Jaime-Marie
Price
*ab,
Phoebe
Allan
*ab and
Peter
Slater
*ab
aSchool of Chemistry, University of Birmingham, Birmingham, B15 2TT, UK. E-mail: p.r.slater@bham.ac.uk; jxp407@student.bham.ac.uk; p.allan@bham.ac.uk
bThe Faraday Institution, Harwell Science and Innovation Campus, Didcot, OX11 0RA, UK
First published on 4th May 2023
LiNiO2 (LNO) was prepared by two synthesis techniques: solid-state (SS-LNO) and coprecipitation (C-LNO). The results showed that C-LNO could be synthesised in as little as 1 hour at 800 °C in O2 to give a pristine material. The layered oxide structures of both materials have been investigated using PXRD, confirming that phase pure samples have been made. Electrochemical properties were explored over a range of voltage windows (2.7–4.1 V, 2.7–4.2 V and 2.7–4.3 V vs. Li+/Li), to analyse how the H2–H3 phase transition impacts the cathode materials’ capacity retention. Electrochemical measurements showed that the initial discharge capacity and cycle stability are improved in C-LNO compared to SS-LNO, achieving 221 mA h g−1 and 199 mA h g−1 respectively in the voltage range 2.7–4.3 V (at 10 mA g−1), with capacity retentions of 47% and 41% after 100 cycles. A Mo doped system, Li1.03Mo0.02Ni0.95O2 (Mo-LNO) was then prepared via the solid-state route. Mo-LNO showed an even higher initial discharge capacity of 240 mA h g−1 between 2.7–4.3 V vs. Li+/Li, with a slightly enhanced capacity retention of 52%. Through the investigation of the different voltage ranges it was shown that capacity fade can be minimised by cycling the materials below 4.2 V, (attributed to avoiding the detrimental H2–H3 phase transition) although this results in a lower discharge capacity. This is shown by the cycling of SS-LNO, C-LNO and Mo-LNO in the voltage window 2.7–4.1 V, where discharge capacities of 144 mA h g−1, 168 mA h g−1 and 177 mA h g−1 were achieved with higher capacity retentions of 84%, 76% and 90% after 100 cycles respectively, the latter system showing promise as a cobalt-free cathode material.
First commercialised in 1991 by Sony, LiCoO2 is a layered oxide cathode material with a hexagonal structure, analogous to α-NaFeO2.3,4 It is an O3 type structure, meaning that lithium ions occupy octahedral sites with three CoO2 layers per unit cell.5 Oxygens stack in an ABCABC arrangement along the c-axis, giving rise to the hexagonal lattice (space group Rm).6,7 This structure allows lithium to move easily in and out of the material between the CoO2 layers.8 This was originally the cathode material of choice as it has a high operating voltage (∼3.6 V vs. Li+/Li) and a theoretical capacity of ∼274 mA h g−1, albeit the usable capacity is significantly lower at around 150 mA h g−1. However, cobalt is expensive at $52000 per ton,9 and unevenly distributed globally. There are also serious ethical concerns over the mining of cobalt, including stories of forced child labour, a lack of personal protective equipment for workers and forced evictions for mining projects.10 This has shifted the research focus to other layered transition metal oxide cathode materials where the Co content is reduced, including LiMO2 (M = Ni, Mn), Li [Ni1−x−yCoxMny]O2 (NMC) and Li[Ni1−x−yCoxAly]O2 (NCA).11–15
With the idea of driving down costs, LNO has been explored as a potential cathode material since the late 1990s. In September 2022, the price of Ni was $22000 per ton, significantly cheaper than Co, with it also being more abundant.9,16 LNO is isostructural to LiCoO2 and so it was only logical to investigate its electrochemical properties. Although it was shown to have a high experimental capacity of around 200 mA h g−1, this material itself faces some drawbacks and so shifted out of focus for many years. These drawbacks include being inherently difficult to synthesise stoichiometrically, predominantly down to cation mixing between Li+/Ni2+, due to similarities in the ionic size of Li+ and Ni2+. As a result of this, antisite defects occur (Li and Ni exchange sites within the layers).17 LNO also commonly contains an excess of Ni2+ in the Li layers due to lithium loss occurring through the long synthesis times and because of the incomplete oxidation of Ni2+ to Ni3+. These issues correlate to rapid capacity fade being observed, which leads to poor cycling stability. These deficiencies have led to LNO currently not being used in mass applications, although Mn and Co doped analogues (NMCs) are widely used.
To overcome the synthetic challenges of LNO, various synthesis techniques have been employed in the literature including solid-state reactions (SS), coprecipitation methods and hydrothermal processes, however, details of the optimisation of these synthetic processes have not previously been reported. The hydrothermal process is regarded as an inherently complicated process and is difficult to scale up and so is not studied here.18 This work therefore focuses on optimising the synthesis of LNO via two methods-solid-state and coprecipitation. Synthesis conditions including reaction time, temperature, and Li-excess level were optimised to synthesise stoichiometric LNO with minimal numbers of antisite defects. It is shown herein that the coprecipitation route has the advantage of synthesising phase pure LNO in a 1 hour sintering step, saving energy, time, and money. The work is then extended to prepare a new Mo-doped LNO material via SS, where we show a promising performance.
During cycling, LNO undergoes multiple phase transitions enabled by the electrostatic interactions from the ionic bonds that stabilize its structure. On delithiation, LNO is observed to transition through a series of phases; hexagonal (H1) → monoclinic (M) → second hexagonal (H2) → third hexagonal (H3), all of which possess an O3 sequence, with a symmetry reduction from Rm to C2/m for the monoclinic phase.19–23 On charging LixNiO2, the material undergoes a solid-solution reaction until x = 0.75, when the material undergoes its first phase transition from H1 → M. The M phase then remains stable until x = 0.4 to 0.36, when the phase changes again from M → H2. Finally, at 4.2 V (vs. Li+/Li), where the lithium content ranges from x = 0.26–0.16, a transition from H2 → H3 occurs, with the c-lattice parameter of the H3 phase being considerably smaller. Currently in the literature, there is debate around what state of charge (SoC) these transitions occur at, however, the discrepancies observed could be a result of different cycling conditions, cell fabrication and also the amount of cation mixing occurring within materials.19,21 Nevertheless, it is agreed in the literature that when LNO cells are cycled at 4.2 V or above (vs. Li+/Li), rapid capacity fade is seen. This coincides with the H2 → H3 transition and the drastic changes in the c-lattice parameter.19,23,24 As expected, this strains the LNO particles, leading to cracking and stacking faults in primary particles, thus reducing the stability of the material. In particular, cracks can lead to the cathode material becoming permanently electronically disconnected from the bulk, decreasing the capacity the material can achieve. Cracks also lead to fresh cathode material being exposed to electrolyte, allowing further side reactions to occur.19,20,23 Therefore, in this work we have also investigated the effect of limiting the upper cut-off voltage to 4.1 V, 4.2 V and 4.3 V (vs. Li+/Li) during electrochemical cycling. This is to evaluate how the H2–H3 transition affects the capacity fade of the materials made here and to determine whether high capacities can be achieved without severe capacity fade.
We also investigated the possibility of doping LNO to improve the performance. Mo was chosen as a dopant for a few reasons; it has been shown in the literature to enhance the electrochemical performance of NMC type materials, and Mo has a high charge and so allows extra Li introduction into the structure for charge balance.25–32 Research conducted by Park et al. indicated that Mo may have a ‘pillaring effect’ in Ni-rich NMC type materials.33 This is achieved by doping Mo into the layered oxide structure, where Mo occupies Ni sites. On high levels of delithiation, it is thought the Mo ions then act as pillars to hold the TM layers apart, thereby reducing the effect of the H2–H3 phase transition, as the difference between the H2 and H3 c-lattice parameter is no longer as great, allowing for the re-intercalation of the Li ions more easily back into the layered oxide structure. This then in turn reduces the capacity fade seen in the material. Other groups have also reported that low level doping with Mo in NMC materials can enhance the capacity and electrochemical stability.25–30,32 Sun et al. indicated that 1% Mo6+ doping in TM layers of Ni-rich Li[Ni0.91Co0.09]O2 led to an increased capacity retention with in situ XRD measurements showing smaller volume and c-lattice parameter changes during charging and discharging, compared to their pristine material. This is thought to be down to the stabilisation of the micro and atomic structure via pillaring effects.34 Yet to the best of our knowledge, there are limited reports on the effect of Mo-doping on pristine LNO. In this research, low level doping Mo was therefore employed via solid-state synthesis to try and overcome the capacity fade seen in LNO and diminish some of the effects of the H2–H3 phase transition. The results of Mo doping indicated a 20% improvement in discharge capacity as well as improved cycling stability. This is discussed later in the paper.
The cathode coating was then used to construct LIR2032 coin cells. These contained around 3–5 mg of active material cut into 14.8 mm discs. CF/C glass fibre separator was used with Li-metal as a counter electrode, cut into 16 mm and 15 mm discs respectively. 80 μL 1M LiPF6 in 1:1 ethylene carbonate:dimethyl carbonate (EC:DMC) with 2 wt% of vinyl chloride was used as the electrolyte. The coin cells were assembled in a dry room (temperature controlled at 20 °C, with an average dew point of −50 °C Td). The cycling performance of the cathode materials was analysed in the voltage window 2.7–4.3 V (vs. Li+/Li) at 25 °C (Biologic BCS-COM cell tester used, operated by BT-Lab software). For cells tested over the different voltage ranges 2.7–4.1 V, 2.7–4.2 V and 2.7–4.3 V, the current density was kept constant at 10 mA g−1 for 100 cycles to allow the stability of the materials at different cut-off voltages to be examined. To test the rate capability of the materials (between 2.7–4.3 V vs. Li+/Li) the current density was changed every 5 cycles (10 mA g−1, 20 mA g−1, 30 mA g−1, 40 mA g−1, 50 mA g−1, 10 mA g−1). These cell tests were all done in triplicate.
Sample | Synthesis conditions | c (Å) | a (Å) | Volume (Å3) | I (003)/I(004) | R w |
---|---|---|---|---|---|---|
SS-LNO | 700 °C/12 h in O2 | 14.1948(5) | 2.8782(1) | 101.84(1) | 1.66 | 4.1 |
SS-LNO | 900 °C/12 h in O2 | 14.2151(7) | 2.8878(3) | 102.66(1) | 1.19 | 5.0 |
SS-LNO | 650 °C/12 h in O2 | 14.1800(8) | 2.8853(2) | 102.23(6) | 1.15 | 4.8 |
Prior literature suggests at least 12 hours of heating is needed to form LNO via SS methods.19,22,41 Therefore, SS-LNO was first synthesised at 950 °C in air with no excess lithium for 12 hours respectively (Fig. 1). However, analysis of the PXRD pattern gave an I(003)/I(104) ratio of 1.15 (less than the accepted 1.2 for a well ordered lattice) and no clear splitting between the (006)/(102) and (108)/(110) peaks is observed. Site occupancy factors from Rietveld refinement of the LNO structure against PXRD data indicated that more than 7% cation mixing was present (ESI,† Table S2). Previous reports in the literature indicated that the best conditions to synthesise near-stoichiometric LNO was under a constant flow of oxygen and with a small amount of lithium excess.4,20,24,40 A constant flow of O2 hinders the reduction of Ni3+ to Ni2+, which in turn lowers the number of antisite defects and cation mixing. The lithium excess is used to counteract the volatility of Li at higher temperatures, as such Li loss can again give a suboptimal material.4,24,35,42 LNO synthesised in oxygen flow using 5% excess lithium at a slightly lower temperature of 900 °C (Fig. 1) for 12 hours displayed an I(003)/I(104) ratio of 1.19, which was still indicative of Li/Ni site mixing. Li/Ni site mixing was minimised by heating the sample to 700 °C for 12 hours under a flow of oxygen, using 5% excess lithium to counteract the lithium volatility during the long duration heat treatment. The I(003)/I(104) ratio achieved was 1.66 with clear peak splitting observed between (006)/(102) and (108)/(110) and the (101) reflection is more intense than the (006)/(102) reflections, indicating that cation mixing is minimal. The Mo-doped LNO material was also synthesised using these optimised solid-state conditions (XRD shown in the ESI,† Fig. S1).
Fig. 2 XRD pattern of SS-LNO synthesised at 700 °C for 12 hours in O2 (blue), and C-LNO synthesised at 800 °C for 1 hour in O2 (green). |
Sample | c (Å) | a (Å) | c/a | Volume (Å3) | Li (3b) | Ni (3b) | I (003)/I(104) |
---|---|---|---|---|---|---|---|
SS-LNO | 14.1948(5) | 2.8782(1) | 4.932 | 101.84(1) | 0.98(1) | 0.02(1) | 1.66 |
C-LNO | 14.2021(4) | 2.8795(3) | 4.932 | 101.98(1) | 0.99(1) | 0.01(1) | 2.12 |
To try and improve the capacity fade and maintain a high capacity of LNO, the material was doped with low levels of MoO3 to form Li1.03Mo0.02Ni0.95O2 (denoted Mo-LNO herein). Due to the fact that it was not possible to make a Mo doped Ni(OH)2 precursor (since Mo6+ precipitates under acidic rather than basic conditions), this material was prepared by the solid-state route under the same conditions used for LNO (for XRD and refinement see Fig. S1 and S6 in ESI†). The Li content was increased, and the Ni content reduced to charge balance the introduction of the higher charge of Mo.
After 100 cycles, the discharge capacities are 82.2 mA h g−1 (cut-off 4.3 V), 77.0 mA h g−1 (cut-off 4.2 V) and 121.2 mA h g−1 (cut-off 4.1 V), giving capacity retentions of 41%, 42% and 84% respectively. This shows that the lower cut-off voltage of 4.1 V leads to slower capacity fade in the material and significantly, a higher capacity after 100 cycles than for higher cut off voltages. From the differential capacity plots shown in Fig. 3 (dQ/dV vs. V), above 4.1 V the plots in Fig. 3(d) and (e) show a well-defined peak which has been assigned to the H2–H3 phase transition, with the H3 phase transitions occurring at 4.2 V. The dQ/dV plots and the discharge capacities reported here match well with previous literature.22,23,46 This H2–H3 transition leads to the abrupt shrinkage of the c-lattice parameter, as highlighted in the introduction.19–21,23,24 This translates to severe strain across the material, facilitating electrode pulverisation and leading to the observed large irreversible capacity loss. This can be seen in dQ/dV plots in Fig. 3(b) and (d) by the lowering of peak intensities and the slight shift in peaks as the cells are cycled up to 100 times. Cycling to the lower cut off voltage of 4.1 V means that the H2–H3 phase transition does not occur, and so only 16% of the initial capacity was lost after 100 cycles (with peak intensities also remaining consistent in the dQ/dV plot), emphasising that it is the H2–H3 transition that is mainly responsible for the detrimental performance loss of the material.
The highest initial capacity for any of the materials discussed in this paper was achieved for this Mo-LNO phase in the voltage window 2.7–4.3 V (vs. Li+/Li). This gave a value of 240 mA h g−1, compared to 221.4 mA h g−1 and 199.1 mA h g−1 for C-LNO and SS-LNO in the same voltage window (Fig. 3(a) and 5(a) respectively). The Mo-LNO also had a higher initial capacity in the 2.7–4.2 V and 2.7–4.1 V voltage windows, being 207 mA h g−1 and 177 mA h g−1 respectively (Fig. 6(c) and (e)). After the Mo-LNO had completed 100 cycles, the discharge capacity dropped to 120 mA h g−1, 156 mA h g−1 and 160 mA h g−1 for cut off voltages 4.3 V, 4.2 V and 4.1 V respectively. This corresponds to a capacity retention of 50% (4.3 V), 75% (4.2 V) and 90% (4.1 V). From the dQ/dV plots in Fig. 6(b) and (d), at cut-off voltages of 4.3 V and 4.2 V, the H2–H3 phase transition is still present in this Mo-LNO material. However, the capacity fade seen in Mo-LNO is less than in the corresponding SS-LNO and C-LNO at all the cut-off voltages. Nevertheless, it should also be noted that the peaks for the H2–H3 phase transition in Fig. 6(b) and (d) are slightly broader compared to the dQ/dV plots for C-LNO and SS-LNO under the same electrochemical conditions (Fig. 5(b), 5(d) and 3(b) and 3(d) respectively), while the high voltage “plateau” starts to slope more. This may give an early indication that the Mo could be having some pillaring type effect within the layered oxide Mo-LNO material to reduce the large volume change effect of the H2–H3 phase transition (as the H2–H3 transition is still occurring) and stabilising the structure, which in turn helps to reduce the capacity fade seen within the material.
From Fig. 4, making a direct comparison of the dQ/dV plots for the Mo-LNO and LNO samples also shows significantly that the H2–H3 transition is shifted to a slightly higher voltage, which is consistent with more Li being removed before the transition occurs, and therefore allowing for a greater Li removal and hence higher capacity.
The results here thus agree with previous research where Mo doping has been utilised for Ni-rich NMC cathode materials, and the improved performance attributed to this pillaring effect allowing facile movement of Li+ between these layers during charge and discharge.25–34
In terms of the significantly improved overall capacity on Mo doping, this can be attributed to several factors: the introduction of Mo6+ in place of Ni3+ requires further charge balance by the introduction of extra Li+ in place of Ni3+ leading to a Li rich phase where Li:Ni ratio >1, increasing the theoretical capacity of the material. The consequence of this will be also likely to help suppress detrimental Li/Ni antisite defects. Therefore, the increased Li content, reduction in Li/Ni antisite defects, Mo pillaring effect and influence of the shifted H2–H3 transition, all allowing for more Li removal, can most likely account for the significantly improved performance.
During the course of this paper being under review, similar beneficial results for Mo doping were reported by Li et al.47 In this work, the authors examined higher Mo contents, Li1+yNi(3−5y)/3Mo2y/3O2 0.03 ≤ y ≤ 0.12) with samples prepared by a sol gel-type route. This work suggested the coexistence of a Mo doped Li rich LiNiO2 phase and Li4MoO5 with the former refined as Li1.04Ni0.93Mo0.03O2, which is close to the composition reported here, which we observed as the Mo doping limit before a rock salt impurity was observed (ESI†-Fig. S15). Interestingly this work by Li et al., showed that the presence of Li4MoO4 appears to help to reduce capacity fade by providing an”epitaxial stabilization” effect for the Li rich LNO domain. It is therefore possible that some of the improved capacity retention in our systems compared to undoped LNO may relate to a small amount of local Li4MoO4 domains, although the retention we observe (for the comparable 4.3 V cut off) is lower than observed for the higher Mo contents studied by Li et al., therefore suggesting that any Li4MoO4 levels are significantly lower. Also, given the observed changes in peak ratios I(003)/I(104) (reported in Table S1, ESI†) and the shift to a higher voltage of the H2–H3 transition, we believe that the evidence points to most of the Mo being incorporated into the LNO phase in our material.
Here the H2–H3 phase transition is occurring, showing the detrimental effect this has on the capacity the material can achieve. When the current density was increased to 50 mA g−1, a capacity of 91 mA h g−1 was achieved – 45.5% of the initial capacity, representing a substantial drop in performance. A similar effect is seen for rate tests performed between 2.7–4.2 V for SS-LNO, shown in Fig. 7(d). Again, the H2–H3 transition occurs in these cells. In this case, 26% of the initial capacity was lost in the first 5 cycles (from 185 mA h g−1 to 137 mA h g−1). However, the capacity was more stable at higher rates, still maintaining 137 mA h g−1 at a current density of 50 mA g−1. When the current density returned to 10 mA g−1, the capacity returned to 157 mA h g−1, giving a capacity retention of 85% at a cut off voltage of 4.2 V. The rate performance tests were shown to be most stable for SS-LNO between 2.7–4.1 V, where the H2–H3 phase transition is not occurring. Although the initial capacity was lowest in this voltage range at 144 mA h g−1, after the first 5 cycles the capacity was still 137 mA h g−1, giving a capacity retention of 95%. At 50 mA g−1, a capacity of 128 mA h g−1 was achieved (88% of initial capacity) and then when the current density was returned to 10 mA g−1, the capacity increased to 141 mA h g−1, indicating a capacity retention of 98%. These rate tests further confirm that it is the occurrence of the H2–H3 transition in SS-LNO that detrimentally affects the rate capability and capacity retention of the LNO cathode material.
As with SS-LNO, C-LNO was also electrochemically tested at multiple current densities between voltage limits of 2.7–4.1 V, 2.7–4.2 V and 2.7–4.3 V (vs. Li+/Li) to evaluate the effect the synthesis technique and the cut-off voltage had on the rate capability of the material. This is also shown in Fig. 7. The initial discharge capacity between 2.7–4.3 V for C-LNO was 224 mA h g−1, dropping to 201 mA h g−1 after 5 cycles giving a capacity loss of 11% (Fig. 7(b)). This is lower than the capacity lost for SS-LNO over the first 5 cycles in the sample voltage range, which lost 31.5% of its original capacity. When C-LNO was cycled at a current density of 50 mA g−1 between 2.7–4.3 V, a discharge capacity of 163 mA h g−1 was achieved. On returning the current density to 10 mA h g−1, a discharge capacity of 189 mA h g−1 was retained, corresponding to 84% of the initial discharge capacity. This was a higher capacity retention of 84% for C-LNO compared to 65% for SS-LNO under the same testing conditions, with C-LNO also having a higher discharge capacity, once more illustrating the better capacity performance of the coprecipitation material. Similar effects are seen for C-LNO tested between 2.7–4.2 V and 2.7–4.1 V, as shown in Fig. 7(d) and 7(f) respectively. The C-LNO sample tested between 2.7–4.2 V achieved an initial discharge capacity of 196 mA h g−1, decreasing to 173 mA h g−1 after the first 5 cycles. However, SS-LNO only achieved 185 mA h g−1 decreasing to 137 mA h g−1 under the same conditions. For C-LNO, the capacity appears to be more stable at higher rates when the cut off voltage is limited to 4.2 V compared to 4.3 V, giving a discharge capacity of 155 mA h g−1 at 50 mA g−1 (79% of initial capacity). When returning the current density to 10 mA g−1, 87% of the original discharge capacity was achieved. Although the initial discharge capacity is lowest for C-LNO when the voltage is set between 2.7–4.1 V, the material is most stable when performing rate tests within these voltage limits, as the H2–H3 phase transition is not occurring in this voltage range. From Fig. 7(f) you can see an initial discharge capacity of 158 mA h g−1 is achieved, decreasing by 7% in the first 5 cycles. At 50 mA g−1, a discharge capacity of 131 mA h g−1 is seen and when the current density is returned to 10 mA g−1, the capacity increases back to 139 mA h g−1 giving 88% capacity retention. As for the other voltage ranges, C-LNO shows a higher capacity of 158 mA h g−1 compared to 144 mA h g−1 for SS-LNO when comparing rate testing between 2.7–4.1 V. Overall, these results indicate that although the H2–H3 phase transition is occurring in C-LNO tested at cut-off voltages 4.2 V and 4.3 V, the effects are not as severe as in SS-LNO, which is evident in the reduced capacity fade. The higher capacities and the reduction in capacity fade seen in C-LNO can be attributed to the more optimised synthesis of C-LNO. Thus, synthesising C-LNO at 800 °C for 1 hour only in O2 yields a higher quality material with less cation mixing (Table 2) and a more perfectly ordered layered oxide, which can be supported by the higher capacities seen for the C-LNO material.
Fig. 7 also shows the rate tests performed on the Mo-LNO material within the voltage windows of 2.7–4.3 V, 2.7–4.2 V and 2.7–4.1 V. As shown for the undoped LNO materials, the voltage window 2.7–4.1 V proved to be the most stable during cycling at varied rates, due to the H2–H3 phase transition not occurring here (Fig. 7(f)). An initial discharge capacity of 181 mA h g−1 was achieved, losing only 2% of this initial capacity in the first 5 cycles. At 50 mA g−1, the material had a discharge capacity of 158 mA h g−1, giving 81% capacity retention. When the current density was returned to 10 mA g−1, the Mo-LNO showed a capacity of 178 mA h g−1, meaning 98% of the overall capacity was retained. This is not only a higher capacity achieved compared to SS-LNO and C-LNO, but also a better capacity retention. For Mo-LNO tested between 2.7–4.2 V, an initial discharge capacity of 200 mA h g−1 was seen, decreasing to 190 mA h g−1 in the first 5 cycles. At 50 mA g−1, 172 mA h g−1 was achieved and when the rate was returned to 10 mA g−1, the capacity also increased back to 192 mA h g−1, giving an overall capacity retention of 96%. In the voltage window 2.7–4.3 V, an initial capacity of 240 mA h g−1 was observed, reducing by 6% on the first 5 cycles and further reducing to 188 mA h g−1 at a rate of 50 mA g−1. When the rate was returned to 10 mA g−1, a capacity of 216 mA h g−1 was achieved, giving a 90% capacity retention.
Overall, the data in Fig. 7 shows that the Mo-LNO material proved to be more stable with a higher capacity and less capacity fade than the SS-LNO and C-LNO during rate testing and long-term cycling, suggesting that the detrimental effect of the H2–H3 phase transition is slightly suppressed by the incorporation of small amounts of Mo into the LNO material. It is also evident that low level Mo doping enhances the discharge capacity seen in LNO, possibly relating to the corresponding increase in Li content for charge balance, the small shift in the voltage for the H2–H3 phase transition and a pillaring effect by Mo.
Overall, this work highlights that LNO can be prepared in very short heat treatment times, and that low level Mo doping can have a significant effect on improving the cathode materials performance. These finding can be employed to form new Ni-rich doped materials (e.g. co-doping with Mo and other transition metals) to propel us ever closer to optimising this phase to reach the demands needed to satisfy the expanding EV industry.
Footnote |
† Electronic supplementary information (ESI) available. See DOI: https://doi.org/10.1039/d3ya00046j |
This journal is © The Royal Society of Chemistry 2023 |