Articles
Reduction of Electrical Resistance of Active Materials for Lithium-ion Secondary Batteries by Making Contact with Carbon Materials
2022 Volume 90 Issue 8 Pages 087006
Details
2022 Volume 90 Issue 8 Pages 087006
In the active material of a lithium-ion secondary battery, electrons and lithium-ions react at the same time. The fact that the active material can transfer electrons with a small reaction overvoltage leads to a reduction in the internal resistance of the battery. Therefore, when the electrical resistance was measured by changing the electrode material in contact with the active material, it was found that the electrical resistance of the active material was reduced depend on the carbon materials make contact with the active material.
Lithium manganese oxide has been used as the active material for lithium-ion secondary batteries, and the reported conductivity of lithium manganese oxide has a range of double digits. Although the measurement method is not described, the conductivities of lithium manganese oxide are 10−5, 10−6 S/cm.1,2 The conductivities measured by the two-probe AC method are 3.2 × 10−5, 2.84 × 10−5 S/cm for lithium manganese oxide prepared by solid state reaction3,4 and 1.84 × 10−4 S/cm for lithium manganese oxide nanorods.4 The conductivities measured by the four-probe method are 1.21 × 10−5, 5.10 × 10−5 S/cm.5 The conductivities measured using the thin-film electrode and electrolyte are 1.1211 × 10−5 to 5.3922 × 10−5 S/cm.6 As described above, there are various methods for measuring conductivity. In view of this, one of the factors that change the electrical properties of lithium manganese oxide is considered to be the material that comes into contact as an electrode.
A carbon material is blended in the positive electrode mixture of the lithium-ion secondary battery. The carbon material is used as a conductive additive or a current collector coating material, and is well known as a method for strongly reducing resistance.7–9 Thus, the active material is in contact with the carbon material in the batteries. The authors applied electrodes by strongly hammering active material into metal foil without using conductive additive or binders in order to elucidate the characteristics of only the active material in the lithium-ion secondary battery.10 However, this method cannot evaluate the pure contact resistance between the electrode material and the active material because this method deforms the metal foil and destroys the coating layer such as the carbon coated foil. Therefore, we tried to evaluate the contact resistance by suppressing the contact pressure within the range that does not deform the metal foil.
The aim of this study is to confirm the influence of electrical resistance of the active material changes when the active material makes contact with the carbon materials or other materials.
The samples of the lithium manganese oxide, lithium iron phosphate, and lithium nickel manganese cobalt oxide used were fresh samples without any charging/discharging reaction. Therefore, it should not be called as active materials. However, the term should not be increased unnecessarily because this paper is greatly related to the reaction overvoltage in the lithium-ion secondary battery, so we chosed to describe them as active material.
The lithium manganese oxide (LiMn2O4, Taiyo Nissan Corp., test sample was manufactured by spray pyrolysis method), the lithium iron phosphate (LiFePO4, Aleees, M121), and the lithium nickel manganese cobalt oxide (LiCo1/3Ni1/3Mn1/3O2) were used as test samples.
The electrical resistance of the active material was measured by the two-probe DC method.
Gold foil (Niraco Corp., purity = 99.95 %, thickness = 0.02 mm), the carbon coated gold foil (home made), the carbon nanotube sheet (Taiyo Nissan Corp., Test sample: Multi-walled nanotubes), carbon-coated aluminum foil (Showa Denko K.K., SDX®), and aluminum foil (Japan Capacitor Industrial Co., Ltd., A1085 equivalent) were prepared as electrode materials. The carbon coated gold foil were made by dropping the carbon dispersion (provided by Mikuni-Color Ltd., acetylene black slurry) on the gold foil and dried for 1 hours at 40 °C.
Figure 1 shows a schematic diagram of the conductivity cell for contacting the electrode material with the active material. The active material (Fig. 1a) was sandwiched between both electrode materials (Fig. 1b) and soft-contacted at near zero pressure. A push pin (Fig. 1c) (φ11) equipped with a flat rivet head was used for the terminal that comes into contact with the electrode material. A cellophane tape (Fig. 1d) (Nichiban Cellotape™, 24 mm) was used as a spacer. The distance d between the two electrodes was adjusted to 0.35 ± 0.1 mm according to the thickness of the spacer. A hole with a diameter of 6 mm was made in the spacer. The spacer was placed on the electrode material and the holes were filled with active material. Both pushpins were sandwiched between hand clamps (Fig. 1e) (Komeri Co., Ltd., JAN 4920501652).
Schematic diagram of the conductivity cell for contacting the electrode material with the active material. a: Active materials were placed in 6 mm diameter hole between electrodes, b: Electrode to make contact with the active material for collecting current, c: Pushpin as terminals, d: Cellophane tape laminated as spacer, and e: Hand clamp arranged to hold the conductivity cell. Cell constant is ca. 12 m−1.
A potentiostat (Hokuto Denko Corp., HAB-151) was connected to the terminal, and the potential was swept from −5 V to +5 V at a sweep speed of 100 mV/s. The current at that time was measured and a current-voltage curve was obtained. The electric resistance R was obtained from the slope of the current-voltage curve at the voltage V of 0 V. Further, the electric resistance R was set to 1 when the electrode material was gold, and the ratio to the electric resistance R, when another electrode material was used, was obtained as the normalized resistance RN.
Figure 2 shows the current-voltage curves of lithium manganese oxide when the electrode material is changed. Only the aluminum showed non-linear response at around 2 V, and the electrical resistance R, which was obtained from the slope at 0 V, was very large at 7000 kΩ. The others showed linear response.
The electrical resistance R for the others were calculated similarly, and the results were summarized in Table 1. The electrical resistance R were 200 kΩ for gold, 20 kΩ for carbon nanotube sheet, carbon coated gold foil, and carbon-coated aluminum foil. It is worthy to note that the electrical resistance changed drastically as a function of the electrode material even for the same active material. Generally, gold is used as an electrode material for resistance measurement since gold has no oxide film on its surface and results in good contact. It was convenient to introduce the normalized resistance RN as a dimensionless value by dividing individual resistance by the electrical resistance for gold to eliminate the effect of cell dimensions. Thus, measured electrical resistances were transformed to the normalized resistance RN and also shown in Table 1.
| Electrode materials | Mode of R /kΩ |
Range of R /kΩ |
RN |
|---|---|---|---|
| Gold foil | 200 | 150–300 | 1 |
| Carbon nanotube sheet | 20 | 5–45 | 0.1 |
| Carbon coated gold foil | 20 | 10–30 | 0.1 |
| Carbon-coated aluminum | 20 | 15–25 | 0.1 |
| Aluminum | 7000 | 5000–8000 | 35 |
The large RN value of 35 for aluminum was easy to understand since aluminum has surface oxide film as a passivation film. However, it was somewhat strange for the small RN value of 0.1 for carbon related electrodes and we called this phenomena “carbon material effect” in this paper. Generally, gold is used as an electrode material for resistance measurement since gold has no oxide film on its surface and results in good contact. When carbon nanotubes were used as the electrode material, the electrical resistance increased with increasing the distance d between both electrodes. When gold foil was used as the electrode material, the electrical resistance tended to decrease with increasing the distance d between the both electrodes. If we accept this standard concept, smaller RN values than 1 should be attributed to the electrical resistance on the active material side.
To elucidate these phenomena, we measured RN as a function of active materials for carbon material effect, and the results are shown in Table 2. The resultant RN values were 0.8 for lithium iron phosphate and 0.5 for lithium nickel manganese cobalt oxide. Thus, all active materials used in this study showed smaller RN values than 1. Lithium iron phosphate has a carbon coat layer on its surface to prevent oxidation and to stabilize the surface, it still had a smaller RN of 0.8 than 1 even for the gold electrode. Thus, it was concluded that the normalized resistance RN became smaller than 1 when the carbon material was in contacted on the surface of active materials.
| Active material | Mode of RN |
Range of RN |
|---|---|---|
| Lithium manganese oxide | 0.1 | 0.05–0.2 |
| Lithium nickel manganese cobalt oxide | 0.5 | 0.4–0.8 |
| Lithium iron phosphate (carbon coated) | 0.8 | 0.7–1.1 |
To understand these phenomena, we introduced a depletion layer to the side of the electrode or the active materials based on some assumption as shown in Fig. 3. In general, electrical resistance is proportional to the distance between electrodes. When a carbon material is used as the electrode material, the measured electrical resistance is considered to be dominated by the bulk resistance of lithium manganese oxide, because the electrical resistance of lithium manganese oxide was proportional to the distance between both electrodes. When gold foil was used as the electrode material, the electrical resistance was higher than that of carbon material since the measured electrical resistance was dominated by the interface between the gold foil and lithium manganese oxide. However, a gold foil has no oxide film, the contact resistance could be attributed to the electrical resistance on the lithium manganese oxide side of the interface. Moreover, when gold foil is used as the electrode material, the electrical resistance of lithium manganese oxide tends to decrease as the distance between the electrodes increases. Therefore, when the electric field on the lithium manganese oxide side of the interface decreases, the electrical resistance decreases. This can systematically explain the measured electrical resistance, assuming that a depletion layer is formed on the side of lithium manganese oxide at the interface, and the thickness of the depletion layer increasing with increasing the electric field.
Idea drawing of the depletion layer generated from the electrode material making contact with the active material.
Figure 3A is the case for the aluminum electrode. Aluminum is used as the positive electrode current collector for lithium-ion secondary batteries.11 Aluminum has an oxide film on its surface. The oxide film has an oxygen deficiency and is considered to be an n-type semiconductor of non-stoichiometric compound. On the other hand, an active material such as lithium manganese oxide is supposed to be locally negatively charged because its surface is an oxide ion. When an electric field is generated in the aluminum oxide due to the negative charge, a depletion layer of electrons is formed on the aluminum oxide side, it is thought that it is a large contact resistance.12 The electrical resistance R of lithium manganese oxide contacted with the aluminum obtained from Fig. 2 was mostly due to the large contact resistance of the aluminum surface.
Figure 3B is the case for gold electrode. Gold is free of oxides on its surface,10 so the growth of a depletion layer at the electrode side is impossible. However, for the active material side, the bulk of lithium manganese oxide assumes a non-metal, its surface has metal defects, and this present the lithium manganese oxide as a p-type semiconductor of non-stoichiometric compound. Generally, in the contact between the metal and the p-type semiconductor, if the work function of the semiconductor is large, the semiconductor forms a schottky junction with the metal and forms a depletion layer near the semiconductor. For such a case, an electric field is generated inside the lithium manganese oxide, and this introduced a depletion layer of holes near the surface of lithium manganese oxide. It seemed that this was an origin of contact resistance for gold. For this case, increasing the distance d between the electrodes decreases the electric field, the thickness of depletion layer becomes thin, and thus the contact resistance tended to decrease. In the solution, it is thought that there is no significant difference in the reaction resistance between the gold and carbon materials because the electrolyte solution penetrates to the surface, so the electric field generated is very small and such a depletion layer does not exist. The electrical resistance R of lithium manganese oxide making contact with gold is attributed to the large contact resistance on the surface of lithium manganese oxide.
Figure 3C is the case for carbon materials. In this paper, the non-metal indicates that the density of states is 0 even though it had the no bandgap at wavenumber of 0.13 Carbon materials are well known as such non-metals. For the case of carbon electrode material, a depletion layer is not formed on both sides of the carbon electrode and the active materials because holes are injected into the lithium manganese oxide even if the carbon material is positively polarized. For this case, the electrical resistance R should be proportional to the distance d between the electrodes. Thus, the electrical resistance R is considered to be mostly due to the resistance of the lithium manganese oxide bulk, and the accurate conductivity of lithium manganese oxide can be estimated. Only in this case, the conductivity of lithium manganese oxide can be estimated. Ignoring the effects of porosity and pressure, the apparent conductivity of lithium manganese oxide was 10−7 S/cm calculated from the electrical resistance R, the distance d between the electrodes, and the electrode area.
According to these considerations, it was possible to explain the reason why there is no significant difference in the reaction resistance between the gold and carbon materials in the electrolyte solution since the electrolyte solution penetrates to the contact surface, and this makes the electric field very small due to prevention of the growth of the depletion layer. The widely dispersed reported conductivity of lithium manganese oxide is due not only to the contact state of powder particles between the active material, but also to the contact of the electrode material with the active material.
The electrochemical reaction equation for lithium manganese oxide is generally written as Eq. 1.
| \begin{equation} \text{LiMn$_{2}$O$_{4}$}\rightleftarrows \text{Li$_{(1-x)}$Mn$_{2}$O$_{4}$} + \text{$x$Li$^{+}$} + \text{$x$e$^{-}$} \end{equation} | (1) |
| \begin{equation} \text{LiMn$_{2}$O$_{4}$} + \text{$x$h$^{+}$}\rightleftarrows \text{Li$_{(1-x)}$Mn$_{2}$O$_{4}$} + \text{$x$Li$^{+}$} \end{equation} | (2) |
The electrochemical reaction of the active material is considered to be a charge transfer between the delocalized π orbitals of the carbon material and the delocalized eg orbitals of lithium manganese oxide. Therefore, the contact between the carbon material and lithium manganese oxide in the active material mixture is hindered by the dispersant or binder, and if a depletion layer is formed on the surface of lithium manganese oxide, the reaction overvoltage may increase. The addition of long carbon nanotubes is effective in reducing internal resistance, and the lithium iron phosphate is coated by using the carbon. These are considered to reduce the reaction overvoltage by facilitating a charge transfer between the delocalized π orbitals of the carbon material and the delocalized eg orbitals of active materials.
The electrical resistance R of the active material varies depending on the material making contact with an active material. The electrical resistance R of the active material making contact with carbon material is reduced as compared with aluminum or gold.
The authors are grateful to S. Higuchi, A. Kikuchi and S. Fujita for comments on this paper. This work is supported and funded by Taiyo Nippon Sanso Corp. and Mikuni-Color Ltd. Also, this work is supported by Japan Capacitor Industrial Co., Ltd., the Showa Denko K. K., Advanced Lithium Electrochemistry (Cayman) Co., Ltd. (Aleees) and Marubeni Corp.
Tomohiro Ito: Data curation (Equal), Investigation (Lead), Project administration (Supporting), Resources (Equal), Supervision (Supporting), Visualization (Equal), Writing – original draft (Lead)
Kazuhiro Tachibana: Conceptualization (Lead), Data curation (Supporting), Funding acquisition (Lead), Project administration (Lead), Resources (Lead), Supervision (Lead), Visualization (Lead), Writing – original draft (Equal)
Yoshihisa Yamamoto: Data curation (Lead), Investigation (Lead), Visualization (Equal)
Tatsuo Nishina: Validation (Lead), Writing – review & editing (Lead)
The authors declare no conflict of interest in the manuscript.
T. Ito, K. Tachibana, and T. Nishina: ECSJ Active Members
Y. Yamamoto: ECSJ Student Member
