Performance and Cost Evaluation of Lithium Ion ... - ACS Publications

4 downloads 0 Views 3MB Size Report
Abstract. In order to increase the energy content of lithium ion batteries (LIBs), researchers worldwide focus on high specific energy (Wh/kg) and energy density ...
Article pubs.acs.org/cm

Best Practice: Performance and Cost Evaluation of Lithium Ion Battery Active Materials with Special Emphasis on Energy Efficiency Paul Meister,† Haiping Jia,† Jie Li,† Richard Kloepsch,† Martin Winter,*,†,‡ and Tobias Placke*,† †

MEET Battery Research Center, Institute of Physical Chemistry, University of Münster, Corrensstr. 46, 48149 Münster, Germany Helmholtz Institute Münster, IEK-12, Forschungszentrum Jülich GmbH, Corrensstraße 46, 48149 Münster, Germany



S Supporting Information *

ABSTRACT: In order to increase the energy content of lithium ion batteries (LIBs), researchers worldwide focus on high specific energy (Wh/kg) and energy density (Wh/L) anode and cathode materials. However, most of the attention is primarily paid to the specific gravimetric and/or volumetric capacities of these materials, while other key parameters are often neglected. For practical applications, in particular for large size battery cells, the Coulombic efficiency (CE), voltage efficiency (VE), and energy efficiency (EE) have to be considered, which we point out in this work by comparing numerous LIB active materials. For all presented active materials, energy inefficiency is mainly caused by a voltage inefficiency, which in turn is affected by the voltage hysteresis between the charge and discharge curves. Hence, this study could show that materials with larger voltage hysteresis such as the ZnFe2O4 (ZFO) anode or the Lirich cathode material exhibit also a lower VE and EE than for instance graphite and LiNi0.5Mn1.5O4. Furthermore, from the accumulated EE losses the resulting “extra energy costs” are calculated based on industry and domestic electricity costs in Germany, in Japan and in the U.S.A. In particular, in countries with higher electricity costs such as Germany, the accumulated extra energy, which is necessary to compensate the energy inefficiency while retaining a certain energy level in the electrode material, has a stronger impact on the extra energy costs and thus on the total cost of ownership of the battery cell system. reduction of the energy content on a practical level.2 Besides, there are several obstacles limiting the practical realization of Li/O2 batteries, such as severe side reactions upon cycling, involving the electrode materials, electrolyte, intermediate, and final discharge products, as well as a large voltage hysteresis yielding a low energy efficiency (EE).10,11 Therefore, it seems to have become a common consensus in the battery community that this system will not be commercially realized in the near future.5,12 In the field of PLIBs, specifically the lithium/sulfur (Li/S) system, providing a high theoretical specific energy and energy density (2567 Wh kg−1, 2199 Wh L−1), has emerged as one of the most promising candidates to compete with LIBs in terms of gravimetric energy.2,3,8 So far, however, prototype Li/S cells, e.g., the cell presented by Sion Power, exhibit a significantly reduced practical energy (350 Wh kg−1, 300 Wh L−1), which, furthermore, can be only achieved at the beginning of the cycle life.8 Currently, the Li/S technology is still at the research stage, as it mainly suffers from a relatively short cycle life, low Coulombic efficiency, and pronounced self-discharge rate, limiting its practical impact.5,8 Another next-generation battery approach is the lithium and nonlithium-based “all-solidstate battery”, which utilizes a solid electrolyte and raised

1. INTRODUCTION A variety of different, commercially and technologically promising battery chemistry approaches exist that, depending on the respective storage technology, are suitable for either automotive or stationary application purposes. In particular, lithium ion batteries (LIBs) as well as the systems of the “next generation”, the so-called post lithium ion batteries (PLIBs), are considered as the most promising technologies for electrochemical energy storage. Currently, there is a strong academic and industrial interest in an in-depth evaluation and comparison of these different battery systems in terms of theoretical and practical gravimetric (Wh/kg, also called specific energy1) and volumetric energy densities (Wh/L, simply referred to as “energy density” in many reports) as well as costs from the material to the system level.2−7 A particular focus lies on the evaluation of the practical gravimetric and volumetric energy contents of the next generation battery systems and their potential to reach commercialization in order to replace the state-of-the-art LIBs. Among the next generation systems, the Li/air or Li/O2 redox couple has gained significant interest in research, owing to the high theoretical specific energy and energy density (3458 Wh kg−1, 3445 Wh L−1) for the Li/Li2O2 couple.8,9 Since Li/O2 cells suffer from ingress of contaminants such as water and carbon dioxide, either a purification system or an oxygencontaining pressure vessel has to be used, both leading to a © XXXX American Chemical Society

Received: July 15, 2016 Revised: September 23, 2016

A

DOI: 10.1021/acs.chemmater.6b02895 Chem. Mater. XXXX, XXX, XXX−XXX

Article

Chemistry of Materials t

enormous expectations according to operational safety and flexible cell geometry as well as high energy density. However, also for this battery system, there are still numerous remaining challenges, including, e.g., the interface stability of metallic lithium with the solid electrolyte, the mechanical integrity of thin electrolyte films, for instance against Li dendrites, and the poor rate capability of the cells, in particular at lower temperatures, as well as achieving cost parity with the established liquid organic solvent-based electrolyte system.13,14 Besides the lithium-based battery systems, also Na-based systems, i.e., sodium ion batteries (NIBs), have gained attention in research, even though the main application may be restricted to stationary energy storage due to the decreased gravimetric and volumetric energy densities compared to LIBs.5 Since the market introduction of PLIBs is not realized yet, LIBs will dominate the battery market in the immediate future.5 The ever-growing demands to increase the specific energy/ energy density as well as the high power performance of LIBs have prompted widespread research to develop novel electrode materials for both the anode and the cathode.15 The academic search for these alternative anode and cathode materials is, however, strongly focused on electrode capacity improvements rather than specific energy and energy density improvements, and accordingly, only rate capability and not the more important characteristic, power capability, is the measure that counts in most academic reports. Many research studies on novel materials report on promising high specific capacities but seem to neglect other fundamental electrochemical material characteristics, including in particular the Coulombic efficiency (CE) and the delithiation potentials, which will both directly determine the gravimetric and volumetric energy densities of the cell. Furthermore, and not surprisingly in this context, less attention is being paid to the energy efficiency of the materials, which represents the usable fraction of energy during the charge/discharge process.16 During cycling of LIB cells, different unwanted/parasitic reactions occur that affect the capacity fading as well as the CE of the cells. Various processes including SEI growth and repair,17 electrolyte oxidation, transition metal dissolution, etc. consume additional charge, which in turn results in a CE decrease.18 Hence, the CE is a measure of the reversibility of the redox reactions and can be generally expressed as the ratio of the discharged capacity to the capacity necessary to charge the material/system:19,20 CE =

Qd Qc

EE =

EE =

Id ∫ Ud dt 0

tc

Ic ∫ Uc dt

(4)

0

The ratio of EE to CE is given by the combination of eqs 2 and 4. Additionally, the voltage efficiency (VE) can be determined by the following expression: t

∫0 d Ud dt

EE = CE

td t

∫0 c Uc dt

= CE

Ud = CE·VE Uc (5)

tc

Equation 5 illustrates that the EE is directly connected to the CE and the VE. Here Ud and Uc represent the average voltage of the battery during discharge or charge, respectively. Consequently, factors that affect the VE and/or the CE will also have an impact on the EE. The charge/discharge current, stateof-charge, internal resistance, operating temperature, and state of health of the battery have been identified as main parameters that influence the VE and CE.19,21 Beside the above-mentioned methods to calculate the EE of electrochemical energy storage systems, also different approaches exist in the literature. For instance, the PNGV (Partnership for a New Generation of Vehicles) Battery Test Manual of the U.S. DOE (U.S. Department of Energy) presents a method to determine the round-trip EE given as the fraction of discharge energy that is transferred to the regen energy returned during the profiles:16,22 Round‐trip efficiency =

watt ·hours (discharge) ·100% watt ·hours (regen) (6)

As this approach only considers the EE of the discharge reaction, Kang et al. presented a novel way to calculate the EE under charge (EEcharge), discharge (EEdischarge) and with regard to the ratio of charge and discharge (EEbattery).16 Since we compare different anode and cathode materials in this work and intend to give a specific overview of the EE on the material level, only the EE under the charge−discharge regime will be further discussed. The ratio of the discharge energy (Eout) to the charge energy (Ein) corresponds to the EE of the battery cell under charge−discharge (EEbattery):16

t

(1)

In this equation, Id/Ic are the discharge/charge currents and td/tc are the respective discharge/charge times. To make the situation simple, during constant current cycling, eq 1 can be written as It CE = d d Ictc

(3)

td

∫0 d Id dt ∫0 c Ic dt

t

∫0 c UcIc dt

This equation, where Ud/Uc are the discharge/charge cell voltages, can be simplified, when a constant current is applied during charge and discharge:

t

=

∫0 d UdId dt

EE battery = (2)

Eout E in

(7)

The calculation of Eout is performed by means of the following equation:

Analogously to CE, EE represents the ratio of the discharge energy to the charge energy. EE is a key performance indicator of a battery cell. Since energy cannot be lost but is rather transformed into another form of energy, the “losses” of electrical or chemical energy point, for instance, into a conversion into thermal energy.16 A general definition of EE is given by the following equation:19,20

SOC(t )

Eout =

∫SOC(0)

UdischCn dSOC

(8)

Analogously, the charge energy Ein is determined via the following equation: B

DOI: 10.1021/acs.chemmater.6b02895 Chem. Mater. XXXX, XXX, XXX−XXX

Article

Chemistry of Materials

Figure 1. Comparison of (a) gravimetric capacities, (b) volumetric capacities, and (c) the operating potential range and the average delithiation potential of different lithium ion battery anode materials including intercalation/insertion materials, alloying materials, and conversion materials. The average delithiation potential is defined as the potential where half of the discharge capacity has been reached.

In order to compensate for the accumulated energy inefficiency, while maintaining a discharge energy of 90 kWh (as installed, e.g., in the Tesla Model S25), extra charge energy is needed. Based on the accumulated extra energy, the domestic and industrial extra energy costs were calculated exemplarily for Germany, Japan and the U.S.A., giving a clear indication of how much the total cost of ownership (TCO) of a battery system (based on the individual electrode material chemistry) is affected by the EE.

SOC(t )

E in =

∫SOC(0)

UchargeCn dSOC

(9)

SOC(0) represents the initial state of charge, while SOC(t) is the terminal state, when the charge process is completed. The terms Ucharge and Udisch correspond to the voltage of the battery cell during the charging and discharging process, respectively, and Cn is the nominal capacity of the battery cell. Usually, the battery cell cycling equipment records parameters such as current, voltage/potential, and time. Hence, the charge energy Ein and the discharge energy Eout can be determined from the data provided by the cycling equipment.16 Recently, the importance of energy efficiency was highlighted and re-recognized as relevant in the field of ultra-/super- and pseudocapacitors. For example, it was demonstrated that an evaluation of the electrochemical performance only on the basis of the CE resulted in an overestimation of CE. For supercapacitor systems where Faradaic reactions are involved in the charge storage mechanism (= pseudocapacitors), a low EE ranging only between 50% and 80% is observed, while the CE accounted for 90−95%.23,24 In this contribution, the influence of the LIB anode material characteristics, such as first cycle CE and average delithiation potential of different types of anodes, including insertion/ intercalation, “alloying”, and conversion materials on the specific energy of LIB full cells is critically evaluated and discussed. Furthermore, we evaluate different anode and cathode materials with regard to their CE, VE, EE, and reversible discharge capacity. To get an understanding of the impact of the EE on the practical application, the accumulated EE losses of the respective electrode materials are determined. On purpose, we focus on the individual electrode materials in half cell measurements and not on the EE of a full battery cell, which is in line with most reports in the battery materials field.

2. EXPERIMENTAL SECTION Table S1 and Table S2 (Supporting Information) show an overview of the compositions of the prepared active material electrode tapes and the average mass loadings, respectively. The detailed electrode preparation process is also described in the Supporting Information. The electrochemical performance was evaluated on a Maccor 4300 battery test system at 20 °C. Depending on the active material, different parameters and electrolytes have been used for the constant current charge/discharge experiments. For the electrochemical investigations either lab-scale Swagelok type T-cells with a threeelectrode configuration or coin cells with a two-electrode setup have been used. The reference electrode and counter electrodes were made from high-purity metallic lithium foil (Rockwood Lithium). Two types of polypropylene-based separators were used: Freudenberg FS2190 for the Swagelok cells and Celgard-2400 for the coin cells. The cells were assembled in an argon-filled glovebox (UniLab, MBraun) with oxygen and water contents of less than 1 ppm. The electrochemical characterization of graphite, soft carbon, and NMC-111 was carried out in a three-electrode arrangement, while the respective measurements for silicon/graphite, ZnFe2O4, LiNi0.5Mn1.5O4, and 0.5 Li2MnO3·0.5 LiNi0.4Mn0.4Co0.2O2 were performed using a twoelectrode setup. The cutoff potentials for the carbonaceous materials (graphite and soft carbon) have been set to 0.02 V vs Li/Li+ for the lithiation and 1.5 V vs Li/Li+ for the delithiation process. Five formation cycles were performed, in which the cells were discharged (lithiation process) to C

DOI: 10.1021/acs.chemmater.6b02895 Chem. Mater. XXXX, XXX, XXX−XXX

Article

Chemistry of Materials 0.02 V vs Li/Li+, followed by a constant potential (CP) step at this potential either for 1 h or until the current drops below 0.02 C. Afterward, the cells were charged (delithiation process) to 1.5 V vs Li/ Li+. In the subsequent cycles, a current of rate 1 C (= 372 mA g−1, assuming a theoretical capacity of 372 mAh g−1) was applied for discharge/charge in the above-mentioned potential range, but without a constant potential step. A mixture of ethylene carbonate (EC) and diethyl carbonate (DEC) (3:7 by weight) has been used with 1 M LiPF6 as electrolyte. The silicon/graphite electrodes were cycled between 0.04 and 1.5 V. One formation cycle was carried out, where the lithiation took place with an applied current rate of 0.1 C, followed by a constant voltage (CV) at 0.04 V until the current rate dropped below 0.005 C. Thereafter, the cells were delithiated at 0.2 C until the electrode potential reached 1.5 V vs Li/Li+. In the subsequent cycles, a current of 1 C (= 1100 mA g−1, assuming a theoretical capacity of 1100 mAh g−1) was applied until the lower voltage criterion is reached. Additionally, a CV step was applied until the current rate fell below 0.05 C. Then, the cells were charged (delithiated) with 1 C to 1.5 V. A mixture of 1 M LiPF6 in EC/DEC (3:7 by weight) with 10 wt % fluoroethylene carbonate (FEC) was used as electrolyte. One formation cycle was conducted with the ZnFe2O4 (ZFO) electrodes using a current rate of 0.1 C until the lower voltage limit of 0.01 V was obtained. Then, the cells rested for 30 s, before they were charged at 0.1 C to 3.0 V. Once again the cells rested for 30 s. In the following cycles, the same procedure, but with a current rate of 1 C (= 940 mA g−1, assuming a theoretical capacity of 940 mAh g−1) was executed. The used electrolyte was 1 M LiPF6 in EC/DEC (3:7 by weight). Two formation cycles with charge/discharge current rates of 0.1 C were performed for the LiNi0.5Mn1.5O4 electrodes in a voltage range of 3.5 and 5.0 V. The following cycling was conducted at a rate of 1 C (= 147 mA g−1, assuming a theoretical capacity of 147 mAh g−1) in the same voltage range. 1 M LiPF6 in EC and dimethyl carbonate (DMC) (1:1 by weight) was used as electrolyte. The Li-rich material 0.5 Li2MnO3·0.5 LiNi0.4Mn0.4Co0.2O2 (LRNMC) was cycled at 1 C which corresponds to a specific current of 250 mA g−1 (assuming a practical capacity of 250 mAh g−1) in the voltage range of 2.0 and 4.8 V. For the cells, a mix of 1 M LiPF6 in EC and DMC (1:1 by weight) was selected as electrolyte. Prior to cycling at 1 C (= 150 mA g−1, assuming a theoretical capacity of 150 mAh g−1), three formation cycles using a current rate of 0.2 C for charge and discharge have been performed for the NMC111 electrodes. Within the first three cycles, charging proceeded either until the potential reached the upper limit of 4.3 V vs Li/Li+ or after a step time of 10 h. With regard to the discharge, either the cells were charged to 3.0 V vs Li/Li+ or charging was aborted after 10 h. In the subsequent cycles, the same potential limits were chosen; however, the time criterion was decreased in both cases to 2 h. The utilized electrolyte consisted of 1 M LiPF6 in EC and ethyl methyl carbonate (EMC) (1:1 by weight). All measurement data (capacities and efficiencies) shown in Figures 3 and 6 are average values based on three different cells for each material.

In general, there are three main types of anode materials for LIBs according to their different lithium storage mechanisms: (I) intercalation/insertion materials such as graphite/amorphous carbon,26 Li4Ti5O12 spinel oxides or TiO2 anatase,27−29 (II) the high-capacity “lithium alloying” materials, such as Si, Sn, and Ge,30−33 which have various operating delithiation potentials, with Si showing the lowest potential,31 and (III) conversion materials, such as transition metal oxides, sulfides, or nitrides, which store charge via a conversion reaction.34 The conversion reaction can be generalized by the equation MaXb + (b·n)Li+ + (b·n) e− ⇌ aM + bLinX, where M is the transition metal, X is the anion (most commonly oxide), and n is the formal valence state of X.35,36 In this work, we only want to focus on host materials for LIBs and do not consider lithium metal batteries (such as Li/S, Li/O2, etc.); thus, metallic lithium as high-capacity anode material will not be discussed. Currently, there is a great search for alternatives to graphite anode materials as they possess relatively low specific capacities (theoretically 372 mAh g−1) and not satisfactory rate capabilities, in particular during charge (lithiation).37−39 However, considering that the energy content of a battery cell is the product of capacity and cell voltage, it needs to be taken into account that there is no LIB anode material other than graphite, which possesses such a low (and constant) delithiation potential of ca. 0.2 V vs Li/Li+ (see Figure 1c). In comparison to graphite, alloying materials and conversion materials obviously provide increased gravimetric and volumetric capacities (Figure 1a,b). The alloying materials Si, Sn, and Ge exhibit these high capacities, while still having a moderately low average delithiation potential of ca. 0.43 to 0.5 V vs Li/Li+.31−33 Conversion materials, in particular those using lightweight metals M, typically show relatively high gravimetric capacities in the range of 600 to 1000 mAh g−1 and also high volumetric capacities in the range of 3000 to 5000 mAh cm−3, at least at the material level.40 Among the conversion-type anode materials, ternary metal oxides, such as ZnFe2O4 (ZFO), are currently in the focus of research and are further considered as very promising materials to replace graphitic carbons due to their high specific capacity.41−44 However, for the majority of these materials, this enhanced capacity is achieved at the expense of an average delithiation potential, which is enhanced to above 1 V vs Li/Li+, e.g., for ZFO at ca. 1.5−1.7 V vs Li/Li+ (Figure 1c). Hence, the specific energy/energy density of a LIB full cell, which is calculated as the product of cell voltage and gravimetric/volumetric capacity, will be negatively affected by the resulting decreased cell voltage (see eq 10, Supporting Information). In addition, in a battery system, consisting of a series of battery cells, more cells are needed to achieve a certain battery voltage and accordingly more cell interconnectors and battery management system components are required, resulting in larger costs on the battery system level. It has to be considered that, especially, conversion materials exhibit a relatively poor first cycle CE, which leads to a poor anode/cathode capacity balancing ratio to counterbalance the Coulombic inefficiency of mainly the first as well as the subsequent cycles. Table 1 illustrates the results of a simple model calculation, where LiFePO4 (LFP) is used as a virtual cathode material in combination with graphite, silicon/graphite composite, or ZFO as anode material in a full cell setup. A detailed description of the energy calculation is given in the Supporting Information. Since LIB manufacturing usually involves a capacity-oversized anode (10−15% excess capacity compared to the capacity of the cathode), different anode/

3. RESULTS AND DISCUSSION 3.1. Analysis of Anode Materials in Terms of Capacity, Operating Potential, and Energy. Figure 1 displays the gravimetric and volumetric capacities (a and b) and the operating potential as well as the average delithiation potential (potential at half of the delithiation capacity for anode materials) (c) for different intercalation/insertion materials, alloying materials, and conversion materials. In terms of the gravimetric or volumetric energy density of a LIB full cell, not only the specific capacities of both the anode and cathode but also the cell voltage, which is the difference between the cathode potential and anode potential, plays an important role (see eq 10, Supporting Information). D

DOI: 10.1021/acs.chemmater.6b02895 Chem. Mater. XXXX, XXX, XXX−XXX

Article

Chemistry of Materials Table 1. Calculation of the Cell Voltage as Well as the Specific Energy of LIB Full Cells with Different Anode/ Cathode Balancing Ratiosa graphite −1 b

360 practical capacity (mAh g ) potential vs Li/Li+ 0.2 cell voltage (V) 3.20 Coulombic efficiency in the 1st cycle (%)b 95 Qa/Qc = 1/1 specific energy (Wh/kg) 324 Qa/Qc = 1.1/1 specific energy (Wh/kg) 315 Qa/Qc = 1.2/1 specific energy (Wh/kg) 306

Si/C 1100 0.4 3.00 85

Table 2. Calculation of the Cell Voltage as Well as the Specific Energy of Full Cells with Different Anode/Cathode Capacity Balancing Ratiosa

ZFO

graphite −1 b

1000 1.7 1.70 75

374

205

369

202

365

199

360 practical capacity (mAh g ) potential vs Li/Li+ 0.2 cell voltage (V) 3.20 Coulombic efficiency in the 1st cycle (%) 100 Qa/Qc = 1/1 specific energy (Wh/kg) 354 Qa/Qc = 1.1/1 specific energy (Wh/kg) 344 Qa/Qc = 1.2/1 specific energy (Wh/kg) 334

a

Si/C

ZFO

1100 0.4 3.00 100

1000 1.7 1.70 100

419

234

414

231

409

228

a

Either graphite, silicon/graphite (Si/C), or ZFO were employed as anode and LiFePO4 (practical capacity: 160 mAh g−1; operating potential: 3.40 V vs Li/Li+) as cathode. The energy values are only based on the active material weight without any inactive materials. b Assumed capacity and CE values.

Either graphite, silicon/graphite (Si/C), or ZFO were employed as anode and LiFePO4 (LFP) as cathode. The values are only based on the active material weight without any inactive materials. The displayed data considers a pre-lithiation step; hence, the Coulombic efficiency of each material is set to 100% in the 1st cycle. bAssumed capacity values.

cathode capacity balancing ratios (based on the practical capacity) are also considered in these calculations. In addition, the respective first cycle CE of the three anode materials as well as the one of the LFP cathode (see Supporting Information) have been considered with regard to the mass loading of each material. Typically, graphite-based anodes can display high initial CEs, i.e., in the range of 95%. In contrast, the Si/C and in particular the ZFO anode display a remarkably low first cycle CE, i.e., 85% and 75%, respectively. At first sight, due to the different anode delithiation potentials, the resulting cell voltages of the LIB full cells with a LFP cathode would show a clear deviation from each other. Due to the low average delithiation potential of graphite at 0.2 V vs Li/Li+ as compared to the average delithiation potential of ZFO of 1.7 V vs Li/Li+, a higher cell voltage would be obtained for the graphite/LFP model cell system. The difference of 1.5 V in cell voltage has a significant impact on the specific energy. In all scenarios (compare Table 1), the ZFO-based full cells clearly show the lowest specific energy, despite the high gravimetric capacity of ZFO, and particularly due to the low initial CE and the high delithiation potential. Even though graphite has the lowest theoretical capacity of the three anode materials, the specific energy on the full cell level is still higher than for ZFO. However, the model calculations also show that the use of silicon/graphite composite anodes still results in enhanced energy contents, pointing out the high interest in Si-based materials to replace graphite anodes. In general, it can be stated that higher anode/cathode capacity balancing ratios yield in a decrease of the specific energy of the full cells due to the introduction of extra active anode material that is not utilized (= lithiated). Table 2 presents an analogous approach for the calculation of the specific energy of LIB full cells. Here, a prelithiation step of the anode is assumed, which leads to a CE of 100% for each material. It has to be noted that the additional weight of the lithium from the prelithiation step is neglected in the calculations, since we wanted to only show the general trend. A consideration of the additional weight would lead to a decreased specific energy for all anode materials, but especially for ZFO, which provides the lowest CE. Compared to the data shown in Table 1 (without theoretical prelithiation), higher energy values are obtained for all different anode/cathode

capacity balancings, but the specific energy still follows the order ZFO < graphite < silicon/graphite. As mentioned above, the specific energy decreases by higher anode/cathode capacity balancing ratios. Consequently, the impact of the delithiation potential overrules the influence of the first cycle CE, since the lower CE of ZFO (75%) in comparison to graphite (95%) has only a minor influence on the specific energy. The total cell capacity is limited by the cathode material, which specific capacity is usually lower than that of the anode material. This capacity imbalance is even enlarged when an anode material with a higher specific capacity is deployed. As a result, the increase in the anode potential and, thus, lower cell voltage can only be partially compensated by a higher specific capacity, yielding an overall lower specific energy. Based on these simple calculations, one can say that silicon/graphite composite anode materials are of greatest importance to replace graphite and thus to increase the energy of LIBs. 3.2. Analysis of Anode Materials in Terms of Coulombic Efficiency, Voltage Efficiency, and Energy Efficiency. Beside the specific energy, the energy efficiency (EE) plays a so far completely underrated role for the final cell performance. Since the EE is directly proportional to the Coulombic efficiency (CE) and voltage efficiency (VE), both parameters will have a major impact on the final EE. In general, the CE depends on various fators, such as • Material and electrode properties that influence SEI and CEI formation during the first charge/discharge cycles (e.g., specific surface area, particle size, electrode porosity, etc.) • Material stability and mechanical electrode stability during ongoing cycling (e.g., aging effects such as electrolyte degradation or particle disconnection, etc.) • Lithium storage properties of materials (e.g., selfdischarge behavior) • Operating parameters such as temperature, voltage range, and charge/discharge rate (e.g., electrolyte stability at certain conditions, dwell time at highly oxidizing potentials, etc.) In addition, also the VE is reliant on various factors, which include in particular: E

DOI: 10.1021/acs.chemmater.6b02895 Chem. Mater. XXXX, XXX, XXX−XXX

Article

Chemistry of Materials • Mechanism of lithium storage (intercalation/insertion materials vs alloying materials vs conversion materials) and material properties (e.g., specific surface area, particle size) and material stability (aging effects such as structure/phase changes) • Electrode properties such as the composition (e.g., amount of conductive carbon), mass loading, coating thickness, and porosity • Electrolyte formulation and the influence on the electrode/electrolyte interface (e.g., resistance increase during cycling) • Operating parameters such as temperature and charge/ discharge rate Some of these parameters will be discussed in more detail for certain materials in the following. Since the comparison of anode and cathode materials for high-energy LIBs is in the focus of this study, we kept other parameters such as the operating conditions (charge/discharge rate and temperature) as well as electrode properties (e.g., mass loading) as similar as possible in order to ensure a suitable comparison on the material level. Figure 2 displays the potential vs normalized capacity plots for four different LIB anode materials, namely, graphite, soft

Furthermore, considering the fact that the voltage hysteresis of conversion materials is not well understood up to date, it may be improved in the future. For example, the comparison of different Co-based conversion electrodes showed a trend in the overpotentials following the order fluoride > oxide > sulfide > nitride > phosphide. For these materials the charge/discharge polarization accounted for around 0.4−1.1 V.45−47 The voltage hysteresis for different electrochemical systems has been discussed in literature; however, a potential offset in most cases is caused by multiple reasons.48 For example, the potential offset during intercalation and deintercalation of lithium ions in hydrogen-containing amorphous carbon was attributed to the formation of a lithium−carbon bond on hydrogen terminated edges of hexagonal carbon fragments, leading to an sp2 → sp3 bond transition.48,49 In general, factors that might decrease the voltage hysteresis are the lessening of the charge and discharge rates as well as minimization of particle size and electrode thickness. The main reason for this influence is attributed to a decrease of the polarization.50 However, kinetic factors are not the only origin of hysteresis. A gap between the charge and discharge curve is maintained even when only small currents are applied during cycling. Dreyer et al. termed this potential offset at low currents as “zero-current gap”. They postulated that the appearance of hysteresis in many-particle systems of insertion electrodes might be a consequence of the presence of several equilibria located at different branches.50 For Si−Sn anodes, the voltage hysteresis was assumed to occur by the differences in energy dissipated during the changes in the local atomic environment of the host materials.51 The electrochemical lithiation and delithiation kinetics in silicon were analyzed by Sethuraman et al. For the LixSi system, they concluded that the voltage hysteresis between the lithiation and the delithiation reactions is related to a very large kinetic resistance.48 Upon lithiation a significant expansion of the silicon particles takes place resulting in high stresses, which affect electrochemical lithiation and delithiation. Lu et al. reported that the uptake of lithium is impeded by compressive stress in the surface layer of active materials. Hence, an extra overpotential is required to overcome the stress-induced reaction barrier.52 The origin of the voltage hysteresis of conversion materials has not been fully understood, yet. Different factors such as structural reorganization during charge/discharge accompanied by phase transformations and volume changes have been reported as possible causes; however, the impact of these factors seems to be mainly limited to the first cycle, as no significant changes of, e.g., the surfaceto-bulk sites and the crystalline/amorphous character have been observed in the subsequent cycles.35 Bond energy changes, as inevitably occurring in each lithiation/delithiation cycle, where (partially) covalent chemical bonds are formed/broken, have to be considered, in particular, but not only for conversion materials as possible contribution to the apparent losses of energy observed in the EE plots. Figure 3 shows the course of the delithiation capacity as well as the interplay of CE, EE, and VE of the different anode materials during charge/discharge cycling. The comparison of the intercalation/insertion materials graphite and soft carbon (Figure 3a,b) shows almost comparable values for the CE; after the first formation cycles, the CE rises to almost 100%. However, a comparison of the EE shows an average value (cycle 7−50) of 93.0% for soft carbon and 93.8% for graphite. Generally, the difference in EE and VE is very small after formation, whereas always slightly higher values (∼0.1−0.2%)

Figure 2. Representative potential vs normalized capacity plots of graphite, soft carbon, silicon/graphite (Si/C) composite, and ZFO negative electrodes.

carbon, silicon/graphite (Si/C) composite, and ZFO. Hence, at least one representative of the different anode material categories presented in Figure 1 is also regarded in Figure 2. Since the EE is not only influenced by the CE but also by the VE, the voltage hysteresis found for the four different anode materials will inevitably have an impact on the energy efficiency of the respective material. A large voltage hysteresis will result in a poor VE and, consequently, the EE will also be poor. In the Supporting Information, a detailed description as well as a graphical illustration (see Figure S1) of the EE calculation is given. In comparison of the four anode materials, the biggest potential offset is observed for the conversion anode material ZFO, while the hysteresis is significantly lower for the silicon/ graphite composite. In the case of soft carbon and graphite only a small hysteresis is visible. In general, it should be noted that the voltage hysteresis of conversion materials strongly depends on the respective material and may differ remarkably. F

DOI: 10.1021/acs.chemmater.6b02895 Chem. Mater. XXXX, XXX, XXX−XXX

Article

Chemistry of Materials

Figure 3. Delithiation capacity, Coulombic efficiency, energy efficiency, and voltage efficiency vs cycle number of (a) graphite, (b) soft carbon, (c) silicon/graphite composite, and (d) ZFO. After formation, the cells were charged/discharged at 1 C.

ranging between 1143 mAh g−1 at the beginning of cycling and 816 mAh g−1 in cycle 100. The larger voltage hysteresis of Si/C as compared to graphite or soft carbon that was illustrated in Figure 2 results in a lower VE. The maximum value accounts for 91.8% in the fifth cycle and continuously decreases to 88.6%. Figure S2 (see Supporting Information) illustrates that the voltage hysteresis continuously increases from cycle 20 to cycle 100, which is in accordance with the decline of the VE. In

are obtained for the VE. Both materials exhibit a clearly higher VE of almost 99% within the formation cycles as compared to the subsequent cycling at 1 C, which should be mainly attributed to the polarization effects that occur at higher current rates.50 The delithiation capacity levels off at an average value (cycle 6−100) of 351 mAh g−1 and 226 mAh g−1 for graphite and soft carbon, respectively. The silicon/graphite composite electrodes provide a significantly higher delithiation capacity G

DOI: 10.1021/acs.chemmater.6b02895 Chem. Mater. XXXX, XXX, XXX−XXX

Article

Chemistry of Materials

Figure 4. (a) Energy efficiency and delithiation energy vs charge/discharge C-rate and (b) potential vs normalized capacity curves of ZFO at different C-rates.

of the delithiation energy shows a slightly different picture. Up to 1 C an almost comparable trend for both EE and delithiation energy is observed, but when the rate is further increased the delithiation energy levels off, while the EE decreases almost linearly. The reason for the different trends of EE and delithiation energy is related to the development of the voltage hysteresis. In other words, the charge and discharge potential profiles recorded at different rates change in a different way. Figure 4b displays the potential profiles vs normalized capacity at different rates ranging between 1 and 5 C. Generally, for a high specific energy of a full cell, the anode delithiation potential should be as low as possible to result in a maximum cell voltage. Having a closer look at the potentials for lithiation and delithiation at 1 and 2 C shows that up to a normalized capacity of 0.25 the charge energy is higher at 2 C, while this trend is reversed when the normalized capacity ranges from 0.25 to 1. In summary, a lower charge energy is attained at 2 C. In contrast, a lower discharge energy is recorded at 2 C over the whole range. Taking into account the area of the potential curves at 1 and 2 C, it is obvious that the charge energy decreases to a lower content than the discharge energy. In contrast to that, the charge energy at 5 C shows a net increase, whereas the discharge energy further decreases. However, the decrease of the discharge energy is more pronounced from 1 to 2 C than from 2 to 5 C. Consequently, a stronger drop of the EE is observed at higher rates, while the discharge energy seems to level off. 3.3. Analysis of Cathode Materials in Terms of Coulombic Efficiency, Voltage Efficiency, and Energy Efficiency. In the last decades, the most commonly used LIB cathode material for small LIB cells for consumer applications has been LiCoO2. As larger LIB cells have become of interest for automotive and stationary applications LiNi1/3Mn1/3Co1/3O2 (NMC-111), and other NMCs with varying Ni:Co:Mn ratios are very often used in these applications. This material provides a specific capacity of ∼170 mAh g−1 at an average discharge potential of ∼3.7 V vs Li/Li+.5 In order to further improve the specific energy of LIBs either the capacity of anode and cathode materials and/or the cell voltage have to be improved. With regard to the cathode materials either materials operating at high potentials such as spinel LiNi0.5Mn1.5O4 (LNMO) or materials with a higher specific capacity, e.g., Li-rich xLi2MnO3·(1 − x)LiMO2 (M = Mn, Ni, Co) bears the potential to contribute to a higher energy content of LIBs. Owing to the high working potential of

consequence, the EE will also be negatively affected by the low values of the VE. With regard to the CE of Si/C, a slight decrease from 98.9% after formation to 98.4% in cycle 34, followed by an increase in the CE until the end of cycling to a value exceeding 99%, is monitored. The relatively poor CE, in comparison to the carbon materials, is related to the large volume changes of Si and ongoing SEI formation/electrolyte decomposition, which decreases with ongoing cycling as also the reversible capacity declines. The course of the CE is reflected by the difference of VE and EE; after a growing gap between both parameters until cycle 34, this offset decreases toward the last cycles. Among the investigated anode materials, ZFO shows the lowest EE as a result of a low VE. The average values for the EE and VE (in cycle 3−50) are 62.2% and 64.8%, respectively. However, due to the fluctuating course of the CE, which is most likely related to ongoing SEI formation due to the volume changes of ZFO,44 the gap between VE and EE varies from cycle to cycle. Even though the CE, the VE, and the EE of ZFO alternate stronger than for all the other anode materials, the behavior of the delithiation capacity provides no significant observable fade. After formation, a reversible capacity between 930 mAh g−1 at the beginning and 890 mAh g−1 is attained within the 100 cycles. With the strong focus of academic reports on solid electrolyte interphase (SEI) formation at the anode occurring in the first cycle, where the CE is usually much lower than 100%, it is understandable that the CE has found much more attention than the VE, so far. However, while the EE in the first cycles is dominated by the CE, and in the later cycles the VE has the biggest influence on EE and therefore cannot be neglected for all materials.53,54 In general, it is well-known that the charge/discharge rate has a strong impact on the EE of battery materials, i.e., the EE decreases with increasing rate.55 This fact is very important for applications where “high power charging” is needed, such as for electric vehicles. Thus, apart from lifetime, also the EE and, in turn, the energy costs (see Section 3.4) are strongly influenced by the rate dependency of anode and cathode materials. As an example, ZFO was studied at different C-rates in order to evaluate the impact on the EE and delithiation energy (Figure 4a) as well as on the voltage hysteresis (Figure 4b). By Figure 3d, it was already shown that ZFO exhibits an average EE of around 62.2% at 1 C. The decrease of the applied current to 0.2 C results in an EE of 66.9%, whereas at a higher rate of 10 C a drop of the EE to only 39.2% is recorded. However, the course H

DOI: 10.1021/acs.chemmater.6b02895 Chem. Mater. XXXX, XXX, XXX−XXX

Article

Chemistry of Materials ∼4.7 V vs Li/Li+ of LNMO, conventional organic carbonate based electrolytes suffer from their anodic decomposition.5,7 Kasnatscheew et al. have reported that the low CE of high voltage cathode materials, in particular for NMC-111 at 4.6 V vs Li/Li+, is only partially caused by irreversible electrolyte oxidation but is related to poor Li transport rates into the cathode material during lithiation (= discharge).56 This may also be true for other high voltage cathode materials, such as LNMO, but is currently under investigation. Also, side reactions including transition metal dissolution, especially of Mn2+, as well as surface destabilization occur.57 Hence, the development of more stable electrolytes and additives is crucial for this material.5 Another approach is to move to Li2MnO3-stabilized LiMO2 (M = Ni, Mn, Co) electrodes with a complex “layered−layered” composite structure.58 This class of materials exhibits an average operating potential of ∼3.4 V vs Li/Li+ and a reversible capacity of 200 to 250 mAh g−1 depending on the composition.59,60 During the first charge, a pronounced irreversible capacity loss is observed. This behavior is related to the electrochemical activation of the Li2MnO3 component at ∼4.5 V vs Li/Li+ and is usually accompanied by the irreversible release of gaseous oxygen from the surface and bulk as well as destabilizing structural reorganizations at the surface-near regions. Furthermore, the structure is prone to transition metal (TM) migration and the formation of defect spinel domains resulting in a fast fading of the reversible capacity as well as a gradual voltage decay.5,7 Figure 5 compares the potential versus normalized capacity plots of the 20th cycle for three cathode materials that are

voltage hysteresis as compared to LNMO and NMC. The evolution of the voltage hysteresis during charge/discharge cycling for the three cathode materials is illustrated on a comparative basis for cycles 20 and 50 in Figure S3 (see Supporting Information). With ongoing cycling the spinel material maintains the shape and location of the charge and discharge curves, while NMC displays a slight increase in the potential offset. However, in the case of the LR-NMC a clear increase in the potential offset is obtained from cycle 20 to 50. Generally, the phase transitions occurring within the charging step are accompanied by the extraction of Li ions from the host structure and the withdrawal of electrons from the d orbitals of the transition metal ions (cationic redox). Recent findings also attribute a portion of the capacity of LR-NMC cathode materials to a reversible redox activity of oxygen p electrons (anionic redox),61 which has been confirmed by calculatory and experimental findings of model compounds that incorporate 4d metals.62−64 Accordingly, during discharge Li ions and electrons are inserted into the relative positions within the host structure and into orbitals of the transition metals/oxygen anions, respectively. However, the corresponding energy changes during the phase transitions upon charge and discharge are not equal. The incorporation of lithium ions into the host structure and the uptake of electrons into the d orbitals of the transition metals and oxygen p orbitals during discharge result in an energy decrease and phase stabilization. Hence, energy is needed to promote Li ions and electrons from the lower energy states. Consequently, a slightly higher amount of energy is consumed during charge than is generated during discharge, resulting in an energy difference that causes the potential gap during charge and discharge.65 Different studies were conducted to identify the parameters that influence the voltage hysteresis as well as to determine the origin of this potential offset of xLi2MnO3·(1 − x)LiMO2 (M = Ni, Mn, Co) cathode materials. Croy et al. revealed that the relative amount of Li2MnO3 correlates with the extent of the hysteresis as well as the voltage fade that is observed for xLi2MnO3·(1 − x)LiN0.5Ni0.5O2.66 By a series of GITT experiments a variation of the open-circuit potential for the charge and discharge of different Li-rich materials was found, which was postulated to reflect structural changes in the electrode. Hence, different pathways of the lithium ions during charge and discharge are suggested.67 By means of 6Li magic angle spinning (MAS) nuclear magnetic resonance (NMR), a path-dependent lithium site occupation during charge and discharge was demonstrated. In other words, lithium extraction and insertion do not take place in the same order; e.g., the last host sites, from where Li ions are removed, do not fill up at first during discharge.68 To get a better insight into the electrochemical performance of the three cathode materials, the course of the reversible capacity and CE upon charge/discharge cycling is presented in Figure 6. Furthermore, also the EE and VE of these high specific energy materials are highlighted. The high-voltage spinel LNMO (Figure 6a) provides a stable discharge capacity course close to 110 mAh g−1 at 1 C, while the CE maintains a level of around 99% (except for the first cycles). After formation and achieving a stable plateau, the EE and VE provide average values of approximately 97% and 98%, respectively. In comparison, the Li-rich material (Figure 6b) also provides an average CE of more than 99% (after the first cycles). With regard to the discharge capacity a decrease from

Figure 5. Potential vs normalized capacity plots of LiNi0.5Mn1.5O4 (LNMO), 0.5 Li2MnO3·0.5 LiNi0.4Mn0.4Co0.2O2 (LR-NMC), and LiNi1/3Mn1/3Co1/3O2 (NMC-111). The data corresponds to the 20th cycle of the charge/discharge cycling results illustrated in Figure 6.

currently in the focus of research in the field of LIB cathodes. For the spinel material LiNi0.5Mn1.5O4 (LNMO), which provides the highest potentials for both lithium ion intercalation and lithium ion deintercalation, a very low voltage hysteresis is observed. The layered cathode material LiNi1/3Mn1/3Co1/3O2 (NMC-111) exhibits a comparable electrochemical behavior; i.e., a relatively low potential offset between the charge and discharge potential profile is attained. In contrast to that, the Li-rich material 0.5 Li2MnO3· 0.5LiNi0.4Mn0.4Co0.2O2 (LR-NMC) displays a clearly larger I

DOI: 10.1021/acs.chemmater.6b02895 Chem. Mater. XXXX, XXX, XXX−XXX

Article

Chemistry of Materials

Figure 6. Delithiation capacity, Coulombic efficiency, energy efficiency, and voltage efficiency vs cycle number of (a) LiNi0.5Mn1.5O4 (LNMO), (b) 0.5 Li2MnO3·0.5 LiNi0.4Mn0.4Co0.2O2 (LR-NMC), and (c) LiNi1/3Mn1/3Co1/3O2 (NMC-111). After formation, the cells were charged/discharged at 1 C.

207 mAh g−1 to 192 mAh g−1 is visible after 50 cycles. However, significantly lower EE and VE values are attained for this material as compared to LNMO. Within the first cycles, the maximum value for the EE and VE accounts for both 90%, whereas during charge/discharge cycling a continuous decrease to 85% and 86% is monitored, respectively. Since the EE is directly proportional to the VE, a decrease of the VE will also result in a decline of the EE. In Figure S3, an increase in the voltage hysteresis has been shown from cycle 20 to cycle 50. Due to the rise of the potential offset, it can be concluded that the values of the VE will decrease. Beside the voltage hysteresis also the potential decay (decrease of the average discharge potential), which is a common drawback of the Li-rich materials, has a negative impact on the VE. The origin of hysteresis and potential fade has been ascribed to structural causes including path-dependent lithium extraction and reinsertion as well as defect sides that are formed within the electrode structure upon activation.68 Consequently, the decrease of the VE of the Li-rich material upon cycling is a

result of the increasing voltage hysteresis as well as the ongoing potential fade. The third investigated cathode material, NMC-111, displayed an average discharge capacity of 131 mAh g−1 after formation. However, a slight capacity fading was observed, which is also visible for the CE, EE, and VE. Having a closer look at the course of the EE and VE reveals a growing gap between both values during cycling, which is caused by a declining CE. The decrease of the CE during cycling was already observed in other publications;69−71 however, no statement with regard to the origin of this trend was given. We assume that either intrinsic structure transformations or self-discharge reactions are most likely causing the decrease of the CE, which may be even more enhanced for NMC-111 materials that do not have a surface coating and/or do not use certain electrolyte additives (like in this study). Within the first cycles after formation, an EE of 97% and a VE of 98% are obtained, which are comparable to the values of LNMO. However, with ongoing cycle number, both EE and VE lose around 1% after 50 cycles. J

DOI: 10.1021/acs.chemmater.6b02895 Chem. Mater. XXXX, XXX, XXX−XXX

Article

Chemistry of Materials

Figure 7. (a) Accumulated energy efficiency losses and (b) accumulated extra energy vs cycle number of graphite, soft carbon, silicon−graphite composite (Si/C), and ZFO-based electrodes. The accumulated extra energy (a) mirrors the additional charge energy that is needed to maintain a discharge energy of 90 kWh for the first 50 cycles by taking into account the energy efficiency losses of (a).

Figure 8. (a) Accumulated energy efficiency losses and (b) accumulated extra energy vs cycle number of LiNi0.5Mn1.5O4 (LNMO), 0.5 Li2MnO3·0.5 LiNi0.4Mn0.4Co0.2O2 (LR-NMC), and LiNi1/3Mn1/3Co1/3O2 (NMC-111) positive electrodes. The accumulated extra energy mirrors the additional charge energy that is needed to maintain a discharge energy of 90 kWh for the first 50 cycles by taking into account the energy efficiency losses of (a).

3.4. Impact on Practical Applications: Influence of EE Losses on Extra Electricity Costs. The EE of a certain battery material has a major impact on its practicability, since the EE will contribute to the electricity costs during charge of the battery, i.e., a low EE will lead to enhanced extra electricity costs. In addition, a low EE associated with Joule heating formation asks for extra cooling measures, increasing the energy losses and costs on the battery system level. The accumulated EE losses (a) as well as the accumulated extra energy (b) of the presented anode materials are given in Figure 7, while the corresponding illustration for the cathode materials is displayed in Figure 8. In this context, the accumulated extra energy refers to the additional energy that is needed to compensate the energy inefficiency (as given in section a) while a fictitious accessible energy of 90 kWh is retained. This energy value was chosen since it matches the nominal energy of the battery pack installed in the Tesla Model S.25 In general, the accumulated EE losses follow the order graphite < soft carbon < silicon/graphite ≪ ZFO. However, within the first 50 cycles the sum of the energy inefficiency accounts for 315% and 367% for graphite and soft carbon, respectively, while a distinct increase to 525% is registered for the silicon/graphite composite material. In the case of the conversion material ZFO, where as compared to the other

anode materials the lowest EE is attained, the energy inefficiency adds up to 1916%. With regard to the accumulated extra energy (Figure 7b), a similar trend is obtained compared to the accumulated EE losses. Hence, the lowest extra energy is needed for graphite, while this value is significantly higher in the case of ZFO. The comparison of the cathode materials shows that the lowest accumulated EE losses were found for LNMO, which accounted for 176%. Generally, the accumulated EE losses follow the order LNMO < NMC ≪ LR-NMC. For NMC, a slightly higher value of 189% is obtained. It is noteworthy that LNMO showed a higher accumulated energy inefficiency than NMC up to cycle 43; however, in the subsequent cycles this relation was inversed. The reason for this trend can be found in Figure 6 and Figure S3 (see Supporting Information). While LiNi0.5Mn1.5O4 exhibits a stable course of the CE, the VE, and thus the EE, a decrease of the aforementioned values is monitored for NMC. Additionally, in contrast to the voltage hysteresis of LNMO, which stays almost constant during cycling, an increase in the potential offset is recorded for the NMC material. However, the accumulated energy inefficiency of the LR-NMC electrode accounted for 628% and thus shows more than three times higher energy inefficiency of the other cathode materials within the first 50 cycles. In general, it can be K

DOI: 10.1021/acs.chemmater.6b02895 Chem. Mater. XXXX, XXX, XXX−XXX

Article

Chemistry of Materials

Figure 9. Overview of the extra energy costs in Germany, Japan, and the U.S.A. of ZFO, silicon/graphite composite, soft carbon, and graphite anodes as well as LR-NMC, NMC-111, and LNMO cathodes that are needed to compensate the energy inefficiency of the different materials in order to maintain an energy of 90 kWh during discharge. The cost calculation is based on the annual data for the electricity prices in 2014, including taxes of the respective country.

Figure 10. Classification of anode and cathode materials studied within this work according to their energy efficiency. Classification corresponds to charge/discharge cycling at a rate of 1C for each material. Adopted from the European energy labeling.

stated that, for both anode and cathode materials, the difference in the accumulated energy inefficiency will increase with increasing cycle number. To get an impression of the consequences of the energy inefficiency of the aforementioned materials, the extra charging energy costs for 50 charge/discharge cycles in Germany, Japan, and the U.S.A. are compared in Figure 9. Normalized values (USD/cycle) for each material are shown in Figure S4 (Supporting Information). In particular, the household and industry electricity prices (including taxes) of 2014 are considered for the determination of the costs. The industrial electricity prices accounted for 0.17 $/kWh in Germany, 0.16 $/kWh in Japan, and 0.06 $/kWh in the U.S.A., while the domestic electricity prices were 0.36 $/kWh in Germany, 0.23 $/kWh in Japan, and 0.12 $/kWh in the U.S.A.72,73 The exchange rate was determined on the 11th of December in 2015.74 The calculation is based on the accumulated extra energy data presented in Figure 7b and Figure 8b. Basically, the extra energy costs exhibit the same trend as the accumulated extra energy. The highest costs (1022 USD; 20.4 $/cycle, for a household in Germany) are caused by the application of the

conversion material ZFO, which also displayed the highest accumulated energy inefficiency. The residual materials follow the sequence LR-NMC (240 USD; 4.8 $/cycle) > silicon/ graphite (193 USD; 3.9 $/cycle) > soft carbon (132 USD; 2.6 $/cycle) > graphite (112 USD; 2.2 $/cycle) > NMC (65 USD; 1.3 $/cycle) > LNMO (61 USD; 1.2 $/cycle). However, these costs can be decreased by a factor of approximately 2.2, when industry electricity prices are assumed for these calculations. The comparison of the electricity prices in Germany, in Japan, and in the U.S.A. shows a distinct lower price level in the U.S.A. The industrial prices show a cost reduction by a factor of around 2.6, while the domestic electricity prices are even more than 3 times less expensive. In consequence, the difference of the extra energy costs based on the domestic electricity prices of LNMO and ZFO account for 304 USD in the U.S.A., while in Germany, where the end consumer has to pay higher electricity prices, this difference add up to 961 USD. Please note that the above cost estimations have been made for only 50 charge/discharge cycles, in order to show a general trend. With a longer cycle life the costs will increase. L

DOI: 10.1021/acs.chemmater.6b02895 Chem. Mater. XXXX, XXX, XXX−XXX

Article

Chemistry of Materials

4. CONCLUSION In this work, we evaluated different anode and cathode materials for lithium ion batteries on a comparative basis with respect to their electrochemical characteristics including the specific discharge capacity and energy, as well as energy efficiency (EE), voltage efficiency (VE), and Coulombic efficiency (CE). Furthermore, we presented the practical impact of the EE, i.e., the requirement of the additional energy which is needed to compensate the energy inefficiency of certain materials and the corresponding consequences on the extra electricity costs. Considering the current research on novel anode materials with respect to their electrochemical properties, there is a predominant focus on capacity improvements rather than energy and/or power density improvements, and accordingly, rate capability and not power capability is the measure that counts in most academic reports. As the specific energy of a full cell is not only determined by the specific capacity of the anode material, but also by the cell voltage, a low delithiation anode potential is required to enhance the cell voltage and thus the specific energy. Even though ZnFe2O4 (ZFO), as one representative of conversion anode materials, displays a high specific capacity, it also shows a relatively high delithiation potential (1.5 to 1.7 V vs Li/Li+) and thus cannot compete with graphite in terms of specific energy (on the full cell material level), while the use of silicon/graphite composites results in an enhanced energy, despite the only moderate delithiation potential (0.4 V vs Li/Li+) compared to graphite (0.2 V vs Li/Li+). Thus, beside cost and abundance considerations, silicon/graphite composites are the most promising candidates to replace graphitic carbons as anode material. Another important, but often less noticed, parameter of battery materials is the EE, which describes the ratio of energy input and energy output and is given by the product of the CE and VE. Based on our results, we classified the anode and cathode materials that have been studied in this work according to their EE in different categories (A+++, A++, etc.), as depicted in Figure 10, similar to the energy consumption labeling of instruments, cars, etc. established by the European Union. General trends of energy efficiency can be summarized as follows for the different anode and cathode materials (at similar cycling conditions (charge/discharge rate: 1C, operating temperature: 20 °C) and comparable electrode mass loadings): • EE (anodes): graphite (∼94%) > soft carbon (∼93%) > silicon/graphite (∼89%) ≫ ZFO (∼62%) • EE (cathodes): LNMO (∼97%) ≥ NMC (∼96%) ≫ LR-NMC (∼85%) It also has to be kept in mind that, besides the material properties (lithium storage mechanism, self-discharge behavior, etc.), the electrode characteristics (electrode composition, mass loading, porosity, etc.) and operating conditions (charge/ discharge rate, operating voltage and temperature) will have an impact on the CE and/or VE and, in turn, on the EE. For example, the EE will decrease with increasing charge/discharge rate, which is attributed to an increase in the voltage hysteresis (=decrease in VE), as shown for ZFO. To illustrate the impact of the energy inefficiency of the different anode and cathode materials on practical applications, we determined the accumulated EE losses of each material, which were used to calculate the accumulated extra energy to maintain a discharge energy of 90 kWh that is currently installed in the Tesla Model S. Depending on the country and

the underlying electricity prices (industry vs domestic), a certain cost difference arises to compensate the energy inefficiency of the respective materials. In particular, in countries with high electricity prices, a low EE will be reflected by high extra energy costs. It should be noted that the effect of energy inefficiency on the absolute electricity costs depends on the size (in kWh) of the battery, i.e., extra electricity costs may be negligible, when charging a smart phone but will be significant when charging automotive or even very large stationary/grid batteries. In summary, the TCO (total cost of ownership) of a battery comprises many more aspects than just the battery cost, which is one of the major justifications for the intensive search for alternative battery systems, such as metal/air, metal/sulfur, magnesium, or sodium chemistries, to mention just a few of the current hot spots of research. In fact, most, if not all, of these alternative cell chemistries show poorer EE than the graphitebased LIB. Future works on these alternative systems should also regard improvements in EE to achieve the actually desired goal, i.e., being a truly meaningful alternative battery chemistry with lower cost, based on TCO considerations.



ASSOCIATED CONTENT

S Supporting Information *

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.chemmater.6b02895.



Experimental methods, calculation of energy efficiency, and supporting figures and tables (PDF)

AUTHOR INFORMATION

Corresponding Authors

*(T.P.) [email protected]. Tel.: +49 251 8336826. Fax: +49 251 83-36032. *(M.W.) [email protected] m.winter@fz-juelich. de. Tel.: +49 251 83-36031. Fax: +49 251 83-36032. Notes

The authors declare no competing financial interest.



ACKNOWLEDGMENTS



REFERENCES

The authors wish to thank the German Ministry of Education and Research (BMBF) for funding this work in the project “BenchBatt” (03XP0047A). J.L. kindly acknowledges the financial support of the Federal Ministry of Education and Research (BMBF), the Federal Ministry of Economic and Technology (BMWi), and the Federal Ministry for the Environment, Nature Conservation and Nuclear Safety (BMU) of Germany within the project “KaLiPat” (03EK3008). H.J. acknowledges the German Research Foundation for funding this work in the project “WeNDeLIB” (Priority Programme 1473; Materials with New Design for Improved Lithium Ion Batteries).

(1) Winter, M.; Besenhard, J. O. Wiederaufladbare Batterien. Teil 1: Akkumulatoren mit wäßriger Elektrolytlösung. Chem. Unserer Zeit 1999, 33, 252−266. (2) Gallagher, K. G.; Goebel, S.; Greszler, T.; Mathias, M.; Oelerich, W.; Eroglu, D.; Srinivasan, V. Quantifying the promise of lithium-air batteries for electric vehicles. Energy Environ. Sci. 2014, 7, 1555−1563.

M

DOI: 10.1021/acs.chemmater.6b02895 Chem. Mater. XXXX, XXX, XXX−XXX

Article

Chemistry of Materials (3) Groeger, O.; Gasteiger, H. A.; Suchsland, J. P. ReviewElectromobility: Batteries or Fuel Cells? J. Electrochem. Soc. 2015, 162, A2605−A2622. (4) Van Noorden, R. A Better Battery. Nature 2014, 507, 26−28. (5) Berg, E. J.; Villevieille, C.; Streich, D.; Trabesinger, S.; Novák, P. Rechargeable Batteries: Grasping for the Limits of Chemistry. J. Electrochem. Soc. 2015, 162, A2468−A2475. (6) Patry, G.; Romagny, A.; Martinet, S.; Froelich, D. Cost modeling of lithium-ion battery cells for automotive applications. Energy Sci. Eng. 2015, 3, 71−82. (7) Thackeray, M. M.; Wolverton, C.; Isaacs, E. D. Electrical energy storage for transportation-approaching the limits of, and going beyond, lithium-ion batteries. Energy Environ. Sci. 2012, 5, 7854−7863. (8) Bruce, P. G.; Freunberger, S. A.; Hardwick, L. J.; Tarascon, J. M. Li-O2 and Li-S batteries with high energy storage. Nat. Mater. 2012, 11, 19−29. (9) Christensen, J.; Albertus, P.; Sanchez-Carrera, R. S.; Lohmann, T.; Kozinsky, B.; Liedtke, R.; Ahmed, J.; Kojic, A. A Critical Review of Li/Air Batteries. J. Electrochem. Soc. 2012, 159, R1−R30. (10) Liu, T.; Leskes, M.; Yu, W. J.; Moore, A. J.; Zhou, L. N.; Bayley, P. M.; Kim, G.; Grey, C. P. Cycling Li-O2 batteries via LiOH formation and decomposition. Science 2015, 350, 530−533. (11) Bieker, P.; Winter, M. Lithium-Ionen-Technologie und was danach kommen könnte. Chem. Unserer Zeit 2016, 50, 172−186. (12) Grande, L.; Paillard, E.; Hassoun, J.; Park, J. B.; Lee, Y. J.; Sun, Y. K.; Passerini, S.; Scrosati, B. The Lithium/Air Battery: Still an Emerging System or a Practical Reality? Adv. Mater. 2015, 27, 784− 800. (13) Oudenhoven, J. F. M.; Baggetto, L.; Notten, P. H. L. All-SolidState Lithium-Ion Microbatteries: A Review of Various ThreeDimensional Concepts. Adv. Energy Mater. 2011, 1, 10−33. (14) Kim, J. G.; Son, B.; Mukherjee, S.; Schuppert, N.; Bates, A.; Kwon, O.; Choi, M. J.; Chung, H. Y.; Park, S. A review of lithium and non-lithium based solid state batteries. J. Power Sources 2015, 282, 299−322. (15) Wagner, R.; Preschitschek, N.; Passerini, S.; Leker, J.; Winter, M. Current research trends and prospects among the various materials and designs used in lithium-based batteries. J. Appl. Electrochem. 2013, 43, 481−496. (16) Kang, J. Q.; Yan, F. W.; Zhang, P.; Du, C. Q. A novel way to calculate energy efficiency for rechargeable batteries. J. Power Sources 2012, 206, 310−314. (17) Winter, M.; Appel, W. K.; Evers, B.; Hodal, T.; Möller, K. C.; Schneider, I.; Wachtler, M.; Wagner, M. R.; Wrodnigg, G. H.; Besenhard, J. O. Studies on the anode/electrolyte interface in lithium ion batteries. Monatsh. Chem. 2001, 132, 473−486. (18) Smith, A. J.; Burns, J. C.; Xiong, D.; Dahn, J. R. Interpreting High Precision Coulometry Results on Li-ion Cells. J. Electrochem. Soc. 2011, 158, A1136−A1142. (19) Lu, R.; Yang, A.; Xue, Y.; Xu, L.; Zhu, C. Analysis of the key factors affecting the energy efficiency of batteries in electric vehicle. WEVJ 2010, 4, 9−13. (20) Ponce de Leon, C. P.; Frias-Ferrer, A.; Gonzalez-Garcia, J.; Szanto, D. A.; Walsh, F. C. Redox flow cells for energy conversion. J. Power Sources 2006, 160, 716−732. (21) Glaize, C.; Genies, S. Lithium Batteries and other Electrochemical Storage Systems; John Wiley & Sons, Inc.: Hoboken, NJ, U.S.A., 2013. (22) USA PNGV Battery Test Manual; DOE/ID-10597; February 2001. (23) Laheaar, A.; Przygocki, P.; Abbas, Q.; Beguin, F. Appropriate methods for evaluating the efficiency and capacitive behavior of different types of supercapacitors. Electrochem. Commun. 2015, 60, 21−25. (24) Akinwolemiwa, B.; Peng, C.; Chen, G. Z. Redox Electrolytes in Supercapacitors. J. Electrochem. Soc. 2015, 162, A5054−A5059. (25) U.S. Department of Energy, Model Year 2015 Fuel Economy Guide - Electric vehicles & Plug-in Hybrid Electric Vehicles. https:// www.fueleconomy.gov/feg/pdfs/guides/FEG2015.pdf (accessed July 2016).

(26) Olivier, J. P.; Winter, M. Determination of the absolute and relative extents of basal plane surface area and ″non-basal plane surface″ area of graphites and their impact on anode performance in lithium ion batteries. J. Power Sources 2001, 97−98, 151−155. (27) Bresser, D.; Paillard, E.; Copley, M.; Bishop, P.; Winter, M.; Passerini, S. The importance of ″going nano″ for high power battery materials. J. Power Sources 2012, 219, 217−222. (28) Leitner, K. W.; Besenhard, J. O.; Winter, M. Layer by layer preparation of electrodes with defined thickness by multiple use of the SIC coating process. J. Power Sources 2005, 146, 209−212. (29) Bresser, D.; Paillard, E.; Binetti, E.; Krueger, S.; Striccoli, M.; Winter, M.; Passerini, S. Percolating networks of TiO2 nanorods and carbon for high power lithium insertion electrodes. J. Power Sources 2012, 206, 301−309. (30) Besenhard, J. O.; Yang, J.; Winter, M. Will advanced lithiumalloy anodes have a chance in lithium-ion batteries? J. Power Sources 1997, 68, 87−90. (31) Winter, M.; Besenhard, J. O. Electrochemical lithiation of tin and tin-based intermetallics and composites. Electrochim. Acta 1999, 45, 31−50. (32) Jia, H.; Kloepsch, R.; He, X.; Badillo, J. P.; Gao, P.; Fromm, O.; Placke, T.; Winter, M. Reversible Storage of Lithium in ThreeDimensional Macroporous Germanium. Chem. Mater. 2014, 26, 5683−5688. (33) Winter, M.; Besenhard, J. O.; Albering, J. H.; Yang, J.; Wachtler, M. Lithium storage alloys as anode materials for lithium ion batteries. M. Prog. Batteries Battery Mater. 1998, 17, 208−214. (34) Besenhard, J. O.; Winter, M. Advances in battery technology: Rechargeable magnesium batteries and novel negative-electrode materials for lithium ion batteries. ChemPhysChem 2002, 3, 155−159. (35) Cabana, J.; Monconduit, L.; Larcher, D.; Palacin, M. R. Beyond Intercalation-Based Li-Ion Batteries: The State of the Art and Challenges of Electrode Materials Reacting Through Conversion Reactions. Adv. Mater. 2010, 22, E170−E192. (36) Tien, B.; Xu, M.; Liu, J. Synthesis and electrochemical characterization of carbon spheres as anode material for lithium-ion battery. Mater. Lett. 2010, 64, 1465−1467. (37) Obrovac, M. N.; Christensen, L. Structural changes in silicon anodes during lithium insertion/extraction. Electrochem. Solid-State Lett. 2004, 7, A93−A96. (38) Hossain, S.; Kim, Y. K.; Saleh, Y.; Loutfy, R. Comparative studies of MCMB and C-C composite as anodes for lithium-ion battery systems. J. Power Sources 2003, 114, 264−276. (39) Winter, M.; Besenhard, J. O. Lithiated Carbons. Handbook of Battery Materials; Wiley-VCH Verlag GmbH: 2007; pp 383−418. (40) Courtel, F. M.; Duncan, H.; Abu-Lebdeh, Y. Beyond Intercalation: Nanoscale-Enabled Conversion Anode Materials for Lithium-Ion Batteries. Nanotechnology for Lithium-Ion Batteries; Springer: 2013; pp 85−116. (41) Bresser, D.; Paillard, E.; Kloepsch, R.; Krueger, S.; Fiedler, M.; Schmitz, R.; Baither, D.; Winter, M.; Passerini, S. Carbon Coated ZnFe2O4 Nanoparticles for Advanced Lithium-Ion Anodes. Adv. Energy Mater. 2013, 3, 513−523. (42) Mueller, F.; Bresser, D.; Paillard, E.; Winter, M.; Passerini, S. Influence of the carbonaceous conductive network on the electrochemical performance of ZnFe2O4 nanoparticles. J. Power Sources 2013, 236, 87−94. (43) Varzi, A.; Bresser, D.; von Zamory, J.; Muller, F.; Passerini, S. ZnFe2O4-C/LiFePO4-CNT: A Novel High-Power Lithium-Ion Battery with Excellent Cycling Performance. Adv. Energy Mater. 2014, 4, 1400054. (44) Jia, H.; Kloepsch, R.; He, X.; Evertz, M.; Nowak, S.; Li, J.; Winter, M.; Placke, T. Nanostructured ZnFe2O4 as Anode Material for Lithium Ion Batteries: Ionic Liquid-Assisted Synthesis and Performance Evaluation with Special Emphasis on Comparative Metal Dissolution. Acta Chim. Slov. 2016, 63, 470−483. (45) Oumellal, Y.; Rougier, A.; Nazri, G. A.; Tarascon, J. M.; Aymard, L. Metal hydrides for lithium-ion batteries. Nat. Mater. 2008, 7, 916− 921. N

DOI: 10.1021/acs.chemmater.6b02895 Chem. Mater. XXXX, XXX, XXX−XXX

Article

Chemistry of Materials (46) Nitta, N.; Yushin, G. High-Capacity Anode Materials for Lithium-Ion Batteries: Choice of Elements and Structures for Active Particles. Part. Part. Syst. Char. 2014, 31, 317−336. (47) Yu, S. H.; Lee, S. H.; Lee, D. J.; Sung, Y. E.; Hyeon, T. Conversion Reaction-Based Oxide Nanomaterials for Lithium Ion Battery Anodes. Small 2016, 12, 2146−2172. (48) Sethuraman, V. A.; Srinivasan, V.; Newman, J. Analysis of Electrochemical Lithiation and Delithiation Kinetics in Silicon. J. Electrochem. Soc. 2013, 160, A394−A403. (49) Zheng, T.; McKinnon, W. R.; Dahn, J. R. Hysteresis during lithium insertion in hydrogen-containing carbons. J. Electrochem. Soc. 1996, 143, 2137−2145. (50) Dreyer, W.; Jamnik, J.; Guhlke, C.; Huth, R.; Moskon, J.; Gaberscek, M. The thermodynamic origin of hysteresis in insertion batteries. Nat. Mater. 2010, 9, 448−453. (51) Beaulieu, L. Y.; Hewitt, K. C.; Turner, R. L.; Bonakdarpour, A.; Abdo, A. A.; Christensen, L.; Eberman, K. W.; Krause, J. L.; Dahn, J. R. The electrochemical reaction of Li with amorphous Si-Sn alloys. J. Electrochem. Soc. 2003, 150, A149−A156. (52) Lu, B.; Song, Y.; Zhang, Q.; Pan, J.; Cheng, Y.-T.; Zhang, J. Voltage hysteresis of lithium ion batteries caused by mechanical stress. Phys. Chem. Chem. Phys. 2016, 18, 4721−4727. (53) Winter, M. The Solid Electrolyte Interphase - The Most Important and the Least Understood Solid Electrolyte in Rechargeable Li Batteries. Z. Phys. Chem. 2009, 223, 1395−1406. (54) Schranzhofer, H.; Bugajski, J.; Santner, H. J.; Korepp, C.; Möller, K. C.; Besenhard, J. O.; Winter, M.; Sitte, W. Electrochemical impedance spectroscopy study of the SEI formation on graphite and metal electrodes. J. Power Sources 2006, 153, 391−395. (55) Kang, J.; Yan, F.; Zhang, P.; Du, C. Comparison of comprehensive properties of Ni-MH (nickel-metal hydride) and Liion (lithium-ion) batteries in terms of energy efficiency. Energy 2014, 70, 618−625. (56) Kasnatscheew, J.; Evertz, M.; Streipert, B.; Wagner, R.; Klopsch, R.; Vortmann, B.; Hahn, H.; Nowak, S.; Amereller, M.; Gentschev, A. C.; Lamp, P.; Winter, M. The truth about the 1st cycle Coulombic efficiency of LiNi1/3Co1/3Mn1/3O2 (NCM) cathodes. Phys. Chem. Chem. Phys. 2016, 18, 3956−3965. (57) Gallus, D. R.; Schmitz, R.; Wagner, R.; Hoffmann, B.; Nowak, S.; Cekic-Laskovic, I.; Schmitz, R. W.; Winter, M. The influence of different conducting salts on the metal dissolution and capacity fading of NCM cathode material. Electrochim. Acta 2014, 134, 393−398. (58) Thackeray, M. M.; Kang, S.-H.; Johnson, C. S.; Vaughey, J. T.; Benedek, R.; Hackney, S. A. Li2MnO3-stabilized LiMO2 (M = Mn, Ni, Co) electrodes for lithium-ion batteries. J. Mater. Chem. 2007, 17, 3112−3125. (59) Andre, D.; Kim, S.-J.; Lamp, P.; Lux, S. F.; Maglia, F.; Paschos, O.; Stiaszny, B. Future generations of cathode materials: an automotive industry perspective. J. Mater. Chem. A 2015, 3, 6709− 6732. (60) Li, J.; Klöpsch, R.; Stan, M. C.; Nowak, S.; Kunze, M.; Winter, M.; Passerini, S. Synthesis and electrochemical performance of the high voltage cathode material Li[Li0.2Mn0.56Ni0.16Co0.08]O2 with improved rate capability. J. Power Sources 2011, 196, 4821−4825. (61) Koga, H.; Croguennec, L.; Menetrier, M.; Douhil, K.; Belin, S.; Bourgeois, L.; Suard, E.; Weill, F.; Delmas, C. Reversible Oxygen Participation to the Redox Processes Revealed for Li1.20Mn0.54Co0.13Ni0.13O2. J. Electrochem. Soc. 2013, 160, A786−A792. (62) Saubanere, M.; McCalla, E.; Tarascon, J.-M.; Doublet, M.-L. The Intriguiging Question of Anionic Redox in High-Energy Density Cathodes for Li-ion Batteries. Energy Environ. Sci. 2016, 9, 984−991. (63) Sathiya, M.; Ramesha, K.; Rousse, G.; Foix, D.; Gonbeau, D.; Prakash, A. S.; Doublet, M. L.; Hemalatha, K.; Tarascon, J. M. High Performance Li2Ru1‑yMnyO3 (0.2 ≤ y ≤ 0.8) Cathode Materials for Rechargeable Lithium-Ion Batteries: Their Understanding. Chem. Mater. 2013, 25, 1121−1131. (64) Yabuuchi, N.; Takeuchi, M.; Nakayama, M.; Shiiba, H.; Ogawa, M.; Nakayama, K.; Ohta, T.; Endo, D.; Ozaki, T.; Inamasu, T.; Sato, K.; Komaba, S. High-capacity electrode materials for rechargeable

lithium batteries: Li3NbO4-based system with cation-disordered rocksalt structure. Proc. Natl. Acad. Sci. U. S. A. 2015, 112, 7650−7655. (65) Liu, C. F.; Neale, Z. G.; Cao, G. Z. Understanding electrochemical potentials of cathode materials in rechargeable batteries. Mater. Today 2016, 19, 109−123. (66) Croy, J. R.; Gallagher, K. G.; Balasubramanian, M.; Long, B. R.; Thackeray, M. M. Quantifying Hysteresis and Voltage Fade in xLi2MnO3·(1-x)LiMn0.5Ni0.5O2 Electrodes as a Function of Li2MnO3 Content. J. Electrochem. Soc. 2014, 161, A318−A325. (67) Croy, J. R.; Gallagher, K. G.; Balasubramanian, M.; Chen, Z. H.; Ren, Y.; Kim, D.; Kang, S. H.; Dees, D. W.; Thackeray, M. M. Examining Hysteresis in Composite xLi2MnO3·(1-x)LiMO2 Cathode Structures. J. Phys. Chem. C 2013, 117, 6525−6536. (68) Dogan, F.; Long, B. R.; Croy, J. R.; Gallagher, K. G.; Iddir, H.; Russell, J. T.; Balasubramanian, M.; Key, B. Re-entrant Lithium Local Environments and Defect Driven Electrochemistry of Li- and Mn-Rich Li-Ion Battery Cathodes. J. Am. Chem. Soc. 2015, 137, 2328−2335. (69) Wagner, R.; Streipert, B.; Kraft, V.; Reyes Jiménez, A.; Röser, S.; Kasnatscheew, J.; Gallus, D. R.; Börner, M.; Mayer, C.; Arlinghaus, H. F.; Korth, M.; Amereller, M.; Cekic-Laskovic, I.; Winter, M. Counterintuitive Role of Magnesium Salts as Effective Electrolyte Additives for High Voltage Lithium-Ion Batteries. Adv. Mater. Interfaces 2016, 3, 1600096. (70) Murmann, P.; Schmitz, R.; Nowak, S.; Ignatiev, N.; Sartori, P.; Cekic-Laskovic, I.; Winter, M. Electrochemical Performance and Thermal Stability Studies of Two Lithium Sulfonyl Methide Salts in Lithium-Ion Battery Electrolytes. J. Electrochem. Soc. 2015, 162, A1738−A1744. (71) Murmann, P.; Monnighoff, X.; von Aspern, N.; Janssen, P.; Kalinovich, N.; Shevchuk, M.; Kazakova, O.; Roschenthaler, G. V.; Cekic-Laskovic, I.; Winter, M. Influence of the Fluorination Degree of Organophosphates on Flammability and Electrochemical Performance in Lithium Ion Batteries: Studies on Fluorinated Compounds Deriving from Triethyl Phosphate. J. Electrochem. Soc. 2016, 163, A751−A757. (72) International industrial energy prices. https://www.gov.uk/ government/uploads/system/uploads/attachment_data/file/487759/ table_531.xls. (accessed December 11, 2015). (73) International domestic energy prices. https://www.gov.uk/ government/uploads/system/uploads/attachment_data/file/487772/ table_551.xls. (accessed December 11, 2015). (74) www.oanda.com. (accessed December 11, 2015).

O

DOI: 10.1021/acs.chemmater.6b02895 Chem. Mater. XXXX, XXX, XXX−XXX