. Highly Conductive In-SnO2/RGO Nano-Heterostructures with Improved Lithium-Ion Battery Performance.Sci. Rep. 2016, 6, 25860.(25) Zhu, Y. G.; Wang, Y.; Han, Z. J.; Shi, Y.; Wong, J. I.; Huang, Z.X.; Ostrikov, K.; Yang, H. Y. Catalyst Engineering for Lithium IonBatteries: the Catalytic Role of Ge in Enhancing the ElectrochemicalPerformance of SnO2(GeO2)0.13/G Anodes. Nanoscale 2014, 6,15020−15028.(26) Wei, W.; Cui, X.; Chen, W.; Ivey, D. G. Manganese Oxide-BasedMaterials as Electrochemical Supercapacitor Electrodes. Chem. Soc.Rev. 2011, 40, 1697−1721.(27) Chang, J. K.; Lee, M. T.; Huang, C. H.; Tsai, W. T.Physicochemical Properties and Electrochemical Behavior of BinaryManganese−Cobalt Oxide Electrodes for Supercapacitor Applications.Mater. Chem. Phys. 2008, 108, 124−131.(28) Adnane, L.; Gokirmak, A.; Silva, H. High Temperature HallMeasurement Setup for Thin Film Characterization. Rev. Sci. Instrum.2016, 87, 075117.(29) Palmieri, A.; Kashfi-Sadabad, R.; Yazdani, S.; Pettes, M. T.;Mustain, W. E. High Performance Bi-Metallic Manganese CobaltOxide/Carbon Nanotube Li-ion Battery Anodes. Electrochim. Acta2016, 213, 620−625.(30) Hummers, W. S.; Offeman, R. E. Preparation of GraphiticOxide. J. Am. Chem. Soc. 1958, 80, 1339.(31) Wang, H.; Cui, L. F.; Yang, Y.; Sanchez Casalongue, H.;Robinson, J. T.; Liang, Y.; Cui, Y.; Dai, H. Mn3O4−Graphene Hybridas a High-Capacity Anode Material for Lithium Ion Batteries. J. Am.Chem. Soc. 2010, 132, 13978−13980.(32) Liu, J.; Pan, Q. MnO/C Nanocomposites as High CapacityAnode Materials for Li-Ion Batteries. Electrochem. Solid-State Lett.2010, 13, A139−A142.(33) Peng, C.; Chen, B.; Qin, Y.; Yang, S.; Li, C.; Zuo, Y.; Liu, S.;Yang, J. Facile Ultrasonic Synthesis of CoO Quantum Dot/GrapheneNanosheet Composites with High Lithium Storage Capacity. ACSNano 2012, 6, 1074−1081.(34) Dubal, D. P.; Dhawale, D. S.; Salunkhe, R. R.; Lokhande, C. D.A Novel Chemical Synthesis of Mn3O4 Thin Film and its StepwiseConversion into Birnessite MnO2 during Super Capacitive Studies. J.Electroanal. Chem. 2010, 647, 60−65.(35) Zhu, X. J.; Hu, J.; Dai, H. L.; Ding, L.; Jiang, L. ReducedGraphene Oxide and Nanosheet-Based Nickel Oxide MicrosphereComposite as an Anode Material for Lithium Ion Battery. Electrochim.Acta 2012, 64, 23−28.(36) Langell, M. A.; Gevrey, F.; Nydegger, M. W. SurfaceComposition of MnxCo(1‑x)O Solid Solutions by X-Ray Photoelectronand Auger Spectroscopies. Appl. Surf. Sci. 2000, 153, 114−127.(37) Bergman, B.; Agren, J. Thermodynimic Assessment of theSystem CoO-MnO. J. Am. Ceram. Soc. 1986, 69, 877−81.(38) Kim, J. C.; Hwang, I. S.; Seo, S. D.; Kim, D. W. NanocompositeLi-Ion Battery Anodes Consisting of Multiwalled Carbon Nanotubesthat Anchor CoO Nanoparticles. Mater. Lett. 2013, 104, 13−16.(39) Sun, Y.; Hu, X.; Luo, W.; Xia, F.; Huang, Y. Reconstruction ofConformal Nanoscale MnO on Graphene as a High-Capacity andLong-Life Anode Material for Lithium Ion Batteries. Adv. Funct. Mater.2013, 23, 2436−2444.(40) Durmus, Z.; Baykal, A.; Kavas, H.; Sozeri, H. Preparation andCharacterization of Polyaniline (PANI)−Mn3O4 Nanocomposite.Phys. B 2011, 406, 1114−1120.(41) Zhong, Y.; Zhang, B.; Luo, S.; Wen, W.; Li, H.; Huang, X.;Chen, L. Investigation on Porous MnO Microsphere Anode forLithium Ion Batteries. J. Power Sources 2011, 196, 6802−6808.(42) Mai, Y. J.; Zhang, D.; Qiao, Y. Q.; Gu, C. D.; Wang, X. L.; Tu, J.P. MnO/Reduced Graphene Oxide Sheet Hybrid as an Anode for Li-Ion Batteries with Enhanced Lithium Storage Performance. J. PowerSources 2012, 216, 201−207.

The Journal of Physical Chemistry C Article

DOI: 10.1021/acs.jpcc.8b00403J. Phys. Chem. C 2018, 122, 7120−7127

7126

(43) Xu, G. L.; Xu, Y. F.; Sun, H.; Fu, F.; Zheng, X. M.; Huang, L.; Li,J. T.; Yang, S. H.; Sun, S. G. Facile Synthesis of Porous MnO/CNanotubes as a High Capacity Anode Material for Lithium IonBatteries. Chem. Commun. 2012, 48 (68), 8502−8504.(44) Zhong, K.; Xia, X.; Zhang, B.; Li, H.; Wang, Z.; Chen, L. MnOPowder as Anode Active Materials for Lithium Ion Batteries. J. PowerSources 2010, 195, 3300−3308.(45) Ji, L.; Zhang, X. Manganese Oxide Nanoparticle-Loaded PorousCarbon Nanofibers as Anode Materials for High-PerformanceLithium-Ion Batteries. Electrochem. Commun. 2009, 11, 795−798.

The Journal of Physical Chemistry C Article

DOI: 10.1021/acs.jpcc.8b00403J. Phys. Chem. C 2018, 122, 7120−7127

7127

S1

Cobalt Doping as a Pathway to Stabilize the Solid-State Conversion Chemistry of Manganese Oxide Anodes in Li-Ion Batteries

Alessandro Palmierib, Sajad Yazdanic, Raana Kashfi-Sadabadd, Stavros G. Karakalose,

Michael T. Pettesc and William E. Mustaina,b*

a Department of Chemical Engineering, Swearingen Engineering Center, University of South Carolina, Columbia, SC

29208

b Department of Chemical & Biomolecular Engineering, University of Connecticut, Storrs, CT 06269-3222

c Department of Mechanical Engineering, University of Connecticut, Storrs, CT 06269-3139

d Institute of Material Science, University of Connecticut, Storrs, CT 06269-3136

e College of Engineering and Computing, Swearingen Engineering Center, University of South Carolina, Columbia, SC

29208

* Phone: 803-777- 4181; Email: mustainw@mailbox.sc.edu

S2

Synthesis Reactants

Multiwall carbon nanotubes (MWCNTs, catalogue number 724769), Potassium permanganate

(KMnO4), hydrogen peroxide (H2O2, 30%), sodium nitrate (NaNO3), sulfuric acid (H2SO4), ethyl

alcohol, (CH3CH2OH), and ammonium hydroxide (NH4OH), manganese(II) acetate tetrahydrate

Mn(CH3COO)2.4H2O and cobalt(II) acetate tetrahydrate Co(CH3COO)2.4H2O were purchased from

Sigma-Aldrich. All chemicals were used as received without any further purification.

Oxidation of Multi-Wall Carbon Nanotubes (MWCNT)

Multiwall carbon nanotubes were oxidized similar to modified Hummers’ method with a lower

concentration of oxidizing agent. First, 60 mL of sulfuric acid was added to 1g of multiwall carbon

nanotubes (MWCNT) in a 250 mL round-bottom flask and stirred at room temperature for 24 h. The

flask was then heated in an oil bath at 40 °C and 0.1 g of NaNO3 was added to the mixture. Then, 1 mg

KMnO4 was slowly added to the solution while keeping the reaction temperature below 20 °C in an

ice-water bath. The reaction was removed from the ice-water bath after 30 min and was transferred to

an oil bath at 45°C and allowed to stir for 30 min. Then, 3 mL of DI-water was added, followed by

another 3 mL after 5 min and 40 mL after 10 min. After 15 min, the flask was taken out of the oil bath

and 140 mL of DI-water followed by 10 mL of 30% H2O2 was added to quench the reaction. The

obtained product was centrifuged and washed with 5% HCl solution two times. The homogeneous

supernatant was collected by centrifugation.

S3

Synthesis of Mn/Co/CNT

In a typical synthesis, 90 mg of the oxidized MWCNTs (for a final product of 10% wt CNTs)

was well dispersed in a 122.5 ml ethanol/DI water solution with a ratio of 48:1 for 1 hr. In the next

step, 3 ml of manganese(II) acetate tetrahydrate (0.6 M in DI-water) solution was added. To achieve

the desired percentage of Co doping in Mn, stoichiometric moles of Mn precursor was replaced with

cobalt(II) acetate tetrahydrate. The mixture was transferred to an oil bath and re-fluxed for 24 hours

after adding 2.5 ml of ammonium hydroxide solution. The solids were collected by centrifugation and

dried in vacuum at room temperature for 72 hrs. Finally, the Mn/Co/CNT anode material was obtained

by annealing the dried sample at 600 °C for 3 hours in an argon atmosphere.

Chemical and Structural Characterization

Powder X-ray diffraction (XRD) patterns were collected on a Bruker D2 Phaser with Cu Kα

radiation (λ=1.54184 Å) at room temperature with an operating voltage and current of 30 kV and 10

mA. X-ray photoelectron spectroscopy (XPS) was conducted on a Kratos AXIS Ultra DLD XPS

system, using Al Kα radiation (λ=1486.6 eV). Transmission electron microscopy (TEM) was

performed using a FEI Talos F200X TEM/STEM at an accelerating voltage of 200 kV. Van Der Pauw

resistivity measurements were carried out using hard pellets of the target materials on a customize 4

probe set up1. The hard pellets were made using a 1.5 cm dice and were obtained by 5 minutes pressing

at 10,000 lbs. The resulting pellet thickness was in a range from 100 to 300 µm.

S4

Anode Fabrication

LIB Anodes were fabricated by preparing inks containing 70% of active material (Mn/Co CNT

supported), 20% conductivity-boosting carbon black (CNERGY Super C65, Imerys), and 10% binder,

polyvinylidene fluoride (PVDF, Kynar blend). The components were dispersed in N-methylpyrrolidone

(NMP, Acros, 99.5% Extra Dry) solvent and the final ink (typically 90 mg of active material dispersed

in 800 µl of solvent) was homogenized through repeated and successive 15 minutes sonication (4

times) and mechanical stirring overnight. A copper foil (Alfa Aesar, 99.999%) was mechanically

roughened and cleaned with isopropanol (Fisher, Optima) before being used as the current collector.

The active material ink was sprayed by hand with an Iwata model sprayer onto the Cu foil to a uniform

thickness, heated under vacuum at 100°C for 24 hours, then pressed at 1500 lbs, calendared with a 0.3

mm gap and massed to obtain the loading. For all electrodes fabricated in this study, the active loading

was held between 1.0 and 1.5 mg/cm2.

Coin Cell Assembly

Coin cells were assembled to investigate the electrochemical behavior of the MnO/Co/CNT

anodes in a half-cell configuration. The materials used were 2.0 cm diameter coin cells (Hohsen Corp.),

lithium metal (Alfa Aesar, 99.9%) as the cathode, and Celgard 2320 tri-layer PP/PE/PP as the

separator. A 1M lithium hexafluorophosphate (LiPF6, Acros 98%) electrolyte was prepared from a

1:1:1 volumetric fraction ethylene carbonate (EC, Acros 99+%): dimethyl carbonate (DMC, Acros

98+%):diethyl carbonate (DEC, Acros 99%). The separator was punched to a diameter of 1.9 cm while

the anode was cut to 1.5 cm diameter disks. In an argon-purged glove box (Labconco), the lithium foil

cathode was also punched to a 1.5 diameter disk. Afterwards, 15 µL of electrolyte was pipetted onto

S5

each side of the separator and placed in between the anode and the cathode. Lastly, the gasket, spacer

disc, spring and upper case were positioned on top of the cathode and all of the components were then

crimped and sealed into the coin cell hardware before being safely removed from the Ar atmosphere-

controlled glove box for electrochemical testing.

Electrochemical Testing

Charge-discharge measurements were carried out at various rates between 0.001 and 3 V using

an Arbin MSTAT battery test system. The MSTAT system was also used to collect cyclic

voltammograms (CVs) at a scan rate of 0.1 mV/s over the same voltage windows as the

charge/discharge cycles. All potentials are reported vs. Li/Li+. Electrochemical Impedance

Spectroscopy (EIS) was conducted between 100 kHz–50 mHz with a 5 mV amplitude at the coin cell

open circuit voltage, using an Autolab PGSTAT302N Potentiostat (Eco Chemie).

Deconvoluting overpotentials from charge-discharge curves

Kinetic overpotentials have been evaluated from the charge-discharge curves, as the mid-point

between the cutoff voltage of the previous redox cycle and the first data point of the following redox

cycle. More specifically, for reduction the overpotential was evaluated as the midpoint between the

oxidation cutoff voltage of 3V and the first data point during the reduction process; for oxidation the

overpotential was evaluated as the midpoint between the reduction cutoff voltage of 0V and the first

data point during the reduction process. An example of visual overpotential evaluation is shown in

figure S13.

S6

Table S1. Electronic conductivity of raw MnO and Mn0.9Co0.1O obtained by the Van Der Pauw method.

Table S2. XRD peak positions for the MnxCo(1-x)O active materials (x=0, 0.05, 0.1, 0.15, 0.2).

MnO Mn0.9Co0.1O

Electronic conductivity (Ω-1cm-1) 1.51*10-8 4.20*10-7

MnO Related Peaks

(2θdegrees)

CoO Related Peaks

(2θdegrees)

0% Co 34.6 40.3 58.5 70.0 73.6 N/A N/A N/A

5% Co 34.7 40.4 58.6 70.1 73.7 N/A N/A N/A

10% Co 34.7 40.4 58.6 70.1 73.7 N/A N/A N/A

15% Co 35.3 41.0 59.2 70.8 74.4 N/A 41.9 N/A

20% Co 34.9 40.6 58.8 70.3 73.9 35.7 41.4 60.2

S7

Table S3. Domain Size obtained from the (200) peak in the XRD pattern for the MnxCo(1-x)O active materials (x=0, 0.05, 0.1, 0.15, 0.2). The increasing domain size for x= 0, 0.05 and 0.1 active materials

suggests a gradual larger inclusion of Co in MnO. The decreasing domain size for x= 0.15 and 0.2 active materials provides additional evidence to that presented in the main paper of phase separation on

the active material surface.

Table S4. Quantification of species mass % based on Figure S4.

Peak Position

(2θdegrees)

Full Width Half

Maximum (rad)

Domain Size (nm)

Sherrer Eqn

0% Co 40.3 0.00741 20.0

5% Co 40.4 0.00757 19.5

10% Co 40.4 0.00534 27.7

15% Co 41.0 0.00665 22.3

20% Co (MnO) 40.6 0.00876 16.9

20% Co (CoO) 41.4 0.0119 12.5

Mn MnO Mn3O4

% ratio 5.6 46.8 47.6

S8

Table S5. Kinetics overpotentials at a current rate of 1600 mA/g for the MnxCo(1-x)O active materials (x=0, 0.05, 0.1, 0.15, 0.2).

Table S6. Summary of the electrochemical properties of MnO and carbon composites in recent literature

Reduction Overpotential (mV) at

1600 mA/g

Oxidation Overpotential (mV) at

1600 mA/g

0% Co 98 53

5% Co 121 80

10% Co 86 57

15% Co 80 49

20% Co 52 47

Active Material Current Rate Reversible

Capacity (mAh/g)

Cycle

Number

Reference

MnO/Graphene 2000 mA/g 840 400 [2] MnO Microspheres 50 mA/g 700 100 [3] MnO/rGO 100 mA/g 650 50 [4] MnO/CNTs 510 mA/g 600 200 [5] MnO/C Composite 50 mA/g 650 150 [6] MnO/C Nanofibers 50 mA/g 580 50 [7] MnO Sputtering Thin Film

40 µA/cm 680 100 [8]

MnO/C Composite 100 mA/g 690 50 [9] MnO/Nanoflakes 246 mA/g 700 200 [10]

S9

Mn2O3 MnO2

% ratio 6.1 93.9

Table S7. Quantification of species mass % based on Figure S8.

Table S8. Quantification of species mass % based on Figure S11.

0 10 20 30 40 500

100

200

300

400

500

600

700

MnO/CNTs

Pure MnO

Cap

acit

y(m

Ah

/g)

Cycle Number

Fig S1. Capacity retention for manganese oxide (black line) compared to MnO/CNTs that have their inter particle conductivity enhanced (blue line). The capacity loss over 50 cycles for pure manganese oxide is 58.2 % whereas MnO/CNTs only loses 4.3 %, confirming the beneficial effect of increasing

the inter-particle electronic conductivity.

Co3O4 CoO

% ratio 52.1 47.9

S10

Fig S2. XPS general survey for Mn0.9Co0.1O. The result displays a Co to Mn ratio of ca. 0.1, which is in good agreement with the nominal value.

S11

Fig S3. Cyclic voltammetry of the first 5 cycles for Mn3O4/CNT

Fig S4. XPS high resolution scan of Mn 2p area for Mn0.9Co0.1O after 300 charge/discharge cycles at 1C.

-0.5 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5

-0.0010

-0.0008

-0.0006

-0.0004

-0.0002

0.0000

0.0002

0.0004

Cycle 1 Mn3O

4/CNTs

Cycle 2 Mn3O

4/CNTs

Cycle 3 Mn3O

4/CNTs

Cycle 4 Mn3O

4/CNTs

Cycle 5 Mn3O

4/CNTs

Cu

rren

t (A

)

E(V) Vs Li/Li+

657 654 651 648 645 642 639 636 633

Metallic Mn

MnO640.8 eV

Mn3O

4

641.7 eV

Mn 2p

Binding Energy /eV

X

PS

Peak In

ten

sit

y /

a.u

.

S12

Fig S5. Charge/discharge curves at different rates for (a) MnO (b) Mn0.95Co0.05O (c) Mn0.9Co0.1O (d) Mn0.85Co0.15O (e) Mn0.8Co0.2O.

0 50 100 150 200 250 300 350 4000.0

0.5

1.0

1.5

2.0

2.5

3.0

E(V

) V

s L

i/L

i+

Capacity(mAh/g)

40 mA/g

400 mA/g

800 mA/g

1600 mA/g

40 mA/g SB

a)

0 100 200 300 400 500 600 7000.0

0.5

1.0

1.5

2.0

2.5

3.0

40 mA/g

400 mA/g

800 mA/g

1600 mA/g

40 mA/g SB

Capacity(mAh/g)

E(V

) V

s L

i/L

i+

b)

0 100 200 300 400 500 600 7000.0

0.5

1.0

1.5

2.0

2.5

3.0

40 mA/g

400 mA/g

800 mA/g

1600 mA/g

40 mA/g SB

Capacity(mAh/g)

E(V

) V

s L

i/L

i+

d)

0 100 200 300 400 500 600 700 800 9000.0

0.5

1.0

1.5

2.0

2.5

3.0

40 mA/g

400 mA/g

800 mA/g

1600 mA/g

40 mA/g SB

Capacity(mAh/g)

E(V

) V

s L

i/L

i+

e)

0 100 200 300 400 500 600 7000.0

0.5

1.0

1.5

2.0

2.5

3.0

E(V

) V

s L

i/L

i+

Capacity(mAh/g)

40 mA/g

400 mA/g

800 mA/g

1600 mA/g

40 mA/g SB

(c)

S13

Fig S6. Tafel slope for MnO during lithiaton (black line) and delithiation (blue line). Applying Eq. 3 in the main text, the calculated value of αeff for the oxidation and reduction reactions were equal to 0.55 (~0.5) and 1.67 (~1.5) respectively.

-3.4 -3.3 -3.2 -3.1 -3.0 -2.9 -2.8 -2.7

0.03

0.04

0.05

0.06

0.07

0.08

0.09

0.10 Reduction

Oxidation

Oxidation (delithiation) Tafel

Slope: 36 mV/Decade

η

η

η

η (( ((V

)

log i (A)

Reduction (lithiation)

Tafel Slope: 107 mV/decade

S14

Fig S7. Charge/discharge curves for cycle 3 (black line) and cycle 300 (red line) at 400 mA/g for (a) MnO (b) Mn0.95Co0.05O (c) Mn0.9Co0.1O (d) Mn0.85Co0.15O and (e) Mn0.8Co0.2O.

0 50 100 150 200 250 300 350 400 4500.0

0.5

1.0

1.5

2.0

2.5

3.0

(a)

Cycle 3 0%

Cycle 300 0%

Capacity(mAh/g)

E(V

) V

s L

i/L

i+

0 100 200 300 400 5000.0

0.5

1.0

1.5

2.0

2.5

3.0

(b)

E(V

) V

s L

i/L

i+

Cycle 3 5%

Cycle 300 5%

Capacity(mAh/g)

0 100 200 300 400 500 600

0.0

0.5

1.0

1.5

2.0

2.5

3.0

(c)

E(V

) V

s L

i/L

i+

Capacity(mAh/g)

Cycle 3 10%

Cycle 300 10%

0 100 200 300 400 500 600 700 800

0.0

0.5

1.0

1.5

2.0

2.5

3.0

(d)

Capacity(mAh/g)

E(V

) V

s L

i/L

i+

Cycle 3 15%

Cycle 300 15%

0 100 200 300 400 500 600 700 800

0.0

0.5

1.0

1.5

2.0

2.5

3.0

(e)

Capacity(mAh/g)

E(V

) V

s L

i/L

i+

Cycle 3 20%

Cycle 300 20%

S15

Fig S8. Mn3O4/CNT capacity retention at 1C over 300 charge/discharge cycles.

Fig S9. XPS high resolution scan of Mn 2p area for Mn3O4 after 300 charge/discharge cycles at 1C displayed in Figure S8.

0 50 100 150 200 250 300

200

400

600

800

1000

1200

Mn3O4/CNTs

Cap

acit

y(m

Ah

/g)

Cycle Number

660 655 650 645 640 635

Binding Energy /eV

XP

S P

ea

k I

nte

ns

ity /

a.u

.

Mn 2p

Mn3O

4

641.3 eV

MnO2

642.5 eV

S16

Fig S10. XPS high resolution scan of Mn 2p area for MnO after 300 charge/discharge cycles at 1C.

Fig S11. XPS high resolution scan of Co 2p area for Mn0.9Co0.1O after 300 charge/discharge cycles at 1C.

660 655 650 645 640 635

Mn 2pMn

3O

4

641.7 eV

Binding Energy /eV

XP

S P

ea

k I

nte

nsit

y /

a.u

.

S17

Fig S12. Capacity retention for lithiation (black line), delithiation (blue line) and Coulombic efficiency (green line) for (a) MnO (b) Mn0.95Co0.05O (c) Mn0.9Co0.1O (d) Mn0.85Co0.15O and (e) Mn0.8Co0.2O at 400 mA/g for 300 cycles. Coulombic efficiency (extent of reaction between charge and subsequent

0 50 100 150 200 250 3000

250

500

750

1000

1250

0

20

40

60

80

100(a)

Co

ulo

mb

ic e

ffic

ien

cy

Cap

acit

y(m

Ah

/g)

Cycle Number

Charge 0%

Discharge 0%

Coulombic Efficiency

0 50 100 150 200 250 3000

250

500

750

1000

1250

0

20

40

60

80

100(b)

Cycle Number

Co

ulo

mb

ic e

ffic

ien

cy

Ca

pac

ity(m

Ah

/g) Charge 5%

Discharge 5%

Coulombic Efficiency

0 50 100 150 200 250 300

0

250

500

750

1000

1250

0

20

40

60

80

100

Co

ulo

mb

ic e

ffic

ien

cy

Charge 10%

Discharge 10%

Coulombic Efficiency

Cap

ac

ity (

mA

h/g

)

Cycle Number

(c)

0 50 100 150 200 250 3000

250

500

750

1000

1250

0

20

40

60

80

100(d)

Charge 15%

Discharge 15%

Coulombic Efficiency

Cycle Number

Co

ulo

mb

ic e

ffic

ien

cy

Ca

pa

cit

y(m

Ah

/g)

0 50 100 150 200 250 3000

250

500

750

1000

1250

0

20

40

60

80

100

Cycle Number

Co

ulo

mb

ic e

ffic

ien

cy

Cap

acit

y(m

Ah

/g)

(e)

Charge 20%

Discharge 20%

Coulombic Efficiency

S18

discharge of the same cycle) has also been evaluated for all MnxCo(1-x)O materials displaying constant values > 99% with the exception of 20% Co doped, which shows large gap between charge and discharge capacity, which led to an instable behavior in coulombic efficiency, with values as low as 60%. This is due to phase separation between cobalt and manganese on the electrode surface and consequent partial decomposition of the SEI with cycling.

Fig S13. Visual evaluation of kinetic overpotentials from charge/discharge curves for a general MnxCo(1-x)O material.

S19

References

1) Adnane, L.; Gokirmak, A.; Silva, H. Review of Scientific Instruments 2016, 87, 075117.

2) Sun, Y.; Hu, X.; Luo, W.; Xia F.; Huang Y. Advanced Functional Materials 2013, 23, 2436.

3) Zhong, Y.; Zhang, B.; Luo, S.; Wen W.; Li H. Journal of Power Sources 2011, 196, 6802.

4) Mai, Y. J.; Zhang, D.; Qiao, Y. Q.; Gu, C. D.; Wang X. L.; Tu, J. P. Journal of Power Sources 2012 216, 201.

5) Xu, G. L..; Xu, Y. F..; Sun, H.; Fu, F.; Zheng, X. M.; Huang, L.; Li, J. T.; Yang, S. H.; Sun, S. G. Chemical Communications 68, 2012, 8502.

6) Zhong, K.; Xia, X.; Zhang, B.; Li, H.; Wang, Z.; Chen, L. Journal of Power Sources 2010, 195, 3300.

7) Ji, L.; Zhang, X. Electrochemistry Communications 2009, 11, 795.

8) Cui, Z.; Guo, X.; Li, H. Journal of Power Sources 2013, 244, 731.

9) Liu, J.; Pan, Q. Electrochemical and Solid-State Letters 2010, 13, A139.

10) Li, X.; Li, D.; Qiao, L.; Wang, X.; Sun, X.; Wang, P.; He, D. Journal of Material Chemistry A, 2012, 22, 919.