Modification on Li(NiCoMn) 1/3 O 2 Cathode Material
5-1 Introduction
In this chapter, instead of electrodes, this time, the surfacemodification was directly conducted on the powders, a time saving and less power consuming process for the surface coating have been usilized to modify the Li(NiCoMn)1/3O2 (NCM) powders.
The organic compounds, polystyrene sulfonate based polymers (PSS) (Fig. 5-1 (a)) was selected as the modified ingredient owing to its electrochemical stability, ionic conductivity due to the electron lone pairs surrounded sulfonyl groups, and aqueous soluble makes it environmentally friendly and easy to process. Compared with the inorganic coatings, the flexibility of polymeric materials enable to expand the coverage of coatings on the surface of materials, and the organic feature of polymers is also expected as an artificial SEI for helping the solvated and de-solvated of Li-ions during the intercalation and deintercalation process at the electrolyte/electrode interface.
In this study, the effects of polymer coating on the electrochemical performance have been investigated and divided into two parts. In the first part, the sodium type PSS, Poly sodium-4-styrene sulfonate (PSSNa) (Fig. 5-1 (b)) was introduced to modify the first batch of NCM powders, and in order to compensate the electron-insulated polymeric coatings, the conductive additive, super P, was adopted onto the NCM material by using electrostatic self-assembling of Poly dially dimethyl ammonium chloride (PDDA) (Fig. 5-1 (d)) and PSSNa, the cationic/aionic polymers. Fig. 5-2 shows the schematic conception of the electrostatic self-assembling mechanism. The performance of each coated NCM powders and pristine NCM were compared to each
doi:10.6342/NTU201601970
84
other in this chapter. In the second part, for attempting to promote the lithium-ion transfer within the polymer layer, lithium type PSS, Poly (lithium-4-styrene sulfonate) (PSSLi) (Fig. 5-1 (c)), was prepared as described in section 3-2 and coated onto the second batch of NCM powders. The performance between the Li and Na type PSS as the polymeric SEI modification were investigated and discussed in this chapter. By the way, a carbon nanotube (CNT) wrapped and polymer-coated NCM powder was concomitant designed (Fig. 5-3) for strengthening both the electronic connections and protective capability of the active material.
Figure 5-1 Chemical formula of (a) polystyrene sulfonate, (b) poly(sodium-4-styrene sulfonate), (c) poly(lithium-4-styrene sulfonate) and (d) polu(diallydimethyammonium chloride)
doi:10.6342/NTU201601970
85
Figure 5-2 Schematic illustration of the fabrication of Super P-PSSNa @ NCM-PDDA powders through electrostatic attraction of ionic polymers
Figure 5-3 Schematic illustration of CNT-wrapped and polymer-coated NCM composite powder.
doi:10.6342/NTU201601970
86
5-2 Polymer Modification on First Batch of NCM Powder:
5-2-1 Materials characterizations
For the first batch of NCM powder, 2 wt. % PSSNa was first coated onto the NCM powder through the water-based coating process as described in section 3-3-2.
Following, the 1 wt. % Super P was adopted into the coating process by the electrostatic self-assembling technique for increasing the electronic conductivity, as mentioned in the section 5-1. The zeta potential analysis was carried out to confirm the surface charge of Super P-PSSNa and NCM-PDDA, the result is shown in Table 5-1. For the bare NCM and Super P particles, the zeta potential are presented in negative state, which means these two particles carried the negative surface charge. However, once the NCM particle was coated with 1wt. % cationic polymer PDDA, instead of negative charge, the zeta potential was then changed from -30.0 to 56.9 mV, indicating the surface charge is altered to positive. In the same way, in the anion polymer PSSNa-coated Super P, the zeta potential became more negative (-31.2 to -61.8 mV). This result is agree with our design for the electrostatic attraction of the particles in the opposite surface charge to accomplish the self-assembling.
To distinguish the intrinsic property of PDDA-blend-PSSNa co-polymer without Super P, NCM coated with the 0.8 wt. % PDDA and 1.2 wt. % PSSNa co-polymer was also prepared. Therefore, in the first batch of NCM powders, three kinds of modified particles : NCM-2%-PSSNa, NCM-2%-PDDA-PSSNa and Super P-PSS @NCM-PDDA have been prepared and studied with the bare NCM (NCM-Pristine).
To further insure that the water-based coating process may not lead to the serious oxidation for sensitive transition metals, which are particularly located at the outside surface of the NCM material. The X-ray absorption spectroscopy (XPS) was conducted both on the NCM-Pristine and NCM polymer-coated particles. The XPS spectra and
doi:10.6342/NTU201601970
87
their binding energy (BE) in 2p3/2 orbital are shown in Fig. 5-4 and listed in Table 5-2.
Both the NCM-Pristine and NCM-polymer modified particles underwent 3 second Ar ion etching, In polymeric modified particle, the BE of Ni, Co, and Mn in 2P3/2 are located at 855.5, 780.4, and 642.2 eV, respectively, which are slightly higher than the NCM-Pristine (855.0, 780.0 and 641.7 eV). The result could be explained that the water-based process indeed caused partial oxidation to the transition metals, however, this slight oxidation we believe may not lead to the serious consequence to the cell performance. The comparison of cell performance between the active material which was prepared with and without the water-based coating process (no polymer additive) are shown in Fig. A-1 in Appendix A. The result displayed no much difference in the cell performance by the water treated process to the material.
The surface morphologies of polymer-coated particles are displayed in Fig. 5-5.
NCM-Pristine showed particles size round 3~5μm with the clean surface (Fig. 5-5 (a)).
Apparent polymer membranes could be observed from the SEM images in the 2%-PSSNa and 2%-PDDA-2%-PSSNa-coated particles (Fig. 5-5 (b) and (c)). The sulfur and chlorine element in the EDS analysis of NCM-2%-PDDA-PSSNa particles are also verified the existence of coated polymer (Fig. 5-6 and Table 5-3). From the Fig. 5-5 (d), it shows that the Super P particles on the NCM tend to agglomerate to each other, this aggregation of particles may happen during the drying process since the interaction between the nano-nano particles is greater than the nano-micron particles. The TEM images shown in Fig. 5-7 (b),(c) demonstrate the thickness of the polymer membrane on the NCM-2%-PSSNa and 2%-PDDA-PSSNa particles is varied from 2 to 10 nm, and for the Super P-PSS @NCM-PDDA particle (Fig. 5-7 (d)), the attached Super P on the NCM particle along with the polymer membrane can also be observed.
doi:10.6342/NTU201601970
88
Table 5-1 Zeta potential of the particles before and after polymer coating.
Sample Zeta potential (mV) Surface charge
Bare NCM -30.0 Negative
Bare Super P -31.2 Negative
NCM-PDDA 56.9 Positive
Super P-PSSNa -61.8 Negative
doi:10.6342/NTU201601970
89
Figure 5-4 XPS profiles of bare NCM (a)~(c) and polymer-coated NCM (d)~(f) after 3 second etching.
Table 5-2 XPS binding energy. of 2p3/2 orbital
Sample
Binding Energy (eV)
Ni 2p3/2 Co 2p3/2 Mn 2p3/2
NCM-Pristine 855.0 780.0 641.7
NCM-Polymer coating 855.5 780.4 642.2
660 655 650 645 640 635 630
200
815 810 805 800 795 790 785 780 775 770 500
885 880 875 870 865 860 855 850 845
400
885 880 875 870 865 860 855 850 845
300
815 810 805 800 795 790 785 780 775 770 400
500 600
Counts(s)
Binding Energy (eV)
660 655 650 645 640 635 630
100
doi:10.6342/NTU201601970
90
Figure 5-5 SEM micrographs of (a) Pristine, (b) 2%-PSSNa, (c) NCM-2%-PDDA-PSSNa, and (d) Super P-PSS @NCM-PDDA particles
doi:10.6342/NTU201601970
91
Figure 5-6 EDS profile of the NCM-2%-PDDA-PSSNa particles
Table 5-3 EDS analysis of the NCM-2%-PDDA-PSSNa particles
Element Weight % Atomic %
C K 5.79 13.78
O K 30.15 53.85
Na K 0.48 0.59
S K 0.33 0.29
Cl K 0.32 0.26
Mn L 18.77 9.76
Co L 14.33 6.95
Ni L 29.83 14.52
Totals 100.00 100.00
doi:10.6342/NTU201601970
92
Figure 5-7 TEM micrographs of (a) Pristine, (b) 2%-PSSNa, (c) NCM-2%-PDDA-PSSNa, and (d) Super P-PSS @NCM-PDDA particles
doi:10.6342/NTU201601970
93
5-2-2 Electrochemical Performance
In the electrochemical performance, the modified and pristine cells were firstly formation at room temperature in the current density of 0.1C (1C = 160mAh/g) and operated from 2.5 to 4.4 V for three cycles, the voltage profiles are shown in Fig. 5-8, and the relevant capacity data are listed in Table 5-4. From the voltage profiles, the onset oxidation of the pristine sample starts from 3.74 V in the first cycle (Fig. 5-8 (a)), which is lower than that of the modified samples (about 3.84 V) (Fig. 5-8 (b)～(d)).
The higher oxidation potential in the charging process may relative to the coated polymer, which delayed the oxidation reaction of the active material with electrolyte from the material surface, leading to the appearance of the plateau at higher potential.
However, after the first cycle, the plateaus in the modified cells returned to the similar but still higher position with the pristine sample in the following two cycles, this is due to the polymer-coated materials have been totally immersed in the liquid electrolyte, thus the channels for the Li-ion diffusion were opened. The specific discharge capacity in the first cycle of the NCM-Pristine, NCM-2%-PSSNa, NCM-2%-PDDA-PSSNa, and Super P-PSSNa@NCM-PDDA samples are 167.35, 142.35, 145.83, and 150.26 mAh/g respectively, the reasons for causing the lower capacity in the modified samples could be described as following two points: (i) The object mass whom the specific capacity (SC) based on is different between the pristine and modified electrode. For the control non-treated sample, the SC is only based on the mass of active material itself, while for the polymer-coated powder, taking NCM-2%-PSSNa electrode for example, its SC here is based on the bare NCM and 2 wt. % PSSNa, where PSSNa is electrochemical inactive but still calculated into the SC; (ii) The relative polarized plateaus presented in the modified samples. As described in earlier, the relaxation of Li-ion diffusion caused by the polymer film would make the temporary polarization of the electrode, hence the
doi:10.6342/NTU201601970
94
early termination of lithiation process sacrificed the SC to the cells. In addition, for the NCM-Pristine cell, an extra capacity contribution was obtained by the uncommon partial reduction of surface metal ions at 3.0 V during the lithiation process, which also being the reason for getting the higher capacity. This tiny plateau appears at 3.0 V in the pristine sample during the discharge is observed in voltage profile and also can be demonstrated from the dQ/dV plots, as displayed in Fig. 5-9 (a) (the peak at 3.0 V).
From the literature [28, 131], this reaction may be associated with the reduction of manganese to the trivalent state. However, for the normal NCM materials, the manganese is generally in the oxidation state of tetravalent throughout the electrochemical process [55, 126, 132], this unexpected reduction of the manganese appeared in the pristine sample can be explained to the defects or oxygen vacancy formed on the NCM surface. Thus, the partial instability of structure boosts this unexpected reaction to occur. In the contrast, after the polymer modification, the samples no longer demonstrate this reduction reaction at 3.0 V, it implies that the polymer coatings somehow stabilize the materials interface and suppress this side reaction to occur or the water-based coating process led to the oxidation of those low valence state of transition metal. Although the initial capacity is higher in the pristine sample, it failed to be maintained in subsequent two cycles, the reversible capacity was decayed 17.22% to 138.53 mAh/g in the third cycle, and for the modified cells, the dropped percentage in discharge capacity are smaller than 10%. The capacity fading might be associated with the metal ion dissolution, the reduction of manganese to trivalent sate was reported that would further lead to the disproportional reaction according to 2Mn(III)Mn(IV) + Mn(II) at the interface between the cathode and electrolyte[91, 133], whereas Mn(II) is electrolyte soluble, hence the dissolution of manganese was believed to come up.
doi:10.6342/NTU201601970
95
The AC impedance was carried out in 4.0 V at room temperature after the formation cycles (Fig. 5-10). As mentioned in chapter 4, the cell resistance could be divided into three kinds of parts, they are the bulk electrolyte resistance (Re), surface film (SEI) resistance (Rs) and the charge transfer resistance (Rct) respectively from the high to the low frequency. From the impedance plot, it is obvious that the NCM-2%-PSSNa and 2%-PDDA-NCM-2%-PSSNa sample demonstrate a higher value in total resistance, where the major contribution comes from the Rct. The higher Rct in these two samples is because of the electronically insulated polymer, which blocked the electron conduction of the NCM. However, this problem can be easily solved by introducing the extra conductive agent such as the sample Super P-PSS @NCM-PDDA. For this sample, only 1% super P addition could considerably lower the Rct. To study the SEI growth after formation cycles, the Rs is an indicator should be discussed. As we can see,the relatively lower surface resistance, shown in the inset plot of Fig. 5-10, can be observed in all the modified samples, it can be attributed to the coated polymer that effectively inhibits the electrolyte decomposition during the low rate formation cycles, the growing of SEI thus could be suppressed.
After formation cycles, the rate test was carried out in 55 oC with the various rate from 0.5 to 10 C in the voltage window of 2.5－4.4 V. The result is displayed in Fig. 5-11 (voltage profiles of the rate performance is shown in Fig A-2). In the rate performance, the relatively higher discharge capacity, this time in the modified cells rather the pristine during the test at elevated temperature is associated with more thermal stability of the active material because of the polymer coatings, which stabilized the interface of the material. For the NCM-2%-PSSNa cell, the single polymer-coated sample, the promoted rate capability is obtained. It could attribute to the lithium transportability anionic sulfonated group on the PSSNa that could also
doi:10.6342/NTU201601970
96
accelerate the de-solvated and solvated process for Li ion. The Super P-PSS @NCM-PDDA owing to its electron conductivity and polymer coating, it demonstrated the highest specific capacity among the cells at 0.5 C, while the capacity retention could not be maintained as good as the single polymer-coated sample once the rate was increased. The similar poor retention was demonstrated in the NCM-2%-PDDA-PSSNa sample as well. Hence, the lower capacity retention in Super P-PSS @NCM-PDDA is inferred to associate with the agglomeration of the super P and the entanglement of polymer PDDA and PSSNa, which constructed a barrier for the lithium ion diffusion.
Cycling test was also conducted at 55 oC under the current density of 0.3 C in the voltage window of 2.5－4.4 V for 100 cycles after rate test as shown in Fig. 5-12 and Table 5-5. It shows that the modified cell could exhibit the significant improvement in the capacity retention > 86 %, while the pristine cell is only left 78.08 % after 100 cycles. The normalized voltage profiles of cycling performance shown in Fig. 5-13 are plotted in every 50 cycles. The apparent curves change in the pristine cell along with the arrows (Fig. 5-13 (a)) indicating the serious voltage polarization and capacity fading.
Among the modified cells, the 2%-PDDA-PSSNa and Super P-PSS @NCM-PDDA cells have the similar tendency in the capacity retention, however, the later demonstrates the relatively higher specific capacity owing to its higher electron conductivity. For the NCM-2%-PSSNa sample, even though its retention behavior is lower than the other modified samples in the earlier stage, the gradually revived tendency in capacity retention in this sample can be observed with the advanced cycles, and it is expected to have the best performance in the extended cycles.
After cycling at 55 oC for 100 cycles, the AC impedance was again employed for measuring the cell resistance (Fig. 5-14). Instead of the higher total resistance, the modified cells show the lower total resistance than the pristine, especially in the
2%-doi:10.6342/NTU201601970
97
PSSNa coated sample, this is attributed to the polymer A-SEI that inhibits the nature SEI growing. The above-mentioned phenomenon can be illustrated to the side reaction inhibition between the electrolyte and the active material.
doi:10.6342/NTU201601970
98
Figure 5-8 Charge/discharge voltage profiles of the (a) NCM-Pristine, (b) NCM-2%-PSSNa, (c) NCM-2%-PDDA-NCM-2%-PSSNa, and (d) Super P-PSSNa@NCM-PDDA cells in 0.1 C at room temperature.
Table 5-4 Electrochemical performance of NCMcells in formation cycles. (the first batch of NCM) NCM-Pristine 167.35 80.93 145.88 138.53
NCM-2%-PSSNa 142.35 81.76 136.76 133.82
NCM-2%-PDDA-PSSNa 145.83 81.02 137.78 133.61 Super
doi:10.6342/NTU201601970
99
Figure 5-9 dQ/dV plots of the (a) Pristine, (b) 2%-PSSNa, (c) NCM-2%-PDDA-PSSNa, and (d) Super P-PSSNa@NCM-PDDA cells.
Figure 5-10 Nyquist plot of NCM cells performed at 4.0 V in the fifth charge cycle at room temperature. (the first batch of NCM)
0 100 200 300 400 500 600 700
NCM-1%-Super P-PSSNa @NCM-PDDA-1stcycle NCM-1%-Super P-PSSNa @NCM-PDDA-2ndcycle
2.7 3.0 3.3
doi:10.6342/NTU201601970
100
Figure 5-11 Rate performance of the different NCM samples in (a) discharge capacity; (b) capacity retentionto the cycle number at various rates from 0.1 C to 10 C (1 C = 160 mA/g) at 55 oC (the first batch of NCM).
doi:10.6342/NTU201601970
101
Figure 5-12 Cycling performance showing (a) Specific discharge capacity; (b) capacity retention versus the cycle number of different NCM samples charge/discharge in 0.3C (1C=160 mAh/g) at 55 oC. (the first batch of NCM)
10 20 30 40 50 60 70 80 90 100
doi:10.6342/NTU201601970
102
Table 5-5 Cycling performance of NCM cells at 55 oC. (the first batch of NCM)
Cell
Figure 5-13 Normalized voltage profiles of (a) NCM-Pristine, (b) NCM-2%-PSSNa, (c) NCM-2%-PDDA-PSSNa, and (d) Super P-PSSNa@NCM-PDDA cells in every 50 cycles of cycling test.
0.0 0.5 1.0
doi:10.6342/NTU201601970
103
Figure 5-14 Nyquist plot of NCM cells performed at 4.0 V after 100 cycles at 55 oC 0 100 200 300 400 500 600 700 800 900 1000
0 100 200 300 400 500 600 700 800 900
NCM-Pristine NCM-2%-PSSNa NCM-2%-PDDA-PSSNa
1%-Super P-PSSNa @NCM-PDDA
-Z ''/Ohm
Z'/Ohm
0 5 10 15 20 25 30 35 40 45 0
4 8 12 16
-Z' '/ Ohm
Z'/Ohm
R
ctR
sdoi:10.6342/NTU201601970
104
5-2-3 Morphology and Surface Structure Evolution of Electrode
The SEI formation on the electrodes was observed by employing the SEM after the electrochemical cycling at high temperature. Fig. 5-15 shows the SEM images of the fresh and cycled electrodes. For the fresh electrodes, the NCM surface is bald and clean in all kinds of electrodes, the small particles on the Super P-PSSNa@NCM-PDDA surface are regarded to the Super P attachments. It is apparent to see that after cycling, the thicker and non-uniform SEI components were deposited on the NCM pristine particles, while the relative thinner SEI layer was seen on the modified electrodes. The results could give the evidence to the seriously interfacial side reaction occurring on the non-treated NCM surface.
The EDS analysis on the cycling electrodes was carried out to quantify the degree of electrolyte decomposition. For the decomposition process, it can be generally described as followings reaction:
1. LiPF6=LiF + PF5 (5-1) 2. PF5 + H2O =POF3 + 2HF (5-2) 3. POF3 + 3Li2O =6LiF + P2O5 (5-3)
As we can see, the fluorides and phosphides are the final products of the electrolyte decomposition. Therefore, the content of the fluorine and the phosphorus in the EDS result can be regarded as the index to the electrolyte decomposition, which would further induce the SEI formation. The EDS result is displayed in Fig. 5-16. For the fresh electrodes, there is no phosphorus content since no components except the electrolyte having the phosphorus in the compounds, while about 14 wt. % of fluorine appearance in the fresh electrodes comes from the binder PVDF, so that every electrode have similar content in fluorine element in the first beginning. During the electrochemical
doi:10.6342/NTU201601970
105
operation, both the content of these two elements are increased with the increased cycles, representing the evolution of the electrolyte decomposition was continued during the cycling. After 100 cycls, the rapidly expanded content in the F and P elements can be observed in the NCM-pristine electrode, which is consistent with the result of SEM observation. The content of the F and P element is found to be dependent on each other, the arrangement of the total content from high to low in the NCM electrodes are NCM-Pristine, NCM-2%-PDDA-PSSNa, Super P-PSSNa@NCM-PDDA and NCM-2%-PSSNa respectively, this result is in accord with the total resistance in the impedance analysis (Fig. 5-14), that is, the SEI growth is one of the factors for causing the high resistance to the cells.
As mentioned in earlier, for the kinds of lithium layered transition metal materials, the correlation between the capacity fading and metal dissolution is strongly relative to the corrosion of HF. In addition, the cation disorder also occurs during the electrochemical operation. The dissolution and the disorder of the transition metals would lead to the structure instability and phase transformation. In order to understand the structure evolution of the materials, the surface structure of the cycled materials was examined by the high-resolution STEM. Fig. 5-17 shows the STEM images of the NCM-Pristine and NCM-2%-PSSNa particles after cycling at 55 oC. The image indicates that there are obviously different surface structures appeared between these two particles Fig. 5-17 (a) and (b). The phase identification of the structures was shown in the selected regions of the fast Fourier transformed (FFT) patterns in Fig. 5-17 (c)~(f).
For the NCM-2%-PSSNa particles, the crystal structure approximate to 2 nm from the
For the NCM-2%-PSSNa particles, the crystal structure approximate to 2 nm from the