注册 登录  
 加关注
   显示下一条  |  关闭
温馨提示!由于新浪微博认证机制调整,您的新浪微博帐号绑定已过期,请重新绑定!立即重新绑定新浪微博》  |  关闭

锂空气电池文献

博士4年的经营,看看会创造多少奇迹

 
 
 

日志

 
 

Oxygen reduction reaction catalyst on lithium/air battery discharge performance  

2011-09-23 10:53:09|  分类: 默认分类 |  标签: |举报 |字号 订阅

  下载LOFTER 我的照片书  |
J. Mater. Chem., 2011, 21, 10118-10125 DOI:10.1039/C0JM04170J (Paper) Oxygen reduction reaction catalyst on lithium/air battery discharge performance - 伯虎 - 锂空气电池文献  PDF Version

Xiaoming Ren *, Sheng S. Zhang , Dat T. Tran and Jeffrey Read
U.S. Army Research Laboratory, RDRL-SED-C, Adelphi, MD 20783-1197, USA. E-mail: xiaoming.ren@us.army.mil; Fax: +1 (301) 394-0273; Tel: +1 (301) 394-0379

Received 30th November 2010 , Accepted 25th January 2011
First published on the web 21st February 2011

Lithium/air batteries have the potential to substantially outperform the best battery system nowadays on the market. Oxygen reduction reaction (ORR) at the cathode in an aprotic organic lithium electrolyte is well-known to limit the discharge rate and capacity of the lithium/air batteries. In this study, the discharge characteristics of Li/air cells with cathodes made of different carbon materials were examined. The results showed that the ORR kinetics in the lithium/air batteries can be drastically improved by using an effective catalyst, achieving higher discharge voltage and rate. The discharge capacity of the lithium/air battery was found to be correlated to the cathode pore volume, to which the mesopore volume of the carbon material has a large contribution. An ORR mechanistic model involving a reaction product deactivating the catalytic sites on the carbon surface is proposed to explain the experimental results.


1. Introduction

Lithium/air batteries have the potential of achieving a higher energy density and are safer than the leading commercial lithium batteries, primary or secondary, by utilizing ambient air as the oxidant, thus removing the need to store any oxidant within enclosed cells. A Li/air battery is expected to have a theoretical charge capacity 5–10 times higher than that of a lithium battery,1 which is limited by the amount of lithium that can be reversibly inserted in the intercalation cathode material such as LixCoO2, 0.5 < x < 1. Since the first introduction of a rechargeable Li/air battery by Abraham and Jiang in 1996,2 there has been much progress being made in enhancing and understanding its performance. For practical applications, electrolyte solutions made with non-hydrolytic lithium salts and aprotic organic solvents with low volatility and high electrochemical stability were used.3 Other electrolyte systems were also reported, such as Deng et al.4 for the use of a hydrophobic ionic liquid–silica–PVdF–HFP polymer composite electrolyte membrane to minimize moisture exposure to the lithium anode and Wang and Zhou5 for an aqueous electrolyte at the air cathode and aprotic organic electrolyte at the lithium anode with the two electrolyte solutions separated by a super-ionic glass conductor. In aprotic organic electrolyte, the cell electrode reactions involved in the discharge process are:

Anode: Li → Li+ + e?

Cathode: 2Li+ + O2 + 2e? → Li2O2 (solid)

and the complete cell reaction:

2Li + O2 → Li2O2 (solid), Vcell0 = 2.96 V.

The lithium anode has shown very little polarization at the discharge current density of interest, at less than a few mA cm?2,6 and the loss in cell voltage from the reversible value (Vcell0) occurs largely at the cathode. Many studies have been devoted to address this cathode voltage loss by exploring oxygen reduction reaction (ORR) catalysts for the cathode and by studying the ORR mechanism in aprotic organic electrolyte solutions. Cathode catalysts, such as carbon supported MnOx,7 α-MnO2 nanowires,8 carbon supported nanosized γ-MnOOH (manganite),9 carbon supported Fe2O3, Fe3O4, CuO, CoFe2O4 and Co3O4,10 carbon supported Pt and Au,11 and carbon supported pyrolyzed Co macrocyles,2 have not shown much improvement to the Li/O2 (air) battery discharge voltage over what has been obtained with using carbon only as the ORR catalyst in the cathode. From such observations, it has been assumed that the ORR in a Li–air cathode is not a catalytically sensitive process or that the carbon itself can provide better catalytic activity than those catalysts of interest which themselves are supported on carbon.11 Furthermore, study on the ORR mechanism in aprotic organic lithium electrolyte solution on glassy carbon electrode by Laoire et al.12 using cyclic voltammetry and rotating disc electrode technique showed the initial formation of lithium superoxide (LiO2), which can further be converted to lithium peroxide either by disproportionation reaction (eqn (4)) or by further reduction reaction:

2LiO2 → Li2O2 (solid) + O2

LiO2 + Li+ + e? → Li2O2 (solid)

Computational studies13,14 indicate that LiO2 is likely very unstable at room temperature at less than 1 atm. O2 pressure, favoring the disproportionation to Li2O2, which has been identified as the major reaction product by ex situ examination of the cathode products from discharged Li/O2 cells using Raman spectroscopy2 and by oxygen consumption stoichiometry in the discharge process.3 At a low discharge current density, Zhang et al.6 found part of the initially deposited Li2O2 at the cathode can be further converted to Li2O at a highly polarized voltage below 2 V.

The precipitation of Li2O2 as a solid product on the carbon surface in the cathode further complicates the ORR process in discharging a Li/air cell by either blocking the oxygen diffusion pathways in the pores or by occupying and deactivating the catalytic sites. Zhang et al.6 found the impedance of the air electrode is progressively increased with polarization cycles, indicating the carbon surfaces are gradually covered by the insoluble products, which prevents oxygen from diffusing to the reaction sites on carbon. Mirzaeian and Hall15 found the discharge capacity and discharge voltage of Li/air cell depend on the morphology of carbon, which exerts a combined effect of pore volume, pore size and surface area of carbon on the storage capacity, with carbon with a larger pore volume and a wider pore size preferred. Williford and Zhang16 analyzed several approaches in designing the air electrode by considering the electrode porosity and catalyst reactivity distributions to minimize diffusion limitations and maximize air electrode material utilization. Tran et al.17 proposed a model mechanism based on gas diffusion electrode passivation by the reaction products in blocking small pores and thus preventing them from further utilization, again emphasizing the use of carbon materials possessing high surface area and large pore diameter in the cathode. On the contrary, Read et al.18 found the BET surface area of the carbon in the air electrode is not a significant factor in determining the discharge capacity, which is correlated to oxygen transport in organic electrolyte. Dramatic decrease in the discharge capacity at a high current density was linked to the rapid decrease in cell voltage and uneven distribution of Li2O2 deposition concentrating at the air interface.6 The uneven Li2O2 deposition may have been one of the factors in explaining the vast difference in the specific discharge capacities reported in literature, up to a high value of 5800 mA h per g carbon,19 when normalized to a low carbon mass in the cathode. Xu et al.20 explored using tris(pentafluorophenyl)borane as a functional additive and co-solvent in electrolytes to dissolve part of Li2O and Li2O2 in order to achieve higher Li/air battery discharge capacity, although other factors introduced by the same additive adversely affect the discharge capacity.

It is evident that the poor ORR cathode performance severely limits the discharge rate and capacity of a Li/air cell. However, no clear and systematic results have been obtained in finding an effective catalyst to improve the ORR kinetics. In addition, it is still not clear on the air cathode performance loss mechanism as whether it is from Li2O2 deposits blocking the narrow pores of the carbon material or Li2O2 covering and deactivating the ORR catalytic sites on the carbon surface. To address these issues, we examined in this study the discharge characteristics of Li/air cells with cathodes made of three carbon materials: Super P carbon (SP-carbon), Ketjen carbon (K-carbon) and pyrolized CuFe macrocycle compounds on Ketjen carbon (CuFe catalyzed K-carbon). These three carbon materials differ in the type and number of the catalytic sites for the ORR, and in the carbon pore volume and pore distribution. Comparison of the Li/air cell discharge behaviors under identical test conditions for the cells made with these cathode materials could provide the opportunity to identify the key contributing factors to the cell performance in terms of cell discharge voltage, rate and capacity, and thus shed light on the complex Li/air discharge process. Based on experimental results, an ORR mechanistic model involving a reaction product deactivating the catalytic sites on the carbon surface is proposed.

2. Experimental

Super P carbon (SP-carbon) was obtained from TIMCAL Graphite Carbon, and Ketjenblack EC-600 JD carbon black (K-carbon) from Akzo Nobel. The CuFe-catalyzed Ketjenblack EC-600 JD (CuFe-catalyzed K-carbon) was a non-precious metal–oxygen reduction catalyst produced on a commercial scale and provided by Acta SpA, Italy. This carbon supported catalyst was made first by absorbing a mixture of iron and copper complexes with phthalocyanine-based ligands onto the carbon support, and then heat-treated at between 800 and 900 °C in Ar atmosphere. The CuFe-catalyzed K-carbon has a Cu content at 1.7 wt% and a Fe content at 1.5 wt%, corresponding to an atomic ratio of Cu to Fe at 1 to 1.

X-Ray powder diffraction (XRPD) patterns of the carbon materials were collected on a Rigaku Ultima III diffractometer using a Bragg–Brentano geometry and Cu Kα radiation (λ = 1.5418 ?) over a 2θ scan range of 10–90° with a resolution of 0.02° at a scan rate of 0.2° per min. The surface areas of the carbon materials were measured with a Micromeritics ASAP 2010 system using N2 gas as adsorptive. The pore size distributions of the carbon materials were obtained from Barrett–Joyner–Halenda (BJH) desorption pore distribution using ASAP 2010 V1.00 software.

Lithium triflate (LiSO3CF3, 96%, Aldrich) was dried at 100 °C under vacuum for 8 h. Electrolyte grade propylene carbonate (PC, Ferro) was used as received. Tris(2,2,2-trifluoroethyl) phosphate (TFP) was synthesized by reacting sodium trifluoroethoxide with phosphorus oxychloride and purified by repeated fractionation under reduced pressure. Detailed descriptions on the synthesis and characterization of TFP are referred to in ref. 21 and 22.

Electrolyte solution with a fixed solvent composition of PC[thin space (1/6-em)]:[thin space (1/6-em)]TFP at a 7[thin space (1/6-em)]:[thin space (1/6-em)]3 weight ratio containing 0.2 M lithium triflate was prepared in a glove-box. Physical properties of this electrolyte solution, such as kinematic viscosity, ionic conductivity, oxygen solubility, and boiling point, are referred to in ref. 23.

Cathodes with a composition of 90 wt% carbon materials and 10 wt% polytetrafluoroethylene (PTFE) were prepared by mixing calculated amounts of carbon materials and PTFE emulsion (Teflon?, solid content = 61.5%, DuPont Co.), and then rolling the resulting paste mixture into a freestanding cathode sheet, which was punched into disks with an area of 0.97 cm2 and dried at 100 °C under vacuum for over 8 h. Li/air cells with an air access window of 0.97 cm2 were assembled in a dry-room having a dew point below ?90 °C by stacking sequentially a Li foil, a Celgard? 3500 membrane, a cathode, a Ni mesh (as cathode current collector), and an air-window frame into a coin cell cap. To activate the cell, a volume of 200 ?L liquid electrolyte solution was added through the air-window, followed by applying vacuum for 20 s to ensure complete filling of the electrolyte into the cell internal space. Excessive electrolyte solution over the cathode outer surface was removed by gently swiping with a filter paper over the Ni mesh. The electrolyte-activated cell was rested for 2 h before commencing discharging tests, which were carried out at room temperature (22 °C) in the dry-room on an Arbin BT-2000 tester galvanostatically from OCV until reaching the cutoff voltage at 1.5 V. Specific capacity of the cell was normalized by the mass of the carbon materials in the air cathode, which was in the range of 6–7 mg cm?2.

3. Results and discussions

 

3.1. Characterization of carbon materials and cathodes

The XRD patterns of the carbon materials shown in Fig. 1 reveal the relative degree of graphitization of the three carbon materials. SP-carbon has the highest degree of graphitization of the three carbon materials as shown by the sharp graphitic basal plane (002) peak at a d-spacing of 3.573 ?. Relative to K-carbon, the heat-treatment at around 800–900 °C in producing CuFe catalyzed K-carbon increased carbon graphitization degree, and shrank basal plane d-spacing (from 3.722 to 3.620 ?) towards that of an ideal graphite (3.354 ?). The sizes of the crystallites parallel to the graphite basal plane for the carbon materials listed in Table 1 are calculated by using the equation L = 1.84λ/(Bcos θ), where λ is the wavelength of the X-ray beam, B is the angular width of the basal plane diffraction (002) peak at the half-maximum intensity, and θ is the Bragg angle. The Brunauer–Emmett–Teller (BET) surface areas obtained using N2 gas adsorption at 77 K for these three carbon materials are listed in Table 1, and the carbon pore distribution and accumulated pore volume as a function of pore diameter obtained using BJH method is plotted in Fig. 2. SP-carbon possesses little internal pore volume and a rather low surface area, as compared to those of K-carbon and CuFe catalyzed K-carbon. The loss of carbon pore volume by the heat treatment used in making CuFe catalyzed K-carbon is consistent with the increase in graphitization degree of the carbon material as revealed by the XRD pattern. Also, the pore volume distribution of CuFe catalyzed K-carbon remains the same as that of the original K-carbon, which suggests a uniform structural collapse occurred during the heat treatment. This evidence does not support the suggestion of CuFe filling up or blocking some pores of K-carbon, especially at a metal loading at less than 3% of the carbon support.

XRD patterns for SP-carbon, K-carbon and CuFe catalyzed K-carbon.
Fig. 1 XRD patterns for SP-carbon, K-carbon and CuFe catalyzed K-carbon.


Table 1 Physical properties of carbon materials, cathode porosity and specific discharge capacity of Li/air cells with cathodes made of the carbon materials


SP-carbon K-carbon CuFe catalyzed K-carbon
a Average and standard deviation from measurements on 6 samples. b Calculated from electrode porosity by solvent method.
Graphitic basal plane d-spacing/? 3.573 3.722 3.620
Crystallite size parallel to basal plane/? 39.6 20.8 30.2
BET surface/m2 g?1 69.3 1413 751
Total pore vol. @ >20 ? dia./cm3 g?1 0.14 2.06 1.23
Electrode porosity by solvent methoda 77.3 ± 1.8% 90.8 ± 0.5% 86.8 ± 0.7%
Electrode pore vol.,b/cm3 g?1carbon 1.89 5.46 3.64
Electrode porosity by thickness methoda 75.3 ± 2.2% 90.9 ± 1.1% 87.9 ± 1.5%
Li/air cell discharge capacity, mA h g?1carbon
@ 0.05/mA cm?2 531 1286 1339
@ 0.20/mA cm?2 356 761 817
@ 0.50/mA cm?2 205 430 597
@ 1.00/mA cm?2 165 390





Incremental pore volume and accumulated pore volumes as a function of pore diameters obtained using BJH method for SP-carbon, K-carbon and CuFe catalyzed K-carbon.
Fig. 2 Incremental pore volume and accumulated pore volumes as a function of pore diameters obtained using BJH method for SP-carbon, K-carbon and CuFe catalyzed K-carbon.


The porosities of the cathodes made of the three carbon materials were measured using two separate methods: solvent filling and wet thickness. With the solvent filling method, the cathodes were vacuum filled with PC solvent, followed by removing excessive solvent on the electrode surfaces by gently dabbling on both sides with filter papers until reaching a stable mass. The carbon electrode porosity obtained using the solvent method is calculated according to the following equation:

ugraphic, filename = c0jm04170j-t1.tif


where Wdry is the dry mass of the electrode, Wwet is the electrode mass wetted with solvent, dc is the density of carbon (2.0 g cm?3), and dsol is the density of solvent (1.206 g cm?3 for PC). With the thickness method, the porosities were calculated from the dry mass and the wet thickness when filled with PC solvent. The porosity obtained using the thickness method is calculated according to the following equation:

ugraphic, filename = c0jm04170j-t2.tif


where A is the electrode area and δ is the thickness of electrode wetted with solvent. Both these methods yielded consistent results for the electrode porosity measurements listed in Table 1. A trend observed is that a cathode made of a carbon material possessing a higher internal volume of mesopores (and consequently a higher surface area) has a higher electrode pore volume.

3.2. Li/air cell discharge curves

In Fig. 3, cell voltage curves during discharge at a constant current density of 0.2 mA cm?2 for the Li/air cells with cathodes made of three carbon materials are compared. Without catalyst, a cell with a K-carbon cathode out-performed a cell with a SP-carbon cathode in terms of cell discharge voltage and capacity. A cell with a CuFe catalyzed K-carbon cathode exhibited two distinguishable voltage plateaus in its discharge voltage curve, with the first discharge plateau at a cell voltage above 2.5 V demonstrating a substantial improvement in cell discharge voltage of over 200 mV higher than that of a cell with a K-carbon cathode, and over 500 mV higher than that of a cell with a SP-carbon cathode; while the second discharge plateau at a voltage window between 1.7 to 1.5 V has a substantially higher discharge capacity than that of a cell with a K-carbon cathode. The discharge curve of the cell with a SP-carbon cathode did not have the second discharge plateau in the voltage window studied.

Li/air cell discharge voltage curves at 0.2 mA cm?2 for cells with cathodes made of (1) SP-carbon, (2) K-carbon and (3) CuFe catalyzed K-carbon.
Fig. 3 Li/air cell discharge voltage curves at 0.2 mA cm?2 for cells with cathodes made of (1) SP-carbon, (2) K-carbon and (3) CuFe catalyzed K-carbon.


Results of further efforts carried out to identify the origin of the second discharge plateau observed for the Li/air cells with K-carbon and CuFe catalyzed K-carbon cathodes, especially from the charge contribution by the possible cathode process of Li2O2 to Li2O conversion, are presented in Fig. 4. As will be shown by the experimental results, there is negligible charge contribution by the Li2O2 to Li2O conversion for the Li/air cells. Shown in Fig. 4 (top), the Li/air cells were first discharged at a constant current density of 0.2 mA cm?2 from point A to point B to reach a state of discharge at a capacity of 300 mA h g?1. After that, the two cells were reassembled in sealed coil cells to close off the air access to the cells, and then discharged as Li/C cells at a current density of 0.05 mA cm?2 from point B until reaching the cutoff voltage at 1.5 V. For the cell with a K-carbon cathode, the Li/C cell discharge step reached a discharge capacity close to 80 mA h g?1; while for the cell with a CuFe catalyzed K-carbon cathode, a discharge capacity over 480 mA h g?1, which is by far higher than what would be expected if the charge were originated from further reduction of Li2O2 produced at a charge capacity of 300 mA h g?1 during the initial Li/air cell discharge step (from point A to point B in Fig. 4a). To further verify if there is any charge contribution by the possible cathode process of Li2O2 to Li2O conversion, hermetic Li/C cells with fresh cathodes were built and discharged in the absence of oxygen access. As shown in Fig. 4b, the charges obtained with the fresh cathodes almost account for all of the charges obtained from the corresponding two cells shown in Fig. 4a in the discharge step starting from point B. The two discharge curves shown in Fig. 4b and that in Fig. 4a starting from point B for cells with CuFe catalyzed K-carbon cathodes are re-plotted in Fig. 4c, where the two curves overlap each other completely, verifying the negligible charge contribution by the Li2O2 to Li2O conversion for the Li/air cells tested. This conclusion could be rationalized by the non-electronic conductivity and low solubility of Li2O2 in the chosen electrolyte solution.



a) Li/air cells with cathode made of (1) K-carbon and (2) CuFe catalyzed K-carbon discharged from point A to point B at 0.2 mA cm?2 for 300 mA h g?1, then the cells containing Li2O2 produced were re-assembled in enclosed coin cells to block off the air access, and then discharged as Li/C cells without oxygen access from point B at 0.05 mA cm?2 to 1.5 V. (b) Li/C cells discharge voltage curves obtained at 0.05 mA cm?2 for cells with cathodes made of (1′) K-carbon and (2′) CuFe catalyzed K-carbon. Initial capacity at voltage greater than 2 V is due to the oxygen absorbed in electrolyte solution. (c) Discharge voltage curves for cells with CuFe catalyzed K-carbon, where curve (2) is obtained in (a) from point B discharge process and curve (2′) in (b).
Fig. 4 a) Li/air cells with cathode made of (1) K-carbon and (2) CuFe catalyzed K-carbon discharged from point A to point B at 0.2 mA cm?2 for 300 mA h g?1, then the cells containing Li2O2 produced were re-assembled in enclosed coin cells to block off the air access, and then discharged as Li/C cells without oxygen access from point B at 0.05 mA cm?2 to 1.5 V. (b) Li/C cells discharge voltage curves obtained at 0.05 mA cm?2 for cells with cathodes made of (1′) K-carbon and (2′) CuFe catalyzed K-carbon. Initial capacity at voltage greater than 2 V is due to the oxygen absorbed in electrolyte solution. (c) Discharge voltage curves for cells with CuFe catalyzed K-carbon, where curve (2) is obtained in (a) from point B discharge process and curve (2′) in (b).


The second discharge plateau observed for the cell with a K-carbon cathode (shown in Fig. 3, 4a from point B, and 4b) could be attributed to the surface adsorbed oxygen on the carbon material, and for the cell with a CuFe catalyzed K-carbon cathode to both of the surface adsorbed oxygen on the carbon material and solvent reduction. The lack of surface adsorbed oxygen on SP-carbon could be attributed to its much lower surface area and a more ordered surface structure. Solvent reduction often occurs on the carbonaceous anode in the formation cycling of Li-ion for Li rechargeable batteries, albeit at a lower voltage plateau around 0.9 V without using catalyzed carbon materials at the electrode. To understand the reaction mechanism of the CuFe catalyzed carbon in catalyzing the solvent reduction further investigation is needed. With purposeful selection and optimization of the solvent and catalyst pairs, the solvent redox process catalyzed by the cathode material could be utilized as new cell chemistry in designing rechargeable Li batteries.

3.3. Li/air cell discharge rate and capacity

The discharge voltages obtained for cells with K-carbon and CuFe catalyzed K-carbon cathodes at various discharge current densities are shown in Fig. 5, and these results are summarized in Table 1 and in Fig. 6 in a plot of cell energy density as a function of discharge current density. The discharge voltage of the cell with a CuFe catalyzed K-carbon cathode is over 200 mV higher than that of the cell with a K-carbon cathode at a low discharge current density, and over 1 V higher at a high discharge current density. The Li/air cell with a CuFe catalyzed K-carbon cathode discharged at a relatively high current density of 1 mA cm?2 demonstrated an energy density over 0.8 Wh g?1, which is more than twice of that of the cell with a K-carbon cathode.

Li/air cell discharge voltage curves obtained at various discharge current densities for cells with cathodes made of (1) K-carbon and (2) CuFe catalyzed K-carbon.
Fig. 5 Li/air cell discharge voltage curves obtained at various discharge current densities for cells with cathodes made of (1) K-carbon and (2) CuFe catalyzed K-carbon.




Plot of energy density as a function of discharge current density from data shown in Fig. 5 for Li/air cells with cathodes made of (1) K-carbon and (2) CuFe catalyzed K-carbon.
Fig. 6 Plot of energy density as a function of discharge current density from data shown in Fig. 5 for Li/air cells with cathodes made of (1) K-carbon and (2) CuFe catalyzed K-carbon.


The discharge capacities of the Li/air cells, shown in Fig. 3 and 5a, are plotted as a function of the cathode pore volume in Fig. 7. There is a good correlation between the cell discharge capacity and electrode pore volume for these three carbon materials, especially at a low discharge current density of 0.05 mA cm?2, where it is estimated 10% of the cathode pore volume is filled by the electrode reaction product of Li2O2 at the end of the discharge process. At a higher discharge current density, there is a rapid decrease in the discharge capacity. At the fast discharge rate of 1 mA cm?2, no visible depositions of the reaction products at the cathode out-surface were observed, likely due to the low capacity obtained. For the three types of carbon cathodes, the difference in the obtained discharge capacity is significant: the capacity of SP-carbon is at near zero, and that of CuFe catalyzed K-carbon is about twice that of the original K-carbon. These results would argue against the clogging pore at the cathode surface being the limiting factor for the discharging capacitance at the fast discharging rate. Furthermore, if it were clogging pores at the carbon surface in limiting the discharge capacity, the original K-carbon would have a higher capacitance than the CuFe-catalyzed K-carbon due to its higher carbon pore volume. The rapid build-up of lithium peroxide covering the active sites for ORR on the carbon surface at a higher discharging current density may have caused the fast passivation of the cathode, as shown by the slant discharge voltage plateau, attributing to the early termination of the cell discharge process.



Plots of Li/air cell specific discharge capacities as a function of cathode pore volume.
Fig. 7 Plots of Li/air cell specific discharge capacities as a function of cathode pore volume.


3.4. Micrographs of cathodes after discharge

Because of the low solubility of the Li2O2 produced in discharging a Li/air battery, several factors are expected to affect its production and precipitation during the discharging process, and subsequently, its final location within the cathode, which can be examined with a microscope after the discharge step is completed. It has been observed that as the discharge current density increases, the Li2O2 particles tend to be found within the cathode at a location close to the air interface rather than the electrolyte interface. Even at the same discharge current density, more Li2O2 deposits were observed on the surface of the cathode facing air for a cell with CuFe catalyzed K-carbon than those with K-carbon, as shown in Fig. 8, where large crystals of Li2O2 were found on the cathode surface and in the cracks close to the air surface with CuFe catalyzed K-carbon. The formation of large Li2O2 crystals on the surface and within the cracks of the electrode implies a meaningful solubility and mobility of Li2O2 in the electrolyte solution needed for the crystal growth from the Li2O2 initially produced on the ORR catalytic sites during the discharging process. The fact that Li2O2 was found at a location close to the air surface of the cathode also indicates the rather slow diffusion of dissolved oxygen in the non-aqueous electrolyte solution. At a high discharging current density or with a cathode material possessing a higher density of active sites as offered by the CuFe catalyzed K-carbon cathode material, the ORR reaction zone moves to the air surface of the cathode. The severe decrease in cell discharge capacity at high current density is attributed to both increased cathode passivation and uneven distribution of Li2O2 deposits within the pore volume of the cathode.

Optical microscope pictures of cathodes facing air for cathodes made of (1) K-carbon and (2) CuFe catalyzed K-carbon in Li/air cells discharged at 1.0 mA cm?2 to 1.5 V, rested for 5 h, and then discharged at 0.05 mA cm?2 to a total capacity of 850 mA h g?1. Deposits from cathode reaction of discharging Li/air cells were seen as the white spots on the black carbon surface and as colorless crystals formed (shown by arrows) within the cracks of the cathode.
Fig. 8 Optical microscope pictures of cathodes facing air for cathodes made of (1) K-carbon and (2) CuFe catalyzed K-carbon in Li/air cells discharged at 1.0 mA cm?2 to 1.5 V, rested for 5 h, and then discharged at 0.05 mA cm?2 to a total capacity of 850 mA h g?1. Deposits from cathode reaction of discharging Li/air cells were seen as the white spots on the black carbon surface and as colorless crystals formed (shown by arrows) within the cracks of the cathode.


3.5. Li/air cell polarization

The dynamic Li/air cell discharging polarization curves, shown in Fig. 9 dashed lines, were obtained at a current density scan rate of 0.02 mA cm?2 s?1 for cells with K-carbon cathode and CuFe catalyzed K-carbon cathode after both cells were discharged to a charge capacity of 300 mA h g?1 and then followed by a rest period of two hours. Previous study has shown that the lithium anode has negligible polarization, and the cell voltage polarization is largely attributed to the cell cathode polarization and cell IR drop. Under the dynamic scan conditions, the polarization curves for both cells are similar, with the cell with a CuFe catalyzed K-carbon cathode showing a slightly better ORR kinetics at the low current density range, and the cell with a K-carbon cathode a slightly better mass transport at the high current density range. However, when the cell polarization curves representing the steady state discharge as obtained from data shown in Fig. 5a–d are compared, there are large differences between the dynamic and steady-state conditions for the two cells. For the cell with a CuFe catalyzed carbon cathode, the polarization curve at steady-state shifts downward by about 80 mV from that at dynamic scan; while for the cell with a K-carbon cathode, the corresponding downward shift in polarization curve is over 300 mV, and increasing with the current density to over 500 mV at 1 mA cm?2. It was expected that the ORR polarization from the increased mass transportation limitation from the dynamic state to steady-state is similar for the two cells. The much larger voltage downward shift, especially at a higher current density, for a cell with a K-carbon cathode than that for a cell with a CuFe catalyzed K-carbon cathode indicates that there is added sluggishness in the ORR kinetics under the steady-state conditions. Under the steady-state conditions, there is a continuous formation of Li2O2 on the ORR catalytic sites, and its subsequent removal by dissolution and crystal growth keeps the number of active sites constant, and the number of available free active sites decreases with the increase in discharge current density. The higher ORR catalytic activity provided by a higher level of free site density found in CuFe catalyzed K-carbon material, which contains catalytic sites consisting of CuFe and those on K-carbon, decreases the ORR polarization under the steady-state conditions.

Li/air cell polarizations at steady-state (solid line) and at dynamic scan at 0.02 mA cm?2 s?1 (dashed line), measured after discharged to 300 mA h g?1 for cells with cathodes made of K-carbon (■) and CuFe catalyzed K-carbon (▲).
Fig. 9 Li/air cell polarizations at steady-state (solid line) and at dynamic scan at 0.02 mA cm?2 s?1 (dashed line), measured after discharged to 300 mA h g?1 for cells with cathodes made of K-carbon (■) and CuFe catalyzed K-carbon (▲).


3.6. ORR electrode process

It becomes evident from above test results and observations that the performance of a Li/air cell is limited by the ORR electrode process, which could involve the following key steps:

  O2 (g) → O2 (sol) (1)

  O2 (sol) + *–CS → O2–CS (2)

  2Li+ + O2–CS + 2e? → Li2O2–CS (3)

  Li2O2–CS → *–CS + Li2O2 (sol) (4)

  Li2O2 (sol) → Li2O2 (solid)↓ (5)

In this cathode reaction mechanism, the oxygen gas molecule (O2(g)) from air first dissolves in the electrolyte solution forming dissolved oxygen molecules (O2(sol)), which then adsorbs onto a free ORR catalytic site (*–CS) on the carbon surface. The subsequent ORR reaction produces one Li2O2 molecule attached to the catalytic site (Li2O2–CS) resulting in deactivating the catalytic site for further ORR. The deactivated catalytic site by attached Li2O2 is only freed by removal of the Li2O2 molecule into the surrounding electrolyte solution. The driving forces for such a removal process come from the Li2O2 diffusion into the electrolyte solution within the electrode pore volume, and from the aging-growth process of some Li2O2 crystals at nearby locations. The sustainable Li/air cell discharge rate would thus depend on the initial number of free catalytic sites on the carbon surface and the balancing act of Li2O2 production and its removal in maintaining a sufficient number of free catalytic sites to sustain the ORR. At an excessively high discharge current density, the free catalytic sites diminish rapidly because of the relatively slow rate of Li2O2 removal from the deactivated catalytic sites, resulting in fast increase in ORR polarization. As shown in Fig. 5d, the Li/air cell with a CuFe catalyzed K-carbon cathode has two distinguishable cell voltage plateaus at above 2 V when discharged at a relatively high current density of 1 mA cm?2. The two voltage plateaus reflect the ORR process carried out at two different types of catalytic sites in the cathode material. The first discharge plateau arises from ORR catalyzed by the more active CuFe catalytic sites. After these CuFe sites are deactivated at the end of the first discharge plateau, the ORR proceeds at the less active catalytic sites on the supporting carbon surface, forming a second voltage plateau, which is at a similar voltage to what was observed for a Li/air cell with a K-carbon cathode where only the active sites provided by the K-carbon exist. At the fast discharge rate, all active sites are deactivated and cause rapid termination of the discharge process. For the CuFe-catalyzed K-carbon, the highly active CuFe catalytic sites are deactivated first at a high cathode voltage, followed by the deactivation of the less active sites of the carbon surface at a lower cathode voltage. The charge contribution from absorbed oxygen only occurs at the start of the discharging process, and since oxygen supply is not the limiting factor in an operating Li/air cell, the overall impact of absorbed oxygen on the discharge capacity of a Li/air cell at a cell voltage >2.0 V is small at less than 35 mA h g?1. The charge contribution from solvent reduction occurs at a cell voltage bellow 1.8 V.

As for the Li/air battery discharge capacity obtained at a low discharging current density, available electrode pore volume is a determining factor. Carbon material possessing a high mesopore volume at a pore diameter greater than 20 ? provides additional electrode pore volume accessible for the ORR and for accommodating the Li2O2 deposit. The Li/air cell discharge capacity is also impacted by the cell discharging current density, which affects the rate of cathode catalytic sites deactivation and the distribution of Li2O2 within the cathode. The ability to redistribute the Li2O2 reaction product more evenly throughout the cathode pore volume in the thickness direction could provide substantial improvement in the Li/air cell discharge capacity, from currently 10% of the potential capacity estimated based on the full occupancy of the electrode pore volume by the Li2O2 deposit. There are several possible approaches to address this issue, and these measures include: (1) increasing the solubility of Li2O2 in the electrolyte solution with a better selection of solvents and electrolytes, (2) a further decrease in the lithium electrolyte concentration to move the reaction zone from the air interface towards the electrolyte interface of the cathode and (3) modifying electrode structures to provide less tortuous diffusion paths for the oxygen and Li2O2.

4. Conclusions

Based on the observations and test results of the Li/air batteries with cathodes made of different carbon materials, a mechanism for the ORR electrode process is proposed where the newly formed Li2O2 molecules deactivate the catalytic sites, and the density and activity of free catalytic sites on the surface of carbon and the removal rate of attached Li2O2 from the deactivated catalytic sites determine the ORR polarization, and thus the discharge cell voltage.

It has been demonstrated that by increasing the catalytic site density and activity on the carbon surface for the ORR, as shown from the graphite-like SP-carbon, to the defect-rich amorphous K-carbon, and to the CuFe catalyzed K-carbon, the ORR polarization experienced during the Li/air discharging process can be significantly decreased, thus considerably increase the discharge cell voltage and rate for the Li/air batteries. Cells with a CuFe catalyzed K-carbon cathode demonstrated a higher cell discharging voltage of over 200 mV than that with K-carbon, and of over 500 mV than that with SP-carbon.

It was observed that there is a negligible amount of Li2O2 being converted to Li2O during the Li/air cell discharge process. A cell with a CuFe catalyzed K-carbon cathode has a distinguishable second discharge plateau in the voltage window from 1.7 to 1.5 V attributed to the solvent reduction catalyzed by the CuFe catalyst.

 

References

1 T. Ogasawara, A. Debart, M. Holzapfel, P. Novak and P. G. Bruce, J. Am. Chem. Soc., 2006, 128, 1390–1393 Article  CAS .
2 K. M. Abraham and Z. Jiang, J. Electrochem. Soc., 1996, 143, 1–5 Article  CAS .
3 J. Read, J. Electrochem. Soc., 2002, 149, A1190–A1195 Article  CAS .
4 Z. Deng, L. Ruoshi, H. Tao and Y. Aishui, J. Power Sources, 2010, 195, 1202–1206.
5 Y. G. Wang and H. S. Zhou, J. Power Sources, 2010, 195, 358–361 Article  CAS .
6 S. S. Zhang, D. Foster and J. Read, J. Power Sources, 2010, 195, 1235–1240.
7 H. Cheng and K. Scott, J. Power Sources, 2010, 195, 1370–1374.
8 A. Debart, A. J. Paterson, J. Bao and P. G. Bruce, Angew. Chem., Int. Ed., 2008, 47, 4521–4524 Article  CAS .
9 V. M. B. Crisostomo, J. K. Ngala, S. Alia, A. Dobley, C. Morein, C. H. Chen, X. Shen and S. L. Suib, Chem. Mater., 2007, 19, 1832–1839 Article  CAS .
10 A. Debart, J. Bao, G. Armstrong and P. G. Bruce, J. Power Sources, 2007, 174, 1177–1182 Article  CAS .
11 Y. C. Lu, H. A. Gasteiger, M. C. Parent, V. Chiloyan and Y. Shao-Horn, Electrochem. Solid-State Lett., 2010, 13, A69–A72.
12 C. O. Laoire, S. Mukerjee, K. M. Abraham, E. J. Plichta and M. A. Hendrickson, J. Phys. Chem. C, 2009, 113, 20127–20134.
13 N. Seriani, Nanotechnology, 2009, 20, 1–7.
14 V. S. Bryantsev, M. Blanco and F. Faglioni, J. Phys. Chem. A, 2010, 114, 8165–8169 Article  CAS .
15 M. Mirzaeian and P. J. Hall, Electrochim. Acta, 2009, 54, 7444–7451 Article  CAS .
16 R. E. Williford and J. G. Zhang, J. Power Sources, 2009, 194, 1164–1170.
17 C. Tran, X. Yang and D. Qu, J. Power Sources, 2010, 195, 2057–2063 Article  CAS .
18 J. Read, K. Mutolo, M. Ervin, W. Behl, J. Wolfenstine, A. Driedger and D. Foster, J. Electrochem. Soc., 2003, 150, A1351–A1356 Article  CAS .
19 S. D. Beattie, D. M. Manolescu and S. L. Blair, J. Electrochem. Soc., 2009, 156, A44–A47 Article  CAS .
20 W. Xu, J. Xiao, D. Y. Wang, J. Zhang and J. G. Zhang, J. Electrochem. Soc., 2010, 157, A219–A224.
21 K. Xu, M. S. Ding, S. S. Zhang, J. L. Allen and T. R. Jow, J. Electrochem. Soc., 2003, 150, A161–A169.
22 K. Xu, S. S. Zhang, J. L. Allen and T. R. Jow, J. Electrochem. Soc., 2003, 150, A170–A175.
23 S. S. Zhang, K. Xu and J. Read, J. Power Sources, 2011, 196, 3906–3910.

This journal is ? The Royal Society of Chemistry 2011

  评论这张
 
阅读(776)| 评论(0)
推荐 转载

历史上的今天

评论

<#--最新日志,群博日志--> <#--推荐日志--> <#--引用记录--> <#--博主推荐--> <#--随机阅读--> <#--首页推荐--> <#--历史上的今天--> <#--被推荐日志--> <#--上一篇,下一篇--> <#-- 热度 --> <#-- 网易新闻广告 --> <#--右边模块结构--> <#--评论模块结构--> <#--引用模块结构--> <#--博主发起的投票-->
 
 
 
 
 
 
 
 
 
 
 
 
 
 

页脚

网易公司版权所有 ©1997-2017