composite electrolytes to stabilize metallic lithium anodesceramic interfaces and high conductivity...
TRANSCRIPT
-
Project ID: ES182
Nancy Dudney and Sergiy Kalnaus Oak Ridge National Laboratory
Vehicle Technologies Program Annual Merit Review and Peer Evaluation Meeting
June 2015
Composite Electrolyte to Stabilize Metallic Lithium Anodes
“This presentation does not contain any proprietary, confidential, or otherwise restricted information”
-
2 Presentation name
Overview – Composite Electrolytes to Stabilize Li Metal Anode
• Timeline – Start October, 2011 – End September, 2016
• Technical barriers – Energy density (500-700 Wh/kg) – Cycle life, 3000 to 5000 deep
discharge cycles – Safety
• Budget – $335k FY13 – $400k FY14
• Partners and collaborators – Oak Ridge National Laboratory
(lead) – Center for Nanophase Materials
Sciences, ORNL – Collaborators:
– Jeff Sakamoto, Michigan State University
– Nitash Balsara, UC Berkeley – nGimat, GA – Ohara Corporation, CA
-
3 Presentation name
Overview – Composite Electrolytes to Stabilize Li Metal Anode
• Timeline – Start October, 2011 – End September, 2016
• Technical barriers – Energy density (500-700 Wh/kg) – Cycle life, 3000 to 5000 deep
discharge cycles – Safety
• Budget – $335k FY13 – $400k FY14
• Partners and collaborators – Oak Ridge National Laboratory
(lead) – Center for Nanophase Materials
Sciences, ORNL – Collaborators:
– Jeff Sakamoto, Michigan State University
– Nitash Balsara, UC Berkeley – nGimat, GA – Ohara Corporation, CA
To match Li-ion cathodes, Li cycling must achieve:
20-40 µm Li per cycle no loss to reaction
10-20 nm/sec, pulse no loss to physical isolation
3000 cycles no roughening or dendrites
-
4 Presentation name
Relevance – Our strategy
Premise: To ensure stable and efficient use of high energy dense lithium metal anode requires a protective and robust solid electrolyte. The combination of two or more solid electrolytes is more likely to meet the many materials and manufacturing requirements than any single material.
What single solid has: adequate Li+ conductivity AND
robust mechanical properties AND thin sheet processing AND
no pathways for dendrites AND chemical stability with Li ?
composite of solid
electrolytes
-
5 Presentation name
Relevance and Objectives • Objectives:
– Understand Li+ transport at interface between two dissimilar solid electrolytes, e.g. ceramic/polymer
– Develop composites of electrolyte materials with requisite electrochemical and mechanical properties as guided by simulation
– Fabricate thin membranes to use with a thin metallic lithium anode providing good power performance and long cycle life
– Identify design rules that can be generally applied to composites of other solid electrolyte materials
• Relevance to technical barriers: – Multi-year program plan identifies the Li metal anode and its poor cycling as
the fundamental problem for very high energy Li batteries. Hence, research takes the approach of completely isolating the anode from the electrolyte.
– Success of our composite electrolyte will: • Enable very High Energy Li-S Battery (500 Wh/kg) by 2020 and Li-Air Battery (700
Wh/kg) by 2030
• Fully protect lithium anode for long cycle life (3000 to 5000 deep discharge cycles) • Ensure lithium remains dense and free of dendrites (Safety) • Improve energy density lithium batteries (USABC has targeted a 5X improvement)
-
6 Presentation name
Milestones Milestones: FY14-FY15 Target: Status:
1. Fabricate ceramic-polymer composite sheets with 20-60 vol% ceramic to determine the composition dependence of the conductivity as the structure approaches mechanical stability
Q1 FY14
2. Demonstrate, with at least two ceramic-polymer composites, how the Li conductivity is impacted by either the connectivity or dimensions of the component phases
Q2
3. Cycle the most promising composite electrolyte membranes with a thin lithium metal electrode and quantify interface resistance and coulombic efficiency.
Q3 impedance, limited cycling
4. Probe whether homogeneity of the composite electrolyte contacting the Li electrode is important to the interface stability in order to assess whether an interface coating is needed
Q4 lithium reacts
1. Compare the vapor absorption (rate and equilibrium) of small molecules in a single phase polymer and a corresponding ceramic/polymer composite electrolyte.
Q1 FY15
2. By generating a database with at least 5 compositions, determine if the presence of trace solvent molecules that enhance the ionic conductivity is also detrimental to the stability and cycleability of a lithium metal electrode, and if the effect can be altered by adding an intermediate films. (SMART)
Q3 on schedule
-
7 Presentation name
1. Utilize known solid electrolytes. 2. Determine properties of interfaces
joining different solid electrolytes 3. Use theory and simulation to explore mechanical
stability and ion transport 4. Fabricate simple composites with high loading of
dispersed particles and no solvents 5. Fabricate refined composites and develop practical
processing methods. 6. Evaluate stability with lithium metal upon cycling.
Approach - Choose ceramic
+ polymer electrolytes
Computer simulation
Bilayer
interfaces
Dispersed composites
Advanced composites
Cycle lithium metal
Key to success - fundamental understanding and control of these interfaces.
polymer
glass ceramic
One of the unique aspects of program – We address mechanical properties, as well as ion transport.
-
8 Presentation name
Background: About a year ago, paths to low resistance polymer - ceramic interfaces and high conductivity composites were at hand.
• Processing was important, but why?
2.7 2.9 3.1 3.3 3.5 3.7
1E-7
1E-6
1E-5
1E-4
1E-3
Lipon
PEO10:LiTFSI
T (°C)
σ (S
cm
-1)
1000/T (K-1)
PEO16:LiCF3SO3
OharaLi7La3Zr2O12
90 80 70 60 50 40 30 20
Ohara & LTAP
Li salt – PEO low R interface ~10 Ω
high R interface ~10 kΩ
depends on processing
Ion transport across interfaces intrinsic extrinsic
available lattice sites
reaction layer
space charge impurity molecules Li+ solvation voids, pores polymer anchors salt segregation
-
9 Presentation name
Background: About a year ago, paths to low resistance polymer - ceramic interfaces and high conductivity composites were at hand.
• Processing was important, but why?
• Need evidence that Li+ ions cross interface in composite.
2.7 2.9 3.1 3.3 3.5 3.7
1E-7
1E-6
1E-5
1E-4
1E-3
Lipon
PEO10:LiTFSI
T (°C)
σ (S
cm
-1)
1000/T (K-1)
PEO16:LiCF3SO3
OharaLi7La3Zr2O12
90 80 70 60 50 40 30 20
.
-
10 Presentation name
Technical progress: Several loose ends raise doubt about basis for highly conductive composite. • Why is conductivity increasing and why so slowly? • Why is increase initiated by crimping and not accelerated by heating?
40-50 vol% ceramic MWPEO is 100k or 1M
LiTf salt
Darker symbols at RT following scans to 90˚C
1.E-09
1.E-08
1.E-07
1.E-06
1.E-05
1.E-04
0 50 100
Cond
uctiv
ity a
t RT
(S/c
m)
Time from casting of composite (days)
seal in coin cell
-
11 Presentation name
Technical progress: Several loose ends raise doubt about basis for highly conductive composite. (milestone) • Why is conductivity increasing and why so slowly? • Why is increase initiated by crimping and not accelerated by heating? • Why does this change with coarser ceramic particles?
1.E-09
1.E-08
1.E-07
1.E-06
1.E-05
1.E-04
0 50 100
Cond
uctiv
ity a
t RT
(S/c
m)
Time from casting of composite (days)
seal in coin cell
-
12 Presentation name
Technical progress: Further analysis showed effect of glove box.
• Alternative explanations – Crimping into coin cells composite percolated conductive path – Exposure to standard glove box for crimping
automatic crimp
weigh, mill, melt/press pellets manual crimp
standard glove box with liquid electrolytes
heat, added seal,
contacts
dry glove box – NO electrolytes
-
13 Presentation name
25°C
Technical progress: Further analysis showed effect of glove box.
• Alternative explanations – Crimping into coin cells composite percolated conductive path – Exposure to standard glove box for crimping
automatic crimp
weigh, mill, melt/press pellets manual crimp dry glove box – NO electrolytes
standard glove box with liquid electrolytes
heat, added seal,
contacts
• Reproduced with several side by side sets
-
14 Presentation name
Technical progress: ~ 15 minutes in a standard glove box makes a huge difference in the ionic conductivity. Due to contamination?
• New investigation (milestones FY15) – Treat composite and polymer electrolytes with selected vapors – Revisit planar interface studies
1.E-08
1.E-07
1.E-06
1.E-05
1.E-04
1.E-03
1.E-02
2.7 2.9 3.1 3.3 3.5
cond
uctiv
ity S
/cm
1000/T(K)
σOhara σPEO-LiTf ternary 0:50:50 R104 crimp ternary 0:50:50 G49 crimp
Ohara ceramic
PEO16 Li(CF3SO3)
composite, dry
composite +impurity crimped in: standard box dry glove box
-
15 Presentation name
Technical progress: Intentional exposure to likely vapor phase ‘contaminants’ or additives give little to large enhancement
• Some additives inhibit polymer crystallization, others no effect. • Others provide enhanced conduction
1.E-07
1.E-06
1.E-05
1.E-04
1.E-03
1.E-02
2.7 2.8 2.9 3 3.1 3.2 3.3 3.4 3.5
cond
uctiv
ity S
/cm
10^3/T(K)
Ohara ceramic
PEO16 Li(CF3SO3)
different additives
composite, dry
-
16 Presentation name
Technical progress: DMC and water vapor treatments have large effects for polymer-alone and composite
• Treatment in DMC saturated Argon enhanced conductivity • Higher conductivity for composite, but ion path still uncertain
1.E-07
1.E-06
1.E-05
1.E-04
1.E-03
1.E-02
2.7 2.8 2.9 3 3.1 3.2 3.3 3.4 3.5
cond
uctiv
ity S
/cm
10^3/T(K)
Ohara ceramic
PEO16 Li(CF3SO3)
composite
polymer
composite, dry
-
17 Presentation name
Technical progress: Controlled dose of DMC indicate that DMC is readily adsorbed, but trace quantities have small effect. (milestone)
• Plot assumes all available DMC is adsorbed. • Conductivity does not change with time, differing from previous
observation • Short 15 minute treatment with DMC slight increase
10-7
10-6
10-5
10-4
0.001 0.01 0.1 1 10
15 hour exposure15 min exposure
Con
duct
ance
(oh
m-1
)
DMC molecules per Li+ (available)
40 vol% Ohara ceramic with PEO16-LiCF3SO3 measured at 25˚C
-
18 Presentation name
Technical progress: Less dramatic results for water exposure of composite samples. Extending to other vapors. (milestone) • In response to reviewer comment, and also because water would be likely
effect of a poor seal. • Need to separate dissolution from adsorption kinetics
– DMC & H2O similar vapor pressure, different Hansen distances
-
19 Presentation name
• Tri-layer of PEO(salt) – sintered LLZO disk – PEO(salt), compared to PEO(salt) of same thickness – LLZO only adds ~300 ohms, outside frequency window at 25˚C – Interface impedance ≈ polymer layer impedance – Both interface and polymer impedance reduced with DMC exposure
Technical progress: Trilayer samples have significant interface resistance, reduced by DMC exposure
LLZO PEO (salt)
SS contacts
0 5e5 1e6
-1000000
-500000
0
Re (Z) (ohms)
Im(Z
)
Trilayer sample with fitTri-layer with fit
ZPEO Zi
0 1e6 2e6
-2e6
-1e6
0
Re (Z) (ohms)
Im(Z
)
Trilayer and sister polymer
Tri-layer sister polymer
0 100000 200000 300000
-300000
-200000
-100000
0
Re (Z) (ohms)Im
(Z) Trilayer before/after DMC
Tri-layer before after DMC
-
20 Presentation name
Technical progress: For composite with high ceramic loading, cycling with Li proved unstable. (milestone)
• Composites studied after Li foil pressed to one face – High and increasing interface resistance – Failed when Li plated to SS contact
0 100000 200000 300000 400000 500000
-300000
-200000
-100000
0
Z'
Z''
0 1000 2000 3000 4000 5000
-3000
-2000
-1000
0
Z'
Z''
Example for DMC treated composite SS/sample/SS SS/Li/sample/SS SS/Li/sample/(Li)/SS
-
21 Presentation name
Response to previous reviewer comments - Many positive comments, but a few concerns addressed here.
Reviewer comments last AMR (June 2013) and US DRIVE tech team (Sept 2014)
Response
“… the technical accomplishments have shown the need for low interfacial impedances. Effort is now needed to lower these interfacial impedance values.“
Small additives may be key to lowering the interface impedance. More investigation is needed.
“… developing a dense and thin electrolyte will be challenging.“
This is true , but expect composite to be easier to form and use than ceramic plates. Others will assist.
“Water contamination may be an issue, better experimental set up to account for outside influences and better preplanning and experimental design “
Following this recommendation, water treatment was specifically investigated.
“The critical path to meeting project goal was not very clear. Risks and mitigation strategies are unclear.”
Next slide.
-
22 Presentation name
Challenges, Barriers and Future Work
• Challenges, Risks and Mitigation – Subtle differences in composite processing have large impact on the bulk conductivity.
Risk - we may not identify root causes of this effect. – Mitigation - provide additives that overwhelm the variable processing effects without
compromising the mechanical properties and Li reactivity. – With better understanding, this may indicate path to form stable, conductive
composite. • Remainder of FY15
– Publish DMC/H2O treatment results of composite – Complete and publish trilayer with LLZO. Vary thickness of PEO layers. – For Li tests, reduce ceramic loading for full density. Test barrier coating & DMC
effects. – Identify vapor species that prove to be most beneficial additive.
• FY16
– Improve composites, substitute LLZO for LATP particles. Bimodal size distribution.
-
23 Presentation name
Collaborations and coordination
• Collaborations which include coordinated sample preparation, sample exchange for analysis. We have joint publications; several are in preparation.
– Jeff Sakamoto (University of Michigan) inside VT
• Coordination with a BES program at ORNL is growing; both programs focus on solid electrolytes, particularly the bulk and interfacial ion transport.
• Ceramic electrolytes supplied by: Ohara Corp. and nGimat and Jeff Sakamoto
• Industrial communications, outside VT. – Corning Corp., Paul Johnson & K P Reddy, tapped for discussion regarding ceramic
electrolytes – A similar contact regarding polymer electrolytes would be helpful.
-
24 Presentation name
Summary • Relevance Success of our composite electrolyte will isolate the anode from any liquid electrolyte,
enabling very high energy batteries, with thousands of deep cycles, negligible consumption of lithium and good safety.
• Approach Premise: the combination of two or more solid electrolytes is more likely to meet the many materials and manufacturing requirements than any single material. Interface is critical.
– Use model structures to simplify evaluation of the ceramic –polymer interface. – Following simulation, fabricate composites with maximum ceramic loading. – Evaluate interface to lithium metal with and without barrier coating. Cathode is not considered.
• Accomplishments and progress - new understanding that will address technical barriers – A subtle variation in processing conditions for composite electrolytes can profoundly alter the
ionic conductivity of ceramic-polymer composites. This may well have been overlooked. – Adsorption of some organic vapors and water can enhance Li+ ion transport in ceramic-polymer
dispersed composites, multi-layers, and polymer electrolytes. – Enhancement has been quantified for different molecules, including dimethyl carbonate and
water. Controlled exposure may be a new path to improve properties of composites. – Contact and cycling with Li metal led to reaction with DMC and/or LATP ceramic particles.
• Future work – Publish DMC/H2O treatment results of composites and trilayer samples. – For Li tests, reduce ceramic loading for full density. Test barrier coating & DMC effects. – Identify vapor species that prove to be most beneficial additive. – Improve composites; substitute LLZO for LATP; increase ceramic loading with bimodal size.
• Collaborations and coordination – key for past year are: Sakamoto and Ohara
-
25 Presentation name
Technical backup slides
-
26 Presentation name
Technical progress: Analysis of standard versus dry (no volatile) glove box reveals only small differences
• Mass spec and FTIR of gas samples – nothing obvious
• XPS of high surface area carbon – possible F surface contamination
• G49 is dry glove box; R103 is standard glove box. 650660670680690700710720
NoritG049R103
Inte
nsity
(nor
m.)
Binding Energy (eV)
In
In
F
10-15
10-14
10-13
10-12
0 50 100 150 200
Mass Spec of gas samples G49 glove boxR103 glove box
Log
curre
nt
mass units
Composite Electrolyte to Stabilize Metallic Lithium Anodes�Overview – Composite Electrolytes to Stabilize Li Metal AnodeOverview – Composite Electrolytes to Stabilize Li Metal AnodeRelevance – Our strategy Relevance and ObjectivesMilestonesApproach - Background: About a year ago, paths to low resistance polymer - ceramic interfaces and high conductivity composites were at hand. Background: About a year ago, paths to low resistance polymer - ceramic interfaces and high conductivity composites were at hand. Technical progress: Several loose ends raise doubt about basis for highly conductive composite. Technical progress: Several loose ends raise doubt about basis for highly conductive composite. (milestone) Technical progress: Further analysis showed effect of glove box. Technical progress: Further analysis showed effect of glove box. Technical progress: ~ 15 minutes in a standard glove box makes a huge difference in the ionic conductivity. Due to contamination?Technical progress: Intentional exposure to likely vapor phase ‘contaminants’ or additives give little to large enhancementTechnical progress: DMC and water vapor treatments have large effects for polymer-alone and compositeTechnical progress: Controlled dose of DMC indicate that DMC is readily adsorbed, but trace quantities have small effect. (milestone)Technical progress: Less dramatic results for water exposure of composite samples. Extending to other vapors. (milestone) Technical progress: Trilayer samples have significant interface resistance, reduced by DMC exposureTechnical progress: For composite with high ceramic loading, cycling with Li proved unstable. (milestone) Response to previous reviewer comments - Many positive comments, but a few concerns addressed here.Challenges, Barriers and Future WorkCollaborations and coordinationSummaryTechnical backup slidesTechnical progress: Analysis of standard versus dry (no volatile) glove box reveals only small differencesReviewer Only SlidesPublications and presentations (March 2014-5)Critical assumptions and issues