The anionic redox activity: From fundamental understanding to practical challenges

Transcription

1 1st International Conference of Young Scientists TOPICAL PROBLEMS OF MODERN ELECTROCHEMISTRY AND ELECTROCHEMICAL MATERIALS SCIENCE September 17 th, 2017 The anionic redox activity: From fundamental understanding to practical challenges J.M. Tarascon

2 Contents General remarks on insertion compounds Setting up the problem Emergence of the anionic redox prcess The science underlying the anionic redox process Widening the spectrum of oxides showing anionic redox Practicality of Li-rich oxides: a mixed blessing Conclusions and perspectives

3 The Chemistry and Physics of Insertion Reactions [Ti +4 S [H] 2 ] + + x Li A n+ + n++ + xn e x - ne e- - [A H [LiTi ] +3 S 2 ] x Matériau hôte Espèce Invité e - e - e - e - e - e - e - 3D 1D Cristallographic structure 2D Open structures (tunnels, layers,...) The crystallographic structure Duality ions/electrons e - e - Li + Li + e - Li + e - Li + s i s e e - Li + e - Li + e - Li + e - e - Li + Li + e - e - Li + Li + e - Li + E (ev) E f s, p p s e - from MO* Oxy + e - Red t 2g Red Oxy + e- N (E) The electronic band structure

4 The anionic redox activity: a transformational change LIB has relied on cationic redox reactions This is no longer true since 2013 (2D Layered oxide) (2D Li-rich layered oxide) Cationic framework Anionic framework 150 mah/g 280 mah/g Li M O x in LiCoO 2 NATURE MATERIALS, 2013, 12, NATURE MATERIALS VOL 14 FEBRUARY 2015, SCIENCE, 2015, 305 (6267)

5 Anionic redox : It s emergence Ligand-hole chemistry in chalcogenides Rouxel (1991) Oxygen activity in LiCoO 2 proposed / theory confirmed it Tarascon / Ceder (1999) Oxygen activity in LiCoO 2 - Experiments Yoon / Dedryvère ( ) Synthesis of Li-rich Li 2 MnO 3 - based cathodes Dahn / Thackrey ( ) The Emergence of Anionic Redox Oxygen redox in Li-rich cathodes Theory Ceder / Doublet (2016) Oxygen redox in Li-rich NMC Direct proofs Bruce (2016) Oxygen redox in Li 2 RuO 3 and Li 2 IrO 3 Direct proofs Tarascon ( ) Suspecting oxygen redox in Li-rich NMC Delmas (2012) vs.

6 Anionic redox in chalcogenides: Rouxel s pioneering work in the 1990 s Layered TiS 2 Metal electronegativity Pyrite FeS 2 Rouxel, Chem. Eur. J., 2, 1996 Creation of ligand holes and S 2 2- pairs 2 S -- (S 2 ) e - M m+ M (m-1)+ - e - 6

7 Experimental evidence of the S participation to the redox process by XPS x= 0 x= 1 Li x TiS 3 Ti 4+ (S 2- ) (S 2 ) 2-2- Li 1 Ti 4+ (S 2- ) 2 (S 2 ) 0.5 Dimère S 2 2- Anion S 2- x= 1.8 Li 1.8 Ti 4+ (S 2- ) 2.8 (S 2 ) x= 2.2 Li 2.2 Ti Ti S 3 M.H. Lindic et al Solid State Ionics, 176 (2005)

8 Potentiel (V vs. Li + /Li ) Existence of CoO 2? No longer a mystery Energie Energy Chemical/electrochemical approaches LiCoO 2 CoO 2 Chemical oxidation (Br 2, NO 2 BF 4 ) Li 1-x (CdI 2 Structure) Electrochemical oxidation CoO x in Li x CoO 2 Co 4-x (O 2- ) 2-2x (O 2-2 ) x d G.G. Amatucci, J.M. Tarascon, et al., J. Electrochem. Soc., Vol. 143, N 3, March 1996 J.M. Tarascon et al. Jr. of Solid State Chemistry 1999 sp sp E F DOS N( ) DOS N( )

9 J. Dahn et al. Journal of The Electrochemical Society, 2002 A few intriguing results which were overlooked P. Kalyani et al, Journal of power source (1999) Li(Li 0.16 Ni 0.25 Mn 0.58 )O 2 Li 2 MnO V 240 mah/g V Extra-capacity triggers by oxygen deficiency Removal of O 2- plus exchange of Li + by H +

10 O 1s core peaks in delithiated Li x CoO 2 compounds Deintercalation from LiCoO 2 to CoO 2 investigated by XPS LiCoO 2 CoO 2 R. Debryvére, et al. Chem. Mater. 2008

11 Voltage (V vs. Li + /Li 0 ) LiCoO 2 has been the stellar material for numerous years The Li x CO 2 system and its evolution Capacité (mah/g) Li 0.5 CoO mah/g x in in Li LixCoO 1-x 22 Improvements via chemical substitutions: Dahn et al., Chemistry of materials 1 Li[Co]O 2 [150 mah/g] T.Ohzuku, Y. Makimura; Chem Lett. 642 (2001) Replace partially Co by Mn et Ni Li[NiMnCo]O 2 [180 mah/g] NMC phases

12 New Li-rich high-capacity layered oxides Voltage (Volts vs. Li + /Li) Potentiel (Volts vs. Li + /Li) Potentiel (Volts vs. Li + /Li ) Li[LiCoNiMn]O 2 Li-excess Solid Solution Li 1 [Li 0.2 Ni 0.15 Mn 0.48 Co 0.17 ]O 2 Co Ni Mn (Dahn:3M) Composites (1-z) [Li 1/3 Mn 1/3 ]O 2 (z)li[mn 0.5-y Ni 0.5-y Co 2y ]O Capacité en mah/g Double fundamental problem Origin of capacity Origin of drop in voltage Dahn et al., Chemistry of materials 15, (2003), High capacity (> 280 mah/g) 250? y in Capacity Voltage fade? mah /g Capacité en mah/g M Thackeray. J/ Mat. Chem 15, 2257 (2005) 50 0

13 Voltage (V vs. Li + / Li 0 ) Potentiel (Volts vs. Li + /Li ) Potentiel (Volts vs. Li + /Li ) How to simplify this problem??? a new chemical approach.. Li[Li 0.2 Mn 0.53 Co 0.14 Ni 0.13 ]O 2 [3 redox centers] Layered structure preserved 3 rd périod 4 th périod Li 2 [Ru 1-x Sn x ]O 3 [redox center] x in Capacity Capacité Li x Ru 0.75 in en SnmAh/g 0.25 O 3 M. Sathiya et al., Nature Materials, 12, 827 (2013) Capacité Identicalidentique capacity 250 mah/g > 260 mah/g Voltage (V vs. Li + /Li 0 ) th cycles 10 th cycles Chute Potential de fading potentiel Nearly n très canceled limité e Capacité x Capacity in x Li in Li Ru O x Ru en in x mah/g Sn O 3 50 th cycles 1 st cycle M. Sathiya, et al. Nature Materials 14, (201

14 M. Sathiya et al., Nature Materials, 12, 827 (2013) Cell Voltage (V) d ''/db M. Sathiya et al., World patent Origin of the redox activity in Li 2 Ru 0.75 Sn 0.25 O 3 X-ray photoemission X-ray spectroscopy photoemission(xps) spectroscopy XPS (fully charged sample) Signal (O 2 ) n- 2 (H 2 O 2 ) (O 2 ) (H 2 O) (O 2 ) n- 2O -- Ru Capacity in mah/g Ru 4+ Electron paramagnetic Electron Resonance paramagnetic (EPR) resonanc RPE (fully charged sample) Signal (O 2 ) n 4.6 V Energiede Binding energy liaison (ev) First direct evidence for an anionic redox process (2O -- (O 2 ) n- ) in Li-rich lamellar compounds Champs Magnetic magnétique field in [mt] [mt]

15 d /db (O 2 ) n- + Ru 4+ Monitoring redox processes in Li 2 Ru 0.75 Sn 0.25 O 2 by operando Electron Paramagnetic Resonance (EPR) Charge (O 2 ) n V Discharge (O 2 ) n- Ru 5+ + O 2- Ru 4+ + O 2- Ru V Vs. Li + /Li 0 Magnetic field strength (mt) mm 4 V x in Li x Ru 0.75 Sn 0.25 O A A Magnetic field strength A A A (mt) Ru 5+ d /db The exacerbated capacity of Li-rich layered materials is nested in the cumulative cationic Ru 4+/5+ and anioinic (2O -- (O 2 ) n- ) redox species. M. Sathya, et al. (Nature communications, Février 2015) A A

16 mm mm mm mm mm mm mm mm mm V vs. Li + /Li 0 mm mm Cu Current collector c -2 Metallic lithium Separator Li 2 Ru 0.75 Sn 0.25 O 3 +C Al Current collector b (a) species Al Current collector by insitu -5-4EPR Imaging: mm mm A first (d) a d x in Li x Ru 0.75 Sn 0.25 O 3 mm Localisation of the cationic/anionic redox Ru mm 2 3 mm mm c (a) 4.0 b a 3.5 Li foil + plating V vs. Li + /Li (d) x in Li x Ru 0.75 Sn 0.25 O (b)(d) -1-2 Li foil O Ru mm -2 (c) mm mm (b) (b) (c) (c) mm EPR imaging: -4 Key tool for monitoring -4 electrode homogenity and kinetics M. Sathya, J.B. Leriche, E. Salader, mmd. Gourrier, H. Vezin and J.M. Tarascon mm (Nature communications, Fevrier 2015) d 4 5 Ru +5 + (O 2 ) n- (O 2 ) n- mm mm

17 Charged till 4.6V 1 nm What s going at the atomic scale?. Can we visualize the O-O dimers.. Visualize the material at the microscopic scale (HAADF-STEM) Pristine Li 2 Ru 0.75 Sn 0.25 O 3 1 st Discharge Li Ru/Sn Ru/Sn 1 nm Octa. Li Ru/Sn Li 1 nm A massive cationic migration which is reversible M. Sathiya, et al. Nature Materials 14, (2015) Collaboration avec A. M. Abakumov et G. Van Tende

18 Potential (V) How to prevent this cationic migration? Play with chemistry Explore members of the Li 2 MnO 3 family (M=Ir) O 1 -type Li O x in Li 2-x IrO 3 Ir b c a O 3 -type [001] Z 2 Z 1/3 (O 2 ) n- First visualization of (O-O) dimers - in Li-rich layered oxide electrodes E. McCalla, A. Abrakusov,.. and J.M. Tarascon Science, Décembre 2015

19 Can we directly measure the (O-O) dimer lengths? Use of the D1B Neutron line at ILL Pristine sample (O 3 ) Pristine sample (O 3 ) Charged sample (O 1 ) Rp = 7.7% Rp = 7.7% Li Ir O (O-O) 2.84 Å Charged sample (O 1 ) Charged sample (O 1 ) 2.4( A (O-O) 2.45 Å Direct evidence for shorter (O-O) bonds in Li-rich layered oxides electrodes E. McCalla, A. Abakumov, G. Rousse, M.L. Doublet and J.M. Tarascon, Science ; December 2015

20 Recent evidence of (O-O) dimers in Li-rich layered oxides: LiRhO 2 Li y Rh 3 O 6 Pristine LiRhO Å 2.82Å Shortening of the bonds between O belonging to two octahedras D. ikhailova, 0.M. Karakuina, D. Batuk, A. Abrakusov, et al; Inorganic chemistry, July 2016

21 What abput the origin of the voltage fade? From Sn (4d 10 ) to Ti(3d 0 ) substitute

22 Voltage (V vs. Li + /Li 0 ) Capacity (mah/g) Voltage (V vs. Li + /Li 0 ) Voltage (V vs. Li + /Li 0 ) Studying the Li 2 Ru 0.75 Ti 0.25 O 3 system 800 C Li 2 CO RuO TiO 2 Ball mill + 24 hours at 800 C Li 2 Ru 0.75 Ti 0.25 O 3 Capacity/cycling Voltage fade Cycle number x in Li x Ru 0.75 Ti 0.25 O x in Li x Ru 0.75 Ti 0.25 O 3 Ti +4, seems to be the worst substitute V x in Li x Ru 0.75 Ti 0.25 O 3 M. Sathya, A.M. Abakumov, K. Ramesha, D. Foix, G. Rousse, and J.M. Tarascon, Nature Materials (2015) 80 1

23 M. Sathya, A.M. Abakumov, K. Ramesha, D. Foix, G. Rousse, and J.M. Tarascon, Nature Materials 1er Décembre Li 2 Ru 0.75 Sn 0.25 O 3 Suppression of the potential fade upon cycling by Tin: Why? Li 2 Ru 0.75 Ti 0.25 O 3 Octa Octa Octa tetra tetra 100 Cycles Cations trapped in tetrahedral sites Voltage fade upon cycling is linked to the capturing of small cations ions tetrahedral in sites Ti 4+ < Sn Å 0.69 Å 50 Cycles Provide chemical clues to enable the development of Li-rich NMC for the next generation of Li-ion batteries 100 Cycles

24 From modem materials to Li-rich NMC

25 2 nd 1 charge st discharge bulk 2 nd discharge surface bulk surface 1 st charge 2 nd charge bulk surface a 1 st cycle at 6.9 kev O 1s Pristine surface deposits 4.46 V O n O 2 b 2From nd cycle at model 6.9 kev compounds c 1 st charged, to Li-Rich 4.80 V Li 1.2 Ni 0.14 Mn 0.54 Co 0.13 O 2 : What s the story? O 1s Anionic redox in Li-rich NMC via bulk-sensitive hard-xps 3.75 V surface deposits 4.25 V O n O 2 Li-rich NMC O 1s kev surface deposits O n O 2 What are the cationic / anionic redox potentials? 3.0 kev 34% 34% b 4.80 V sat V O 1s 2 nd cycle at 6.9 kev c 1 st charged, 4.80 V 4.10 V O 1s 6.9 kev 4.80 VDirect evidence of bulk anionic sat. redox 33% sat Binding energy (ev) d 1 st discharged, 2.00 V O 1s V kev 3.60 V 34% kev 10% surface deposits V O n O kev V surface deposits O n O 2 34% surface O n O 2 deposits 3.0 kev 9% VV sat Binding energy (ev) A.S Gaurav 4.10 et al; V(Submitted 2017) 6.9 kev 2.00 V sat. 33% Binding energy (ev) d 1 st discharged, 2.00 V 6.9 kev 9% Binding energy (ev)

26 The science underlying the anionic redox process The source of controversial debates (need to go back to basics)

27 MO 6 Metal (M) Metal (M) t 1U * Some basic recalls. From Molecular Orbitals to the Band Structure Ligand (O) Ligand (O) 4p a * 1g 4s 3d e* g * t* 2g t 2g p s e g t 1U a 1g

28 The iono-covalence: Its influence on band positions Ionic bond 100% M character MO* M S 2 /D Increasing O character M M Covalent bond MO* S 2 /D D D O S 2 /D S 2 /D MO 100% O character O Increasing M character MO O Energy difference MO/MO* given by D + 2S 2 /D

29 How this can happen? What is the underpinning science LiMO 2 Li 2 MO 3, Li(Li 1/3 Mn 2/3 )O 2 O 2 O O 2 E E (M-O)* U (M-O)* (M-O) D (M-O) O(2p) non bonding O(2s) Classical situation: 1 Redox band Y. Xie, M. Saubanère, M.-L. Doublet, Energy Environ. Sci. 10 (2017) O(2s) An illicit situation: 2 Redox bands D.-H. Seo, G. Ceder et al. Nature Chemistry (201

30 Cationic redox (One band) E MO* U v U << D D e Anionic redox depends upon the competition between U and D D U/2 Anionic redox (two bands) Extra capacity E e + e - U D D << U Anionic redox (One band but irreversible) E e U D O 2 O 2 2 release MO al distorsion -O bonds (O 2 ) n- Charged Pearce, Tarascon et al, Nature Mater. (2017) P. Bruce, Nature chemistry (2016) 30 Saubanère, Tarascon et al, EES 9 (20

31 To which extreme can we push the capacity using the anionic redox concept? O 2 More Li? O 2 Li 2 MO 3 Li 3 IrO 4 or Li(Li 1/2 Ir 1/2 )O 2 Pure Li planes Disordered Li 2 Ir planes Increase the O/M ratio New phase: Li 3 IrO 4 New Li 3 IrO 4 phase Ir Li 2 CO C, 24h Disordered Li 2 Ir planes Air Li 3 MO A. Perez et al. Submitted 2 (, = Å) Li 3 IrO 4

32 Electrochemical performance of Li 3 IrO 4 vs. Li + /Li while monitoring pressure V vs Li + /Li Pressure sensor 5 4 O 2- /(O 2 ) n- Mass Spect. 3 2 Limit the charge at x = 1 Cell x in x Li in x IrO Li x IrO e - are exchanged by transition metal. (340 mah/g) Li 1 IrO 4 Li 4.5 IrO 4 Highest value ever reported for insertion cathode materials in Li-ion batteries A. Perez et al. submitted Increasing the O/M ratio: a trade-off would have to be found between extra-capacity and structural stability

33 Reaction enthalpy (ev/fu) 3d 4d 5d (O 2 ) n- stability against recombination into O 2 : a DFT approach Ti V Cr Mn Fe Co Ni Zr Nb Mo Tc Ru Rh Pd Hf Ta W Re Os Ir Pt Network stability w.r.t O 2 release Li x MO 3 Li x MO 3 δ + δ 2 O 2 Positive enthalpy => formation of O 2 vacancies is not favorable nd-shell : M-O covalence is to stabilize the network zone Li-content : all M(3d)-based materials are unstable w.r.t. O 2 release once the first Li is removed! Bad news for applications Therefore 5d metals with n> 3 are stable configuration in, in agreement with the use of Ir Exploit such findings to design Li 2 MM O 3 phases stable against (O 2 ) n- recombination Xie et al. Energy & Environ. Sci. Asap 2016

34 The Li 3 MO 4 family: a rich crystal chemistry J. Quentin et al. Chemistry of materials 2017 Yabuuchi et al, PNAS 112 (2015) Li 3 RuO 4 Li 3 SbO 4 Li 3 MO 4 phases Li 3 NbO 4 Li 3 Nb y Sb 1-y O 4 Creation of a platform for identifying key indicators to manipulate anionic redox 34

35 Effect of dimensionality: is anionic redox limited to 2D compounds?

36 Yabuchi et al; PNAS, 2016 tructure -3 m Effect of modifying the crystal structure on the anionic redox reactivity Layered rocksalt structures as a funstion of increasing Li/M ratio. M Li M Li M LiMO 2 R -3 m 2D LiM 2 Li LiM 2 Li LiM 2 Li 2 MO 3 C 2/m Nb Li Li 3 NbO 4 I -4 3 m Increases disorder Li Mo Li 4 MoO 5 P -1 Disordered phases x in Li 1.3-x Nb 0.3 Mn 0.4 O 2 x in Li 1.2-x Ti 0.4 Mn 0.4 O 2 3D

37 Is the anionic redox process only limited to 2D layered compounds?: IrO2 + Li2CO3 (10% excess) < 950 C (24h) α-li2iro3 2D structure Li3 LiIr2 IrO2 + Li2CO3 (10% excess) > 1050 C ( > 90h) β-li2iro3 3D structure Li2Ir 100 µm Li2Ir

38 Electrochemical performance of the 3D Li-rich polymorph All Li can be removed "IrO 3 " (~ 2 e - per Ir) 3 rd cycle 2 nd cycle 1 st cycle X-ray photoemission Anionic + cationic Neutron Powder diffraction + ABF-STEM 4 V - charged -LiIrO 3 O-O distortion D (x10 4 )= 25.2 <Ir-O> = Å Anionic redox activity is not solely limited to 2D systems: its implementation to 3D oxides opens wide the research field for high capacity electrodes P. Pearce et al. Nature materials February 2017

39 Anionic redox A transformational approach to design high energy cathodes Classical Layered Oxides LiMO 2 cationic redox only Li-rich Layered Oxides Li 1+y M 1 y O 2 combined cationic + anionic redox Can we use Li-rich cathodes in real-world batteries Any practical roadblocks??? 500 Wh kg Wh kg 1 Poor kinetics Hysteresis Voltage decay

40 dq / dv dq / dv Poor kinetics of Li-rich NMC: (Li 2 Ru 0.75 Sn 0.25 O 3 ) Deconvolution via operando XAS Cationic redox Anionic redox Summation Voltage (V) G. Assat, J-M. Tarascon et al., J. Electrochem. Soc., (2016). Anionic redox shows sluggish kinetics and triggers hysteresis C/5 C/10 Ru 4+/5+ peak C/2.5 O 2 /O n peak Voltage (V)

41 Consequences of poor kinetics/large polarisation: Energy efficiency. Poor round-trip energy efficiency in LR-NMC:_Role of anionic redox G. Assat, J-M. Tarascon et al., J. Electrochem. Soc., (2016).

42 Potentiel (Volts vs. Li + /Li) dq/dv dq / / dv Poten Capacité y in en mah/g Voltage decay (> 280 mah/g) y in Capacité Capacity en in mah/g mah/g Practical roadblocks in Li-rich cathodes and their consequences Voltage fade 100 th 50 1 st 0 Poor kinetics Voltage (V) Are these practical issues surmountable? C/5 Ru 4+/5+ C/2.5 peak O 2 /O n peak Voltage (V) C/10 Importance of substituent (chemical elements with large ionic radii) Td A promising direction being pursued Design of materials within which cationic and anionic redox occur at the same potential Sathya, Tarascon et al., Nature Materials 2015,vol. 14, no. 2, pp G. Assat, J-M. Tarascon et al., J. Electrochem. Soc., (2016).

43 Importancy of anionic redox beyond electrode materials N. Yabuushi et al. PNAS May 2015 M. Freire et al. Nature materials Novembre 2015 A. Grimaud,,,, Tarascon Nature materials 15 (201 New perovskites with tailored O-O bonds

44 Acknowledgments 44