Primary Particle Level Diagnostics by Transmission Electron

Microscopy

Huolin L. Xin

Center for Functional Nanomaterials, Brookhaven National Laboratory, Upton, NY

TEM Grid in a coin

cell: quasi in-situ

method

“Dry-cell” in-situ

approach

Electrochemical liquid

cell approach

Scanning Transmission Electron Microscopy

1 atom wide (1 Å) beam is scanned

across the sample to form a 2-D image100 keV

Electron beam

Lenses

ADF

Detector

Electron

Spectrometer

Sample

Elastic Scattering ~

Z1.7

NdGaO3 La0.7Sr0.3MnO3

200 pixels

68 p

ixels

Inelastic Scattering

EELS: Chemical Imaging

Nd

O

Mn

LaO K edge

Mn L2,3 edge

La M4,5 edge

Nd M4,5 edge

What are the remaining challenges in lithium ion battery?

Limited by cathode/electrolyte

interface stability

Capacity fading and thermal

stability of cathode

Surface reconstruction and chemical evolution of stoichiometric layered cathode materials for lithium-ion batteries, Feng Lin,

Isaac Markus, Dennis Nordlund, Tsu-Chien Weng, Mark Asta, H. L. Xin*, M. M. Doeff, Nature Communications, 5, 3529

LiNi0.4Mn0.4Co0.2O2

LUMO

HOMOAnode/SEI/electrolyte

Self-passivation by reducing

the electrolyte

(Li2CO3, LiF, etc…)

Cathode

Materials and voltage

dependent

Still under debate

Degraded Surface Layers of NMC442 = LiNi0.4Mn0.4Co0.2O2

• Surface structures were altered after exposing to electrolyte and

the surface layer was thickened after 1cycle of discharge-charge

Electrolyte Immersed Sample After 1 cycle

Feng Lin et al Xin*, and Doeff, Nat. Comm. 5, 3529 (2014)

20 Cycles (NMC 442)

NMC442 = LiNi0.4Mn0.4Co0.2O2

4

“4+”

“2+”

3.7 nm

Feng Lin et al Xin*, and Doeff, Nat. Comm. 5, 3529 (2014)

The Degradation is Facet DependentFeng Lin et al Xin*, and Doeff, Nat. Comm. 5, 3529 (2014)

Radiation Damage of Layer-Layer Compound

• LiNixMnxCo(1-2x)O2

300 keV

~36 pA

1024x1024 pixels,

pixel size 0.153

angstrom

pixel dwell time 6 us

frame time 7.55

sec/frame

Feng Lin et al, H. L. Xin*, Scientific Reports, 4, 5694 (2014)

Faster scanning spreads dose

• 4x faster scanning

• 300 keV

• ~36 pA, 512x512 pixels

• pixel size 0.306

angstrom

• pixel dwell time 6 us,

frame rate 1.89

sec/frame

Feng Lin et al, H. L. Xin*, Scientific Reports, 4, 5694 (2014)

High-energy Electron vs. X-ray

Henderson, Quarterly Reviews of Biophysics 28, 171 (1995)

It’s not the cross-section, but

How many damaging events

per useful imaging event.

Least DamageElastic Imaging – Electrons win

Inelastic imaging – Soft X-rays win

Electron Elastic

X-ray Elas.

Spectroscopy: What’s the True Carbon K edge?

X-ray Absorption

Ade and Stoll, Nature Mats, 8 (2009) 281Ade and Hitchcock, Polymer 49 (2008) 643

EELS of Li2CO3

500 electrons/Å2/spectrum

F Lin et al Xin*, Nat. Comm. 5, 3356 (2014)

Imaging of Pristine NMC

• No Damaging in EELS and Imaging to the Pristine Surface5

nm

5 n

m

Sp

ectru

m Im

age

1 n m1 n m

Subplixel Scanning Collage

Feng Lin et al Xin*, and Doeff, Nat. Comm. 5, 3529 (2014)

1. Gold index grid (working electrode)

2. Reversible Hydrogen Electrode (reference electrode)

3. Platinum foil (counter electrode)

4. Argon gas inlet

5. Argon gas outlet

1

2

3

45

3-D imaging for the first time the

same region before and after

electrochemical cycling

50 nm

Before After

Ex-Situ: Coarsening During Voltage CyclingY. Yu*, H. L. Xin* et al, Nano Lett.,12 (9),4417-4423 (2012)

TEM Grid-in-a Coin Cell Experiment

• Grids need be dissembled (only quasi-in situ)

• Need to wash the grids with EC; i.e. dissolvable components in the SEI are gone.

• Not in situ and needs protected transfer

• Materials are 3-D submerged in a

commercial organic electrolyte

• Provides a uniform and conductive

support

• The grid can be used in

synchrotrons and TEMs

Pros Cons

F Lin et al Xin*, Nat. Comm. 5, 3356 (2014)

Use NiO Pallets as a Model System

Anodic Conversion Reaction 2Li + NiO Ni + Li2O

• Metallic Ni (green) is nanoporous.

Before cycling 0.25 cycle

Ni

NiO

F Lin et al Xin*, Nat. Comm. 5, 3356 (2014)

Heterogeneous Nucleation/Phase Transition in Battery Anodic

Reactions

• Multiple nucleation sites– Difference in electric/ionic contact

– Difference in phase transition kinetics

0.25 cycle (being lithiated) 0.75 cycle (being delithiated)

F Lin et al Xin*, Nat. Comm. 5, 3356 (2014)

What are the Preferential Sites for Lithiation

Speculation: grain boundaries could have lower activation energy for lithiation.

Needs in-situ experiments

【111】

Channeling Contrast (brighter if on zone aixs) F Lin et al Xin*, Nat. Comm. 5, 3356 (2014)

A Dry Cell Experiment

• dKai et al, Nano Lett., 2015, 15, 1437

Real-time in situ nano-indentation

In-Situ Lithiation of NiO

Kai et al, Nano Lett., 2015, 15, 1437

Kai et al, Nano Lett., 2015, 15, 1437

An incubation period limits your battery charging speed

revealed by real-time observations

finger

shrinking core

T = ~100 sec

Take home:

cannot charge

faster than

this time

scale.

Transitions from Near-Surface to Interior Redox upon Lithiation in NiOConversion Electrode Materials

There exist and incubation time for passivity break down. Which places

an intrinsic limit on how fast you can charge your conversion materials.

Kai et al, Nano Lett., 2015, 15, 1437

Lithiation Pathways on NiO

100 Sec

Passivation

Passivity break down

Tincubation

Critical Thinking

?

Tomography Experiment

Slab Geometry:

• One image every 1-2° from ±70°

• 70 - 140 images

Acquisition

e-

Reconstruction

Requirement:

• Image intensities vary

monotonically with thickness

(STEM is superior to TEM for tomography because TEM

suffers from phase contrast and diffraction contrast

which breaks the monotony of the signal )

200 nm

Pristine c

Cycled d

Voids Formation in LMR-NMC

Grain boundaries are both migration channels and sinks for

vacancies

• Voids form along grain boundaries.

Mass-contrast

Fe

Co

Fe2/3+ vs. Fe0

• Inelastic scattering artifacts removed.

• Many more channels of signals are under processing.

5D Microscopy (x, y, z, vs. Eloss vs. t)

Live Recorded Tomography Tilt Series

• Sub-10 min electron tomography achieved in 2014

115 kx (field of view 673 nm)

1 frame/sec, a total of 600 frames

-70-70 degrees, 2 degree tilt interval

2 days

Acquiring multiple tilt series within a minutes

500 nm

Comparison between single-series reconstruct and the merged

reconstruction

• Merging multiple reconstructions reduces noise and artifacts

A cross section of

Single-series Reconstruction

A cross section of

Merged Reconstruction (4 series)

Electrochemical liquid in S/TEM

• Battery applications underdevelopment

Electrochemical liquid

cell approach

500 nm

Acknowledgement

Feng LinEric Stach

Jing Li

Xiqian Yu

Dong Su

Xiaoqing Yang

Kai He

Marca Doeff

Huolin Xin

Dennis Nordlund

Yijin LiuDeli Wang

U.S. DOE Office of

Science Facility, at

Brookhaven National

Laboratory under

Contract No. DE-

SC0012704

Students: Lili Han*,

Ruoqian Lin, Wei-Yu Pan,

Sai Vishnubhotla*,

Guangzhong Huang*, Jing

Li*, Yujing Liu*

Center for Functional Nanomaterials• Located on Long Island• Fully operational since April 2008• Two-fold mission:

– Enable nanoscience researchby external users

– Discover and understand energy-related nanomaterials

• Concentration of world-class tools and scientific expertise under one roof

• Integration with BNL’s other world-class facilities and research programs

• Huolin Xin (hxin@bnl.gov)