In Situ NMR Insights into the Electrochemical Reaction of Cu3P

Article pubs.acs.org/cm

In Situ NMR Insights into the Electrochemical Reaction of Cu3P Electrodes in Lithium Batteries Fabrizia Poli,*,† Alan Wong,‡ Jugeshwar S. Kshetrimayum,†,¶ Laure Monconduit,§ and Michel Letellier*,†,⊥ †

Centre de Recherche sur la Matière Divisée (CRMD), FRE 3520 CNRS, 1b rue de la Férollerie, 45071 Orléans, France NIMBE, CEA, CNRS, Université Paris-Saclay, CEA Saclay, Gif-sur-Yvette, 91191 Paris, France § Laboratoire des Agrégats Moléculaires et Matériaux Inorganiques, UMR 5072 CNRS, Institut Charles Gerard, Université Montpellier, 2 Cedex 05, 34095 Montpellier, France ‡

S Supporting Information *

ABSTRACT: This study reports a multinuclei in situ (real-time) NMR spectroscopic characterization of the electrochemical reactions of a negative Cu3P electrode toward lithium. Taking advantage of the different nuclear spin characteristics, we have obtained real-time 31P and 7Li NMR data for a comprehensive understanding of the electrochemical mechanism during the discharge and charge processes of a lithium battery. The large NMR chemical shift span of 31P facilitates the observation of the chemical evolutions of different lithiated and delithiated LixCu3−xP phases, whereas the quadrupolar line features in 7Li enable identification of asymmetric Li sites. These combined NMR data offer an unambiguous identification of four distinct LixCu3−xP phases, Cu3P, Li0.2Cu2.8P, Li2CuP, and Li3P, and the characterization of their involvement in the electrochemical reactions. The NMR data led us to propose a delithiation process involving the intercalation of metallic Cu0 atomic aggregates into the Li2CuP structure to form a Cu0-Li2−xCu1+xP phase. This process might be responsible for the poor capacity retention in Cu3P lithium batteries when cycled to a low voltage.

■

INTRODUCTION In a quest for high capacity electrodes for lithium-ion secondary batteries, copper phosphide (Cu3P) has been investigated as possible negative electrode and demonstrated its good electrochemical performances.1−9 Several routes for the synthesis of Cu3P are possible,1,3 and among them solid-state synthesis at high temperature (HT Cu3P) provides well-crystallized samples. In an attempt to investigate the reaction mechanism involving various LixCu3−xP phases, HT Cu3P samples were studied by X-ray diffraction (XRD).5−7 Through these XRD studies, Cu metal, LixCu3−xP (with x < 0.5), and Li2CuP were identified during the discharge,5 and a progressive Li/Cu substitution with the formation of LixCu3−xP (0 < x < 3) phases associated with copper extrusion was proposed as mechanism.6,7 Although Li3P is generally identified as the final phase formed during the discharge of phosphide-based electrodes, in the case of Cu3P electrodes, the conversion into Li3P was not observed by XRD. Likewise, while Li2CuP was reversibly identified during the charge, LiCu2P was not. These phases (LiCu2P and Li3P) may either not form or be formed with a low crystalline order or, most likely, with crystallites of small size that cannot be detected by XRD. For this reason, different analytical techniques should be considered to obtain complementary information and deduce © XXXX American Chemical Society

the mechanism of the electrochemical reaction of Cu3P with lithium. NMR spectroscopy is sensitive to atomic short-range ordering regardless of the crystalline order; therefore, it can provide additional insights into the electrochemical mechanism of Cu3P in lithium batteries. Although static NMR gives broad and complex spectra compared to high-resolution magic-angle spinning (MAS) NMR, this is the only in situ NMR technique available today for characterizing lithium batteries and supercapacitors8,10−17 and providing real-time information. In a previous work, we presented the in situ 31P NMR measurement on a Cu3P/Li battery;8 however, only a brief data interpretation was reported. In the present study, we report extensive in situ NMR analyses performed on two different NMR-active isotopes, 7 Li and 31P. The results offer valuable information that is not available from previous XRD6,7 and NMR8 studies such as the possible insertion of Cu0-atomic clusters into the Li2CuP phase during the delithiation process and the formation of the fully lithiated Li3P phase. We note that Cu3P is a conversion material Received: December 15, 2015 Revised: February 1, 2016

A

DOI: 10.1021/acs.chemmater.5b04802 Chem. Mater. XXXX, XXX, XXX−XXX

Article

Chemistry of Materials and that therefore metastable LixCu3−xP phases are expected to form during cycling; therefore, in situ real-time NMR characterization is a preferred approach for preventing unwanted electrochemical relaxations when the battery is stopped and dismantled for ex situ NMR characterizations.

■

EXPERIMENTAL SECTION

Sample Preparation. Crystalline Cu3P powders were prepared at high temperature by ceramic route4,6 with a Cu0/P(red) ratio of 2.8:1. The Cu3P powder electrodes were electrochemically cycled using a 10 mm cylindrical cell designed for in situ solid-state NMR measurements. A description of the custom cell can be found in our previous work.8 The cell was assembled under inert atmosphere of argon using a metal Li foil as negative electrode (380 μm thickness, Li-99.9% Aldrich), a glass microfiber sheet as separator (675 μm thickness, Whatman GF/D), and Cu3P powders containing 15% of black carbon as positive electrode. One molar solution of LiPF6 in 1:1 mixture of ethylene carbonate (EC) and dimethyl carbonate (DMC) was used as electrolyte (Selectipur 30, Merck). The Cu3P/LiPF6 1 M (EC/DMC 1:1)/Li cells were cycled between 2.80 and 0.02 V using a multichannel VSP Biologic instrument under galvanostatic conditions and with a cycling rate of C/20 (1 Li inserted or extracted every 20 h). In Situ Static NMR Experiments. In situ 31P and 7Li NMR measurements were carried out on a Bruker Avance DSX360 spectrometer operating at a frequency of 145.78 and 139.98 MHz, respectively (8.46 T). A Bruker static probe was modified to accommodate an external galvanostat connection via a second-order low pass filter. The filter damps the high frequencies collected by the galvanostat and prevents the NMR frequency from being sent to the galvanostat. The battery cell is placed inside a 10 mm NMR detection coil8 and is connected to the galvanostat. 31 P and 7Li NMR spectra were recorded separately with two individual Cu3P/LiPF6/Li cells. Both sets of spectra were acquired using a solid-echo sequence (90°−τ−90°) experiment with a 5 μs pulse length. The 31P and 7Li spectra were recorded with, respectively, a spectral width of 500 and 150 kHz, a recycle delay of 0.5 and 20 s, and a total of 1024 and 168 signals averaging. It should be noted that a short recycle delay (0.5 s) in the 31P experiment was strategically chosen for recording good quality spectra in a restricted 1 h acquisition; however, the signal intensity of the Li3P final product was sacrificed since the T1 of the fully lithiated phase is much longer (∼60 s) than Cu3P ( V > 0.9 (i.e., 0.05 < x < 0.2). This shift and the absence of new signals throughout the process are consistent with a solidsolution phase transformation (single-phase reaction). At 0.90 V (early process 1, x = 0.2), the 31P spectrum exhibits a defined CSA pattern with Ω = 240 ppm and δiso = −167 ppm, indicating the formation of a single phosphorus site Li0.2Cu2.8P. The corresponding 7Li spectrum shows a small shoulder at around −9 ppm. Unfortunately, there are no structural data available for the Li0.2Cu2.8P phase to support the 31P and 7Li NMR results. The 31P and 7Li spectral features at 0.82 V (early process 2, x = 0.5) are similar to those at 0.90 V suggesting the presence of a phase similar in structure to Li0.2Cu2.8P. Conversely, at 0.74 V, the 31P and 7Li signals have evolved significantly (end process 2, x = 1.75). The 31P spectrum displays CSA features (Ω ≈ 474 ppm and δiso ≈ −190 ppm) that could be associated with the early formation of the stable Li2CuP phase identified by XRD.6,7 Figure S1 details the 31P transformation from Li0.2Cu2.8P to Li2CuP (processes 1 and 2): a gradual decrease and disordering of the Li0.2Cu2.8P phase occurs between 0.91 ≥ V ≥ 0.81 (0.2 ≤ x ≤ 0.9) followed by the Li2CuP phase between 0.81 ≥ V ≥ 0.41 (0.9 ≤ x ≤ 2.0) formation. Li2CuP is identified also by the 7Li spectra displaying a quadrupolar splitting with a quadrupolar coupling frequency (vq) of 8.7 kHz. Moreover, a second Li site in Li2CuP is visible as a shoulder at −9 ppm (see the line-fitting in Figure S4). The spectral evidence of two Li sites is in agreement with the XRD crystal structure of Li2CuP (16248-ICSD). The Li2CuP unit cell (Figure 4) consists of a series of phosphorus planes on the a−b

Figure 2. In situ (a) 31P and (b) 7Li NMR spectra (at 8.46 T) of a Cu3P/ LiPF6/Li cell cycled between 2.8 and 0.02 V. Blue spectra correspond to the discharge process, and red correspond to charge. The real-time NMR spectra were collected every hour during the first 1.5 cycles. The total displayed spectra are about 50% of the collected spectra.

Figure 3. Selected in situ 31P (left) and 7Li (right) NMR spectra of a Cu3P/LiPF6/Li cell at different voltages (a) during the first discharge and (b) during the charge. The voltage at which the spectra were collected is marked by circles in Figure 1.

Figure 4. A crystal structure of Li2CuP represented in a 2 × 2 × 2 unit cell (16248-ICSD). It shows the P plane arrangements and the locations of the in-plan Li and interplan Li.

2.353−2.535 Å. The 31P δiso is confirmed by high-resolution MAS spectra (Figure S5). Note that the sharp septet 31P signal at −142 ppm with a 19F-31P J-splitting of 700 Hz corresponds to the liquid LiPF6 electrolyte. A large singlet 7Li peak at −0.7 ppm from the electrolyte is also observed in the starting in situ 7Li spectrum. These 31P and 7Li electrolyte signals are present in all in situ spectra. At 0.95 V (early process 1 with Li content x = 0.15), the 31 P signal of Cu3P has completely disappeared, and it has been replaced by a broad signal at about −97 ppm. The significant 31 P upfield shift (Δ = −254 ppm) suggests that Cu3P has

plane with one crystallographic P site linked to five Li ( 2.2 in the charge. Likewise, the in situ XRD data6,7 showed that the crystalline Li2CuP transforms much faster in the charge than that in the discharge. Interestingly, although the 7Li spectral profile at 1.06 V is similar to that at 0.87 V, a significant upfield shift is observed: the 7 Li quadrupolar signal moves from −1 ppm to −9 ppm (Δ = −8 ppm) and the shoulder from −9 ppm to −15 ppm (Δ = −6 ppm). These changes indicate that the Li2CuP phase has changed form. The same conclusions can be drawn by the 31P spectrum that shows an intense signal at 151 ppm. One hypothesis that may explain the changes in the NMR signals is a possible intercalation of small Cu0 atomic clusters between the P planes of the Li2CuP structure (Figure 4), forming a delithiated phase denoted here as Cu0-Li2−xCu1+xP. The close proximity of tiny metallic Cu0-atomic aggregates to the P and Li sites would cause a shift and shortening the nuclear magnetic relaxations D

DOI: 10.1021/acs.chemmater.5b04802 Chem. Mater. XXXX, XXX, XXX−XXX

Article

Chemistry of Materials

Figure 5. 31P δcg profile of the observed signal throughout the 1.5 cycles. The δcg is obtained by line-fitting with a single Gaussian−Lorentzian line and with the assumption of one P-site (blue circle). However, a second Gaussian line is used in the presence of the well-resolved signal at 150 ppm (red circle).

Figure 6. 7Li intensity profile of the observed narrow peaks at 9 ± 2 ppm (solid blue circle), − 9 ± 2 ppm (open red circle), and −15 ± 2 ppm (open blue circle) corresponding to the interplane Li of Li3P, the interplane Li of Li2CuP, and the tetrahedral Li of Cu0-Li2−xCu1+xP, respectively. The nonzero baseline is due to the overlapping signals including the large electrolyte signal at −0.7 ppm.

the three stable lithiation phases, Li0.2Cu2.8P, Li2CuP, and Li3P, respectively. Even though the overall Cu3P ⇋ Li3P reaction is reversible, the δcg profile of the charge process is not the reversed profile of the discharge, indicating the presence of an unsymmetrical electrochemical reaction as already pointed out by the NMR spectra analysis. A large 31P downfield shift (∼500 ppm) is found during process 3′, immediately after the formation of Li2CuP. This shift is attributed to the presence of Cu0 clusters inside the Li2CuP phase: clusters that in fact might have prevented the formation of Li0.2Cu2.8P in the charge reaction. Despite this, the electrochemical mechanisms of the discharge processes are repeatable as suggested by the similarities between the δcg profile of the first and second discharge. 7 Li Peak Intensity Profile. Unlike 31P, the 7Li chemical shift sensitivity to the structural and compositional changes in Cu3P/ Li cells is small. However, the spectral analyses of 7Li benefit from the small CSA. The 7Li offers resolved narrow lines in the static spectra for the symmetric sites. This enables us to assess the peak intensity evolution of the crystalline LixCu3−xP phases. Figure 6 displays the intensity plot of the interplane Li in Li2CuP (−9 ± 2 ppm), the interplane Li in Li3P (9 ± 2 ppm), and the interplane Li in Cu0-intercalated Li2−xCu1+xP (−15 ± 2 ppm). The plot shows a gradual formation of Li2CuP (solid blue circle) in process 2, followed by the Li3P formation (open red circle) in process 4. The plot also shows the formation of Li0.2Cu2.8P at process 1. The 7Li signal of Li0.2Cu2.8P appears at the same chemical shift (−9 ppm) of the interplane Li of Li2CuP, suggesting that the Li site in Li0.2Cu2.8P might have a similar local coordination environment as that in Li2CuP. A large intensity profile is visible in process 3′, which corresponds to Li2CuP (solid blue circle) and Cu0-intercalated Li2−xCu1+xP (open blue circle). As mentioned above, the significant increase in signal intensity in 31P is the result of a short T1 (possibility from tenths of second to millisecond), which might be caused by the intercalation of small Cu0 atomic clusters in the Li2CuP phase. It is noteworthy that the nonzero baseline observed in the 7Li intensity plot is related to the fact that the deduced intensity is a contribution of overlapping signals including the large electrolyte signal at −0.7 ppm. This has prevented us from observing any

parallel process in the two-phase transformations (increasing one Li phase while decreasing another). Finally, as mentioned above, the overall increase level of the signal intensity in the second discharge is probably related to the interruption of the voltage cycling at the end of the first cycle. Overall NMR Data Interpretation. On the basis of the NMR results, the electrochemical mechanism involved in a Cu3P lithium cell, cycled between 0.02 to 2.8 V, can be summarized as follows: Cu3P + 0.2Li → Li 0.2Cu 2.8P(cryst) + 0.2Cu 0

(process 1)

Li 0.2Cu 2.8P + 1.8Li → Li 2CuP(cryst) + 1.8Cu 0

(process 2)

Li 2CuP + Li → Li3P + Cu

0

(processes 3 and 4′)

Li3P + Cu 0 → Li 2CuP(cryst) + Li

(process 4′)

Li 2CuP(cryst) + Cu 0 − cluster + xCu 0 → Cu 0 − Li 2 − xCu1 + xP + x Li

(process 3′)

Cu 0 − Li 2 − xCu1 + xP(disord) + 2Cu 0 → Cu3P + x Li (processes 2′ and 1′) 1

In a recent report by Stan et al., the mechanism in a Cu3P/Li battery was revisited, and it was demonstrated that when the battery is cycled to 0.02 V, the conversion to Li3P and Cu ions reduces the capacity retention. The authors explain this with the greater energy barrier needed to reconvert Li3P to Cu3P. Our NMR study, in which the battery is cycled to a low potential value 0.02 V, shows that the electrochemical mechanism in the charge is different than that in the discharge. The charge involves a process of Cu0-atomic cluster intercalation into a Li2CuP phase, which most likely requires higher energy and thus diminishes E

DOI: 10.1021/acs.chemmater.5b04802 Chem. Mater. XXXX, XXX, XXX−XXX

Chemistry of Materials

■

the capacity retention performance of the battery. This is also in agreement with the higher polarization (potential difference between the charge and discharge) of the 4/4′ and 3/3′ processes compared to processes 1/1′ and 2/2′.6 Stan et al.1 have demonstrated that a smaller energy barrier is needed for converting an intermediate LixCu3−xP phase to Cu3P when the cycling potential is cutoff at 0.5 V. At this potential, the Li3P phase does not form and thus the formation of free Cu0 is limited (i.e., eliminating the processes 3 and 4 in discharge). One can speculate that the reduced amount of free Cu0 in the cell cycled to 0.5 V might also prevent the formation of Cu0-Li2−xCu1+xP during the charge (process 3′) and hence lower the energy barrier for the conversion into Cu3P and improve the capacity retention of the battery. Although the overall electrochemical reaction pathway has shown to be reversible (Cu3P ⇋ Li3P),21 a nonsymmetrical mechanism between discharge and charge is observed. In the charge, the back conversion of Li2CuP into Cu3P occurs through the insertion of Cu0-atomic clusters between the P-layer of Li2CuP, a reaction that is not observed in the discharge as the Cu0-clusters are likely more favored to be extruded than inserted. The observed nonsymmetrical reaction is in agreement with the electrochemical behavior that shows that the reinsertion of Cu0 into the phosphorus matrix is a more resistive process than the others. Impedance measurements are required to rationalize these features.

AUTHOR INFORMATION

Corresponding Authors

*E-mail: [email protected]. *E-mail: [email protected]. Present Addresses ⊥

Institut Universitaire de Technologie (IUT GEII), Université François Rabelais, Avenue Monge, 37200 Tours, France. ¶ Talent Development Center, Indian Institute of Science, Karnataka 577536, India. Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS We would like to acknowledge the French Agence Nationale de la Recherche (ANR-07-BLAN-0172) for the financial support of the project of in situ NMR on Li battery. We would also like to specially thank Dr. Pierre Florian (CEMHTI, CNRS-Orleans) and the CEMHTI NMR team for assisting the MAS NMR measurements, and also Dr. Thibault Charpentier (CSA-Saclay, LSDRM) for computing the 7Li EFG values.

■

REFERENCES

(1) Stan, M. C.; Klopsch, R.; Bhaskar, A.; Li, J.; Passerini, S.; Winter, M. Cu3P Binary Phosphide: Synthesis Via a Wet Mechanochemical Method and Electrochemical Behavior as Negative Electrode Material for Lithium-Ion Batteries. Adv. Energy Mater. 2013, 3, 231−238. (2) Crosnier, O.; Nazar, L. F. Facile Reversible Displacement Reaction of Cu3 P With Lithium at Low Potential. Electrochem. Solid-State Lett. 2004, 7, A181−A189. (3) Pfeiffer, H.; Tancret, F.; Bichat, M. P.; Monconduit, L.; Favier, F.; Brousse, T. Air Stable Copper Phosphide (Cu3P): A Possible Negative Electrode Material for Lithium Batteries. Electrochem. Commun. 2004, 6, 263−267. (4) Bichat, M. P.; Politova, T.; Pfeiffer, H.; Tancret, F.; Monconduit, L.; Pascal, J. L.; Brousse, T.; Favier, F. Cu3P as Anode Material for Lithium Ion Battery: Powder Morphology and Electrochemical Performances. J. Power Sources 2004, 136, 80−87. (5) Bichat, M. P.; Politova, T.; Pascal, J. L.; Favier, F.; Monconduit, L. Electrochemical Reactivity of Cu3P with Lithium. J. Electrochem. Soc. 2004, 151, A2074−A2081. (6) Mauvernay, B.; Doublet, M. L.; Monconduit, L. Redox Mechanism in the Binary Transition Metal Phosphide Cu3P. J. Phys. Chem. Solids 2006, 67, 1252−1257. (7) Mauvernay, B.; Bichat, M. P.; Favier, F.; Monconduit, L.; Morcrette, M.; Doublet, M. Progress in the Lithium Insertion Mechanism in Cu3P. Ionics 2005, 11, 36−45. (8) Poli, F.; Kshetrimayum, J. S.; Monconduit, L.; Letellier, M. New Cell Design for In-Situ NMR studies of Lithium-Ion Batteries. Electrochem. Commun. 2011, 13, 1293−1295. (9) 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. (10) Chevallier, F.; Letellier, M.; Morcrette, M.; Tarascon, J.-M.; Frackowiak, E.; Rouzaud, J.-N.; Béguin, F. In Situ 7Li-Nuclear Magnetic Resonance Observation of Reversible Lithium Insertion Into Disordered Carbons. Electrochem. Solid-State Lett. 2003, 6, A225−A228. (11) Letellier, M.; Chevallier, F.; Morcrette, M. In Situ 7Li Nuclear Magnetic Resonance Observation of the Electrochemical Intercalation of Lithium in Graphite; 1st Cycle. Carbon 2007, 45, 1025−1034. (12) Chevallier, F.; Poli, F.; Montigny, B.; Letellier, M. In Situ 7Li Nuclear Magnetic Resonance Observation of the Electrochemical Intercalation of Lithium in Graphite: Second Cycle Analysis. Carbon 2013, 61, 140−153. (13) Bhattacharyya, R.; Key, B.; Chen, H.; Best, A. S.; Hollenkamp, A. F.; Grey, C. P. In Situ NMR Observation of the Formation of Metallic

■

CONCLUSIONS The in situ 31P and 7Li NMR data presented here complement well the previous in situ XRD studies performed on Cu3P/Li cells.6,7 At high potential, the 31P spectra show a quick decline of Cu3P content in the discharge, and followed by the formation of Li0.2Cu2.8P. A gradual conversion of Li2CuP and the final phase Li3P in discharge is visible in both 7Li and 31P spectra. In contrast to the in situ XRD, 7Li NMR clearly identifies the Li3P phase. Both 31P and 7Li NMR data also suggest a different electrochemical pathway for the Cu3P → Li2CuP conversion (in discharge) and its reversed reaction, Li2CuP → Cu3P (in charge). A delithiation process involving the insertion of metallic Cu0 atomic aggregates into the Li2CuP structure is proposed; this process may contribute to the poor capacity retention obtained when Cu3P/Li cells are cycled to a low potential of 0.02 V.1 The in situ NMR data, combined with previous XRD data, permit us for an in-depth characterization of the various steps of the electrochemical reaction of Cu3P with Li through the identification of LixCu3−xP phases and to ascertain that the overall electrochemical pathway of the Cu3P/Li batteries is not a complete reversible Li/Cu insertion. This in situ study illustrates that the different characteristics of NMR-active nuclei can be used to obtain comprehensive information on the Li insertion/ desertion mechanism in lithium batteries.

■

Article

ASSOCIATED CONTENT

S Supporting Information *

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.chemmater.5b04802. Incremental 31P spectra of the 1st discharge and the 1st charge; 7Li spectral evolution of the different LixCu3−xP phases; spectral line-fitting; MAS spectra of Cu3P and Li3P; spectral comparison of the Cu3P phase in 1st and 2nd discharge; table of 31P and 7Li NMR parameters (PDF) F

DOI: 10.1021/acs.chemmater.5b04802 Chem. Mater. XXXX, XXX, XXX−XXX

Article

Chemistry of Materials Lithium Microstructures in Lithium Batteries. Nat. Mater. 2010, 9, 504− 510. (14) Trease, N. M.; Koster, T. K-J.; Grey, C. P. In Situ NMR Studies of Lithium Ion Batteries. Electrochem. Soc. Interface 2011, 20, 69−73. (15) Blanc, F.; Leskes, M.; Grey, C. P. In Situ Solid-State NMR Spectroscopy of Electrochemical Cells: Batteries, Supercapacitors, and Fuel Cells. Acc. Chem. Res. 2013, 46, 1952−1963. (16) See, K. A.; Leskes, M.; Griffin, J. M.; Britto, S.; Matthews, P. D.; Emly, A.; Van der Ven, A.; Wright, D. S.; Morris, A. J.; Grey, C. P.; Seshadri, R. Ab Initio Structure Search and In Situ 7Li NMR Studies of Discharge Products in the Li−S Battery System. J. Am. Chem. Soc. 2014, 136, 16368−16377. (17) Salager, E.; Sarou-Kanian, V. M.; Sathiya, M.; Tang, M.; Leriche, J.-B.; Melin, P.; Wang, Z.; Vezin, H.; Bessada, C.; Deschamps, M.; Tarascon, J.-M. Solid-state NMR on the Family of Positive Electrode Materials Li2Ru1‑ySnyO3 for Li-Ion Batteries. Chem. Mater. 2014, 26, 7009−7019. (18) Massiot, D.; Fayon, F.; Capron, M.; King, I.; Le Calvé, S.; Alonso, B.; et al. Modeling One and Two-Dimensional Solid-State NMR Spectra. Magn. Reson. Chem. 2002, 40, 70−76. (19) Ni, S.; Ma, J.; Lv, X.; Yang, X.; Zhang, L. The Fine Electrochemical Performance of Porous Cu3P/Cu and the High Energy Density of Cu3P as Anode for Li-Ion Batteries. J. Mater. Chem. A 2014, 2, 20506−20509. (20) 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. (21) Fullenwarth, J.; Darwiche, A.; Soares, A.; Donnadieu, B.; Monconduit, L. NaP3: a promising negative electrode for Li- and Naion batteries. J. Mater. Chem. A 2014, 2, 2050−2059.

G

DOI: 10.1021/acs.chemmater.5b04802 Chem. Mater. XXXX, XXX, XXX−XXX