Tao Liuab,
Shimei Suna,
Zhao Zangc,
Xichao Lia,
Xiaolin Suna,
Fengting Caoa and
Jianfei Wu*a
aQingdao Institute of Bioenergy and Bioprocess Technology, Chinese Academy of Sciences, Qingdao, 266042, PR China. E-mail: wujf@qibebt.ac.cn
bUniversity of Chinese Academy of Sciences, Beijing, 100049, PR China
cCollege of Mechanical and Electrical Engineering in China University of Petroleum, Qingdao, 266580, PR China
First published on 12th April 2017
This paper aims to demonstrate whether graphene nanosheets (GN) with different sizes as conductive additives are able to affect the electrochemical performance of a LiFePO4 (LFP) cathode. The results of electrochemical measurements present that graphene nanosheets (GN) and Super-P (SP) used as conductive additives simultaneously could construct an effective electronic conducting network and achieve excellent electrochemical performance when compared with traditional carbon materials. A LFP with small-size GN shows better specific capacity and rate performance than those with medium-size and large-size GN, and the LFP with large-size GN displays a poor rate performance. The results indicate that the specific capacity and rate performance tend to worsen with the increase of the size of graphene, owing to the length of the lithium ion transport path being prolonged and the ionic conductivity decreasing greatly by the “barrier effect” of graphene.
Graphene is a type of novel and powerful planar conductive additive, and have been considered as one of promising conductive additives for both positive and negative electrodes in lithium ion batteries to partly or even completely replace the existing carbon black (CB), due to its unique physical and chemical properties, high aspect ratio, chemical tolerance, excellent electrical conductivity and effective conducting network at even a trace amount.15,16 Therefore, the electronic conductivity of electrode can be greatly improved with much less amount of graphene as conductive additive. Su Fang-Yuan and co-workers3 directly added graphene into active materials LiFePO4 to constitute an efficient GN-based conducting network with only 2% mass fraction, displaying better electrochemical performance than the system with 10% commercial carbon additives, which demonstrated that the “plane-to-point” mode of GN-based additives exhibited more excellent electronic conducting properties and flexible conducting network than the “point-to-point” mode of conventional carbon additives. Tang Yu-feng and co-workers17 used 5% 3D graphene instead of acetylene black as conductive additives. The experimental results suggested that the 3D graphene could form an ideal conducting network for LiFePO4 to provide a significant performance boost on capacity and rate performance, demonstrating graphene have potential application in high rate lithium ion batteries. Wei Wei and co-workers18 found a partial graphene wrapping could be able to maintain a good balance between improving electron transport and lithium ion diffusion, while the opposite result occurred in full graphene wrapping, which is due to redundant graphene isolates LiFePO4 from the electrolyte and hinders lithium ion diffusion.
Unfortunately, one serious challenge in using graphene as conductive material in lithium ion batteries is that the planar structure of graphene shows a steric hindrance effect for lithium ion diffusion,10 as well as difficulties in homogenous dispersion, having become the most predominant factors to hinder the application of graphene in lithium ion batteries which have been well accepted by many researchers.3,10,16 It is widely accepted that graphene could constitute an efficient conducting network but also hinder lithium ion transport channel since lithium ion diffusion only occurs through several defect sites. Although lithium ion diffusion through an edge plane-enriched graphitic flakes can be easily facilitated but further complicated by the presence of functional groups.19 Meanwhile, uniform dispersion in LFP is difficult to achieve. Therefore, it is a difficult but highly significant task to take full advantage of the conductivity of graphene as well as to reduce or evenly avoid hindering the diffusion of lithium ion. Therefore, it is urgent to evaluate and investigate the effects of different types of graphene on electrochemical performance of LFP battery, finding more effective and suitable types of graphene to improve the electrochemical properties of LFP. With best of our knowledge, there is no report regarding the influence of different sizes of graphene on the electrochemical performance of lithium ion batteries. The present paper is trying to demonstrate whether graphene with different sizes be able to affect the electrochemical performance of LiFePO4, and promote a final commercial application of GN additive in high-performance LIB. In this work, GN and SP are used as conductive additives simultaneously to reduce cost and alleviate the agglomeration of GN.
Fig. 1 SEM images of LFP with (a) 10% SP (×50.0k); (b) 9% SP and 1% GN1 (×50.0k); (c) 9% SP and 1% GN2 (×50.0k); (d) 9% SP and 1% GN3 (×28.0k). |
We have investigated the effects of various GN fractions on the discharge plateau performance and specific capacity of LFP/GN to optimize the component of different conductive additives. As shown in Fig. 3, apparent and great enhancement of the charge/discharge profile are found for the three LFP/GN system at 0.1C rate, as compared to the cases of the LiFePO4 with only SP. We can also find a very sound difference between the three cases with different sizes of GN (Fig. 3d), which fell out as we had anticipated. It can be drawn from Fig. 3a and b that the performance of the LFP/GN1 and LFP/GN2 electrodes become better with the increase of the GN fraction from 0.5 to 1.0 wt%, and with the specific capacity of LFP with 1.0 wt% GN1 and LFP with 1.0 wt% GN2 mostly maintained about 165 mA h g−1 and 163 mA h g−1 respectively. The LFP with 10 wt% SP for the reference had lower specific capacity with about 155 mA h g−1. This is mainly up to the 0.5 wt% GN1 and GN2 fraction are too low in the cathode materials to take full advantage of the “plane-to-point” mode of GN-based additives, which is more advantage than the “point-to-point” mode of SP-based additives to establish an effective conducting network. While, the opposite result occurred when a further increase is employed from 1.0 to 3.0 wt%, owing to GN would more tend to agglomerate together into thickly aggregated structures and be difficult to disperse well in the active materials,3 which greatly increase the probability of hindrance of lithium ion transport and result in a heavy polarization. However, the specific capacity of LFP with 1.0 wt% GN3 was about 161 mA h g−1, slightly worse than that of LFP with 0.5 wt% GN3 (162.5 mA h g−1). Even, when GN3 fractions increased to 3%, the capacity was 148 mA h g−1. Which is considered to prolong more times of the distance for lithium ion diffusion than that with GN1 and GN2, leading to more heavy polarization and longer distance for lithium ion diffusion with more GN3 fractions. The results once again proved that the large-size GN is much easier to hinder lithium ion transport for nanometer grade cathode materials. Based on this point of view, small-size GN may be more suitable for nanometer grade cathode materials. Generally, the LFP electrode with 1.0 wt% GN1, 1.0 wt% GN2 and 0.5 wt% GN3 show the better charge/discharge performance among their respective tested cases. From the preliminary results, we can arrive at a preliminary conclusion that the optimal content of GN varies according to the size of GN, that LFP with small-size GN shows better electrochemical performance than that with large-size GN at 0.1C rate under this experimental condition. Which on account that the large-size GN would more tend to prolong the lithium ion transport path and increase polarization in the active materials.
Further studies on the relationship between electrochemical properties of LiFePO4 and different size GN were done in this paper. We have investigated the effect of the same content of LFP/GN with different sizes of GN on the rate performance comprehensively. Fig. 4 presented the rate performances of the LFP/GN with different sizes of GN, LiFePO4 with 10 wt% SP for the reference. It can be seen that the rate performance of the LFP/GN electrode becomes worse with the increase of GN sizes. Regarding each specific capacity at 0.1C as 100%, the rate performance of the electrode with those three different sizes of graphene improve significantly and have comparable capacity at 0.1–1.0C rate. However, a sudden capacity loss at 2C, 5C discharge rate appeared for LFP/GN3, while that obvious loss at high discharge rate did not appear for LFP/GN1 and LFP/GN2 case, and the specific capacity at 5C of LFP/GN1 and LFP/GN2 electrode are 112 mA h g−1 and 104 mA h g−1, and capacity retention of 5C compared to 0.1C are 67.9% and 63.8%, superior to LFP/GN3 79 mA h g−1 and LFP/SP 85 mA h g−1 (48.5% and 54.8%, respectively). The results suggest that the rate performance of the electrode become worse with the increase of the GN size. This may be due to the different electrode structure shown in Fig. 5, which represents possible conducting modes in LiFePO4 cathode based on the SEM images. From the mechanism map we can deduce that the poor rate performance of the LFP/GN3 is caused by poor lithium ion transport. It is easily understood that, with the increase of GN size, the length of ion transport path is prolonged more and the ionic conductivity decreases more greatly, which can be attributed to the fact that lithium ion transport path is more efficient in SP case than in GN case since the size of SP is much smaller than GN.3,16,21 In a complete charge/discharge process, electrons and lithium ion must reach the same electrode material simultaneously. From this view point, two factors should be considered for structuring an optimal conducting network. One is the conductivity, with the increase of the GN fractions, the electronic conductivity of the poorly conductive LiFePO4 improved greatly, which is due to that GN is a one-atom-thick allotrope of carbon, π-electrons are “free” and move more easily, which have great advantages in electronic conductivity over other conductive carbons.22 On the other hand, the “plane-to-point” mode of GN-based additives could bridge the active materials as effectively as possible, and more efficient than the “point-to-point” mode of conventional carbon additives where high fractions and intimately contacted particles are needed to form an effective electron network. Another important factor can be the lithium ion diffusion. As shown in Fig. 6, lithium ion diffusion through basal plane of GN is rather limited. Lithium diffusion may only occur through some defect sites and grain boundaries,19 which would block the most efficient and shortest paths for the lithium ion transport and become worse with the increase of the GN size. Therefore, the large-size GN are more likely to prolong the lithium ion diffusion path and amplify the “barrier effect” of GN, which were proved out from the above results (as Fig. 3 shown).
Fig. 4 Comparison of rate performance of LFP with different sizes of GN at different discharge rates. |
Fig. 5 The mechanism map of lithium ion transport paths in different systems utilizing GN with different sizes and SP as conductive additives. |
Fig. 6 Schematics of lithium ion diffusion mechanism on the surface of graphene with different defect population. |
We measured cyclic voltammograms to shed light on the polarization effect of various LFP samples, Fig. 7a show the cyclic voltammograms of the LFP/GN and LFP/SP electrodes (scan rate is 0.1 mV s−1) under ambient temperature. More symmetric and poignant shape of redox peaks of LFP/GN than the LFP/SP cases suggest a higher electrochemical reactivity and lower ohmic resistance of LFP/GN than that of LFP/SP, which is due to the “plane-to-point” mode of GN-based additives bridge active LiFePO4 particles in a more effective and efficient way than the “point-to-point” mode of SP cases. Moreover, bigger interval between the oxidation and reduction peak was observed with the increase of GN sizes, which denoted increasing of polarization with the increase of GN sizes. This may be due to GN is barrier for lithium ion transport to cut off the most efficient transmission path,9,10 the more bigger GN sizes, the longer the length of ion transport path. Which is consistent with the results of our experiment.
Fig. 7 Cyclic voltammograms (a) and EIS spectra (b) of LFP with different sizes GN and LFP/SP as a reference. |
To deeply understand the remarkably enhancement of performance of LFP/GN with different sizes of GN, compared with the original materials without GN respectively, EIS measurement was also employed, the EIS data is simulated through the equivalent circuit in Fig. 7b. The EIS profiles consist of a partially quasi-semicircle at the range from high to middle frequency and a sloping line in the low frequency region. The intercept of the real axis at the high frequency range can be attributed to the ohmic resistance (RO), which refers to the sum of resistance of the electrodes, electrolyte and separator. The semicircle at the high and middle frequency range stands for the charge transfer resistance (Rct). The slope line at low frequency range represents the Warburg impedance, which is attributed to lithium ion diffusion in the electrode.23,24 As shown in Fig. 7b from the intersection of the real axis, it can be drawn that the bulk ohmic resistance (RO) of the LiFePO4 with three different sizes of graphene approximate very much, and much lower than that of the LiFePO4 with 10% SP. Therefore, the significant difference in the bulk ohmic resistance (RO) suggests the superiority of the electronic conductivity of GN over SP as the conductive additive. Graphene nanosheet (GN), together with Super-P (SP) could build a more effective electron transport network in lithium ion batteries. The semicircle of electrode LFP/SP was found to be larger than that of electrode LFP/GN1, LFP/GN2, and slightly smaller than electrode LFP/GN3 from the Nyquist plots. The EIS results are consistent with our above electrochemical measurements, and further confirming that the graphene with different sizes could affect electrochemical performance of LiFePO4. That is, LFP/GN with small size GN1 and GN2 show much better interface contact efficiency due to the more efficient “plane-to-point” mode to “bridge” active LiFePO4 particles sufficiently, resulting in better charge transfer and satisfied electronic conductivity which promotes the electrochemical reaction.3,10 The Rct values increase with the increase of GN sizes, which have been proved above. This is because with the increase of the GN sizes, GN tends to prolong the length of ion transport path and hinder dispersion between the electrolytes and active materials, leading to the ionic conductivity decreases substantially.
This journal is © The Royal Society of Chemistry 2017 |