“The dose makes the poison.“ ‐‐‐Paracelsus (Philippus Theophrastus Aureolus Bombastus von Hohenheim) (Swiss Medic, Alchemist, Theologist, Astrologist and Philosopher, 1493 ‐ 1541) 31st PSI Electrochemistry Symposium - Electrochemical Energy Storage: A Key for Future Energy Systems, Villigen, CH, 6 May 2015 Electrolytes as Key to New Battery Technologies D. R. Gallus, R. Wagner, J. Kasnatscheew, M. Amereller, B. Streipert, A. Reyes Jiménez, S. Roeser, R. Kloepsch, B. Hoffmann, R. Schmitz, R. Schmitz, A. Berken, V. Kraft, W. Weber, S. Wiemers-Meyer, S. Nowak, I. Cekic-Laskovic and M. Winter MEET Battery Research Center, Institute of Physical Chemistry, University of Muenster, GER & Helmholtz Institute Münster (HI MS), GER Acknowledgments Funding: BMW Group German Ministry of Education and Research (BMBF) within the project “Electrolyte Lab 4E” German Research Foundation (DFG) within the project “High Performance Batteries” Material support by BASF From the Beginning: Electrolytes Enable Batteries The medic Luigi Galvani (IT) generated electricity using a “galvanic cell” with a frog leg as electrolyte between two metal electrodes Galvani(1737‐1798) Volta (1745‐1827) Ritter (1776‐1810) Cell Pile The physicist Allesandro Volta (IT) built the 1st battery (“Volta pile”), using Zn and Cu/air as electrodes and saltwater/cardboard as electrolyte/separator system. The failure of the Volta‐Pile was related to the electrolyte: Drying out, because of H2O evaporation The chemist Johann Wilhelm Ritter (GER) constructed a rechargeable battery, where the electrolyte also participated in the cell reaction: via electrolysis of water during charge; and recombination of H2 and O2 during discharge Electrolyte in the Center of Electrochemical Energy Storage Electrolyte is a system component, that is in contact with many other cell components Electrolyte is a system consisting out of components Electrolytes are decisive for power, life and safety of the battery cell In many cases electrolytes are (co‐) determining the energy, as well From 1791 Until Today From Aqueous to Non‐Aqueous Electrolytes From 1 V to >5 V Batteries 1800 1900 1.1 V Zn-O2 (Volta) 1 V H2-O2 (Ritter) 2V Pb-Acid (Sinsteden, Planté) 1791 1803 1854/59 Primary Recharge- Rechargeable able 2000 1 V 3 V Aqu. Ion Li-Metal Transfer Systems System (Rüdorff, Hoffmann) 1938 1960ies Rechargeable Primary Today + Up to 4.2 V 3 V Li-Ion Li-Ion Transfer Systems Systems 1980ies 1990ies Recharge- Rechargeable able >4.2V Li-Ion Systems Today & Tomorrow Rechargeable High Voltage Cathodes: ‘High Voltage’ is Relative Lightning: Several 10.000.000 Volts High voltage grid: Several 100.000 Volts Static electricity: Several 10,000 Volts Humans: Up to 30.000 Volts Sr/Sr+ Li/Li+ NHE F2/F‐ KrF2/KrF+ ‐4.1 ‐3.040 0.0 3.070* 3.27* E / V vs. SHE *in acidic solution Batteries: Possible: <8V Practical: ≤5 V Typical: 1.2 – 4V The challenge: >4.3V Non‐Aqueous Liquid Electrolyte Economics • Standard non‐aqueous electrolytes are blends of linear and cyclic carbonates High permittivity cyclic carbonate Low viscosity linear carbonates • Blends are formulated to achieve specific physical & chemical electrolyte properties • Typical range of Li salts, such as LiPF6 is 0.8 to 1.2 molar (10 ‐ 15% by weight) • Production of nonaqueus electrolytes in 2013: ca. 30,000 tons • With ca. 12$/kg, electrolyte contributes ca. 6‐8 % of the lithium ion battery material costs • With a mass fraction of <15%, the salt costs are up to 90% of the electrolyte costs depending on the salt concentration, purity, and supplier and use of electrolyte additives • Bulk solvents add only 5 to 15% of the total cost. Additives (0.1 – 5wt.‐%) increase costs • Moderate changes in solvent and additive chemistry are tolerable in view of costs Lower salt content and lower purity requirements can reduce costs Towards High Energy Density LIB with High Voltage (HV) Cathodes E=C·V 2V 1V 0V HV Cathodes NMC at HV NMC Cathode • Energy density can be elevated by: higher specific capacity and higher cell voltage (via cathode potential increase) • The use of high voltage cathodes materials presents a major challenge to the oxidation stability of the electrolyte e.g., organic carbonate solvents: > 4.2 ‐ 4.3 V Cathode CellCell Voltage Voltage 3V Electrolyte Stability 4V ? SEI Potential vs. Li/Li+ 5V Graphite Anode LiNi0.33Mn0.33Co0.33O2 (“1/3‐NMC”) at HV: Enhanced Potential and Capacity • NMC can be charged to different upper cut‐off potentials Lithium Li layer • Higher cut‐off potential HV application higher specific energy Transitionmetal oxide metal oxide layer B. Xu, D. Qian, Z. Wang, Y.S. Meng, Materials Science and Engineering: R: Reports 2012, 73, 51-65 48% Li+ 68% Li+ NMC: Failure Mechanisms at 4.6 V Overall cell impedance rise Oxygen release from cathode Transition metal dissolution NMC de‐ gradation at 4.6 V Oxidative electrolyte degradation Irreversible phase changes [1] H. Zheng, Q. Sun, G. Liu, X. Song, V.S. Battaglia, J Power Sources, 207 (2012) 134-140. [2] V.A. Godbole, J.-F. Colin, P. Novák, J Electrochem Soc, 158 (2011) A1005-A1010. [3] G. Yan, X. Li, Z. Wang, H. Guo, C. Wang, J Power Sources, 248 (2014) 1306-1311. [4] A. Riley, et al., J Power Sources, 196 (2011) 3317-3324. NMC: Failure Mechanisms at 4.6 V Overall cell impedance rise Oxygen release from cathode Transition metal dissolution NMC de‐ gradation at 4.6 V Oxidative electrolyte degradation Irreversible phase changes [1] H. Zheng, Q. Sun, G. Liu, X. Song, V.S. Battaglia, J Power Sources, 207 (2012) 134-140. [2] V.A. Godbole, J.-F. Colin, P. Novák, J Electrochem Soc, 158 (2011) A1005-A1010. [3] G. Yan, X. Li, Z. Wang, H. Guo, C. Wang, J Power Sources, 248 (2014) 1306-1311. [4] A. Riley, et al., J Power Sources, 196 (2011) 3317-3324. High Voltage Application of NMC Use of LiPF6 in Electrolyte at HV Metal Dissolution (Promoted by HF) 180 -1 WE: NMC, CE, RE: Li LiPF6 160 1000 140 120 Upper cut-off potential vs. Li/Li 4.2 V 4.4 V 4.6 V 100 80 1500 Concentration / g L Specific capacity / mAh g -1 2000 0 10 20 30 Cycle number + 500 Ni Co Mn NMC storage in electrolyte for 28 days 0 40 50 3.0 3.5 4.0 4.5 + Electrode potential vs. Li/Li / V • Enhanced average discharge potential • Electrolyte: 1M LiPF6 in EC/DMC (1:1) • Higher specific capacity • Ni, Co, and Mn dissolution • Lower Coulombic Efficiency • Large dissolution at 4.6 V vs. Li/Li+ • Insufficient cycle life D.R. Gallus, R. Schmitz, R. Wagner, B. Hoffmann, S. Nowak, I. Cekic-Laskovic, R.W. Schmitz, M. Winter, Electrochim Acta, 134 (2014) 393-398. Protection of the Cathode vs. Dissolution HF HF Scavenger CEI* Forming Electrolyte Additive Alternative Conducting Salt Cathode Material Coating/ Doping Salt Stabilizer to avoid HF *CEI: Cathode Electrolyte Interphase Alternative Conducting Salt: LiBF4 Less Sensitive to HF formation Enables HV application 160 -1 LiBF4 Concentration / g L Specific capacity / mAh g -1 400 180 WE: NMC, CE, RE: Li 140 120 100 Upper cut-off potential vs. Li/Li 4.2 V 4.6 V 80 60 40 0 10 20 30 Cycle number + 40 • Higher capacity at 4.6 V vs. Li/Li+ • Enhanced cycling stability • Enables HV application Ni Co Mn 300 200 100 NMC storage in electrolyte for 28 days 5 x l o w e r 0 50 3.0 3.5 4.0 + 4.5 Electrode potential vs. Li/Li / V • Electrolyte: 1M LiBF4 in EC/DMC (1:1) • Metal dissolution in the presence of LiBF4 5 times lower than with LiPF6 D.R. Gallus, R. Schmitz, R. Wagner, B. Hoffmann, S. Nowak, I. Cekic-Laskovic, R.W. Schmitz, M. Winter, Electrochim Acta, 134 (2014) 393-398. Protection of the Cathode vs. Dissolution HF HF Scavenger CEI* Forming Electrolyte Additive Alternative Conducting Salt Cathode Material Coating/ Doping Salt Stabilizer to avoid HF *CEI: Cathode Electrolyte Interphase New HF (and H2O) Scavenging Electrolyte Additives: TMS (Trimethylsilyl‐) Based • HF Si F + NH Si N H2O Patent Claim by Saidi et al.*: TMS diethylamine can reduce HF induced transition metal dissolution* Proposal of mechanism by Zhang** Diethylamine = Leaving group (LG) Si OH + NH • • • However: Not stable at high cathode potentials -2 1600 NMC storage at 4.6V vs. Li/Li+ for 28 days 1200 800 400 0 1M LiPF6 in EC/DMC + 1wt% TMS diethylamine Current density / mA cm Mn concentration g L-1 **Mechanism S.S. Zhang, J Power Sources, 162 (2006) 1379-1394 1 M LiPF6 in EC/DMC (1/1) 0.6 + 1 wt% TMS diethylamine 0.4 0.2 WE: LMO; CE, RE: Li Scan rate: 0.1 mV s‐1 0.0 3.0 3.5 4.0 4.5 5.0 5.5 + Potential vs. Li/Li / V *M.Y. Saidi, F. Gao, J. Barker, C. Scordilis‐Kelley, U.S. Patent 5,846,673 (1998) 6.0 Why Using a TMS Group in the Additive? Single bond energy with F / kJ mol-1 H B C Si N P O S 570 646 489 595 278 496 214 368 • Si‐F bond is one of the strongest bonds • Si is a much better electrophile for F‐ than carbon TMS group offers a more attractive electrophilic center than organic carbonate solvents • TMS based additives are known to enhance the effectiveness of the CEI[1‐3] • Structural modification of TMS diethyl amine by variation of the Leaving Group (LG) to achieve better oxidation stability: Si LG Or Or LG Si [1] X. Liao, Q. Huang, S. Mai, X. Wang, M. Xu, L. Xing, Y. Liao, W. Li, J Power Sources, 272 (2014) 501-507. [2] Y. Liu, L. Tan, L. Li, J Power Sources, 221 (2013) 90-96. [3] H. Rong, M. Xu, B. Xie, X. Liao, W. Huang, L. Xing, W. Li, Electrochim. Acta, 147 (2014) 31-39. Design of the HF Scavenger The Choice of the Leaving Group LG by pKa Considerations • Ka: Acid strength • The smaller the pKa value, the higher the acidity of HA and the higher the affinity of A‐ to attract negative charge • Good LGs stabilize the additional negative charge before (in the activated complex) and after heterolytic fission of the Si‐LG‐bond. • This is indicated by a low pKa of the corresponding acid of the LG • Charge stabilization supports the F‐‐scavenging capability of the TMS‐LG additive → “LG Ability” Correlation of pka with HOMO Energy** Oxidation Stability and LG Ability Can Be Estimated by pKa Value HOMO* energy pKa value ~ • A low pKa value is proportional to a low HOMO energy level**, as both are related to a high electron affinity. • Only an approximate, but still useful correlation: The interaction of the LG with the TMS group is unattended by the pkA. • pKa values are listed in tables and unlike HOMO energies do not have to be calculated. *Highest Occupied Molecular Orbital **H.J. Soscún Machado, A. Hinchliffe, Journal of Molecular Structure: THEOCHEM, 339 (1995) 255-258. Effect of TMS Additives on NMC Cycling at HV 105 WE: NMC; CE, RE: Li; 3.0-4.6V vs. Li 1st -3rd cycle: 0.2C; 4th-50th cycle: 1C 200 Coulombic efficiency / % Specific capacity / mAh g -1 240 100 160 120 80 1M LiPF6 in EC/DMC (1/1) + 1wt-% TMS diethylamine + 1wt-% TMS trifluoroacetate 40 0 0 10 20 30 Cycle number 40 50 95 90 1M LiPF6 in EC/DMC (1/1) 85 80 + 1wt-% TMS diethylamine + 1wt-% TMS trifluoroacetate 0 10 20 30 Cycle number TMS diethyl amine: TMS trifluoro acetate: • Better capacity retention • Better capacity retention • Low Coulombic efficiency • Higher Coulombic efficiency • Oxid. decomposition during cycling • Enables HV application 40 50 Capacity retention / % Lower pKa of the LG Better Capacity Retention 100 8 9 4 7 6 80 1 2 3 5 60 Capacity retention: Discharge capacity 5th/50th cycle Ox. stability / HF‐scavenger / More effective CEI 40 20 1 TMS diethylamine 2 TMS isocyanate 3 TMS acetate 4 TMS trifluoroacetate 5 Methoxy TMS 6 N,O-Bis(TMS)acetamide -10 0 7 N,O-Bis(TMS)trifluoroacetamide 8 TMS methanesulfonate 9 TMS TFSI 10 pKa value of the leaving group Based on an initial capacity of ca. 180 mAh g‐1 20 500 1 TMS diethylamine 2 TMS isocyanate 3 TMS acetate 4 TMS trifluoroacetate 5 Methoxy TMS 400 Cum. irr. cap. loss / mAh g-1 Cum. irr. cap. loss / mAh g-1 Cumulated Irreversible Capacity Losses vs. pKa Value and HOMO Energy 6 N,O-Bis(TMS)acetamide 7 N,O-Bis(TMS)trifluoroacetamide 8 TMS methanesulfonate 9 TMS TFSI 300 1 200 7 5 100 9 8 0 -15 -10 -5 6 4 0 23 5 10 15 pKa value of the leaving group • • 20 500 400 1 TMS diethylamine 2 TMS isocyanate 3 TMS acetate 4 TMS trifluoroacetate 5 Methoxy TMS 6 N,O-Bis(TMS)acetamide 7 N,O-Bis(TMS)trifluoroacetamide 8 TMS methanesulfonate 9 TMS TFSI 300 1 200 7 3 100 2 4 5 9 6 8 0 -13.0 -12.5 -12.0 -11.5 -11.0 -10.5 -10.0 HOMO energy level / eV Irreversible capacity losses accumulated in the 4th – 50th cycle: • Are an indication for the oxidation stability of the electrolyte and/or the effectiveness of the CEI. • Can be correlated with the pKa value and the HOMO energy level. Better correlation with the HOMO energy, as TMS group is considered. -9.5 CEI Film Formation? Reverted Trend of Impedance Development During Cycling after Addition of the Additive WE: NMC; CE, RE: Li 1M LiPF6 in EC/DMC (1:1) • Higher Coulombic efficiency with TMS additive may be a hint for a change of the properties of the CEI film. • LiF is known to increase the charge transfer resistance. [1] + 1 wt‐% TMS trifluoroacetate With additive • TMS trifluoro acetate leads to a decrease of charge transfer resistance (3rd → 50th cycle) → Less LiF? [1] X. Zuo, C. Fan, J. Liu, X. Xiao, J. Wu, J. Nan, J Power Sources, 229 (2013) 308-312. Methods to Understand the Mechanisms 19F & 29Si NMR Storage of the electrolytes w/wo additives @60°C, 7 days Determination of the HF content Determination of reaction products of TMS additives GC-MS SEM Post mortem ana‐ lysis of the electro‐ lyte after CC cycling Post mortem ana‐ lysis of the electro‐ des after CC cycling Storage of a water‐ doped electrolyte @ 60°C, 7 days Determination of reaction products of TMS additives Study of the elec‐ trode surface/CEI Proposal for the Mechanism for Reaction of TMS Trifluoro Acetate GC-MS HF Si F Si O Si + Si OH Condensation GC-MS Storage experiment O H2O O Si O CF3 main r Si OH + HO NMR oute HF Si F + HO CF3 Post mortem O CF3 NMR GC-MS • HF scavenging route predominant in LIB, due to higher HF concentration and higher Si affinity towards HF (compared to H2O) • Products after “H2O scavenging“ below the limit of detection (LOD) in the post mortem analysis data. TMS Trifluoroacetate Performance with Graphite Anode • Compatible with graphite • No negative influence on capacity retention • Slightly enhanced Coulombic efficiency Al Passivation in TMS Electrolyte at Smaller HF Amount • Sufficient amounts of HF in the electrolyte are beneficial in order to passivate the Al current collector*: a Al2O3 + 2HF 2AlOF + H2O 2AlOF + 2HF Al2OF4 + H2O Al2OF4 + 2HF 2AlF3 + H2O b Al : Constant voltage @ 4.6 V vs. Li/Li+, 24 h c SEM of Al foils after polarization to 4.6 V vs. Li/Li+ for 24 h a.) 1M LiPF6 in EC/DMC, b.) + 1 wt.‐% TMS trifluoro acetate, c.) 1M LiTFSI in EC/DMC *Krämer, et al. J. Electrochem. Soc. 2013, 160 (2), A356-A360 Krämer, et al. Electrochem. Lett., 2012, 1(5), C1 - C3. Protection of the Cathode vs. Dissolution HF HF Scavenger CEI* Forming Electrolyte Additive Cathode Material Coating/ Doping Alternative Conducting Salt Salt Stabilizer to avoid HF *CEI: Cathode Electrolyte Interphase Previous Reports [1‐3]: Negative Influence of Metal Cations on Cycling Performance Metal dissolution or addition of metal salts Transition metals are incorporated in the SEI Impedance rise at the anode Capacity fading [1] H. Zheng, Q. Sun, G. Liu, X. Song, V.S. Battaglia, J Power Sources, 207 (2012) 134-140. [2] S. Komaba, N. Kumagai, Y. Kataoka, Electrochim Acta, 47 (2002) 1229-1239. [3] J.C. Burns, A. Kassam, N.N. Sinha, L.E. Downie, L. Solnickova, B.M. Way, J.R. Dahn, J Electrochem Soc, 160 (2013) A1451-A1456. Previous Reports [1]: Positive Influence of Metal Cations on Cycling Performance on Graphite 1 M LiTFSA in EC:DEC:TMP (3:3:4, by vol.) Addition of Ca(TFSA)2 [1] S. Takeuchi, S. Yano, T. Fukutsuka, K. Miyazaki, T. Abe, J Electrochem Soc , 159 (2012), A2089. TFSA- = TFSI- Summary of the Investigated Additives Additive No additive LiTFSI NaTFSI Mg(TFSI)2 Al(OTf)3 Ca(TFSI)2 Sc(TFSI)3 Cr(TFSI)3 Ni(TFSI)2 Cu(TFSI)2 Zn(TFSI)2 Supplier, Purity 3M, battery grade Solvionic, 99.5% Solvionic, 99.5% Aldrich, 99.9% Solvionic, 99.5% abcr, 99% Synthesis: In‐house Solvionic, 99.5% Synthesis: In‐house Solvionic, 99.5% 5th cycle Discharge Cap. [mAh g‐1] 167.9 157.1 156.3 144.5 121.0 149.4 168.9 143.3 165.2 167.1 161.9 50th cycle Discharge Cap. [mAh g‐1] 146.4 139.2 136.4 149.8 114.8 119.6 150.7 141.0 146.8 158.1 142.6 Cap. Retention [50th/5th] [%] 87.2 88.6 87.3 103.6 94.9 80.0 89.2 98.4 88.8 94.6 88.1 Electrolyte: 1M LiPF6 in EC:EMC (1:1, by wt.), WE: NMC, 3-electrode cell with Li as CE and RE 3 formation cycles with C/5 and 50 cycles with 1C, Cut-off: 3.0 V to 4.6 V vs. Li/Li+ Effect of Mg(TFSI)2 on NMC Cycling Performance in a Full Cell Discharge capacity / mAh g-1 200 180 1 M LiPF6 in EC:EMC (1:1) 160 1 M LiPF6 in EC:EMC (1:1) + 1% Mg(TFSI)2 140 120 100 after 250 cycles without Mg2+ 80 after 750 cycles with Mg2+ 60 • Coulombic efficiencies in 1. and 5. cycle 40 Electrolyte 20 0 • 2‐Electrode Cell: NMC/MAGD20 15% Anode Excess Cut‐off: 2.8 – 4.5V End of life (EOL) criterion, which indicates 80% of the initial cell capacity after formation, is reached: 0 LP50 (1 M LiPF6 in 100 200 300 400 500 600 700 800 900 1000 EC:EMC, 1:1) Cycle no. LP50 + 3 formation cycles: 15 mA g‐1 (C/10) and CV step at 4.5 V down to C ≤ 0.02C 1000 cycles: 150 mA g‐1 (1C) and CV step at 4.5 V down to C ≤ 0.2C 1% Mg(TFSI)2 1st 5th 78.5% 99.1% 87.9% 99.7% Effect of Mg(TFSI)2 on NMC Self‐Discharge Potential vs. (Li/Li+) / V 4.7 WE: NMC; CE, RE: Li; Cut‐off 4.6 V vs. Li/Li+ 3 formation cycles at C/5 120h storage at OCP 4.6 4.5 4.4 4.3 4.2 1 M LiPF6 in EC:EMC (1:1) 1 M LiPF6 in EC:EMC (1:1) + 1% Mg(TFSI)2 0 20 40 60 80 Time / h • Reduced self‐discharge with Mg(TFSI)2 100 120 Mg(TFSI)2 as HF Scavenger? 19F‐NMR 10 spectroscopy after storage at 60 °C under inert atmosphere in NMR tubes for up to 12 days. 19F‐NMR 1 M LiPF6 in EC:EMC (1:1) 1 M LiPF6 in EC:EMC (1:1) + 2% Mg(TFSI)2 Concentration / mmol 8 6 • HF concentration in the case of the benchmark electrolyte is increasing during storage. 4 • With addition of Mg(TFSI)2, the HF concentration remains below the limit of detection (LOD). 2 0 0 2 4 6 time / days 8 10 12 Mg(TFSI)2 is an effective HF‐scavenger (no intuitive mechanism) or prevents HF formation Mechanism? • We assume that the Mg(TFSI)2 additive has the following functions: Mg(TFSI)2 HF‐Scavenger HF‐Prevention CEI‐film‐ forming additive Stabilization of the host structure ? Reduced HF formation: MgF2 formation Formation of a Mg‐containing film on the surface or sub‐surface by in situ coating: EIS, XPS, SEM, TOF‐SIMS, LA‐ICP‐MS Reduced Li/Ni mixing: XRD, Rietveld refinement Very probable: no bulk‐doping of NMC Investigations on Capacity Fading by EIS: Reverted Trend of Impedance Development During Cycling after Mg2+ Addition 200 300 1 M LiPF6 in EC:EMC (1:1) + 1% Mg(TFSI)2 after 5 cycles after 30 cycles after 50 cycles 150 200 -Zimag / ohm -Zimag / ohm 250 150 100 1 M LiPF6 in EC:EMC (1:1) 50 after 5 cycles after 30 cycles after 50 cycles 0 0 200 400 600 800 Zreal / ohm 100 50 WE: NMC; CE, RE: Li 0 0 100 200 300 400 Zreal / ohm • Impedance increase in Rct during cycling • With Mg2+: Higher Rct after 5 cycles, Growth of a passivation layer on the cathode by electrolyte decomposition? But decrease in Rct during cycling Future Work • Better mechanistic elucidation • Not optimized with regard to additive amount and additive combination • Use with of other cathode materials, e.g. LiNi0.5Mn1.5O2 (LMNO) 1C 1C 5.0 Cycling 110 Self‐discharge 100 Potential vs. (Li/Li ) / V 90 4.8 + Discharge capacity / mAh g -1 C/5 120 D-rate test recovery 80 70 WE: LMNO; CE, RE: Li; Cut‐off: 4.95 – 3.0 V vs. Li/Li+ 60 50 40 1 M LiPF6 in EC:EMC (1:1) 30 20 + 0.1% Mg(TFSI)2 10 + 0.2% Al(Otf)3 20 40 60 WE: LMNO; CE, RE: Li; Cut‐off: 4.95 V vs. Li/Li+ 3 formation cycles: C/5 120h storage at OCP 4.4 1 M LiPF6 in EC:EMC (1:1) 4.2 1 M LiPF6 in EC:EMC (1:1) + 1% Mg(TFSI)2 0 0 4.6 80 100 120 140 Cycle no. 3 formation cycles at C/5, then C/2 D‐rate tests at: D/5, D/3, D/2, 1D, 2D, 3D und 5D 160 0 10 20 30 40 50 60 70 time / h 80 90 100 110 120 NMC: Lithiation vs. De‐lithiation Kinetics: Effect of Rate Switching During Cycling Charge Discharge Specific capacity / mAh g-1 220 16.5% CL 200 180 160 140 120 100 0 2 0.2 C 0.2 D 1.0 C 1.0 D 0.2 C 0.2 D 4 6 50 52 WE: NMC; CE, Re: Li 1M LiPF6 in EC/EMC (1/1) Cut‐off: 4.6–3.0 V vs. Li/Li+ Cycles 1‐3: 0.2C, 0.2D Cycles 4‐53: 1.0C, 1.0D Cycle 54: 0.2C, 0.2D 54 Cycle No. Capacity loss (CL) in the 1st cycle and after rate increase in the 4th cycle Rate switch in cycle No. 54 results in a partial capacity regain Explanation: Lithiation (discharge) of NMC slower than de‐lithiation (charge) Facilitating Lithiation (I): Lowering the Discharge Rate to 0.01 D 1st Cycle Potential vs. Li/Li+ / V 4.5 WE: NMC; CE, Re: Li 1M LiPF6 in EC/EMC (1/1) Cut‐off: 4.6–3.0V vs. Li/Li+ Rate: 0.2 C, ?D 0.2 C - X D 0.2 0.01 4.0 3.5 D‐rate decrease gains 12.8 mAh g‐1 in the 1st cycle 3.0 0 40 80 120 160 200 -1 Specific Capacity / mAh g 240 1. Charge capacity: 222.7 mAh g‐1 2. Discharge capacity (0.2 D): 185.9 mAh g‐1 3. Discharge capacity (0.01 D): 198.7 mAh g‐1 4. Coulombic Eff: 83.5 % 89.2 % Evidence for kinetics effect Achievable discharge capacity and Coulombic efficiency (CE) depend on D‐rate Lithiation (= discharge) of NMC slower than de‐lithiation (= charge) Facilitating Lithiation (II): Constant Potential Holding Step during Discharge at 3.0 V vs. Li/Li+ Potential vs Li/Li+ / V 4.5 WE: NMC; CE, Re: Li 1M LiPF6 in EC/EMC (1/1) Cut‐off: 4.6–3.0 V vs. Li/Li+ Constant Potential Step: 3.0 V vs. Li/Li+ for 24 h Rate: 0.2C, 0.2D 1st Cycle 4.2 3.9 3.6 CP step at 3.0V vs. Li/Li+ gains extra 21.3 mAh g‐1 in the 1st cycle 3.3 3.0 0 40 80 120 160 1. Charge capacity: 221.4 mAh g‐1 2. Discharge capacity (without CP step): 184.9 mAh g‐1 3. Discharge capacity (with CP step): 206.2 mAh g‐1 4. Coulombic Eff.: 83.5 % 93.0 % 200 Specific Capacity / mAh g-1 Evidence for kinetics effect Lithiation (= discharge) of NMC slower than de‐lithiation (= charge) Facilitating Lithiation (III): Lowering the Cut‐off Down to 2.0 V vs Li/Li+ Potential vs Li/Li+ / V 4.5 1st Cycle WE: NMC; CE, Re: Li 1M LiPF6 in EC/EMC (1/1) Cut‐off: 4.6–2.0V vs. Li/Li+ Rate: 0.2C, 0.2D 4.0 3.5 Cut‐off potential decrease gains 2.3 mAh g‐1 in the 1st cycle 3.0 2.5 2.0 0 40 80 120 160 Specific Capacity / mAh g-1 200 1. Charge capacity: 222.7 mAh g‐1 2. Discharge [email protected] vs Li/Li+: 185.9 mAh g‐1 3. Discharge [email protected] vs Li/Li+: 188.2 mAh g‐1 4. Coulombic Eff: 83.5 % 84.5 % Achievable discharge capacity depends also on discharge cut‐off potential !!! Be careful in full cell: At (too) low cell voltages, SEI and/or Cu oxidation may take place at the anode Facilitating Lithiation (IV): Constant Potential (CP) Holding Step in Discharge at 2.0 V vs. Li/Li+ Potential vs Li/Li+ / V 4.5 1st Cycle WE: NMC; CE, Re: Li 1M LiPF6 in EC/EMC (1/1) Cut‐off: 4.6–2.0V vs. Li/Li+ Constant Potential Step: 2.0V vs. Li/Li+ for 24 h Rate: 0.2C, 0.2D 4.0 3.5 Cut‐off potential decrease gains 2.3 mAh g‐1 3.0 2.5 CP step at 2.0 V vs Li/Li+ gains 21.7 mAh g‐1 2.0 0 40 80 120 160 Specific Capacity / mAh g-1 1. Charge capacity: 222.7 mAh g‐1 2. Discharge capacity (without CP step): 188.2 mAh g‐1 3. Discharge capacity (with CP step): 209.9 mAh g‐1 4. Coulombic Eff.: 84.5 % 94.3 % 200 In summary 24 mAh g‐1 (2.3 mAh g‐1 + 21.7 mAh g‐1) of total capacity loss of 36.8 mAh g‐1 in the first cycle could be regenerated by simple kinetic measures 12.8 mAh g‐1 irr. capacity loss Residual capacity loss of 12.8 mAh g‐1 attributable to electrolyte oxidation? 1st Cycle Capacity Losses at Different Upper Cut‐Off Potentials 4.3 V vs. Li/Li+ WE: NMC; CE, Re: Li 1M LiPF6 in EC/EMC (1/1) Cut‐off: ?V–3.0 V vs. Li/Li+ Constant Potential Step: 3.0 V vs. Li/Li+ for 24 h Rate: 0.2C, 0.2D 4.4 Stable electrolyte environment 4.0 40 3.8 3.6 3.9 V 4.0 V 4.3 V 3.4 3.2 Total capacity loss 3.0 0 40 80 120 Specific Capacity / mAh g-1 160 Capacity loss / mAh g-1 Potential vs. Li/Li+ / V 4.2 Total capacity loss Irr. capacity loss 30 20 10 0 Total capacity loss constant up to 4.3 V vs. Li/Li+ 3.9 4.0 4.1 4.2 4.3 4.4 4.5 4.6 After CP step: Irr. capacity loss: constant up to 4.3 V vs. Li/Li+ + Potential / V vs. Li/Li Intrinsic Material problem, e.g., Li/Ni Mixing? Intrinsic Irreversible Capacity Loss: ca. 5 mAh g‐1 Effect of Ni2+ (Trapped on Li+ Sites) on Capacity Loss Li+/Ni2+ mixing is well known for Ni based layered oxide cathodes. M. Hu et al. / Journal of Power Sources 237 (2013) 229-242 Labrini et al.: Irreversible oxidation of the “trapped” Ni2+ to Ni3+ during charge (= delithiation)1 5 mAh g‐1 (observed by us) can be assigned to 1.8% of Li/Ni mixing Literature reports 1‐6% Li/Ni mixing for NMC2 1Labrini, M. Electrochim Acta 2013, 111, 567 Mater 2010, 22, 691. 2Ellis, B. L.; Lee, K. T.; Nazar, L. F. Chem Determination of the Irrev. Charge Loss in the Overall Charge Loss, e.g.: in the 1st Cycle Charge Potential vs Li/Li+ / V 3.8 WE: NMC; CE, Re: Li 1M LiPF6 in EC/EMC (1/1) Cut‐off: 4.6–3.0 V vs. Li/Li+ Constant Potential Step: 3.0V vs. Li/Li+ for 24 h Rate: 0.2C, 0.2D Discharge 3.6 3.4 10 mAh g‐1 parasitic reactions, such as electrolyte oxidation Ca. 5 mAh g‐1 intrinsic material loss 3.2 21.3 mAh g‐1 kinetics effect 36.5 mAh g‐1 initial charge loss 3.0 0 10 20 30 40 50 Specific Capacity / mAh g-1 Note: For a full cell: Apparent “Li loss” at the NMC cathode happens in parallel to Li loss in the SEI at the graphite anode 60 4 4 3 3 CV step at 2.4 V gains 15.1 mAh g‐1 2 2 Cell voltage Graphite Potential NMC Potential 1 0 1 Potential vs. Li/Li+ / V Voltage / V Full Cells with NMC/MCMB WE: NMC; CE: MCMB (30 % oversized); Re: Li 1M LiPF6 in EC/EMC (1/1) Cut‐off: 4.55–2.0 V Constant Voltage Step @Discharge cut‐off voltage for 24 h Rate: 0.2C, 0.2D 1. 2. 3. 4. Charge capacity: 222.7 mAh g‐1 Discharge capacity @ 2.0 V: 186.3 mAh g‐1 (with CP step): 201.4 mAh g‐1 Coulombic Eff.: 83.7 % 90.4 % 0 0 40 80 120 160 Specific Capacity / mAh g-1 200 In summary 15.1 mAh g‐1 of an overall capacity loss of 36.4 mAh g‐1 in the first cycle could be regenerated using a CV step In full cells, the max. possible capacity regain is limited by the CE of the specific graphite. “Extra” Li (capacity regain) serves as reservoir for SEI repair and other Li loss mechanisms. Influence of a Higher Cut‐off Potential on Apparent (Kinetic) Capacity Loss of NMC Specific capacity / mAh g-1 40 30 Specific capacity losses: Parasitic reactions Kinetic inhibition Intrinsic NCM loss 36.5 mAh g-1 10.2 mAh g-1 22.1 mAh g-1 ( 4.2 mAh g-1) 17.1 mAh g-1 21.3 mAh g-1 5.0 mAh g-1 5.0 mAh g-1 3.9 4.6 20 Parasitic reactions lead to larger kinetic inhibition (= kinetic capacity loss) 10 0 Potential vs. Li/Li+ Parasitic reactions increase the kinetic inhibition probably due to the formation of passivation layer caused by electrolyte oxidation 1st Cycle Coulombic Efficiencies (CE) at Different Upper Cut‐Off Potentials Coulombic Efficiency / % 100 CE (after CP step): 97% Intrinsic NMC loss 90 Add. parasitic reactions 80 70 4.0 4.1 4.2 CE (conventional): 87% WE: NMC; CE, Re: Li 1M LiPF6 in EC/EMC (1/1) Cut‐off: ?V–3.0 V vs. Li/Li+ Constant Potential Step: 3.0 V vs. Li/Li+ for 24 h Rate: 0.2C, 0.2D CE (conventional) CE (after CP-step) 3.9 +10% 4.3 4.4 4.5 4.6 Potential / V vs. Li/Li+ Highest possible 1st cyle CE (97%) for NMC at a potential of 4.4 V vs. Li/Li+ Capacity loss / mAh g-1 Irreversible and Total Capacity Losses in the 1st Cycle vs. C‐Rate 60 Total capacity loss Irr. capacity loss 45 WE: NMC; CE, Re: Li 1M LiPF6 in EC/EMC (1/1) Cut‐off: 4.6–3.0 V vs. Li/Li+ + Constant Potential Step: 3.0 V vs. Li/Li+ for 24 h Rate: XC, 0.2D 30 15 0 0.01 0.1 1 C-rate High C‐rate: low irr./total capacity loss Low C‐rate: high irr./total capacity loss Why? Potential vs. Li/Li+ / V C‐Rate Decrease Longer Operation Time in Unstable Potential Environment 4.6 0.3 h 0.6 h 1.7 h 3.5 h Unstable electrolyte environment 4.4 4.2 4.0 1.0 C 0.5 C 0.2 C 0.1 C 3.8 3.6 0 4 8 12 Time / h High C‐rate: Shorter exposition in HV region WE: NMC; CE, Re: Li 1M LiPF6 in EC/EMC (1/1) Cut‐off: 4.6 V vs. Li/Li+ Rate: ?C 16 Low C‐rate: Longer exposition in HV region More parasitic reactions in HV region Discharge capacity / mAh g-1 Discharge Capacity in the 1st Cycle vs. the C‐Rate (= Rate during Charge) 195 WE: NMC; CE, Re: Li 1M LiPF6 in EC/EMC (1/1) Cut‐off: 4.6–3.0 V vs. Li/Li+ Rate: ?C, 0.2D 180 165 150 0.01 0.1 1 C-rate High C‐rate: Low parasitic reactions and low additional kinetic hindrance vs. a low charge capacity Low C‐rate: More parasitic reactions and higher additional kinetic hindrance vs. a high charge capacity Compromise: 0.1 – 0.5 C Minimize Effects of Impedance on Capacity loss: CP Step During Charge AND Discharge CP step at 4.6V vs. Li/Li+ gains extra Potential vs. Li/Li+ / V 4.5 4.0 13.3 mAh g‐1 +6.8 mAh g‐1 ‐6.5 mAh g‐1 (Rev. Li extraction) (EL oxidation) 3.5 3.0 0 WE: NMC; CE, Re: Li 1M LiPF6 in EC/EMC (1/1) Cut‐off: 4.6–3.0 V vs. Li/Li+ Constant Potential Step: 4.6 V vs. Li/Li+ for 5 h 3.0 V vs. Li/Li+ for 24 h Rate: 0.2 C, 0.2 D 1. Discharge capacity (No charge CP step): 206.2 mAh g‐1 +6.8 2. Discharge capacity ‐1 mAh g (Charge CP step): 213 mAh g‐1 CP step at 3.0V vs. Li/Li+ gains extra 3. Irreversible capacity 24.8 mAh g‐1 (No charge step): 15.2 mAh g‐1 ‐6.5 4. Irreversible capacity 80 120 160 200 240 mAh g‐1 ‐1 (Charge CP step): 21.7 mAh g -1 40 Specific Capacity / mAh g Using a CP step during charge AND discharge discharge capacity 213 mAh g‐1 ! Li1‐xNi1/3Co1/3Mn1/3O2 x = 0.78 (at 4.6 V vs. Li/Li+) Additional Li extraction “Extra” capacity May lead to phase changes in NMC* *S. C. Yin, Y. H. Rho, I. Swainson, L. F. Nazar, Chemistry of Materials 2006, 18, 1901. Future Research Comparison of Specific Energies With and Without CP Step at Various Cut‐Off Potentials and Temperatures WE: NMC; CE, Re: Li 1M LiPF6 in EC/EMC (1/1) Cut‐off: ?V vs. Li/Li+ Constant Potential Step: 3.0 V vs. Li/Li+ for 24 h Rate: 0.2C, 0.2D 1000 800 600 Operation temperatures: 25 °C and 60°C 400 200 0 48% 48%54% + Li+ Li+ Li 4.2 4.2 (CP) 68% 54% + + LiLi 4.6 75% 68% + + LiLi 80%75% Li+ Li+ 4.6 (CP) 4.6(CP;60°C) Upper cut-off potential / V vs. Li/Li+ Ca. 18% energy loss due to kinetic effects For Comparison: Capacity Loss in LiCoO2, LCO (at 4.3 V) 4.4 WE: LFP; CE, Re: Li 1M LiPF6 in EC/EMC (1/1) Cut‐off: 4.3–3.0 V vs. Li/Li+ Constant Potential Step: 3.0 V vs. Li/Li+ for 24 h Sp. current: 30 mA g‐1 Potential vs. Li/Li+ / V 4.2 4.0 1. Charge capacity: 161.4 mAh g‐1 2. Discharge capacity (without CP step): 154.3 mAh g‐1 3. Discharge capacity (with CP step): 159.1 mAh g‐1 4. Coulombic Eff.: 95.6 % 98.6 % 3.8 3.6 3.4 CP step at 3.0V vs. Li/Li+ gains extra 4.8 mAh g‐1 in the 1st cycle 3.2 2.3 mAh g‐1 irr. cap. loss! 3.0 0 40 80 120 Specific Capacity / mAh g-1 160 161.4 mAh g‐1 of possible 276 mAh g‐1 were extracted (58%) For Comparison: Capacity Loss in LiNi0.5Mn1.5O2, LNMO WE: LNMO; CE, Re: Li 1M LiPF6 in EC/EMC (1/1) Cut‐off: 4.95–3.0 V vs. Li/Li+ Constant Potential Step: 3.0 V vs. Li/Li+ for 24 h Rate: 0.2C, 0.2D Potential vs. Li/Li+ / V 5.0 4.5 1. Charge capacity: 143.9 mAh g‐1 2. Discharge capacity (without CP step): 138.3 mAh g‐1 3. Discharge capacity (with CP step): 140.1 mAh g‐1 4. Coulombic Eff.: 96.1 % 97.4 % 4.0 3.5 CP step at 3.0V vs. Li/Li+ gains extra 1.8 mAh g‐1 in the 1st cycle 3.0 3.8 mAh g‐1 irr. cap. loss! 0 40 80 120 Specific Capacity / mAh g-1 160 Comparison to NMC: Despite higher charging potentials (!) lower irr. capacity loss (3.8 vs. 15.2 mAh g‐1 ) For Comparison: Capacity Loss in LiFePO4, LFPO Potential vs. Li/Li+ / V 4.2 WE: LFP; CE, Re: Li 1M LiPF6 in EC/EMC (1/1) Cut‐off: 4.1–3.0 V vs. Li/Li+ Constant Potential Step: 3.0 V vs. Li/Li+ for 24 h Spec. current: 30 mA g‐1 4.0 3.8 3.6 3.4 CP step at 3.0V vs. Li/Li+ gains extra 5.9 mAh g‐1 in the 1st cycle 3.2 1. Charge capacity: 163.1 mAh g‐1 2. Discharge capacity (without CP step): 155.0 mAh g‐1 3. Discharge capacity (with CP step): 160.9 mAh g‐1 4. Coulombic Eff.: 95.0 % 98.7 % 2.2 mAh g‐1 irr. cap. loss! 3.0 0 40 80 120 Specific Capacity / mAh g-1 160 Capacity Loss: NMC is an Exception Capacity Loss / mAh g‐1 25 3.9‐4.3 V 20 15 10 4.1 V 4.3 V 5 0 LFP 1D LCO Kinetic cap. loss Irr. cap. loss 4.95 V LNMO NMC 2D 3D 2D? Li Transport in Cathode Structure Conclusions • “1/3 NMC” has been used as model electrode for investigations on HV electrolytes using novel classes of electrolyte additives: Trimethylsilyl (TMS)‐ based additives can be selected according to pKa considerations. Additives with inorganic metal cations, such as Mg2+, Al3+, etc. • The additives are enabling NMC operation at 4.6 V and have HF‐preventing properties, probably due to formation of an effective protecting CEI layer on the surface or in the sub‐surface of NMC (maybe also by an additional HF‐scavenging effect) • The CEI formed in the presence of the additives shows a higher impedance at the beginning of cycling, but a lower impedance at higher cycle number (cf. reference electrolyte). So far, we do not see a bulk effect. • Kinetics affect capacity loss and Coulombic efficiency Capacity regains • For NMC, the contribution of irreversible reactions such as electrolyte oxidation to the “irreversible charge loss” at 4.6 V in the 1st cycle can be estimated to be less than 30%. • NMC shows a “unique“ capacity loss behavior campared to LFP, LCO and LNMO “God made the bulk; surfaces were invented by the devil.” ‐‐‐Wolfgang Pauli (Austrian‐Swiss Physicist, 1900‐1958) Back-Up NMC at 4.6V vs. Li/Li+: Capacity Fade vs. Power Fade During Cycling Charge Discharge 220 Specific capacity / mAh g-1 WE: NMC; CE, Re: Li 1M LiPF6 in EC/EMC (1/1) Cut‐off: 4.6–3.0 V vs. Li/Li+ Cycles 1‐3: 0.2C, 0.2D Cycles 4‐53: 1.0C, 1.0D Cycle 54: 0.2C, 0.2D (with Constant Potential Step at 3.0V vs. Li/Li+ down to <0.2 C) 200 180 160 Only 83.5 % of initial capacity could be achieved in the 53rd cycle 140 120 100 0 0.2 C 0.2 D 1.0 C 1.0 D 0.2 C 0.2 D 10 20 30 Cycle No. 40 50 60 Material degradation: Capacity fade Impedance increase: Power fade Most of the “lost capacity” can be re‐ gained at low rates (in the 54th cycle) 184 mAh g‐1 could be achieved in cycle 54, despite previous significant “capacity fade” Power fade is dominant, not capacity fade! Facilitating Lithiation (IV): Increasing the Temperature Potential vs. Li/Li+ / V 4.8 WE: NMC; CE, Re: Li 1M LiPF6 in EC/EMC (1/1) Cut‐off: 4.6–3.0 V vs. Li/Li+ Rate: 0.2 C, 0.2 D 4.4 4.0 Temperature increase in 1st cycle 1. Lowers Polarizations 2. Increase charge capacity 3. Increase discharge capacity 4. Lowers total and irr. capacity loss 5. Increase Coulombic efficiency 3.6 20°C 40°C 60°C 3.2 2.8 0 50 100 150 200 -1 Specific Capacity / mAh g 250 Electrochemical performance strongly depends on temperature NMC Investigations | Johannes Kasnatscheew | Page 63
© Copyright 2026 Paperzz