For cell performance comparison, Ketjen Black EC-600JD (KB, Mitsubishi Chemical, Japan), which has the highest capacity among many commercial carbon materials, was chosen. To fabricate the air cathode, 0.2g of CNBs and Ketjen black EC600JD was added to a mixture of 0.12 g of polyvinylidene fluoride (PVDF) and 10g of N-methyl pyrrolydone (NMP). Circular disks were prepared from sheets of high-porosity copper foam (1.4 mm thickness, 6000 g of N-methyl pyrrolidone (NMP)). wing Ar atmoMP/PVDF/Carbon slurry. The copper disc was sonicated for 20 min to enhance slurry penetration into the copper foam structure. To remove NMP, the cathodes were heated in a dryer at 110 odes were heated in into the copper foam structure. To rch carbon material was loaded on the copper foam, and the weight ratio to PVDF was 60 to 40. The electrolyte was 1 M LiClO₄ in an ethylene carbonate/ propylene carbonate (EC/PC=1:1) solution. Lithium metal and nonwoven polypropylene fabric were used as the anode and separator, respectively. Type 2032 coin cell kits (Miclab, Japan) were used to fabricate cell devices, and the cap of the coin cells were drilled with twenty 1 mm pinholes. The Li-air cell was assembled in an argon-filled glove box. Before discharging, the air-cathode was soaked in the electrolyte for 5 min to permit permeation of the electrolyte into the carbon materials. The cells were discharged at a rate of 0.1 mA/cm² and 2 V in an ambient environment.

In order to understand the chemical reaction by the radicals in benzene solution, a detailed liquid analysis was carried out by GC air electplasma process at time intervals of 2 s, 40 s, 60 s, and 3 min. All the intermediates that were identified by GC-MS are listed in Table 1. The nucleation process, where carbon atoms aggregate to produce larger spheres, is unclear. Kroto and Mckay proposed that nucleation begins from a pentagonal carbon ring, followed by spiral shell growth (Figure 2). The nucleation process can be considered to involve four stages: (a) nucleation of a pentagon, (b) growth of a quasi-icosahedral shell, (c) formation of a spiral shell carbon particle, and (d) growth of a large carbon sphere [4]. It is proposed that CNB formation occurs in a similar manner. In order to confirm the growth of a quasi-icosahedral shell (pentagonal, hexagonal, and heptagonal carbon rings) during the discharge, a GC method was used, and replicate samples were obtained after 5 s and 10 min of discharge. Figure 3 shows the gas chromatogram of a discharged solution and Table 1 shows the chemical identification of the peaks (by mass spectrometry) in the chromatogram displayed in Figure 3. Table 1 shows that benzene is found in the solution during the initial stage.

We now consider how the pores of CNB can be converted into mesopores with sizes of more than 10 nm. As shown in Figure 4, the porous properties of carbon black are governed by carbon aggregates and their agglomeration. Generally, the properties of carbon black aggregates can be understood by using the diffusion limited cluster aggregation (DLCA) model. Diffusion-limited aggregation (DLA), as proposed by Witten and Sander, is the process whereby particles go through a random walk due to Brownian motion, form clusters, and finally form aggregates of such particles. The clusters formed in the DLA process are referred to as Brownian trees. The Brownian trees generated from conventional plasma methods have numerous sprigs because the plasma simultaneously generates a large number of particles. Thus, there is probability of formation of mesopores smaller than 10 nm with a ramulose structure and a high degree of chaining (high structure in Figure 3). However, when using solution plasma, the plasma zone is very small and the Brownian trees formed are quickly, circulated by flux of solution. In addition, the void volume between the particles of the aggregates (intra aggregate voids) and that between individual aggregates is filled with liquid as soon as the Brownian trees leave the plasma zone. Therefore, the structures of the Brownian trees are simple, (i.e., the low structure in Figure 3) because of the insufficient growth time. For this reason, CNBs produced by the solution plasma method are mesopores.

Figure 2: 
Formation mechanism

Figure 3: 
Chromatograms of the discharged solution by SPP with structural formula

Voltage-capacity discharge experiments were carried out to evaluate the electrochemical performance of CNBs and were compared to KB performance (Figure 4). The performance of Li/air batteries are dependent on several factors, including carbon source, carbon loading, and electrolyte choice, and electrolyte volume [5][6]. Therefore, the experiments were carried out under identical conditions, as mentioned in the experimental section, excluding the carbon source.

The discharge capacity significantly increased with increasing annealing temperature because the number of electro-chemical sites that can react with oxygen increased by the improved conductivity. By comparison, the discharge profile in Figure 5 shows that KB-based micro-mesopores (less than 10 nm in size) having a large specific surface area, high pore volume, and high conductivity do not give a higher capacity battery than do the CNBs. These findings show that the critical factor for the high capacity is a good mesopore network (more than 10 nm) along with a large pathway for oxygen diffusion and a large pore diameter to prevent clogging of the pore entrances and afford the space to accommodate the deposition of Li oxide during discharge. Therefore, the structure of carbon in the air electrode is a critical factor that affects the performance of Li-air batteries. These features strongly imply that the CNBs have a suitable mesopore diameter with a large pore volume for the Li-air battery or fuel cell [7].

Figure 4: 
Structure growth of carbon black

Figure 5: 
Discharge curves of the Li-air cell