xas studies of battery materials · xas studies of nanocatalysis and chemical transformation co...

XAS Studies of Nanocatalysis and Chemical Transformation

XAS Studies of Battery Materials

Faisal M. AlamgirNational Institute of Standards and Technology


Acknowledgements• Ela Strauss• Marten denBoer• Chi-Chang Kao - In-situ of LiCoO2 – based battery• Jay F. Whitacre

• Dan Fischer• Zugen Fu• Jason VanSlutman - Local strain of LiCoO2• Lydia Nemirovskaya

• Esther Takeuchi • Amy Colon• Krista Martocci - Role of V and Ag in SVO batteries• Tom Reddy• Steve Geenbaum


LiCoO2-based Cathodes Set the Benchmark in Performance in a Variety of UsesThese ultra-compact power sources are being

developed for:

• A new generation of spacecraft• Automotive power• Portable electronics • Implantable medical devices, such as drug

delivery systems


Uses of LixAg2V4O11-based Batteries

• Particularly useful in medical applications• High power delivery• Non-toxic


LiCoO2-based Microbattery design

2 µm

A microbattery is comprised of the following components:

• An Electrolyte - Lithium phospho-oxynitride glass (LIPON)

• Anode - Metallic Lithium • Cathode - LiCoO2• The cathode is deposited

directly onto Ti current collectors by sputtering, annealed at low temperature (~500oC compared to ~600oC for commercial batteries).

The total cell thickness is less than 20 µm.


In-situ XAS of LiCoO2-based Battery

• We expect the following things to change as Li is de-intercalated:– Cell parameter (coupled changes in atomic positions)– Oxidation state of Co and/or O for charge compensation– Local ordering (decoupled changes in atomic positions– Scattering amplitude (fewer Li ions)


Electrochemical behavior

3.85 3.90 3.95 4.00 4.05 4.10 4.15 4.20 4.25 4.30-0.00005

0.00000

0.00005

0.00010

0.00015

0.00020

0.00025

0.00030

0.00035

Capacity(Ah)

Voltage (V)

Charge_Capacity(Ah)

3.85 3.90 3.95 4.00 4.05 4.10 4.15 4.20 4.25 4.30-0.0005

0.0000

0.0005

0.0010

0.0015

0.0020

0.0025

0.0030

0.0035

0.0040

Derivative

(dQ/dV)

Voltage (V)

0 2 4 6 8 10 122.8

3.0

3.2

3.4

3.6

3.8

4.0

4.2

4.4

discharging

rest

charg ing

Volta

ge (V

)

t im e (hours)

minimum

Electrochemically, the charge/discharge behavior corresponds to that of a bulk LiCoO2 cell.

For example, we find the same features on the dQ/dV vs. V plot that has been reported by Reimers et al.

a

bc


X-ray absorption near-edge structure of Co

0.0

0.5

1.0

1.5

2.0

2.5

0.0

7.7 7.72 7.74

Nor

m. a

bsor

ptio

n [a

.u.]

Photon energy [keV]

1.3 eV

•The white line of the near-edge undergoes a shift of ~1.3 eV

• similar shift for a LixNi0.8Co0.2O2 cell has been reported [Johnson and Kropf].


Real-space manifestation of cycling

2.0

4.0

0 2 4 6

FT

( χ(k

)*k

2)

R [Å]

0.5

1.0

1.5

2.0

2.5

0 2 4 65 10 15 20 25 30 35

FT

( χ(k

)*k

2)

R [Å]

scan number

Charging

DischargingRest

•The amplitude of the Fourier transform peaks change as a function of charge/discharge•The two ovals in the inset show the regions of the layered structure that correspond to the peaks. •The inter-/intra-layer atoms distances contract as a function of charging.


PCA Formulation

⎥⎥⎥⎥

⎦

⎤

⎢⎢⎢⎢

⎣

⎡

•

⎥⎥⎥⎥

⎦

⎤

⎢⎢⎢⎢

⎣

⎡

•

⎥⎥⎥⎥⎥⎥⎥

⎦

⎤

⎢⎢⎢⎢⎢⎢⎢

⎣

⎡

=

⎥⎥⎥⎥⎥⎥⎥

⎦

⎤

⎢⎢⎢⎢⎢⎢⎢

⎣

⎡

nnnn

n

n

nn

ij

mnmm

n

n

n

mnmm

n

n

n

www

wwwwww

vv

vv

mmm

cccbbbaaa

.....

.

.

.00...0.00.0

............

..

..

..

............

..

..

..

21

22221

11211

22

11

21

33231

22221

11211

21

33231

22221

11211

µµµ

χχχβββααα

[A] = [B] • [V] • [w]t


PCA: Components and Reconstruction

7.68 7.69 7.70 7.71 7.72 7.73 7.74-0.2

0.0

0.2

0.4

0.6

0.8

1.0

1.2

1.4

1.6

a

principal component 1 principal component 2 principal component 3 Co XANES at x = 0.12N

orm

aliz

ed a

bsor

ptio

n

Energy (KeV)

0.0

0.4

0.8

1.2

1.6

7.69 7.70 7.71 7.72 7.73 7.74

0.0

0.4

0.8

1.2

1.6

0.0

0.4

0.8

1.2

1.6

Energy (KeV)

x = 0.37

******** experimental___ reconstructed using principal components

Nor

mal

ized

abs

orpt

ion

x = 0.7

b

x = 0.12


PCA: Possible Reaction Pathways

P

Increasing x

P

Q

R

QR


7.685 7.690 7.695 7.700 7.705 7.710 7.715 7.7200.0

0.2

0.4

0.6

0.8

1.0

1.2

1.4

1.6 C

B

A

x in Li(1-x)CoO2 =

s to d

Nor

mal

ized

Abs

orpt

ion

Energy (keV)

0.12 0.18 0.25 0.37 0.40 0.47 0.55 0.63 0.70

Co K-edge in Li(1-x)CoO2

The Co K-edge XANES in Li(1-x)CoO2 as a function of x.

The s to d transition and s to p transition (A, B and C) are labeled.

Co does appears to be involved in charge recompensation

…because

7.688 7.6900.00

0.01

0.02

0.03

0.04

0.05

0.06

if the d-band occupancy of Co changes with increased delithiation, then this peak amplitude should increase


Theoretical fit to real-space data

0 1 2 3 4 5 60.0

0.5

1.0

1.5

2.0

2.5x in Li(1-x)CoO2 =

Second clusterCo-Co&Co-Li

Co-O

First cluster

|F

T [χ

'(k)]|

R (Å)

0.12 0.18 0.25 0.37 0.40 0.47 0.55 0.63 0.70


0.1 0.2 0.3 0.4 0.5 0.6 0.7

1.4

1.6

1.8

2.0

2.2

2.4

2.6

Co neighbor Oxy neighbor 1 Oxy neigbor 2

Nea

rest

nei

ghbo

r dis

tanc

es o

f Co

(Å)

x in Li(1-x)CoO2

Theoretical fit to real-space data

The atomic structure around Co shows charge compensation occurs first by formation of O holes (for x<0.25).

XANES shows us that for x>0.25, charge compensation occurs through Co d-hole formation.


Spectro-electrochemical Cell

Spectro-electrochemical cell for in-situ XAS at the O K-edge being developed.

Half-cell built and tested.

appropriate


How Optimal is the Structure in Thin-film LiMCoO2 Cathode?

• The process of thin-film deposition may render the structure to be far from equilibrium.

• Non-equilibrium structure will affect the performance of the cell.


LiNiO2 cathodes: thin-film and bulk

0.0

5.0

10.0

0.0

-5.0

-10.0

-15.0

5 10 15

χ(k

)*k

3

k [Å -1 ]

thin-film

bulk powdera

0.05

0.1

0.15

0.2

0.25

0.3

0 2 4 6 8

FT

( χ(k

)*k

2)

R [Å]

thin-film

bulk powder

b

•The a) oscillating function χ(k) and its Fourier transform, b) FT[ χ(k)*k2 ] function for LiNiO2, as we go from bulk (blue) to thin-film (red).

•The amplitude of the Ni-Ni peak of the thin-film (in the plane of the substrate) is reduced and a shoulder is observed. This total effect cannot be explained by Jahn-Teller distortions.


Theoretical simulation of metal environment

0.05

0.1

0.15

0.2

0 2 4 6

FT

( χ(k

)*k

3)

R [Å]

Individual scattering paths with 0.14 Å splitting

Experimental

Model calculated by FEFF

0 2 4 6 8

R [Å]

Ni Environment in LiNiO2 thin filmNi Environment in LiNiO2 bulk powder

Experimental

Model calculated by FEFF

• Our simulation shows that the set of metal-metal neighbors break away from a six-fold symmetry to a two-fold symmetry; i.e., the six atoms are split into a set of two (at a slightly shorter distance) and four (at slightly larger distance).

• The same effect is observed in LiCoO2 thin films.


EXAFS Experiments Designed to Study the Local Strain in LiCoO2 Thin Films

Samples:

• Multiple 700nm of LiCoO2 was deposited on 2” Si wafer• Samples annealed to different temperatures• Fluorescence EXAFS data collected at X23B and X11B


Bulk Stress Relaxation With Annealing on Isostructural LiCoO2

100 150 200 250 300 350 400 450 500

0.5

1.0

1.5

2.0

2.5

3.0

3.5

4.0

4.5

5.0

1 hr.1 hr.

1 hr.

5 hrs.Intri

nsic

Stre

ss (G

Pa)

Anneal Temp ( 0C)

Wafer curvature measurement of stress, shows bulk stress being annealed out with heat treatment


Co Atomic Neighborhood with Annealing

0 1 2 3 4 5 6

475 oC, 10 hrs.

0 1 2 3 4 5 6

350 oC

0 1 2 3 4 5 6

As deposited

0 1 2 3 4 5 6

100 oC

0 1 2 3 4 5 6

200 oC

0 1 2 3 4 5 6

150 oC

0 1 2 3 4 5 6

700 oC, 2 hrs.

0 1 2 3 4 5 6

bulk powder

R (Å)

Co-O Co-Co

Relaxation of stress takes place in a complex set of steps.

Stress on the Co-O bonds relaxed by 10 hrs. anneal at 475 0C

Stress remains on the Co-Co plane after annealing for 10 hrs at 475 0C.

Stress released by annealing at 700 0C for 2 hrs. However a c-axis expansion remains.


Angle-resolved XAS

0.0

0.5

1.0

1.5

7.68 7.7 7.72 7.74 7.76

0 -10 -20010

Nor

m. a

bsor

ptio

n [a

.u.]

Photon energy [keV]

Angle/ deg

• No texturing effect picked up by the plane-polarized X-rays.

• Orientation of the thin film with respect to the polarized X-ray beam is not the cause of this effect.


The same effect is observed in LiCoO2

• The effect of the amplitude change at the Co-Co bond can be modeled by changing the level of strain on the Co-Co bond.

0 2 4 6

Thin-film LiCoO2annealed at 100 deg C for 1 hr.

R (Å)

|FT

{k2 χ

(k)}

|

Thin-film LiCoO2annealed at 200 deg C for 1 hr.

Bulk LiCoO2 theoreticalexperimental


Experiments at the O K-edge

U7 endstation at the NSLS:

•1200 line grating mono;

•Focused beam (~1.5mm x1.5mm);

•Fluor. and P.E.Y. detector;

•motorized sample manipulator (4 degrees of motional freedom)


O K Fluorescence Yield from LiCoO2 on Si wafer

0.5

1.0

1.5

2.0

5.4 5.6 5.8

abso

rptio

n [a

.u.]

photon energy [keV]Photon energy (10-1 KeV)

σ*

150, 1 hr250, 5 hrs400, 5 hrs

475, 10 hrsSurface oxide state

The degree of covalent σ*

bonding increases with annealing

at the expense of

surface oxide states.


O K Electron Yield from LiCoO2 on Si wafer

0.5

1.0

1.5

5.4 5.6 5.8

abso

rptio

n [a

.u.]

photon energy [keV]Photon energy (10-1 KeV)

150, 1 hr250, 5 hrs400, 5 hrs

475, 10 hrs

σ*

Surface oxide state

Surface oxide states dominate the electron yield signal. They diminish with annealing

while

the degree of covalent σ*

bonding increases with annealing.


Comparison of O K FY data with FEFF8 Theory


Ex-situ Study of Li Intercalation in Ag2V4O11

Ag and V K-edges are well separated in energy and this cathode is not optimized for in-situ studies.

We expect the following things to change as Li is intercalated:- atomic positions (coupled/de-coupled)- oxidation state of V and Ag with charge compensation


Change in the Oxidation State of V

5.464 5.466 5.468 5.470 5.472

0.1

0.2

0.3

0.4

0.5

Nor

mal

Abs

orbt

ion

(au)

Energy (keV)

Ag2V4O11 0.72LiAg2V4O11 2.13LiAg2V4O11 5.59LiAg2V4O11 V2O5 VO2 V2O3


Changes in V Atomic Neighborhood

0 2 40.000

0.001

0.002

0.003

0.004

0.005

FT (k

2 *χ(k

))

Real Space (Å)

Ag2V4O11 0.72LiAg2V4O11 2.13LiAg2V4O11 5.59LiAg2V4O11


Ionic Neighbor Distances Around Vanadium

Equilibrium ionic distances will be a function of the local Coulumbic interactions

We observe that V-V ionic distance decreases with even 0.79 mol Li/ mol SVO


0 1 2 3 4 5 60.000

0.001

0.002

0.003

0.004

0.005

FT (k

2 *χ(k

))

Real Space (Å)

LC of 67% Ag2V4O11 & 33% 5.59LiAg2V4O11

2.13LiAg2V4O11

0 2 4 6 80.000

0.001

0.002

0.003

0.004

0.005

FT (k

2 *χ(k

))

Real Space (Å)

LC of 98% Ag2V4O11 and 2% 5.59LiAg2V4O11

0.72LiAg2V4O11

LC of V Ionic Neighborhood

Phase segregation between SVO and the phase at 5.59 mol Li/ mol SVO

However, this phase segregation not observed at lower lithiation


Ag K-edge analysis of LixAg2V4O11

0 1 2 3 4 5 60.000

0.005

0.010

0.015

0.020

FT(k

2 *χ(k

))

R (Å)

Ag foilAg2V4O11Li0.72Ag2V4O11Li2.13Ag2V4O11Li5.59Ag2V4O11

• FT {χ′} at the silver k-edge of the samples. Features of SVO are well resolved from those of metallic Ag.


S02 Determination in LixAg2V4O11

S02 (fcc Ag) = 0.79

S02 (Li5.59Ag2V4O11) = 0.77

S02 (Li2.13Ag2V4O11) = 0.76

0.60 0.65 0.70 0.75 0.80 0.85 0.90 0.95 1.000.006

0.007

0.008

0.009

0.010

0.011

0.012

0.013

0.014

0.015

0.016

σ2

So2

Li2.13Ag2V4O11 k-weighting 0 k-weighting 1 k-weighting 2 k-weighting 3

N (fcc Ag) = 12N (Li5.59Ag2V4O11) = 11.3

N (Li2.13Ag2V4O11) = 11.25


Disorder: Ag K-edge analysis of LixAg2V4O11

0 1 2 3 4 5 60.000

0.005

0.010

0.015

0.020

FT(k

2 *χ(k

))

R (Å)

Ag foilAg2V4O11Li0.72Ag2V4O11Li2.13Ag2V4O11Li5.59Ag2V4O11

The remaining difference in amplitude at the Ag-Ag bond is from disorder: σ2 (fcc Ag) = 89% σ2 (Li2.13Ag2V4O11)


LC XANES Using SVO and Ag as Endpoints

• A different story from Ag EXAFS

0 1 2 3 4 5 60.0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1.0

Per

cent

age

Amount of Lithium (mol)

Ag Ag2V4O11


Concluding Remarks

• XANES and EXAFS tell complementary stories– Interpretation of XANES and EXAFS data should be internally

consistent • PCA is very useful in providing objective global picture prior to more

subjective analysis of XANES and EXAFS• Attempt should be made to look at all the relevant atomic species in your

system– Low energy excitations can provide some very useful information, e.g.

3d transition metal L-edge data.• Linear Combination approaches can simplify an otherwise complicated

picture– Especially useful for systems where the correct structural model is not

known.• In EXAFS analysis, de-coupling steps should be taken to extract information

about correlated properties.

xas studies of battery materials · xas studies of nanocatalysis and chemical transformation co...

Documents