Enhancing Lithium Manganese Oxide Electrochemical Behavior by Doping and Surface Modifications

Abstract

Lithium manganese oxide is regarded as a capable cathode material for lithium-ion batteries, but it suffers from relative low conductivity, manganese dissolution in electrolyte and structural distortion from cubic to tetragonal during elevated temperature tests. This review covers a comprehensive study about the main directions taken into consideration to supress the drawbacks of lithium manganese oxide: structure doping and surface modification by coating. Regarding the doping of LiMn₂O₄, several perspectives are studied, which include doping with single or multiple cations, only anions and combined doping with cations and anions. Surface modification approach consists in coating with different materials like carbonaceous compounds, oxides, phosphates and solid electrolyte solutions. The modified lithium manganese oxide performs better than pristine samples, showing improved cyclability, better behaviour at high discharge c-rates and elevated temperate and improves lithium ions diffusion coefficient.

1. Introduction

Since their introduction to market by Sony, in 1990, lithium-ion batteries (LIBs) became the most popular power source for consumer electronics (smartphones, laptops, portable medical devices) and electric vehicles too [1,2,3]. Lithium-ion batteries are preferred for current electrically powered devices because of their advantages over the classic energy storage devices: better gravimetric density, higher working potential, large number of charge/discharge cycles and small self-discharge rate [4,5].

The most studied cathode materials are lithium cobalt oxide (LiCoO₂) [6,7,8], lithium iron phosphate (LiFePO₄) [9,10,11] and lithium manganese oxide (LiMn₂O₄) [5,12,13]. The anode consists of carbonaceous materials [12,14], silicon [15,16] or transition metal oxides [17] and the electrolyte mainly contains LiPF₆ dissolved in a mixture of organic solvents (ethylene carbonate and dimethyl carbonate) [18].

Lithium manganese oxide (LMO) is considered an appropriate cathode material for lithium-ion batteries because of plenty raw materials, environmental friendliness, low cost, relatively facile manufacturing process and higher potential (4.1 vs. Li/Li⁺) than LiCoO₂ (4 V vs. Li/Li⁺) and LiFePO₄ (3.45 V vs. Li/Li⁺) [4,19]. The theoretical capacity of LMO is 148 mAh g⁻¹, lower than LiCoO₂ (274 mAh g⁻¹) and LiFePO₄ (170 mAh·g⁻¹). However, lithium cobalt oxide fails to deliver more than half of the theoretical capacity because of structural deformation [2,8]. Pristine LMO managed to deliver more than 95% of its theoretical capacity [20].

LiMn₂O₄ has a close cubic package (ccp) spinel structure within the Fd3m space group. Manganese ions are located in the 16 d octahedral positions, oxygen ions in 32 e sites, while the lithium ions from the 8a tetrahedral positions can diffusive through the interstitial space of the 3D diamond shaped framework formed by [MnO₆] octahedrals [21,22,23,24].

It is well known that lithium manganese oxide’s performance in organic electrolyte solution at elevate temperatures, suffers from manganese dissolution into electrolyte [25,26,27], Jahn-Teller distortion [28] and electrolyte oxidation on the surface of LMO particles [29]. Another problem related to Li-ion battery functioning is the electrolyte stability. Normally, LIBs are assembled in chambers with controlled atmosphere and humidity. If any water traces remain in the cell’s assembly, it can react with lithium hexafluorophosphate (LiPF₆) to form hydrofluoric acid (HF), which has a harmful effect on LMO particles, thus the cycling ability of the cathode material is drastically affected [30].

Two main strategies have been employed to diminish LMO’s disadvantages: structure doping (cations [31,32], anions [33,34] or multiple ions [29,35,36]) and surface modification of the particles [37,38,39]).

Cation doping stabilizes the LMO spinel structure, by reducing the amount of electrochemically active Mn³⁺ [40], which is responsible for manganese disproportionate reaction into electrolyte [41]. However, in the case of doped LiMn₂O₄ a longer cycle life has been noticed at both room and elevated temperature as well, due to lower capacity loss during charge-discharge cycles, because of the more stable unit cell [42,43]. In addition, the anions inserted into LMO’s lattice do not reduce the content of Mn³⁺, therefore the good discharge capacities of anion-doped LiMn₂O₄ spinels were recorded [44,45].

Jahn Teller distortion (JT) occurs on the particle surface and it is a phase transition from cubic (c/a = 1) to tetragonal (c/a > 1), which reduce the structure stability due to a change in volume [46]. JT occurs especially by the end of the discharge process, when the average Mn oxidation state decreases from 3.5 to 3.0 [47,48]. Surface modification of the LMO particles proved to be an efficient way to hinder the JT’s effect on cathode performance, by shielding the active material particles from the electrolyte solution, where Mn³⁺ can easily disproportionate into Mn²⁺ and Mn⁴⁺ [49].

While in a previous review [50], the most common synthesis methods for obtaining pristine LMO have been discussed, in the current article, an intensive study about the dopant ions and coating materials influence upon the stability improvement and electrochemical behaviour enhancement of lithium manganese ions have been written. This review article covers not only the most common dopants and surface coatings, but also the influence of precursors and synthesis methods upon the structure and electrochemical performance of modified LiMn₂O₄. As a means to ease the results visualisation, the electrochemical parameters (initial discharge capacity, capacity retention, number of cycles, coulombic efficiency and charge-transfer resistance and lithium diffusion coefficient) of the modified LiMn₂O₄ have been tabulated.

For a better and easier visualization of electrochemical results for doped and coated lithium manganese oxide,

Table 1. List of discharge rates abbreviations (ordered from the lowest to highest discharge rate value).

Abbr. *	Discharge Rate Value	Unit of Measurement	Abbr. *	Discharge Rate Value	Unit of Measurement	Abbr.*	Discharge Rate Value	Unit of Measurement
(a)	0.05	(C)	(i)	1	(C)	(q)	100	(mA·g−1)
(b)	0.1		(j)	2		(r)	500
(c)	0.14		(k)	3		(s)	0.1	(mA·cm−2)
(d)	0.2		(l)	5		(t)	0.2
(e)	0.25		(m)	6.75		(u)	0.3
(f)	0.3		(n)	100	(μA)	(v)	0.5
(g)	0.5		(o)	50	(mA·g−1)	(w)	0.55
(h)	0.7		(p)	75		(x)	70	(mA·g−1)

* Abbreviation.

contains the discharge rates used during the electrochemical tests. The values are arranged from the lowest to highest discharge rate, thus, some letters may not appear in alphabetical order in the following tables. Typically, the theoretical discharge capacity of LiMn₂O₄ is considered to be 148 mAh·g⁻¹, thus 1 C equals 148 mAh·g⁻¹, but there are references which consider 1 C = 296 mAh·g⁻¹ [51,52] and 1 C = 120 mAh·g⁻¹ [53], respectively.

2. LiMn₂O₄ Heteroatoms Doping

Lithium manganese oxide’s major drawbacks like structural instability during long time cycling and manganese dissolution can be suppressed by doping the pristine spinel with alkali (Na⁺ [54]), alkali-earth (Mg²⁺ [55]), transition (including lanthanides) (Ni [56,57,58,59], Cu²⁺ [60,61], Zn²⁺ [62,63], La³⁺ [64], Dy⁴⁺ [65,66], Ce⁴⁺ [67]) or triels metals ions (Al³⁺ [31,68]).

Among the studied transitional metals, nickel offered an interesting doping solution, due to high potential oxidation (4.7 V) of Ni²⁺ to Ni⁴⁺ [69]. However, the molar ratio between Ni and Mn must be carefully chosen, because, during cycling, two different regions can appear: one with increased Ni content and the other one with reduce nickel. Those two regions may determine the presence of Jahn-Teller distortion [59]. It was established that the optimal ratio between Li:Ni:Mn ions is 1:0.5:1.5 [69,70]. More aspects related to LiNi_0.5Mn_1.5O₄ will be discussed in other sections of the review (see Future prospects). Strategies employed for enhancing structural stability of the lithium manganese oxide spinel included cation multi doping [71,72,73,74,75,76].

Heteroatoms used to stabilize MnO₆ decrease the discharge capacity, in comparison with undoped LMO, by reducing the content of electrochemically active Mn³⁺, thus, several studies aimed to introduce not only cations, but also monovalent anions, like Cl⁻ or F⁻ [77,78], into the structure of LMO spinel. A study published by Mao et al. describes the doping effects of different cations and anions upon the electrochemical performance of LiMn₂O₄ [79].

2.1. Simple Ions Doping

2.1.1. Cations

Aluminum-Doped LiMn₂O₄

Aluminum is one of the most studied cations for doping lithium manganese oxide [80,81,82]. In the

Table 2. Al-doped lithium manganese oxide and its performances during electrochemical tests.

Nr. Crt.	Material	Synthesis Method	Discharge Capacity	Coulombic Efficiency	Discharge Capacity/N Cycles	Capacity Retention	Ref
			(mAh·g−1)	(%)	(mAh·g−1)	(%)
1	LiAl0.02Mn1.98O4	Hydrothermal reaction	129.8 (d)	95.4	106.05/50	81.7	[85]
2	LiAl0.1Mn1.9O4	Phase-inversion technology	117 (d)	96	112.32/420	96	[84]
			120 (i)	-	102/50	85
3	LiAl0.025Mn1.975O4	Sol-gel	109 (i)	-	88/50	80.7	[88]
4	LiAl0.1Mn1.9O4		113.2 (g)	-	109.9/100	97.1	[89]
5	Al-doped LMO		120.1	99.9	115.9/200	96.5	[42]
			119.6 (l)		111.6/100	93.3
6	LiAl0.1Mn1.9O4		81.6 (j)	-	72/1000	88.7	[90]
			89.9 (j)	-	77.4/500	86.1
7	Li1.08 Al0.09 Mn1.83O4	Spray drying	102 (b)	86	-	-	[91]
8	LiAl0.075Mn1.925O4		111 (j)	-	90.6/1000	81.6	[92]
9	LiAl0.08Mn1.92O4	Polymer pyrolysis method	114 (o)	98.6	114/50	100	[83]
			116 (o)	96.3	115.2/50	99.3
10	LiAl0.1Mn1.9O4	Solid state reaction	105.6 (m)	-	73.9/4580	70	[93]
11	LiAl0.1Mn1.9O4		122.5 (v)	91.4	42.2/75	34.7	[94]
12	LiAl0.04 Mn1.96O4		107.1 (d)	94.3	98.5/100	92	[95]
13	Al doped LMO	Self-template	100 (l)	-	78.4/774	78.4	[96]
			100 (l)	-	72.3/200	72.3
14	LiAl0.1Mn1.9O4		107 (g)	92.8	-	-	[43]
			90 (k)		62/500	70
15	LiAl0.10Mn1.90O4	Solution combustion method	124.8 (d)	-	111.1/20	89	[97]
16	LiAl0.1Mn1.9O4		122 (d)	-	113.3/25	92.9	[98]
17	LiAl0.375Mn1.625O4		113.1 (d)	-	96.1/50	85	[68]
18	LiAl0.4Mn1.6O4		113 (h)	-	104/100	92	[99]
19	LiAl0.05Mn1.95O4	Combustion method	159 (b) *	-	75.7/50	47.6	[51]
20	LiAl0.1Mn1.9O4		146 (b) *	-	56.5/50	38.7
21	LiAl0.5Mn1.5O4		95 (b) *	-	67.3/50	70.8
22	LiAl0.16Mn1.5O4		100.7 (g)	-	92.8/400	93.9	[82]

* 1 C = 296 mA·g⁻¹ in the voltage range 2.4–4.8V vs. Li [51].

, some studies related to aluminum ion employment as a stabilizing cation are presented. The reasons why aluminum was widely used as a dopant for LMO are related to its costs, stability and plentiful raw materials. As it is described in Ref [43,83], Al³⁺ occupies the 16 d octahedral sites, its bond with oxygen is much stronger than the Mn–O bond, and thus the aluminum ions improve spinel’s structure stability during charge-discharge at room temperature and elevated temperature, respectively. LiAl_0.08Mn_1.92O₄ synthesized by a polymer pyrolysis method exhibited amazing results at both room and elevated temperature. Delivering an initial discharge capacity of 114 mAh·g⁻¹, LiAl_0.08Mn_1.92O₄ managed to offer the same capacity after 50 cycles (discharge rate: 50 mA·g⁻¹), while at elevated temperature, a capacity loss of only 0.014% per cycle was recorded. The electrochemical performances of LiAl_0.08Mn_1.92O₄ were strongly connected to its improved crystallinity and uniform particle size. It was considered that an amount of 0.10 wt.% of Al can suppress the structural distortion during cell’s employment. LiAl_0.1Mn_1.9O₄ obtained by a phase inversion technology was long-time cycled (420 cycles) and it still delivered 96% of its initial capacity of 117 mAh·g⁻¹ [84]. One of the highest initial discharge capacities furnished by an Al-doped spinel [85], comparable to the ones showed by undoped spinels [47,86,87], was the result of a hydrothermal synthesis method. Owing 129.8 mAh·g⁻¹ in the first cycle, it displayed a capacity loss of 0.475 mAh·g⁻¹ per cycle, having a capacity retention of 81.7% after 50 cycles.

The amount of aluminum dopant quantity should be carefully controlled because an increased amount of Al³⁺ ions into the LMO framework alter the insertion/deintercalation of Li⁺, by occupying the lithium ions 8a positions inside LiO₄ tetrahedral [43]. Moreover, an increased Al content reduces the lattice parameter, thus an initial lower discharge capacity will be provided by modified spinels, like in the case of LiAl_0.5Mn_1.5O₄ material, when the initial discharge capacity was below 100 mAh g⁻¹ (95.2 mAh g⁻¹). Although, the majority of the studies were conducted using moderate aluminum amount, the LiAl_0.375Mn_1.625O₄ synthesized by solution combustion method performed above average and it granted 96.1 mAh g⁻¹ after 50 cycles (0.2 C) [68]. Other high-content Al-doped spinels were obtained employing a solution combustion technique, with and without the adding of oleic acid [99]. Oleic acid favors the formation of a micelle structure template that influence the lattice parameter and morphology, thus LiAl_0.4Mn_1.6O₄ nanospheres were synthesized in the presence of oleic acid, while in its absence nanorods were formed [99]. Even though a similar initial discharge capacity was noticed for both morphologies, rod-like LMO performed worse and delivered only 87.4 mAh g⁻¹ (78% of its initial capacity), while LMO nanospheres exhibited 104 mAh g⁻¹ (92% of the original capacity).

Al-doped spinel was tested in aqueous 5 M LiNO₃ electrolyte solution. The tests implied a long-time cycling (4580 cycles) and LiAl_0.1Mn_1.9O₄ performed impressively, maintaining 70% of its initial discharge capacity of 105.6 mAh g⁻¹ [93]. The current achievement overcame the cycling ability recorded for other LiAl_0.1Mn_1.9O₄ reported in Ref [88] and [96].

Rare earth-Doped Lithium Manganese Oxide

Another class of single doped lithium manganese oxides is represented by the rare earth elements doped LMO. The main studies reported in the literature are indexed in

Table 3. Rare-earth doped LMO.

Nr. Crt.	Material	Synthesis Method	Discharge Capacity	Coulombic Efficiency	Discharge Capacity/N Cycles	Capacity Retention	Ref
			(mAh g−1)	(%)	(mAh g−1)	(%)
1	LiLa0.01Mn1.99O4	Sol-gel	127.6 (d) 109.8 (i)	92.1	116.8 (c) 92.5 (h)	91.6/0.2 C 84.3/1 C	[64]
2	LiCe0.1Mn1.9O4		101 (i)	-	98/150	97	[19]
3	LiLa0.01Mn1.99O4		122.8 (p)	96.4	113.5/50	92.4	[100]
4	LiCe0.05Mn1.95O4		126 (r)	94.7	117/100	92	[101]
5	LiPr0.2Mn1.98O4		118 (f)	-	115.6/100	98	[103]
6	LiDy0.05Mn1.95O4		54.4 (g)	-	49.1/50	93	[102]
7	LiDy0.2Mn1.8O4		124 (b)	81	85.6/10	69	[65]
8	Li Ce0.02Mn1.98O4	Solid state reaction	119.6 (v)	97.6	108.5/50	91.0	[104]
9	LiLa0.1Mn1.9O4		100 (g)	92.2	96.9/100	96.9	[105]
			101.4 (g)	-	95.8/50	94.5
10	LiCe0.1Mn1.9O4		99.1 (g)	96	92.4/100	93.2
			101.2 (g)	-	93.4/50	92.3
11	LiNd0.1Mn1.9O4		99.4 (g)	91	95.2/100	95.8
			101.3 (g)	-	92.4/50	91.2
12	LiSm0.1Mn1.9O4		99.8 (g)	99.2	93.6/100	93.8
			102.1 (g)	-	94.1/50	92.2

. Lanthanum and cerium are among the most investigated lanthanoids [19,100,101]. The larger La³⁺ ion has been inserted into spinel’s structure and due to stronger bonds between La–O than Mn–O, the lattice parameter and crystals size of doped samples are smaller than for pristine LMO [100].

Preliminary studies related to Dy compatibility as a dopant were realized [65,102]. It was noticed that an extremely low amount of Dy (0.02 wt.%) could not stabilize and improve the LMO performance, while an increased quantity led to enhanced results, offering an initial discharge capacity similar to undoped spinels, but a severe capacity fade of 31% was observed after only 10 cycles. Thus, further studies need to be done, in order to take Dy in consideration as a dopant for LiMn₂O₄.

Other Metal Cations—Doped LiMn₂O₄

Throughout the years, many other metal ions were tested as possible dopants for LiMn₂O₄ (

Table 4. Cations doped lithium manganese oxide.

Nr. Crt.	Material	Synthesis Method	Discharge Capacity (mAh g−1)	Coulombic Efficiency (%)	Discharge Capacity/N Cycles (mAh g−1)	Capacity Retention (%)	Ref
1	LiRu0.25Mn1.75O4	Solid state reaction	125 (c)	85	113.8/25	91	[109]
2	Li1.05Co0.10Mn1.85O4		81.3 (d)	97.3	80/30	98.4	[110]
3	Co-LiMn2O4 nanotubes		126.6 (b)	93.1	100/50	79	[53]
4	Li1.02Nb0.01Mn1.99O4		127.9 (t)		113.2/60	88.5	[111]
5	LiCr0.04Mn1.98O4	Solid State Method	119.6 (i)	-	109.9/350	91.9	[112]
	LiCr0.15Mn1.85O4		147.2 (i)	106.6	-	-	[113]
6	LiCr0.1Mn1.9O4	Sol-gel	123.6 (s)	-	123/10	99.5	[114]
7	LiW0.025Mn1.975O4		77.1 (i)	-	42/50	54.5	[88]
8	LiTixMn2−xO4		111.2 (b)	94.9	89.2/100 d)	83.7	[115]
9	LiSn0.02Mn1.98O4		123.1 (v)	94.6	111.7/100	90.7	[116]
			114.0 (v)	-	92.7/50	81.3
10	Li1.05Cr0.1Mn1.85O4		113.7 (t)	-	108.6/38	95.5	[117]
11	LiCo0.15Mn1.85O4		94.4 (u)		90.6/30	96	[118]
12	LiZn0.10Mn1.90O4		122 (g)	96.1	107/100	87	[119]
13	Li1.03Mg0.2Mn1.77O4		88 (n)	88	87/10	98.9	[106]
14	Li1.03Fe0.2Mn1.77O4		110 (n)	-	109/10	99.1
15	Li1.05Co0.05Mn1.95O4		120.7 (b)	95.7	116.2/20	96.3	[120]
16	LiCr0.2Mn1.8O4		105.4 (r)	100	54.1/10000/2000 mA/g	73	[121]
17	LiGd0.02Mn1.98O4		116.0 (d)	-	98.6/70	85	[86]
			116.5 (i)	-	92.04/70	79
18	LiGd0.05Mn1.95O4		59.4 (g)	-	48.4/50	86	[102]
19	LiFe0.1Mn1.9O4		134.7 (g)	85.4	92.5/1400 100C	92	[52]
			-	-	75.8/1500 100C	73(d)
20	LiBa0.02Mn1.98O4	Combustion method	115.2 (i)	-	111.2/30	96.5	[122]
21	LiMg0.02Mn1.98O4		107.2 (d)	87.4	95/100	88.62	[123]
22	LiNi0.1Mn1.9O4	Solution combustion method	116.4 (b)	-	115.2/50	99	[124]
23	LiNi0.1Mn1.9O4		116.4 (h)	-	115.2/50	99
24	LiCr0.02Mn1.98O4	Molten salt combustion	120.7 (d)	88	94.7/40	78.5	[107]
25	LiSi0.05Mn1.95O4	Freeze-drying method	146 (b) 139 (i)	-	105.5/300	75.9	[108]
26	LiGa0.1Mn1.9O4		99 (b) 95 (i)	-	80.2/300	85
27	LiCo0.05Mn1.95O4	Biomimetic synthesis	106.9 (q)	92.1	93.6/100	87.6	[125]
28	LiCr0.12Mn1.88O4	Ionothermal synthesis	129.6 (g)	96.6	125.5/200	96.8	[126]
29	Co-doped LiMn2O4	Rheological phase reaction	112.5 (d)	88.4	100/100	88.9	[41]
30	Li1.1Nb0.03Mn1.87O4	Ion implanted method	120.5 (i)	-	113.8/50	94.4	[127]
			122.1 (i)	-	112.5/50	92.1

).

Iron doped lithium manganese spinel was successfully synthesized by a sol-gel process [52]. Iron ions improved the electrochemical performances by partially replacing the Mn³⁺ from 16 octahedral sites (fact which suppressed the Jahn-Teller distortion and stabilized the spinel’s structure), by reducing the amount of Mn³⁺, which were responsible for blocking the lithium ions movement channels, in 8a tetrahedral sites and preventing the formation of manganese with lower valance on the particles surface. The mixture of chelating agents used during the sol-gel process, citric acid and glucose, helped the targeting of iron ions to the mentioned sites, by affecting the kinetics of the synthesis reactions. The altered LiFe_0.1Mn_1.9O₄ with an initial discharge capacity of 134.7 mAh·g⁻¹ demonstrated long cycle life at high discharge rate (100 C), mainly because the structural modifications occurred after Fe³⁺ doping. The discharge capacities exhibited by iron doped spines after 1400 cycles at RT and 1500 at 60°, respectively, were by the far better than in the case of dopants like Mg [106], Gd [102], Cr [107] or Ga [108].

A silicon modified LMO with high initial discharge capacity (146 mAh·g⁻¹, 0.1 C) has been obtained by a freeze-drying method [108]. In comparison with divalent or trivalent ions, by doping with tetravalent ion (Si⁴⁺), the quantity of electrochemical active Mn³⁺ was not reduced. It has been demonstrated during the charge-discharge cycles that the entire trivalent manganese oxidized to Mn⁴⁺. After 300 cycles, LiSi_0.05Mn_1.95O₄ exhibited a lower capacity retention (75.9%) than the gallium-doped spinel synthesized by a similar technique, because of the larger Mn³⁺ content.

Chromium influence upon structure, stability and electrochemical properties of LMO spinels was examined [114,121,126]. Because Cr⁶⁺ is toxic, the reagent salts must be carefully picked. A relatively low content of Cr³⁺ stabilized LMO’s structure by replacing Mn in 16 d sites, reduced the charge-discharge resistance, thus favoring the lithium ions diffusion during cycles [126]. Unfortunately, the calculated lithium-ion diffusion coefficient of 9.61·10⁻¹² cm²·s⁻¹ [126] was similar to titanium surface-doped LMO [115] and much smaller than for LMO with incorporated PO₄³⁻ [128].

High content chromium doped lithium manganese oxide (LiCr_0.2Mn_1.8O₄) exhibited an initial low discharge capacity (105.4 mAh·g⁻¹) in an Aqueous Rechargeable Li-ion Battery (ARLB), but it proved a very long cycle ability and after 10,000 cycles, the capacity fade was only 27%. This outstanding result was attributed to stronger stabilizing Cr–O bonds, especially in low lithiated state [121].

Among other ions, pentavalent ions such as Nb⁵⁺ were inserted into LMO spinel’s lattice by different synthesis methods, solid state reactions [111] and ion implanted method [127]. Niobium has been selected as a dopant ion due to its similar ionic radius to manganese ion, which can increase the electrical conductivity of the material due to extra electron and its metal-oxygen bond has higher energy [111]. In the case of Li_1.02Nb_0.01Mn_1.99O₄ obtained by SSR, LiNbO₃ impurities have been noticed, but these impurities favored lithium movement throughout the cycling tests. By increasing the Nb⁵⁺ amount, LiNbO₃ hindered the lithium diffusion by blocking its diffusion paths.

A preliminary study related to wolfram doping of LMO has been published [88]. Due to wolfram oxidation to a higher state (W⁶⁺), its ionic radius (0.62 Å) became shorter than Mn³⁺ ionic radius (0.64 Å), thus the lattice parameter of W doped spinel was smaller than pristine sample. Poor capacity of LiW_0.025Mn_1.975O₄ was mainly connected to the presence of electrochemical inert wolfram oxides and high-valance wolfram state.

Rod-like Sc-doped lithium manganese oxides were synthesized by a solid state reaction route by Bhuvaneswari et al. [129]. Rod length varied between 100–600 mm depending on the weight ratio of scandium ions into the structure and increased Sc content determined longer rods. Among the tested percentage of Sc, the best results were recorded for LiSc_0.06Mn_1.94O₄. A slight decrease in the cell parameter “a” was noticed, but this aspect was attributed to the stronger Sc–O bond. Moreover, by introducing Sc³⁺ ions into the structure in the 16 d sites, the content of Mn³⁺ decreased, thus the resulted crystal structure is more stable, because by taking into consideration that Mn³⁺–O bond is longer than Mn⁴⁺-O bond, the MnO₆ octahedrons have a more compact arrangement and less volume change during cell’s operating. The Sc-doped LMO delivered a slightly lower discharge capacity (114 mAh g⁻¹) than pristine LMO (117 mAh·g⁻¹) but succeeded to maintain ~96% of the initial capacity after a long time cycling of 500 cycles (discharge rate = 1 C), while pristine sample failed more than 75% of the first cycle capacity. Li⁺ ions’ diffusion coefficient was one magnitude order higher for the Sc-doped (1 × 10⁻¹² cm² s⁻¹) sample than for undoped LMO (1 × 10⁻¹³ cm² s⁻¹). However, the calculated D_Li+ for LiSc_0.06Mn_1.94O₄ coefficient was not as large as in other cases like La-Sr-Mn coated LiMn₂O₄ [130], LaF₃ coated LMO [131] or LiAl_0.04Mn_1.96O_3.96F_0.04 [95].

2.1.2. Anions

Anions were also considered as potential doping agents for stabilizing the structure of lithium manganese oxide (

Table 5. Anions doped LMO.

Nr. Crt.	Material	Synthesis Method	Discharge Capacity (mAh g−1)	Coulombic Efficiency (%)	Discharge Capacity/N Cycles (mAh g−1)	Capacity Retention (%)	Ref
1	LiMn2O4−xSx	Plasma assisted solid state method	125.2 (d)	-	122.4/60	97.8	[33]
			123.5 (b)	-	108.2/60	87.6
2	LiMn2O3.94F0.06	Solid state reactions	116 (d)	98.4	107.8/100	93	[95]
3	LiMn2O3.99S0.01	Sol-gel	126 (i)	83.2	116.6/70	92.1	[44]
4	LiMn2O3.99S0.01	Sol-gel	158.9 (b) *	95.5	-	-	[45]
5	LiMn2O3.96F0.4	Combustion method	101.2 (d)	-	74.5/40	73.6	[34]
6	Li1.017Mn1.97O4(PO4)0.015	Hydrothermal method	116 (i)	98.3	-	-	[132]

* in the case of LiMn₂O_3.99S_0.01 [45] the discharge capacity is considered to be after the 5th cycle.

).

Fluorine (F⁻) was regarded as a promising dopant, because, due to its lower ionic radius and higher electronegativity, it partially replaces divalent O²⁻ and decreases the Mn⁴⁺ content [34,95]. By reducing Mn⁴⁺ to Mn³⁺, the amount of electrochemically active Mn³⁺ increased, thus, the initial capacity of the fluorine doped spinel was higher than of aluminum doped LMO, synthesized under similar conditions [95].

Sulfur doped spinels were synthesized by plasma assisted solid state method and sol-gel [33,44]. In both cases, high discharge capacities (above 125 mAh·g⁻¹) were registered in the first cycle and the capacity retention was higher than 92% after 60 cycles or more. This remarkable behavior was attributed to stabilizing character of sulfur, by efficiently preventing the Jahn-Teller distortion at low voltage and by reducing the amount of dissolved manganese into the electrolyte solution, even at elevated temperature. Another important aspect of the improved electrochemical performances of LiMn₂O_4−xS_x can be attributed to a more powerful manganese-sulfur bond, which stabilized the structure.

The phosphate group used as a dopant was assimilated into the LMO lattice and it weakened the manganese-oxygen bond, fact proven by FTIR spectra [132]. As a result of no manganese content change into the modified samples, the initial discharge capacity of the Li_1.017Mn_1.97O₄(PO₄)_0.015 was almost similar to the one for pristine sample (116 mAh·g⁻¹). An excellent benefit of the (PO₄) doping was related to the lithium diffusion coefficient, which increased by three magnitude orders, from 3.49·10⁻¹¹ cm²·s⁻¹ for LiMn₂O₄ to 5.38·10⁻⁸ cm²·s⁻¹ for Li_1.017Mn_1.97O₄(PO₄)_0.015. Such improvement was connected to the PO₄³⁻’s structure, its ionic radius being larger than for O²⁻, favored the formation of a 3D framework, with large channels which facilitated the lithium ions movement during the electrochemical tests.

2.2. Multiple Ions Doping

2.2.1. Cations Multi-Doping

Several attempts for multi-doping of LiMn₂O₄ have been investigated (

Table 6. Cations multi-doping of LMO.

Nr. Crt.	Material	Synthesis Method	Discharge Capacity (mAh g−1)	Coulombic Efficiency (%)	Discharge Capacity/N Cycles (mAh g−1)	Capacity Retention (%)	Ref
1	LiZn0.10Pr0.10Mn1.75O4	Sol-gel	130 (b)	76.5	128.8/10	99.1	[136]
2	LiZn0.10Al0.15Mn1.75O4		122 (a)	71.3	112/10	91.8	[137]
3	LiZn0.10Ho0.10Mn1.75O4		127 (a)	91.0	115/10	90.6	[72]
4	LiCr0.5Mg0.05Mn1.45O4		123 (b)	81.5	112/10	91.1	[74]
5	LiCu0.5Cr0.05Mn1.45O4		124 (b)	79	113/10	91.1	[75]
6	LiCu0.5Al0.05Mn1.45O4		120 (b)	78	114/10	95	[138]
7	Li0.94Na0.06Al0.1Mn1.9O4		115.9 (d)	-	-	-	[71]
8	LiNi0.5Al0.05Mn1.45O4		126 (b)	86	116/10	92.0	[73]
9	LiZn0.01Ce0.01Mn1.98O4		124 (b)	97.6	123/10	99.2	[139]
10	LiMg0.05Si0.05Mn1.9O4		126.9 (g)	-	123.5/100	97.3	[140]
			126.4 (g)		117.1/20	92.7 (g)
11	LiAl0.025W0.025Mn1.95O4		45.6 (i)	-	36.2/50	79.4	[88]
12	Li1.01Co0.05Ni0.099Mn1.83O4		91.5 (u)	-	89.7/30	98	[118]
13	Li1.01Co0.05Zn0.094Mn1.82O4		86.7 (u)	-	85.8/30	99
14	Li1.01Co0.048Cu0.099Mn1.82O4		85.7 (u)	-	84.1/30	98.1
15	Li1.03Mg0.1Fe0.1Mn1.77O4		110 (n)	-	107/10	97.3	[106]
16	LiMg0.01Sn0.04Al0.35Mn1.6O4		114.6 (b)	61.8	113.5/10	99	[35]
17	LiLa0.01Zn0.01Mn1.98O4		118.0 (f)	-	99.1/100	84	[136]
18	Li1.03Zn0.03Mg0.03Mn1.91O4		107.5 (f)	97.4	107.4/40	97.4	[141]
19	LiNi0.01Cr0.01Mn1.98O4		113 (b)	-	98/100	87	[133]
20	Li1.00 Co0.028Ni0.026Mn1.95O4		116.1 (i)	99.4	112.6/200	97	[40]
			115 (i)	-	110.4/100	96
21	Li(GdNi)0.01Mn1.98O4		67.1	99	-	-	[142]
22	LiAl0.0125Si0.0125Ti0.0125Co0.0125O4	Combustion process	111.8 (i)	-	106.7/30	90	[135]
23	LiCo0.025Cr0.025Ni0.025Fe0.025Mn1.9O4		105.6 (i)	91.8	85.5/70	81	[143]
24	LiBi0.005La0.005Mn1.99O4	Solution combustion method	130.2 (i)	97.9	123.7/80	95	[21]
25	LiGe0.1Li0.02Mn1.88O4	Solid state reaction	124 (i)	98.4	63/100	50.8	[144]
26	LiSr0.1Cr0.1Mn1.8O4		128 (i)	93.3	116/100	90.6	[29]
27	LiAl0.125Co0.125Mn1.75O4		107.3 (i)	89.9	90.2/350	84.1	[145]
			108.2 (j)	96.1	82.4/180	85.7
28	LiAl0.1Ni0.1Mn1.8O4		140.5 (v)	95.6	66.2/75	47.4	[94]
29	LiAl0.03Si0.05Mg0.05Mn1.87O4		118.7 (g)	-	115.2/50	97.1	[146]
30	LiNi0.03Mo0.01Mn1.96O4		114 (i)	95.6	/300	91.2	[147]
31	LiCe0.012Nd0.012Mn1.976O4	Pechini process	123.5 *	-	113.2/30	91.6	[148]
32	LiCu0.02Al0.02Ti0.02Mn1.94O4		134 (g)	-	130/400	97	[149]
			131.2 (g)	-	119.2/50	91
33	LiAg0.05Y0.05Mn1.9O4	Co-precipitation method	137 **	99	-	-	[150]
34	LiZn0.15W0.10Mn1.75O4		120 (b)	95.2	99/25	82.5	[76]
35	Li1.088Al0.037Co0.028Mn1.847O4	Hydrothermal method	102.9 (i)	-	97.3/100	94.5	[151]

* Discharge rate not available. ** Discharge rate not available; the sample was also irradiated with 30 kGy (non-irradiated sample delivered 129 mAh·g⁻¹).

). Except the entries marked with ^(g) in Table 5, all the tests have been done at room temperature or 25 °C.

An equal percentage of zinc and praseodymium led do a higher initial discharge capacity (130 mAh g⁻¹), not only in the case of other single- [115] or multi-doped spinels [133], but also for pristine LiMn₂O₄ [134]. Despite of the low initial coulombic efficiency (only 76.5%), it managed to increase several percentages after a short number of cycles (10 cycles) and it is expected to slightly increase, if the cycling tests will last longer. Almost no capacity fade was recorded for the investigated number of cycles.

Even small amounts of bismuth and lanthanum had a beneficial effect on LiMn₂O₄ rate capability and an extremely small capacity loss of only 0.08 mAh g⁻¹ per cycle was recorded [21].

Multi cations (Al, Si, Ti, Co) doped LiMn₂O₄ was suggested by Dokan et al. [135]. In comparison with the pristine spinel synthesized via the same glycine-nitrate combustion process, the doped sample exhibited an initial higher discharge capacity and the dopants stabilized the spinel structure, thus a capacity fade of only 10% was recorded after 30 cycles (1 C).

Mesoporous Al and Ni doped LiMn₂O₄ spherical particles, with high initial discharge capacity (140.5 mAh·g⁻¹), related to stabilizing function of Ni, were synthesized by solid state reactions, using as-synthesized MnCO₃ microspheres, LiOH·H₂O and Ni(NO₃)₂·6H₂O [94]. However, in spite the advantages offered by porous structure and aluminum and nickel ion doping, LiAl_0.1Ni_0.1Mn_1.8O₄ was also affected by the structural distortion from cubic to tetragonal during cycles, and severe capacity loss (52.6%) was noticed after only 75 cycles.

Impressive electrochemical performance was recorded for a LMO doped with low content three ions with different valences of Cu²⁺, Al³⁺ and Ti⁴⁺ through a lactic acid aided Pecchini process [149]. According to the authors, the homogenous distribution of the dopants played an important role in achieving such effectiveness. In the first cycle, 90.5% of LMO’s theoretical capacity (148 mAh g⁻¹) was delivered and after a large number of cycles (400 cycles, RT), the capacity fade was just 3%. High capacity retention of 90.1% (119.2 mAh g⁻¹) has been measured for LiCu_0.02Al_0.02Ti_0.02Mn_1.94O₄ sample at elevated temperature (55 °C) too. The results obtained for the current doped spinel are similar to magnesium and fluoride multi-doping of LiMn₂O₄ [78].

LMO spinel stability has been enhanced by an equimolar structure doping with magnesium and silicon ions [140]. The co-doping with Mg²⁺ and Si⁴⁺ favored two aspects: first of all, by doping with Mg²⁺, the average valence of manganese increased to (IV), which hindered the Jahn-Teller structural modification and resulted in a more stable compound; secondly, the Si⁴⁺ determined the formation of a more lasting [MnO₆] octahedral, which permitted the lithium ions diffusion much easier. By increasing the content of dopants, the discharge capacity diminished (from 139 mAh·g⁻¹ for pristine sample down to 115.6 mAh·g⁻¹ for LiMg_0.07Si_0.07Mn_1.86O₄), but the cycling life has been improved, from 87.1% (bare LMO) to 99.3% (LiMg_0.07Si_0.07Mn_1.86O₄), after 30 cycles. LiMg_0.05Si_0.05Mn_1.86O₄ has been chosen as the optimum doped sample, with an initial discharge capacity of 126.9 mAh·g⁻¹ and a retention rate of 97.3% after 100 cycles. In addition, it showed enriched capacity retention at elevated temperature (92.7%, 0.5 C, 55 °C) and low charge-transfer resistance after 100 cycles (102.9 Ω), which suggested better lithium intercalation/deintercalation, because side compounds did not block Li⁺ paths.

Preliminary studies of double doping of LiMn₂O₄ with Gd and Ni have been attempted in Ref. [142]. The obtained material was tested in aqueous electrolyte (saturated Li₂SO₄ electrolyte) and it delivered an initial fairly poor discharge capacity of only 67.1 mAh·g⁻¹. The doped spinel exhibited a similar behavior like in non-aqueous electrolytes and two pairs of peaks were noticed. Moreover, the doping reduced the charge transfer resistance for almost ten times, from 800 Ω (pristine sample) to 85 Ω (Li(GdNi)_0.01Mn_1.98O₄), this fact suggested improved kinetics for the double-doped spinel.

Nickel and copper multiple metal doping of lithium manganese oxide by a citric acid aided sol-gel process has been realized by Iqbal et al. [152]. In the case of samples with low amount of dopants, Ni-Cu ions tend to occupy the tetrahedral positions 8a, while, by increasing the amount of doping ions, Ni-Cu will reside to 16d octahedral sites. The differences in lattice constant and particles size between pristine, low and high doped LMO resemble in the fact that tetrahedrally coordinated Cu²⁺(0.57 Å) own a slightly lower ionic radius than Li⁺(0.59 Å), while the Mn³⁺ ion, which is normally replaced by dopants, has a radius relatively lower (0.645 Å) than octahedrally coordinated Cu²⁺(0.73 Å) [152]. Regarding the electrochemical performances, all the samples showed two pairs of redox peaks (at 4.1 and 3.9 V, respectively), which are characteristic for bare lithium manganese oxide. While the pristine sample delivered a specific capacity of 122 mAh·g⁻¹ (82.4% of theoretical discharge capacity) but faced drastically capacity fade over a number of 100 cycles (28%). Among all the samples, the best results were recorded in the case of low doped LMO, LiCu_0.01Ni_0.01Mn_1.98O₄. It provided a discharge capacity of 113 mAh·g⁻¹ (c = 0.3 C) and maintained 88% of its initial capacity after 100 cycles. The very low amount of dopants managed to enhance the lithium kinetics during charging by reducing the charge-transfer resistance by more than two times, from 212 to 97 Ω.

2.2.2. Anions and Cations Co-Doping of LMO

Many synthesizes investigated the effect of anion and cation co-doping upon the structure, morphology and electrochemical performances of lithium manganese oxide (

Table 7. Anion and cation multi-doping of LMO.

Nr. Crt.	Material	Synthesis Method	Discharge Capacity	Coulombic Efficiency	Discharge Capacity/N Cycles	Capacity Retention	Ref
			(mAh g−1)	(%)	(mAh g−1)	(%)
37	LiNi0.4Mn1.6O4-δSδ	Combustion Method	126 (t)	96.9	135.5/25	107.1 *	[153]
38	Li1.05Ga0.02Mn1.98O3.98S0.02	Solid state reaction	120	-	115/300	95.8	[154]
39	Li1.05Al0.02Mn1.98O3.98S0.02		126	-	118.4/300	94
40	LiMg0.1Mn1.9O3.8F0.2		121.1 (i)	98.9	116.3/400	96.0	[78]
					108/400	89.2 (g)
41	LiMg0.05Mn1.95O3.9Cl0.1		125.2 (d)	-	111.8/100	89.3	[77]
42	LiLa0.02Mn1.98O3.95F0.05		123.6	-	114.6/30	92.7	[155]
43	Li1.02Co0.1Mn1.9O3.98S0.02		122.1 (g)	86.9	109.8/100	90	[156]
			113 (g)		70/50 (g)	61.9
44	LiMn1.96Al0.04O3.94F0.06		109.1 (d)	93.1	103.6/100	95	[95]
45	Li1.02Mg0.05Mn1.93O3.95F0.05		115.9 (f)	99.7	111.8/40	96.5	[157]
			116.3 (f)	-	105.1/40	90.4
46	Li[Li0.01Al0.1Mn1.89]O3.9F0.1		110 (i)	97.3	101/100	91.8	[158]
			111 (i)	95.9	100.9/100	90.9
47	LiZn0.03Mg0.03Mn1.94O3.95F0.05	Sol-gel	120.1 (t)	93.5	118.3/40	98.5	[159]
48	LiLa0.01Mn1.99O3.99F0.01		126.4 (e)	96.9	116.3/50	92	[100]
49	Li0.99K0.01Mn2O3.99S0.01		117.8 (i)	86.4	108.5/70	92.1	[44]
50	LiCo0.01Al0.05Mn1.94O3.95F0.05		122.9 (f)	96.7	118. 9/40	96.8	[160]
			120.8 (f)	-	111.7/40	92.5
51	LiNi0.03Mn1.97O3.95F0.05		120.3 (b)	-	113.7/100	94.5	[161]
					96.7/100g)	80.4
52	Li1.05Mn1.84Al0.11O3.99F0.01	Spray drying process	113 (b)	98	-	-	[91]
53	Li1.09Mn1.87Nb0.031O3.99(PO4)0.021	Ion implanted method	119.2 (i)	-	115.6/50	97.0	[127]
			121 (i)		111.3/50	92.0

).

A multi doped LMO with Mg and F exhibited a relatively good discharge capacity in the first cycle (121.1 mAh·g⁻¹). Moreover, it had an excellent rate capability at room temperature and elevated temperature (55 °C), delivering 116.3 and 108 mAh·g⁻¹, respectively, after 400 cycles at 1 C [78].

In the case of simultaneous cation and anion doping, the strength of newly formed bonds, like Mg–O and Mn–Cl are better than Mn–O, thus the spinel structure was stabilized and the Mn³⁺ dissolution into electrolyte diminished [77].

Raja et al. produced [153] a nickel and sulfur co-doped LMO by an alanine-assisted low temperature combustion [153]. The nickel ion partially substitutes Mn³⁺ and sulfur, due to its improved catalytic activity, favors the oxidation of Mn³⁺ to Mn⁴⁺, resulting in a more stable [MnO₆]²⁻ and preventing the Jahn-Teller distortion. By increasing the amount of dopants, the LMO’s morphology changes from spherical shaped particles, in the case of pristine sample, to octahedron grain for the sample, which contained a high amount of nickel (LiNi_0.5Mn_1.5O_4−δS_δ). The multifaceted LiNi_0.4Mn_1.6O_4−δS_δ exhibited an interesting electrochemical behavior, because no capacity fade was observed in the first 25 cycles, but, moreover, the discharge capacity increased from 126 to 135.5 mAh·g⁻¹. These phenomena might be related to sulfur catalytic activity and stabilizing nature at high voltage (4.5 V). Solid-state reaction method has been employed to obtain multiple spinel with magnesium and fluorine [78]. According to TGA analysis, prior to spinel formation exothermic process (708 °C for pristine and 687 °C for modified spinels, respectively), characteristic endothermic processes for water elimination (between 50–100 °C) and lithium carbonate discomposure occurred. Low doping level of lithium manganese oxide samples with Mg²⁺ and F^- (LiMg_0.05Mn_1.95O_3.9F_0.1 and LiMg_0.1Mn_1.9O_3.8F_0.2) exhibited better Mn³⁺/Mn⁴⁺ ratio, hence the initial discharge capacities (119.1 and 121.4 mAh·g⁻¹, respectively) were improved in comparison with pristine sample (117.2 mAh·g⁻¹). In addition, 0.2 wt.% of magnesium led to a lower value of initial discharge capacity (~110 mAh·g⁻¹). LiMg_0.1Mn_1.9O_3.8F_0.2 showed good stability and cycle life (90.6% of initial capacity after 400 cycles) and performed well at high c-rate (20 C), delivering 76.1 mAh·g⁻¹, with 40% better than the pristine LMO (45.9 mAh·g⁻¹). Magnesium and fluorine co-doped spinels demonstrate enhanced lithium ions diffusion kinetics and the charge-transfer resistance was drastically minimized by dopants.

3. Improving LiMn₂O₄ by Coating

Surface modification of LiMn₂O₄ particles proved to be an efficient way to diminish the harmful activity of electrolyte solution during cell cycling. As mentioned before, one important drawback of LMO spinel is related to manganese dissolution. The coatings not only improve the material stability by reducing its contact area with the electrolyte solution, but also enhance the electrochemical characteristics by reducing the charge transfer resistance and providing a better samples-cycling stability, for instance [162,163]. Throughout the years, many different coatings, which include organic and inorganic compounds like carbon, oxides, perovskites, solid solutions, have been studied [164,165,166,167].

3.1. Carbon

Lithium ions diffusion and rate capability life of lithium manganese oxide can be enhanced by using high conductive carbon-based materials [168,169]. Different carbonaceous coatings like hard carbon [12], activated carbon [170,171], carbon nanotubes [172,173,174,175,176,177], graphene [178,179], amorphous carbon [177,178,179,180,181,182], carbon nanospheres [183] were studied in the last years.

Reduced graphene oxide (rGO) nanosheets were employed as support for LMO uniform deposition by a low temperature microwave assisted hydrothermal reactions method [184]. The rGO presence prevented LMO agglomeration and low range particle distribution (10–40 nm). High crystalline LMO samples with stable 3D framework behaved remarkably at low and high c-rates, as well. Delivering an initial discharge capacity of 137 mAh·g⁻¹ at 1 C, LMO/rGO hybrid material exhibited 85.4% (117 mAh·g⁻¹) and 73.7 (101 mAh·g⁻¹) of its initial discharge capacity, at 50 and 100 C, respectively. The improvement in material’s performance was strongly related to the properties of rGO: high surface area, protecting barrier against the electrolyte solution and the channels presence for permitting faster lithium ions diffusion [14].

LiMn₂O₄/graphene composite electrodes were proposed as suitable cathode materials for Li-ion batteries [168,169,179,185,186,187,188,189,190]. The unique and amazing graphene properties, such as high electrical conductivity, thermal and mechanical stability and high surface area, provided an efficient method for improving lithium manganese oxide’s electrochemical properties [14,167,168,189,191]. Brief characteristics about the electrochemical behavior of lithium manganese oxide/graphene composite materials are presented in

Table 8. Graphene-coated LMO and its electrochemical properties.

No.	Material	Synthesis	Discharge Capacity (mAh g−1)	Coulombic Efficiency (%)	Discharge Capacity/N Cycles (mAh g−1 N−1)	Capacity Retention (%)	Ref
1	LiMn2O4/graphene	Hydrothermal method	143.6 (x) *	-	137.4/50	95.7	[179]
			141.5 (x) **		133.4/50	94.3
2		Sputtering + CVD	131.6 ***	-	119/750 ***	90.4	[169]
3		Ball-milling method	78.8	93.7	64.6/500	82	[186]
4		Solution method	131.1 (b)	-	125.3/50	96	[178]
5		Electrophoretic deposition	49.6 (r)	-	-	-	[187]
6		Spray drying combustion process	139 (b)	92.5	121	87	[193]
7		Hydrothermal method	139.2 (i)	98.2	117.5/200	84.5	[194]
8		Self-assembly process+Solid-state lithiation	155 (d)	94.9	149.8/80	96.6	[168]

* t = 30 °C, ** t = 45 °C, *** the considered unit of measurement is mAh·cm⁻³.

, while a more advanced description about graphene influence on LIB’s cathode materials has been described elsewhere [192]. The challenges related to graphene-based materials usage in energy storage devices have been discussed in [14].

LiMn₂O₄ oxide microspheres were synthesized by a cyclohexanone hydrothermal reaction method, followed by a coating process with amorphous carbon, resulted after calcination of a LMO and glucose mixture [181]. The thin 1.5 nm coating layer enhanced the initial discharge capacity of the surface modified sample, from 127.9 to 138.5 mAh·g⁻¹. In addition, amorphous carbon reduced the contact area between active electrode material and electrolyte, thus, a superior capacity retention (97.6%) was calculated for LMO/C sample after 100 cycles, 0.1 C.

Zhu et al. [191] published an ample review about novel carbon-based coatings (graphene, carbon nanotubes) for enhancing the behavior of LiMn₂O₄ in Rechargeable Hybrid Aqueous Batteries (ReHAB).

3.2. Oxides

Many researchers drove their attention to the usage of metal oxides as surface coatings for lithium manganese oxide [195].

Several research groups focused their studies on the influence of different lanthanide oxides on LMO particles. Since 2007, several studies were concentrated on CeO₂ behavior as coating for lithium manganese oxides [164,165,196,197].

In order to modify the pristine samples, disparate approaches were made. Arumugam et al. [197] employed a polymeric process followed by a high temperature heat treatment (850 °C for 6 h), in order to obtain the modified samples. As expected, the initial discharge capacity of the pristine sample (135 mAh·g⁻¹) was slightly higher than the one of 1 wt.% CeO₂ (126 mAh·g⁻¹), but the latter sample exhibited improved stability during cell’s cycling in both room and elevated temperature (60 °C) tests, providing 120 and 117 mAh·g⁻¹, respectively, after 100 cycles at 0.5 C. While the pristine samples failed to offer more than 60% capacity retention at 30 °C and 50% at 60 °C, the coated material provided astonishing capacity retention of at least 93%, in both cases. Not only the cycle performance is important, but also the material ability to work at higher current rates is important. The 1 wt.% CeO₂-LiMn₂O₄ was cycled for 100 times at higher current rate (5, 10, 15, 20 C, respectively) and its performance was compelling in all the case, because the initial discharge capacities were relatively high (102 mAh g⁻¹ at 20 C) and small capacity fades were recorded (between 12%–17%). For a better understanding of coating layer importance, the electrochemical impedance spectra of LMO-based electrodes were studied. It was noticed that the charge transfer resistance was greatly decreased by coating, while the side reactions between the electrode material and electrolyte solution were minimized. Despite the fact that the initial R_ct of the bare sample was slightly smaller than the one of coated material (21 Ω vs. 26 Ω), after 100 cycles it increased for more than four times, up to 98 Ω, while a R_ct of only 42Ω was recorded for the coated LMO.

The precipitant influences, namely (NH₄)₂CO₃ and NH₄OH, in a precipitation synthesis of CeO₂ modified LiMn₂O₄ particles were studied by Cho et al. [196]. Small particles size and a uniform coating layer were obtained when NH₄OH was involved in the synthesis process, while the usage of (NH₄)₂CO₃ favored particle agglomeration, when the coating amount increased. The precipitant used during the coated materials synthesis enhanced its capacity retention, from 50% for the pristine up to 88% for 5 wt.% CeO₂ upon cycling for 200 times, at 60 °C and 1 C. Even though the lowest capacity fade was recorded for the 5 wt.% CeO₂ (about 12%), its discharge capacity is still too low to be used in practical batteries (~70 mAh·g⁻¹. Due to uniform coating layer formed when NH₄OH was used, the electrode material was less exposed to electrolyte solution and thus the sample electrochemical characteristics were improved, with decreased recorded capacity fade (12–24%). In contrast, higher capacity fades were observed for the samples with (NH₄)₂CO₃ as precipitant (32–40% of initial discharge capacity).

Michalska et al. [164] employed another investigation on CeO₂ coating layer. The modified samples were obtained through a sol-gel synthesis followed by a low-temperature calcination (350 °C). The resultant particles presented relatively clear surface, with wide size interval (between 100 and 600 nm), which agglomerated into larger clusters. Compared to the previous studies, the initial discharge capacity of the modified sample was higher (115 mAh·g⁻¹) than the pure LiMn₂O₄ (mAh·g⁻¹), thus the lithium intercalation-deintercalation processes were not negatively influenced. In addition, the capacity loss of LMO/CeO₂ sample was just 2% after 100 cycles (1 C), in comparison with 10% for bare LMO. Another benefit of the cerium oxide coating can be correlated with material performance at high current rates (

Table 9. Comparative data of pristine and CeO₂ modified LMO at high current rates.

Material	Initial Discharge Capacity (mAh g−1)					Ref.
	5 C	20 C	15 C	20 C	30 C
Pristine LiMn2O4	88	79	-	-	23	[164]
CeO2 1 wt.%-LiMn2O4	108	100	-	-	35
	125	121	117	102	-	[197]

).

Among the lanthanide oxides, La₂O₃ was another oxide studied by Amurugam et al. [198,199]. Uniform grains with average diameters of 200–300 nm and clear surface were obtained. During the electrochemical tests at 30 and 60 °C, respectively, the coated LMO samples exhibited higher capacity retentions than the pristine sample and among all the samples the performance of LMO/2 wt.% La₂O₃ had been highlighted. 2 wt.% La₂O₃ coated LMO not only showed good discharge capacity, but also possessed a high coulombic efficiency at high c-rates (2 and 5 C), which increased with the number of cycles. In the case of discharge rate of 2 C, a coulombic efficiency of 100% was obtained after 80 cycles and it remained unchanged even after 100 cycles. In the other situation, at 5 C, the highest coulombic efficiency (95%) was noted after 50 cycles and it did not modify until the experiment’s end.

Lithium manganese oxide modified with 5% La₂O₃ was synthesized by a classical solid-state reaction method by Feng et al. [200]. The SEM analysis of the sample showed rough surface for the coated spinel, which tended to form agglomerations. Similar discharge capacity with Ref. [199] (116.4 mAh·g⁻¹) was recorded when the cell was cycled at a temperature of 25 °C and a c-rate of 1 C. In contrast with the pure LiMn₂O₄, there was no sudden capacity loss, and the material was able to provide 90.1% of its initial capacity after 200 cycles. Since LiMn₂O₄ suffers from elevate temperature instability, 5 wt.% La₂O₃ demonstrated that it could be an efficient spinel coating when the cell was tested at 55 °C. Like in similar studies, the capacity retention of the current modified sample was at least 82% after 200 cycles.

Lanthanum oxide synthesized by a combustion method, was coated on LMO particles in neopentyl glycol [201]. In contrast with the previous reported article related to La₂O₃ modified lithium manganese oxide, 3 wt.% La₂O₃ improved the cycling ability of LMO and a capacity loss of 8.3% was seen after 100 cycles at elevated temperature (60 °C).

Due to its interesting properties like thermal and structural stability and fast kinetics, titanium dioxide was employed as coating for lithium manganese oxide [202,203,204,205]. In one of the first studies about the TiO₂ effect upon the improvement of electrochemical properties of LiMn₂O₄, it was noticed that it did not modify the spinel structure, but a spinel LiTi_xMn_2−xO₄ formed on the pristine particles [205]. In contrast, when TiO₂-B particles with nano-belts morphology were coated, no intermediary Li-Ti-Mn-O layer was obtained. The TiO₂ (B) porous coating hindered manganese dissolution and enhanced the lithium ions transport, throughout their network, fact proven by smaller charge-transfer and solid electrolyte interface (SEI) resistances, in comparison with bare sample [202].

Different mass weight percentages for possible coatings were studied, in order to determine de optimal amount of TiO₂ coating for LiMn₂O₄. Zhang et al. [203] synthesized modified spinel with 0.5, 1, 2 and 3 wt.% TiO₂, respectively. The charge-discharge curves and cycling performance emphasized that the optimal amount was 2 wt.% TiO₂, because of its highest initial discharge capacity (129 mAh·g⁻¹) and low-capacity loss after 50 cycles (5.5%). Using the same percentage of TiO₂, Shang et al. [202] studied long time cycling of LMO/TiO₂ materials at elevated temperature (55 °C), in order to determine the improvement done by the oxide coating. After 300 cycles, the modified sample furnished a capacity of 77.4 mAh·g⁻¹ (74% of initial discharge capacity), while the LMO offered only 63% of the initial capacity (66.1 mAh·g⁻¹). Walz et al. [206] used ZrO₂ and TiO₂ as coatings for bare LiMn₂O₄ and proposed a possible mechanism ZrO₂ and TiO₂ protection of LMO during cycling at elevated temperature (55 °C). The HF, which has such a harmful effect on lithium manganese oxide, can appear as a reaction’s product between LiPF₆ and water traces during cell’s assembly [207]. TiO₂ may react with HF and form a compound with the formula of TiO₂·2HF and it prevents the manganese dissolution, improving spinel’s stability over repeated cycling. Having an average diameter of only 4.8 nm and pores size around 0.7 nm, TiO₂ coating reduced to half the internal resistance of the cell (81 Ω), which remained lower than for the pristine LiMn₂O₄ cell (168 Ω). The discharge capacity of the modified LMO decreased with about 10% (from 120 to 103 mAh·g⁻¹), while a serious severe capacity fading of ~75% was recorded for pristine material.

For the TiO₂ (B)-nano belts LiMn₂O₄, two synthesis approaches were made. One of them [203] consisted of facile phase liquid mixing, while the other one was a two-steps synthesis, which employed a preliminary hydrothermal method stage, followed by electrostatic attraction [202]. In the case of the latter suggested synthesis method, the heat treatment for the obtaining of modified sample was also studied and it was clearly noticed that a heat treatment of 2 h would be enough to synthesis a porous coating layer on the pristine LMO. When a longer heat period (4 h) was tested, it was observed that a dense compact layer formed and it did not improve the electrochemical performance of the sample.

LMO/V₂O₅ particles were easily obtained by a solid-state method, followed by a calcination at 500 °C, for 2 h in air atmosphere [208]. The XRD patterns of the synthesized samples showed that the vanadium oxide did not modify LMO spinel structure and no remarkable differences between pristine and modified samples XRD peaks were noticed. The main reason of applying vanadium oxide coating on the pristine sample was, like in other case, the improvement of LMO spinel’s stability. The coating layer was not dense, so the modified sample solid electrolyte interphase resistance (R_SEI) had a slightly higher value (8.52 Ω) than the pristine LMO (8.35 Ω) and the LMO charge-transfer resistance was greatly decreased from 69.72 to 26.51 Ω, demonstrating the development brought by V₂O₅ coating. During cycling at elevated temperature (>55 °C) and since the electrode material is in contact with the electrolyte on wide area, the possible presence of HF in the electrolyte tends to encourage the disproportionate reaction of manganese ions (Mn³⁺). Inductively coupled plasma atomic emission spectroscopy (ICP-AES) studies showed that the manganese concentration in electrolyte solution, in certain conditions (30 days’ samples immersion at 55 °C), was five times smaller (88 ppm) for the coated LMO, than for bare LMO, proving the coating efficiency again manganese dissolution. The best results were for 2.5 wt.% V₂O₅ when discharge capacities of 107.3 and 92 mAh·g⁻¹, respectively, were provided after the materials were cycled for 200 times at room and elevated temperatures. In spite the fact that it was not such a large difference between LMO and 2.5 wt.% V₂O₅/LMO (~1%), the low discharge capacity exhibited by the pristine sample synthesized makes it unsuitable for practical devices.

In one of the latest studies on oxide coating for LiMn₂O₄, Tao et al. [209] coated MoO₃ through a wet chemical method. Throughout this synthesis process, a uniform coating layer with a thickness of about three nm was deposited. The cyclic voltammetry tests revealed the two peak pairs, specific for LiMn₂O₄ spinel, but in the case of LMO/MoO₃, the peaks were slightly shifted to the right, with about 0.05 V. Both pairs of peaks of the modified sample were sharper than the ones of bare LMO. The main characteristic of the modified samples was the ability to provide almost 70% of its initial discharge capacity (101.6 mAh·g⁻¹) after long time cycling (900 times) at a current density of 2 C. All the results mentioned above were recorded for the sample modified with three-weight percentage of molybdenum oxide. A higher oxide amount did not lead to improved electrochemical results, except cycling performance. From EIS spectra, it was confirmed that the lowest slope of LMO/5 wt.% MoO₃ provided the lowest lithium ions diffusion, lower even than pristine sample, owing the highest value of charge-transfer resistance (265.1 Ω).

Lithium manganese oxide thin films coated by manganese oxide exhibited an outstanding life cycle and only 20% capacity fade was calculated after 3000 cycles [210]. The bixbyite-type Mn₂O₃ coating layer was deposited on the LMO/Pt substrate by a spin-coating technique, followed by an annealing treatment at 750 °C for 6 h. The relatively large thickness of Mn₂O₃ porous layer (~600 nm) did not block the lithium ions during tests, and a slightly improved discharge capacity (~3 mAh·g⁻¹) was recorded for coated sample in comparison pristine LMO. In order to demonstrate the efficiency of Mn₂O₃ layer against the electrolyte attack over the LMO modified electrode, pristine and coated LMO were immersed in 1 M LiPF₆ electrolyte solution for seven days. The Mn₂O₃/LiMn₂O₄ electrode barely showed dissolution proofs, while the pristine LMO completely dissolved. While the Mn³⁺ ions inside LiMn₂O₄ spinel structure are exposed for reduction to Mn²⁺, fact that hinders material’s performance, it was determined that Mn³⁺ ions from bixbyite structure are more stable.

Recently, Yao et al. [211] suggested that tert-butanol addition during the synthesis of Al₂O₃ coated LMO enhanced the stability and electrochemical properties of the sample. Due to the fact that Al₂O₃ is inactive in the range between 3.0–4 V, the LMO-Al₂O₃ composite delivered a lower specific discharge capacity in the first cycle (105.46 mAh·g⁻¹, c = 0.2 C), in comparison with the bare lithium manganese spinel (114.69 mAh·g⁻¹), but the elevated temperature (55 °C) test proved a better capacity retention (90.78%). This cyclability improvement is related to the protective coating layer, which is strongly attached to the LMO particles. Al³⁺ has the role of an anchor ion between polymeric functional groups formed by tert-butanol’s hydroxyl groups and the hydrogen bonds from the aluminum salt precursor (aluminum sulfate octadecahydrate). In addition, the ageing tests showed less manganese content in LMO-Al₂O₃ sample (62.3 ppm) compared to bare LMO (218.5 ppm). For the ageing tests, the samples were immersed in electrolyte for 20 days at a temperature of 55 °C.

Al₂O₃-coated LiMn₂O₄ have been synthesized by a co-precipitation method [212]. In contrast with the previous reported LMO-Al₂O₃ composite, the current modified sample showed a higher capacity in the first cycle (133.5 mAh g⁻¹, c = 0.5 C) and the recorded capacity faded with only 0.014% per cycle. Al₂O₃ uniform nano-coating can be considered efficient to prevent the apparition of Jahn-Teller structural distortion and also to hinder the electrolyte attack over the active LMO particles, because of the lower specific area (2.3 m² g⁻¹), four times lower than for pristine LMO (13.6 m²·g⁻¹). In addition, Al₂O₃ strengthened the high discharge rate performance, at 5 C (55 °C), thus, the 100th cycle capacity of LMO-Al₂O₃ remained ~92% of the initial one (~108 mAh·g⁻¹), much better than the bare spinel (73.5% capacity retention and ~90 mAh·g⁻¹ in the first cycle).

In another research article [213], Al₂O₃ was coated on pre-sintered and calcined LiMn₂O₄, through a wet process, followed by heat treatment at 750 and 350 °C for 10 h, respectively. During the electrochemical test, it was noticed an improved cycle life for Al₂O₃ pre-coated LMO compared to other samples. The Al₂O₃ layer acted like an inhibitor layer and prevented large growth of LiMn₂O₄ particles, thus the smaller sized material owned better kinetics during tests. Moreover, during ageing tests, 0.05 mg of manganese from the Al₂O₃ pre-coated LMO dissolved into in electrolyte, which was four times lower than for bare spinel and with 25% less than Al₂O₃-coated LMO.

Guo et al. [214] have reported lithium manganese oxide coated by a wet process implying electrostatic attraction forces method. By adjusting the pH of a mixture of water and ethanol to 8, a Zeta potential difference of 44 mV has been recorded. After that, LMO spinel powder was added, while nano-silica sol was poured drop wise. The resulted sol was heated at 90 °C for 10 h and then calcined at 700 °C for 6 h. Different weight percentages of SiO₂ were tested, but the best results were measured for 2 wt.% SiO₂-LMO. The XRD patterns of coated samples were slightly shifted to the left, which can be attributed to a possible replacement of Mn⁴⁺ from 16d sites by Si⁴⁺ ions. During the cell’s cycling at room temperature (25 °C), 2 wt.% SiO₂-LMO retained 97.6% of the initial capacity (101.2 mAh·g⁻¹) after 100 cycles. Even though a low-capacity loss of only 2.4% was calculated, the achieved discharge capacity was too low for a possible practical application (only 68.73% of LMO’s theoretical capacity), whereas higher discharge capacities were recorded for LMO coated with CeO₂ [197], Li₂ZrO₃ [215] or AlP [30]. SiO₂ had benefic effect over the LiMn₂O₄ performance at elevated temperature (55 °C), which is considered a critical parameter for spinel. Moreover, silicon oxide coating reduced the amount of dissolved manganese into the electrolyte by 36% and increased the lithium diffusion coefficient by one magnitude order, which determined faster lithium ions migration throughout the electrochemical tests. Another improvement brought by the SiO₂ coating was related to reducing the R_ct increasing rate after 100 cycles, thus, in the case of surface modified sample, R_ct increased by ~8 times (from 125.9 to 970.1 Ω), while bare sample’s R_ct value dramatically raised by almost 20 times, reaching an enormous value of 11,800 Ω.

3.3. Other Surface Changes

Coatings that are more complex were studied in the last years (

Table 10. Other surface modification of LMO.

Nr. Crt.	Material	Synthesis Method	Discharge Capacity (mAh g−1)	Coulombic Efficiency (%)	Discharge Capacity/N Cycles * (mAh·g−1)	Capacity Retetion (%)	DLi × 1010 (cm2·s−1)	Rct (Ω)	Ref
1	3 wt.% FePO4-LiMn2O4	Chemical deposition method	122.0 (d)	-	83/80	68	-	-	[216]
			118.2 (d)		78/80	66		-
2	1 wt.% FePO4-LiMn2O4		124.0 (d)	98%	118.7/100	93.5	-	-	[217]
			127.0 (g)	-	68/100	53.5		-
3	3% LaF3-LiMn2O4		118.7 (i)	99.9	107/100	90.1	-	-	[163]
			117.1 (i)	99.5	98.6/100	84.2		-
4	3% AlP-LiMn2O4		124.2 (i)	-	116.7/100	94		66/100 cycles *	[30]
				-	108.1/100	87
5	2% AlF3-LiMn2O4		103.4 (i)	90	92.9/50	89.8	-	-	[221]
6	1% Li3PO4-LiMn2O4	Ball milling method450 °C (5 h)	114.2(i)		97.1/100	85	-	293	[222]
7	10% LiNi0.5Mn1.5O4- LiMn2O4	Sol-gel	112.0 (q)	-	107.8/50	96.3	-	-	[1]
			109.5 (q)	-	98.6/100	90
8	0.3% Li3BO3-LiMn2O4		118.2 (d)	90	108.7/40	92		14.3/20	[223]
			119.3 (d)	-	104.5/40	87.8		22/20
9	5% FeF3-LiMn2O4		110.7 (d)	-	75.5/200	68.2	-	-	[224]
			108.1 (d)	-	66.5/100	61.5
10	3% YPO4-LiMn2O4		107.0 (d)	-	90.0/200	84.1	-	151.7	[218]
			107.2 (d)	-	86.1/100	80.3
11	3% LaF3-LiMn2O4		115.9 (d)		84.3/200	72.7	-		[225]
			115.3 (d)		81.1/100	70.3
12	LiCoO2-LiMn2O4		117.7 (j)	-	111.1/100	94.4	-	-	[219]
			117.2 (j)	97.6	110.1/100	93.6
13	0.5% Li2CuO2-Li2NiO2 solid solution-LiMn2O4		121.2 (i)	-	113/300	93.2			[166]
			12 1 **	96	98.1/200	81.1	-	49.23
14	3% Li2ZrO3-LiMn2O4		126.7 (i)	97.5	125.5/100	99.0	-	34.2	[215]
			129.5 (i)	99.8	116.8/100	90.2
15	La-Sr-Mn-O-LMO		129.9 (b)120.0 (i)	90.7	108.1/500/1 C	90.6	2.54 × 102	42.1	[130]
			107.3 (i)	-	97.4/130	90.8
16	2% Li1.2Ni0.2Mn0.6O2-LiMn2O4		116.6 (i)	-	111.8/200	95.88	-	192.6/100 cycles	[26]
				-	/200	89.74
17	3% LaMnO3-LiMn2O4		114.0 (g)106.0 (l)	95.0	97/500	91.5	5.12 × 10−1	-	[37]
18	3% Li2MnO3-LiMn2O4		113.3 (i)	-	106.7/500	94.17	-	75.2/50 cycles	[226]
			115.0 (i)	-	103.2/200	89.75
19	LiNi0.5Mn1.5O4-LiMn2O4		101.0 (i)	94.4	-	-	-	-	[227]
20	3% La0.7Sr0.3Mn0.7Co0.3O3-LiMn2O4 *		123 (d)108.0 (x)	95.8	101/100 x)	93.5	2.15	4.05	[220]
21	LiMn1.912Ni0.072Co0.016O4	Co-precipitation	118.0 (i)	97	113.4/200	96.1	8 × 10	45.31	[228]
22	3% LaF3-LMO		118.4 (i)	97.5	118.04/50	99.7	4.67 × 102	-	[131]
23	3% FePO4-LiMn2O4		105.3 (g)	-	93.3/100	88.6	9.85	171/20	[229]
24	3% LiMnPO4-LiMn2O4	Hydrothermal reaction	112.4 (g)	94.5	104/300	92.5	5.32 × 10−2	400.6	[230]
			115.7 (g)	94.2	111.7/100	96.6
			107.0 (g)	83	75.1/500	70.2
25	2 wt.% LaPO4-LiMn2O4	Polymeric process	123.9 (g)103.0 (g)	95.4	-	-	-	-	[231]
				94.5	82/100	81.5

* t = 60 °C ** 1 C = 150 mAh g⁻¹.

). Metal phosphates [216,217,218] and fluorides [163] were among the most investigated coatings. Among other possible coatings, layered material LiCoO₂ [219], spinel type material (LiNi_0.5Mn_1.5O₄) [1] or pervoskite-type oxide [220] were also tested.

Yang et al. [217] studied iron phosphate as a coating candidate for enhancing lithium manganese oxide’s performances. It was observed that, during the tests at room temperature, a low amount of FePO₄ (1%) was able to increase the capacity retention of the pristine sample by almost 20%, from 75% to 93.5%. Unfortunately, at elevated temperature (60 °C), 1 wt.% FePO₄ was not sufficient to diminish the structural modifications of the material, but by adding a double amount of iron phosphate, the modified sample was more stable and lesser capacity loss was calculated. The quantity of 1 wt.% FePO₄ reduced the charge-transfer resistance, there, lithium ions movement was greatly increased.

Amorphous iron phosphate proved to be a promising coating material for LiMn₂O₄ due to its 3D channels that enhance the lithium ions diffusion during cycles and by stabilizing the LMO structure by a small amount of iron ions incorporation in the spinel’s lattice [229]. The heat treatment during the coating process was the key parameter for obtaining the right FePO₄ porous structure. The samples were calcined at different temperatures, in the range of 300–700 °C. The lowest calcination temperatures favored the formation of porous structure, while when the temperature increased, the FePO₄ become crystalline and less defects were noticed. In terms of electrochemical performances, the highest lithium diffusion coefficient (9.85·10⁻¹⁰ cm²·s⁻¹) was calculated for the 3 wt.% FePO₄-LMO sintered at 400 °C and D_Li decreased as the calcination temperature increased.

Han et al. [1] studied the composite material resulted from coating spinel LiNi_0.5Mn_1.5O₄ (LNMO) on LiMn₂O₄ using a sol-gel process. Although, a coating with a similar structure as the substrate material is not very common, LNMO influenced the capacity retention of the pristine sample, especially at high temperature. While the bare LMO exhibit 82.8 mAh·g⁻¹ (66.7% of its initial discharge capacity) after 40 cycles, the LNMO/LMO composite had 98.6 mAh·g⁻¹ (~90% of the first cycle capacity). Changes in the structure of LMO cycled at 60 °C were noticed and peaks corresponding to tetragonal Li₂M₂O₄ were identified after analyzing the sample by XRD. No similar peaks were observed for the modified sample, fact that proved the efficiency of 10% LNMO to suppress the Jahn-Teller distortion. The LiNi_0.5Mn_1.5O₄ structure, which greatly favors the Li⁺ diffusion, was an important factor in reducing the charge-transfer resistance of the composite.

LNMO/LMO composite material has been synthesized trough a citric acid sol-gel route [227]. The core-shell material presented an initial lower discharge capacity (101 mAh·g⁻¹) and only after five cycles, the structure has been stabilized along with the formation of LNMO’s SEI layer. Charge/discharge tests at different rates (from 1 to 65 C), demonstrated the high-rate ability of the modified sample, due to its discharge capacity of approximately 25 mAh·g⁻¹ at 65 C. In comparison, the pristine sample exhibited a similar capacity, but for a lower discharge rate (45 C).

LiCoO₂ established as a good cathode material since the appearance of the Li-ion batteries. In the last years, it was suggested as a potential coating material for spinel type LMO, in order to improve the high-temperature behavior of the latter [219]. Commercial LMO sample was modified with LiCoO₂ throughout a sol-gel process, employing a two-stage heat treatment, at 350 and 650 °C, respectively. The initial discharge capacities of LiCoO₂-LMO sample were higher than the ones for unmodified sample, both at room and elevated temperature. The reason for this aspect is related to the electrochemical activity of the lithium cobalt oxide in the tests’ voltage area (3.4–4.2 V). LiCoO₂ improved the capacity retention of LMO by 6%, from 87.5% to 93.6%, after 100 cycles at 55 °C. In addition, the LiCoO₂ coating layer acted as a barrier and halved the manganese dissolution into the electrolyte; the dissolved quality of manganese ions (determined by ICP-AES analysis) from LiCoO₂-LiMn₂O₄ electrode was 10.17μg/cm², in comparison to 22.54 μg/cm² for pristine LiMn₂O₄ electrode.

Among the newly studied core-shell material, LMO coated with LiMnPO₄ (LMP) indicated suitable performance for log lasting heavy-duty devices [230]. During the test at room temperature, the LMO coated with 3% LMP showed good capacity retention of ~96% after 100 cycles. 3 wt.% LMP was not enough to suppress the structural modifications, which occurred at elevated temperature (55 °C), thus high-capacity fade (30%) was recorded after 500 cycles. The LMP layer notably improved LiMn₂O₄′s cyclic performance, especially at high discharge rates of 10 and 20 C, respectively. After long time cycling (1000 cycles) at 20 C, the discharge capacity for 3 wt.% LMP (76.3 mAh·g⁻¹) was 1.6 times higher than for the pristine sample (45.8 mAh·g⁻¹). As a drawback, due to low electronic conductivity of LiMnPO₄ material, the modified samples presented a lower D_Li (5.32 × 10⁻¹² cm²·s⁻¹) than in other reported studies [228,229]. LMP layer did not suppress the disproportionate reaction of Mn³⁺, but it successfully blocked the dissolution of Mn²⁺ into the electrolyte.

Feng et al. [30] developed an aluminum phosphide (AlP) coating. An interesting aspect related to aluminum phosphide is its ability to react with the harmful HF, which promote Mn³⁺ disproportionate reaction, and the resulted product, AlF₃, was also successfully tested as coating for LiMn₂O₄ [95,221].

AlF₃ was deposited on already modified LMO by an Al³⁺ and F⁻ co-doping sample via a co-precipitation method [95]. The aluminum fluoride coating protected the LiAl_0.04Mn_1.96O_3.96F_0.04 particles from electrolyte action, resulting in a more stable structure during charge-discharge cycles. Moreover, it reduced the charge-transfer resistance growth rate calculated, which was calculated to be from 67 to 197 Ω for pristine LiAl_0.04Mn_1.96O_3.96F_0.04, to only 6 Ω (from 87 to 93 Ω) after the same number of cycles (100 cycles). However, a minor improvement related to D_Li+ was noticed for AlF₃-coated LMO (1.4·10⁻⁸ cm² s⁻¹) in comparison with LiAl_0.04Mn_1.96O_3.96F_0.04 (1.2·10⁻⁸ cm² s⁻¹), but the lithium diffusion coefficients of the latter samples was ten times higher than for pristine LiMn₂O₄ (1.1·10⁻⁹ cm² s⁻¹).

Modified LMO samples were tested with good results in ARLB cells. For example, 2 wt.% LaF₃-LiMn₂O₄ exhibited an insignificant capacity fade after 50 cycles [131]. Due to LaF₃′s structure, its large channels enhance the lithium migration to and from the electrode material, while an optimum amount of lanthanum fluoride coating may prevent the side reaction that can occur between LMO and electrolyte. The calculated lithium diffusion coefficient (4.67·10⁻⁸ cm²·s⁻¹) was two magnitude higher than the one obtained for iron phosphate modified LiMn₂O₄ in organic electrolyte [229] and similar to D_Li calculated for La-Sr-Mn-O coating [130]. The protective LaF₃ layer acted like a barrier and protected the electrode surface, by reducing the contact area between LMO and aqueous electrolyte; therefore, the manganese dissolution into LiNO₃ was barely noticed.

Solid electrolytes, like Li₃BO₃ [223] or La-Sr-Mn-O (LSM) [130], with good ionic conductivity demonstrated to be adequate coating materials for enhancing the cycle ability of LiMn₂O₄. The higher ionic conductivity of LSM thin uniform layer played an essential role in the higher initial discharge capacity (129.9 mAh·g⁻¹, 0.1 C) and promoted a capacity retention (90.6%) with up to 50% higher than the pristine sample. LSM coating decreased the charge-transfer resistance, enhanced the lithium-ion diffusion and the calculated diffusion coefficient for lithium ions had the same magnitude order as for LaF₃-coated LiMn₂O₄ tested in an ARLB [132]. Another improvement brought by the LSM was related to electrical conductivity (8.42·10⁻⁴ S/cm), which was larger than for pristine sample (5.31·10⁻⁵ S/cm).

An isostructured Li₂CuO₂-Li₂NiO₂ solid solution has been considered as a coating for lithium manganese oxide [166]. During the electrochemical investigations at room temperature, it was demonstrated that a low amount of solid solution (0.5 wt.%) was enough to increase the rate capability (93.2% capacity retention after 300 cycles) and to maintain a reasonable discharge capacity (113 mAh·g⁻¹ in the first cycle). However, taking in consideration that the elevated temperature performance is a key demand, it was noticed that 0.5 wt.% was not enough to stabilize the LiMn₂O₄ at 55 °C, thus the capacity retention was only 81.2% after 200 cycles. In comparison, 2% w/w Li₂CuO₂-Li₂NiO₂ solid solution exhibited an improved electrochemical performance. By having higher amount of coating, the contact area between the electrolyte and LiMn₂O₄ was reduced, so the manganese dissolution was hindered.

Multiple doped LiCo_0.025Cr_0.025Ni_0.025Fe_0.025Mn_1.90O₄ and pristine LiMn₂O₄ were coated by lithium borosilicate (LBS) through a facile solution method [143,232]. LBS-based glass was proposed as a coating due to its improved ionic conductivity. Surface modified samples exhibited lower initial discharge capacity (109.7 mAh·g⁻¹) than pristine LMO, because the coating layer slightly decreased the lithium diffusion. However, LBS reduced the contact surface between the electrode active material (LMO) with the electrolyte (1 M LiPF₆ solution), and superior capacity retention (93.3%) was observed after 70 cycles (25 °C). At elevated temperature (55 °C), LBS-coated LiMn₂O₄ performed better than bare LMO, showing a higher initial discharge capacity (120 mAh·g⁻¹ vs. 104.2 mAh·g⁻¹). In addition, the capacity fade per cycle was slightly lower for coated LMO spinel (0.708 mAh·g⁻¹ cycle⁻¹ vs. 0.728 mAg·g⁻¹ cycle⁻¹).

Perovskite-type LaMnO₃ was successfully deposited on lithium manganese oxide, without modifying the spinel structure [37]. The improved electronic conductivity of LaMnO₃ determined a higher initial discharge capacity of the modified sample (114 mAh g⁻¹) and better coulombic efficiency (95%) than for pristine lithium manganese oxide (106 mAh g⁻¹ and 89.1%).

Titanium carbide (Ti₃C₂T_x) nanosheets, a MXene 2D layered structured material, has been recently used as a protective coating layer for LMO particles by their encapsulation through facile electrostatic self-assembly method. Similar synthesizes methods can be using for other cathode materials like LiCoO₂ or LiFePO₄ [233]. MXene layered material, with the general formula of M_n+1AX_n (n = 1, 2, 3), where M is a transition metal (Sc, Ti, Zr, Nb, Mo), A corresponds to elements from the 13th or 14th groups (Al, Si, for example) and X represents either C or N atoms. Namely, in the case of titanium carbide, T_x is attributed to functional group –OH and atoms like –O and –F [233,234]. Ti₃C₂T_x is obtained by selective removal of Al from Ti₃AlC₂ in dilute HF solution media. Details about the synthesis procedure of titanium carbide along with other aspects related to MXene materials can be found in Ref [235]. Three samples were tested: pristine LMO, Ti₃C₂T_x-LMO (obtained using CTAB as decorator for charging the surface of LMO) and Ti₃C₂T_x-LiMn₂O₄ (synthesized in similar conditions but replacing CTAB with deionized H₂O). During the electrochemical tests, at both room temperature and 55 °C, Ti₃C₂T_x–coated LMO (with CTAB) performed better than the other two tested samples and delivered higher specific capacity of 114.1 mAh·g⁻¹ (c = 1 C, RT) and 115.9 mAh g⁻¹ (c = 1 C, 55 °C), respectively. In addition, the capacity fade was not as broad as in other case, and at 55 °C, Ti₃C₂T_x–coated LMO (with CTAB) showed a discharge capacity with more than 18% than the pristine sample (63%), after 200 cycles. Titanium carbide layered hampered the manganese dissolution into the electrolyte by reducing the contact area between the working electrode and electrolyte. By coating the pristine LMO with Ti₃C₂T_x, which possesses low Li⁺ diffusion barrier (0.007 eV), good electronic conductivity and high Li⁺ diffusion coefficient, determined lower charge-transfer resistance (133.4 Ω) than 370.7 Ω for Ti₃C₂T_x-LMO (deionized water) and 552.6 Ω for bare spinel (at room temperature).

Li₄Ti₅O₁₂(LTO) is a characteristic material which can be used as anode in Li-ion batteries [236,237], but, lately, it was successfully employed as a coating layer for enhancing the performance of LiMn₂O₄ [237]. LTO–LMO nanorods were manufactured by a three-step synthesis method: in the first stage, β-MnO₂ were obtained by a hydrothermal process, followed a solid stage reactions method, when as-synthesized β-MnO₂ were mixed with lithium hydroxide and calcined to result lithium manganese oxide. Afterwards, the composite material resulted through a sol-gel process, where a key parameter was to control the hydrolysis and condensation of titanium precursor (titanium tetrabutoxide). LTO-coated LMO was electrochemically tested for a long number of cycles (500 cycles, c = 1 C), at room (25 °C) and elevated temperature (55°), respectively. In the first cycles, the tested working electrode provided 119.7 mAh·g⁻¹ and a capacity of 17.8% was recorded after 500 cycles, while an initial discharge of 119.2 mAh·g⁻¹ and a 74.8% capacity retention was calculated after an identical number of cycles, at 55 °C. The results obtained by Li₄Ti₅O₁₂ overcome other coatings like Li₃PO₄ [222], LiMnPO₄ [230] or Li₂CuO₂-Li₂NiO₂ [166], but performed slightly worse than Li₂MnO₃ [226] or La-Sr-Mn-O [130]. Despite the fact that lithium titanium oxide improved the cycling ability of the LMO by preventing manganese solution, no significant enhancement was observed in the case of lithium diffusion coefficient (5.33 × 10⁻¹² m² s⁻¹), which was lower than in other cases: La_0.7Sr_0.3Mn_0.7Co_0.3O₃ [220], FePO₄ [229], La-Sr-Mn-O [130] or LaF₃ [131]. However, electrochemical impedance spectra demonstrated that LTO helped lithium diffusion and composite’s R_ct did not increased so much (from 6.58 to 54.2 Ω) after 100 cycles, as in the case of pristine sample (from 8.45 to 156.27 Ω).

Uniform thin layer of hexafluorotitanic lithium (Li₂TiF₆) with high lithium conductivity has been deposited on commercial LiMn₂O₄ by a co-precipitation method, using precursors like lithium carbonate (Li₂CO₃) and hexafluorotitanic acid (H₂TiF₆) [238]. Low amount of hexafluorotitanic lithium was enough to stabilize and improve the lifetime of LMO. It was showed that high weight amount (10 wt.% Li₂TiF₆) of coating determined the formation of a thick coating layer. It managed to constrain capacity fade, although the delivered discharge capacity was not sufficient. Usually, the coatings do not have excellent lithium conductivity and this drawback affects the performance of the working electrode at high discharge rates, because the adequate lithium diffusion paths are available. However, Li₂TiF₆ improved the high discharge rate performance of bare spinel; thus, the specific capacity for composite material was 99.4 mAh·g⁻¹ in the 200th cycle, much greater than 65.7 mAh·g⁻¹ for pristine LMO. The results were superior to the ones obtained by La_0.7Sr_0.3Mn_0.7Co_0.3O₃-coated lithium manganese spinel [220].

4. Future Prospects

Nickel-doped (LiNi_0.5Mn_1.5O₄) lithium manganese oxide is considered as a suitable cathode material for electronic devices and electrical cars, which demand high operating voltage Li-ion batteries [56,239]. Depending on the chosen annealing temperature, LiNi_0.5Mn_1.5O₄ exhibits two type of structures: face-centered spinel (Fd3m) and primitive simple cubic crystal (P4₃32) [59,240]. Spinel structure is more stable and shows better electrochemical characteristics, which recommends to carefully selecting the synthesis route for obtaining LNMO spinel-type materials [240].

The working mechanism during charge/discharge cycles for a pristine LNMO includes three voltage plateaus, one at 4–4.1 V, which corresponds to Mn³⁺/Mn⁴⁺ redox couple and the other two around 4.75 V, which correspond to Ni²⁺/Ni³⁺ and Ni³⁺/Ni⁴⁺ couples. The low voltage plateau is less visible when the content of Mn⁴⁺ and Ni²⁺ is higher [59,240].

The improvement of LNMO’s performance during battery functioning includes doping, surface modification or the usage of a proper electrolyte. Among the doping ions, the insertion of chromium slightly increased the voltage, up to 4.9 V, because of the Cr³⁺/Cr⁴⁺ redox couple [59,70]. The different ways to overcome LNMO’s drawbacks are described in other studies [59,240].

5. Conclusions

Lithium manganese oxide exhibits two main drawbacks, which hinder its performance during LIB cycling: Jahn-Teller distortion [241,242,243] and manganese dissolution into electrolyte solution [164,244]. In order to hinder these two disadvantages and stabilize the structure of LMO, two main strategies have been adopted: structure doping and particle coating.

The dopants can stabilize LMO structure, by decreasing the amount of Mn³⁺ content, which is highly responsible for disproportion during charge/discharge cycles. However, by reducing the amount of electrochemical active Mn³⁺ species, the discharge capacity decreased with several percentages, but the performance of doped material has been improved, by retaining a higher capacity after a certain number of cycles.

Among the studied cations dopants, aluminum is one of the most studied. Aluminum-doped LMO spinel performs well both in aqueous and organic based electrolyte. In aqueous electrolyte, the calculated capacity retention of LiAl_0.1Mn_1.9O₄ was 70% after an extremely long number of cycles (more than 4000 cycles) [93]. In comparison, Al-doped LMO managed to retain more than 80% of its initial discharge capacity in a cell with LiPF₆ based electrolyte after 1000 charge-discharge cycles [92].

Multiple doping with low amounts of cations and anions, like Mg and F [78] or Al and S [154], led to materials capable to provide adequate discharge capacity for long cycle life.

The anions did not reduce the quantity of Mn³⁺, therefore the anion-doped LMO exhibited higher capacity, even than pristine sample [45].

Coating materials lessen the contact between the electrolyte and active material, thus the cycle life ability has been longer [163]. For example, TiO₂ coated on lithium manganese oxide may react with HF, resulted when LiPF₆ accidentally reacts with moisture, to form a protective layer, which prevents the manganese dissolution [205]. Aluminum sulfate octadecahydrate used as a precursor for aluminum oxide coating layer for LMO helped Al₂O₃ to be strongly bonded to LMO by hydrogen bonds, having Al as an anchor atom [211].

Another improvement provided by coatings is related to enhanced lithium diffusion, by reducing the charge-transfer resistance due to better electronic conductivity. In addition, by using a carbon-based support material, one can achieve even a greater discharge capacity than the theoretical capacity of LMO, because of high concentration of lithium ions into carbon nanotubes network [168].