This work illustrates an optimized preparation process of Li4Ti5O12 electrodes via the aqueous process using sodium carboxymethyl cellulose (CMC) as binder. In order to prevent aluminium current collector corrosion due to the rising pH of slurries, different concentrations of phosphoric acid (PA) are added to buffer the pH during the ball milling process. The LTO electrodes with 4PA-CMC are characterized in terms of surface morphology, and a kind of new ‘rod’ is found. Through XRD, XPS and ATR-IR, the ‘rod’ is proved to be Li1+xTi2(PO4)3. However, from the Rietveld refinement of the LTO electrodes with 4PA-CMC, only Li3PO4 could be detected in the XRD pattern, and LTO/Li3PO4 is estimated to be roughly 9/1. Remarkably, the formation of the rod substantially enhanced the electrochemical performance of LTO electrodes. Moreover, we verify that the acids have influence in the structure evolution of LTO particles, such as lattice parameter, unit cell volume, μstrain and oxygen position through analysis of in-situ XRD patterns.
The surface morphology properties must be controlled due to influence on the electrochemical performance of the materials. Indeed, the ball milling speed and carboxymethyl cellulose (CMC) percentage on the electrode formulation have been proved to be the crucial factors that influence the surface morphology of Li4Ti5O12 (LTO) electrodes during aqueous-based electrode preparation. In this work, phosphoric acid (4PA) has been used as additive to buffer the pH of slurry and it has been observed the formation of rod shape particles when high ball milling speed is used (>1100 rpm), while at lower ball milling speed (900 rpm) the rod shape particles are not forming. The surface has been investigated by means of SEM, EDX, HRTEM, STEM-EDX, XPS and EIS, and their electrochemical performance has been characterized on both half-cell and full-cell level (LiNi0.5Mn0.3Co0.2O2–NMC532–as cathode material). The results reveal that the LTO with not rod shape formation exhibits better cycling performance in both half- and full-cell than the one with rod shape particles on the surface (using 7% CMC electrode formulation), due to the formation of a thicker solid electrolyte interphase (SEI) and Li3PO4 layer around the LTO particles. Li3PO4 has a higher lithium ion conductivity than LTO and thus permits an enhanced electronic conduction, whereas a thicker SEI layer is effective in isolating the active material and electrolyte and decreasing the substantial decomposition of electrolyte over cycling. In addition, the LTO electrodes with 5% CMC and rod shape particles has superior electrochemical performance than without rods (7% CMC on the electrode formulation) due to the smaller polarization. In fact, less amount of CMC binder in the final electrode formulation decreases the viscosity of the slurry and the tunnels resistances related to the gap between the binders and the particle clusters. Therefore, controlled surface morphology, as well as, electrode formulation, is beneficial for enhancing the cycling capability of LTO anode material for Lithium ion batteries. When compared to LTO electrode preparing used only CMC as the binder, the introduction of 4PA in LTO electrode can largely improve the electrical conductivity and relieve the polarization, which could contribute to an enhanced rate capability. From the surface morphology of cycled electrodes, the results prove that adding 4PA is helpful in forming a uniform and dense layer on the surface of LTO particles, which protects the LTO particles from being damaged. Besides the protective layer, Li1+xTi2(PO4)3 and Li3PO4 are formed, these two compounds have good conductivity and are useful in enhancing the cycling capability and reversibility of LTO electrodes, even at 3.5 V cut-off or 4C long-term cycling, the cycling stability is also excellent. CMC electrode has a larger μstrain value than that of 4PA-CMC electrode during the first dis-/charge process. PA have benefits in acquiring a lower μstrain value during charge, whereas FA have advantages during discharge. The reduction of PA amount leads to the increase in the μstrain during both discharge and charge process. The surface morphology of LTO electrodes employing 4PA-CMC as binder can be adjusted through the ball milling speed, by forming Li3PO4/Li1+xTi2(PO4)3. The crystalline size of Li3PO4 increases by the ball milling speed, showing the LTO milled under 900 rpm the lowest crystalline size (54 nm) while the LTO milled under 1100 rpm shows much higher size (78 nm), despite both have similar crystallite size for Li4Ti5O12. The ball milling speed determines the existence of Li3PO4/Li1+xTi2(PO4)3 and modify the surface morphology and in turn influence in the electrochemical performance of LTO. The LTO milled under 900 rpm (4P-no), shows a homogeneous layer of Li3PO4 which acts as protective film, rather than isolated rods shows in the LTO milled under 1100 rpm, enhancing the long-term cycling stability. The XPS analysis of cycled electrodes also provides that the LTO milled under 900 rpm with Li3PO4 protection layer (4P-no) forms thicker SEI than that of 4P-rod (1100 rpm), which is in line with the EIS results. The SEI resistance (R2) shows lower and more stable values for 4P-no electrode resulting in a better long cycling performance at 1C. Additionally, the results of LiNi0.5Mn0.3Co0.2O2|Li4Ti5O12 full-cell also confirmed the superior cycling capability of 4P-no with a high retention of 96.8 % (122 mAh g-1), after 800 cycles. Finally, the Li3PO4 protection layer is not the only parameter that affect the electrochemical performance, the CMC percentage also influence in the capacity and stability mainly at high rates. The use of low percentage of CMC, from 7% to 5%, has advantages in reducing the cell polarization and improving the discharge capacity of LTO electrodes at high rate. As the results show that the 5% CMC with 4P-rod electrode has the highest capacity of 106 mAh g-1 at 7C with a medium voltage difference (ΔE) of around 210 mV. However, the CMC concentration should be enough to provide good adhesion between the active material and binder, as well as, current collector, because if not the electrochemical performance is impoverished as showed the 4P-rod electrode with 3% CMC. In conclusion, two parameters should be at least considered to develop high electrochemical performance in LTO electrode, the ball milling speed and the CMC concentration in the electrode.
The aqueous processing of Li4Ti5O12 (LTO) anode, utilizing water-soluble Sodium carboxymethyl cellulose (CMC) binder is explored since it has the potential to remarkably reduce the battery processing costs and makes a great step forward to an eco-friendly production of lithium-ion batteries (LIBs). Citric acid (CA), phosphoric acid (PA) and their mixture are adopted to modify the surface condition of LTO electrodes. As a consequence, the electrodes prepared in the presence of 1PA-CA show an extremely smooth surface geomorphology, the smallest cell polarization, the improved lithium-ion diffusion and transfer kinetics, contributing to the most outstanding cycling stability over 400 cycles at 4C. The impact of PA and CA on the electrochemical properties of Li4Ti5O12 anodes is investigated by means of FTIR spectra, electrochemical impedance spectroscopy (EIS), XPS analysis, cyclic voltammetry (CV) measurements and galvanostatic intermittent titration technique (GITT). Finally, the Li4Ti5O12|LiNi0.5Mn0.3Co0.2O2 full cell, employing the LTO electrode with 1PA-CA as anode, achieves a discharge capacity of 117 mAh g-1 after 1000 cycles with a capacity retention of 92.1% after 1000 cycles at 1C. Hence, acids might play a significantly important role in the electrochemical properties and surface chemistry of the LTO electrodes.The greatly beneficial contribution of phosphoric acid and citric acid (1PA-CA) as additives upon slurry preparation to permit the aqueous process of Li4Ti5O12 (LTO) anode was exhibited. The combination of 1PA and CA allows for a smooth surface and an acceptable polarization during cycling. The interaction of the acids (PA, CA) with the binder (CMC) was investigated through the FTIR spectra, and the result shows a highly increased C=O bond in 1PA-CA-CMC, confirming the interaction of 1PA with CA. Similar conclusion can be obtained from the XPS analysis of acidified LTO powder. The voltage profiles at 4C and the response for the current pulse in GITT proves the smallest polarization of 1PA-CA, and the value of DLi+ verifies its benefits in enhancing the lithium transfer during delithiation. Moreover, the morphology of the electrodes shows that LTO particles cycled with 1PA-CA well maintain their morphology, and the outline of LTO is quite apparent, and the XPS results suggests that 1PA-CA possesses the smallest amount of fluorine species including LiF. Therefore, 1PA-CA is able to substantially suppress the decomposition of the electrolyte, significantly facilitate the lithium ion transportation and then highly enhance the cycling capability of Li4Ti5O12. Although 1PA still has corrosion on the surface of electrode and possesses the thickest SEI layer containing the most fluorine compounds after cycling, 1PA also demonstrates the lowest amount of leaching lithium and titanium and the smallest impedance in dis-/charge state after 50 cycles. This might clearly explain why 1PA stably delivers a higher capacity than 1PA-CA in the initial cycling, but decays quickly in the long round. Conversely, CA exhibits a pronounced surface damaging, remarkable increase of polarization over cycling, the largest amount of leaching lithium and titanium, the continuous rise of impedance during dis-/charge, a dense but inhomogeneous layer of SEI, and the worst lithium-ion diffusion and transfer kinetics obtained from both DLi+ value and GITT results. Undoubtedly, all of these lead to a much worse cycling capability of CA. In conclusion, acids surely play a significantly important role in the electrochemical properties and surface chemistry of the electrodes.