Revisiting Conversion Reaction Mechanisms in Lithium Batteries: Lithiation-Driven Topotactic Transformation in FeF2

Article abstract of DOI:10.1021/jacs.8b07740

Intercalation-type electrodes have now been commonly employed in today's batteries as such materials are capable of storing and releasing lithium reversibly via topotactic transformation, conducive to small structural change, but they have limited interstitial sites to hold Li. In contrast, conversion electrodes feature high Li-storage capacity, but often undergo large structural change during (de)lithiation, resulting in cycling instability. One exception is iron fluoride (FeF₂), a conversion-type cathode that exhibits both high capacity and high cycling stability. Herein, we report a lithiation-driven topotactic transformation in a single crystal of FeF₂, unveiled by in situ visualization of the spatial and crystallographic correlation between the parent and converted phases. Specifically, conversion in FeF₂ resembles the intercalation process but involves transport of both Li⁺ and Fe²⁺ ions within the F-anion array, leading to formation of Fe preferentially along specific crystallographic orientations of FeF₂. Throughout the process, the F-anion framework is retained, creating a checkerboard-like structure, within which the volume change is largely compensated, thereby enabling the high cyclability in FeF₂. Findings from this study, with unique insights into conversion reaction mechanisms, may help to pave the way for designing conversion-type electrodes for the next-generation high energy lithium batteries.

Full text of DOI:10.1021/jacs.8b07740

Subscriber access provided by University of South Dakota
Article
Revisiting Conversion Reaction Mechanisms in Lithium
2
Batteries: Lithiation-Driven Topotactic Transformation in FeF
Khim Karki, Lijun Wu, Ying Ma, Mark J. Armstrong, Justin D. Holmes,
Stephen H. Garofalini, Yimei Zhu, Eric A. Stach, and Feng Wang
J. Am. Chem. Soc., Just Accepted Manuscript • DOI: 10.1021/jacs.8b07740 • Publication Date (Web): 20 Nov 2018
Downloaded from http://pubs.acs.org on November 20, 2018
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a service to the research community to expedite the dissemination
of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in
full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully
peer reviewed, but should not be considered the official version of record. They are citable by the
Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore,
the “Just Accepted” Web site may not include all articles that will be published in the journal. After
a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web
site and published as an ASAP article. Note that technical editing may introduce minor changes
to the manuscript text and/or graphics which could affect content, and all legal disclaimers and
ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or
consequences arising from the use of information contained in these “Just Accepted” manuscripts.
is published by the American Chemical Society. 1155 Sixteenth Street N.W.,
Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the course
of their duties.

Page 1 of 11
Journal of the American Chemical Society
Revisiting Conversion Reaction Mechanisms in Lithium Batteries:
Lithiation-Driven Topotactic Transformation in FeF
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
2
*
,†,Ψ
‡,Ψ
¥
£,δ
£,δ
Khim Karki,
Lijun Wu, Ying Ma, Mark J. Armstrong, Justin D. Holmes, Stephen H.
φ
‡
§,
*, †
Garofalini, Yimei Zhu, Eric A. Stach, and Feng Wang
†
Sustainable Energy Technologies Department, Brookhaven National Laboratory, Upton, New York, 11973, United
States
‡
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Department of Condensed Matter Physics and Materials Science, Brookhaven National Laboratory, Upton, New
York, 11973, United States
¥
Materials Science & Engineering, University of Wisconsin-Eau Claire, Wisconsin, 54701, United States
£
School of Chemistry and the Tyndall National Institute, University College Cork, Cork, T12 YN60, Ireland
δ
AMBER@CRANN, Trinity College Dublin, Dublin 2, Ireland
φ
Department of Materials Science and Engineering, Rutgers University, 607 Taylor Rd, Piscataway, NJ 08854,
United States
§
Center for Functional Nanomaterials, Brookhaven National Laboratory, Upton, New York, 11973, United States

Department of Materials Science and Engineering, University of Pennsylvania, Philadelphia, PA 19104, United
States
KEYWORDS: Lithium batteries, Conversion reaction mechanisms, Topotactic transformation, Iron fluorides (FeF )
2
Intercalation-type electrodes have now been commonly employed in today’s batteries due to their capability of
storing and releasing lithium reversibly via topotactic transformation, conducive to small structural change, but
they have limited interstitial sites to hold Li. In contrast, conversion electrodes feature high Li-storage capacity, but
often undergo large structural change during (de)lithiation, resulting in cycling instability. One exception is iron
fluoride (FeF
report a lithiation-driven topotactic transformation in a single crystal of FeF
spatial and crystallographic correlation between the parent and converted phases. Specifically, conversion in FeF
2
), a conversion-type cathode that exhibits both high capacity and high cycling stability. Herein, we
2
, unveiled by in situ visualization of the
2
+
2+
resembles the intercalation process but involves transport of both Li and Fe ions within the F-anion array,
leading to formation of Fe preferentially along specific crystallographic orientations of FeF . Throughout the
process, the F-anion framework is retained, creating a checkerboard-like structure, within which the volume
change is largely compensated, thereby enabling the high cyclability in FeF . Findings from this study, with unique
2
2
insights into conversion reaction mechanisms, may help to pave the way for designing conversion-type electrodes
for the next-generation lithium batteries.
Conversion-based transition metal (TM) compounds,
TM) (TM = Fe, Co, Ni, Cu; X = O, S, F, P, N), particularly
oxides (e.g., Fe , CoO, NiO) and fluorides (FeF , FeF
CuFeF ) – have recently received enormous attention for use
from the parent host structure, leading to inhomogeneous
phases, e.g. metallic TM and insulating LiX and, consequently,
capacity loss of the electrode during subsequent reconversion
(
a b
X
3
O
4
2
3
,
1
1, 12
2
and cycling processes.
In addition, the large volume
in lithium batteries because of their extremely high specific
expansion that occurs during the conversion reaction leads to
capacity (500–1000 mAh/g), 3–4 times greater than the
intercalation compounds. In contrast to intercalation-based
pulverization of the electrodes, and contributes further to the
irreversibility of the reaction. The low cycling reversibility of
1–6
6
electrodes, which function by shuttling lithium ions between
conversion electrodes has thus been a major impediment to
their commercialization.
11–16
the tunnels of the host through a topotactic transformation
7
process, with causing significant structural changes,
Few conversion-based electrodes such as metal fluorides
electrochemical reactions in conversion-type electrodes are
complex, involving migration/re-ordering of the TM ions, and
8,9
have shown exceptional cycling reversibility. Notably, up to
17
1
2
000 cycles has been demonstrated in FeF , which, along with
n+
+
y
eventually the extrusion of the metallic TMs (TM X + nLi +
the findings that voltage hysteresis is largely kinetic in
-
0
6,8–10
ne = yLin/yX + TM ; n ≥ 2), leading to significant .
The
9,12,14,18
nature
, indicates potential viability of the fluoride based
converted products are spatially distributed and decoupled
battery systems for commercial use. High cycling reversibility
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 2 of 11
1
0
has also been reported in some conversion oxides and
Supporting Information; Section I). The progression of the
lithium reaction in FeF is tracked by the in situ TEM method
1
5
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
displacement-type intermetallic compounds . For instance,
2
Zhang et al. observed retention of the cubic close-packed (ccp)
with individual particles brought in point contact with the
lithium source (See Supporting Information, Section II, Figure
S2, and Movie M1). TEM images were recorded at a frequency
of 5 frames per second from a local area that is approximately
33 nm wide, and 33 nm long (as marked by red frame in Figure
1a). Some of the sequential images extracted from Movie M1
are shown in Fig. 1c. Upon applying the bias, the conversion
occurs first from the surface region due to fast Li diffusion
along the surface. Those converted Fe nanoparticles, being
discerned by dark contrast under the bright field (BF)
condition, form a percolating network, and so serve as the
O-anion framework of Fe
3 4
O during the electrochemical
cycling, thereby enabling multiple lithium intercalation and
1
0
3 4
conversion reaction. The electrochemical reaction in Fe O
shows high similarity to that in displacement-type Li-Cu-Sb
compounds, wherein remarkable cycling reversibility was
observed and attributed to retaining the face-centered cubic
(
fcc) Sb structure through the conversion/reconversion
15
processes. Results from these studies indicate that the
retention of the original structural framework is crucial to
achieving high reversibility of electrodes. Despite these
observations, the exact mechanisms that lead to the structural
retention between the parent and product phases during the
conversion reaction are not well understood.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
3
4, 35
conducting pathway for electronic transport.
As the
reaction proceeds further into the bulk, a well-aligned and
repeating fringe-like pattern grows as an overlayer on the
<
110> FeF
indicated by arrows in the images taken following 2s, 4s, 5s,
s, 10s, 17s and 30s of lithiation, respectively. The reaction
front propagates gradually from surface to the bulk, layer-by-
2
. The Fe domains are formed periodically, as
Here, we present real-time observation of a conversion-
driven topotactic phase transformation in a single crystal of
6
FeF
crystallographic orientations of the parent FeF
transmission electron microscopy (TEM) techniques.
2
, with preferential growth of Fe phase along specific
2
, using in situ
3
4,35
9,19,20
layer (Figure 1c; as illustrated by blue arrows).
Topotactic-mode of phase transformation is usually observed
at high temperatures and during the process the product
phase evolves from the reactant parent substrate by
To identify the local structural change, high-resolution
TEM (HRTEM) images were taken in the pristine state from
selected small area (t = 0 s; marked by a blue frame in Figure
1a), and after full lithiation (t = 30 s; provided in Figure d and
2
1
maintaining structural symmetry between the two. Some
other cases where topotactic reaction mechanisms have been
2
f respectively). The pristine FeF is highly crystalline, as shown
22–24
reported are oxidation-reduction of metal oxides,
by the fast Fourier transform (FFT; inset) of the HRTEM
image, and its lattice is highly-ordered as seen from the
inverse FFT image (Figure 1e; from the local area labelled by
21,24
dehydration
transformations,
reactions,
nanostructured synthesis,
polymeric
phase
2
4,25
26,27
and growth
2
8
of thin films (EuO from the reduction of Eu
2
O
3
). .
the dotted box in Figure 1d). Upon lithium conversion, FeF
turns into a network of nanoscale Fe domains, similar to what
2
Electrochemical intercalation via shuttling lithium ions
between the tunnels of the host is another special case of
was observed in the previous studies on small FeF
2
7
8,9,36
topotactic transformation. However, such a conversion-
nanoparticles
but they appear to arrange in a fringe-like
driven topotactic growth of Fe from FeF
2
is totally unexpected
pattern, as shown in Figure 1f. At this stage, the initial periodic
arrangement of FeF is disrupted and forms a highly strained
because the two have a completely different structure, and the
reaction occurs at room temperature (RT). Although extensive
2
interconnected network of Fe nano-domains surrounded by
random dislocations (denoted as “T”), as clearly shown in the
inverse FFT image in Figure 1g (extracted from the dotted
frame area in Figure 1f). The characteristic fringe-like pattern
of newly formed Fe suggests Fe grows with preferential
in situ studies have been performed on FeF
2
(or its
derivatives), such a conversion-driven topotactic process has
8,9,17,29–32
never been reported.
On one hand, conversion-type
reaction is complex, involving multiphase transformations
and local transport of cations and lithium ions within nano-
sized domains. On the other hand, powder or polycrystalline
materials have mostly been employed in studies of these
processes, and the random orientation of constituent
crystallites obscures the topotactic relationships as they
orientation with respect to the underlayer FeF
2
– in this case
along [110] direction of the parent FeF
below).
2
(as to be discussed
To elucidate the structural correlation between Fe and
the parent FeF , time-resolved in situ electron diffraction
2
8
, 9
2
develop. The use of single-crystalline FeF particles in this
measurements were performed (See Supporting Information,
Section II). Figure 2a presents a time-sequence of electron
diffraction patterns, captured from a local area on the same
study allows tracking of the crystallographic evolution of the
involved phases, i.e. the progression of the lithium conversion
reaction along specific crystallographic orientations.
2
single-crystalline FeF particle but further away from the area
For this study, plate-like FeF
2
single-crystalline particles
for in situ HRTEM measurements (see also Supporting
Information, Movie M2). The corresponding azimuthal
projection of the diffraction patterns (x-axis=angle, y-
axis=radius) was extracted (Figure 2b) to determine all of the
peak parameters via a precise profile analysis (intensity vs.
diffraction) of the selected area diffraction pattern using the
were synthesized by a supercritical fluid method (See
Supporting Information, Figure S1).33 Figure 1a shows the
bright-field TEM image of a typical FeF
oriented along the [001] crystallographic direction, as verified
by electron diffraction (Figure 1b). The FeF single-crystal
appears to have a primary growth axis along the [110]
2
particle, which is
2
3
7
PASAD-tools software.
Prior to lithiation (t = 0 s), characteristic diffraction spots
for single-crystalline FeF particle are observed along the [001]
̅
̅
direction, with two small facets along (020) and (200), and a
̅
longer facet running along (110). Because FeF
2
is an electrical
2
4
insulator, to improve the electrical conductivity, the surface
zone axis. As the lithiation progresses, four new, broad, arc-
shaped diffraction spots appear, faint initially and becoming
of FeF
2
particle was coated with a thin layer of carbon (See
ACS Paragon Plus Environment

Page 3 of 11
Journal of the American Chemical Society
stronger with time (Supporting Movie M2). By t = 8 min, the
arc-shaped spots can be clearly seen at positions adjacent to
the spots of {200}FeF (as labeled by green arrows in Figure 2a)
(200)FeF
2
, indicating no lattice distortion or expansion within
9
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
the resolution limits of the diffraction technique, although
there have been speculations regarding the possibility of
lithium intercalation during the initial lithiation.
While the (110)Fe peaks shift towards higher angles in the early
stage, likely due to the formation of small Fe crystallites with
slight lattice expansion, as observed previously.
2
3
2,34,35,39
and better resolved in the corresponding projected patterns in
Figure 2b. The arc-shaped diffraction spots represent a mosaic
2
4
of tiny Fe crystallites, with a four-fold symmetry (indexed to
the {110}Fe family of planes), thus indicating a topotactic
9
2
1
relationship between the converted Fe and the parent FeF
2
.
To corroborate the in situ TEM observations, we
performed ex situ TEM analysis on electrochemically
discharged FeF via a quasi in situ TEM coin-cell technique
2
Although the coherent nature of Fe crystallites was observed
8
,9
previously, topotactic growth of Fe from the parent FeF
2
is
40
surprising and, indeed unexpected, considering that the two
have a completely different structure (bcc vs. tetragonal).
Furthermore, the topotactic conversion reaction occurs at RT
(See Supporting Information, Section III and Figure S4). A
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
typical BF TEM image from a partially-discharged FeF
particle, and the corresponding diffraction pattern taken from
the local area (as marked by dotted circle) are given in Figures
2
3
8
-- starkly different from the traditional epitaxial
topotactic
or
2
1,24
growth at high temperatures. It should be noted
3a, b. Clearly the diffraction pattern shows the similar features
that the use of a single crystal is crucial to deconvoluting this
information in this in situ study, thereby allowing us to
observe the evolving crystallographic relationship in detail.
The majority of the Fe that is formed (as labeled with green
arrows; Case – I from hereafter), tends to have a strong
as that in the in situ measurements (Figure 2), further
validating the in situ observation of the topotactic growth of
Fe from the parent FeF . Figure 3c presents the integrated
2
intensities from the (110)Fe (Case-I) and (101)Fe (Case-II)
̅
̅
̅
reflections. The higher intensity of the (110)Fe (Case-I)
2
orientation relationship with the parent FeF , such that
reflection indicates that this orientation relationship is more
frequent in the sample. Figure 3d is a magnified image
extracted from the yellow framed area of Figure 3a. There is a
similar fringe-like pattern as found in Figures (1c, 1d). This
pattern was interpreted to be a result of preferential formation
̅
[
001]Fe-I || [001]FeF
2
, [110]Fe-I || [010]FeF
2
and [110]Fe-I ||
[100]FeF (Fe-I and FeF
2
2
represent Case-I Fe domains and FeF
2
,
respectively). Interestingly, there are also additional set of arcs
associated with Fe, as indicated by red arrows (Case – II from
hereafter), demonstrating the presence of a second form of
2
of Fe along specific orientation of FeF . The FFT of the given
topotactic Fe growth, aligned with FeF
2
by a new relationship:
area, which is shown in the inset, is similar to the diffraction
patterns extracted from the large area (Figure 3b). However,
the fringe-like pattern observed in the ex situ studies is similar
to that that observed at 30s during the in situ experiment
(Figure 1g), which should be due to the relaxation effect, after
removal of the electric bias in the coin cell.
̅
[101]Fe-II ||[001]FeF , [010]Fe-II || [110]FeF
2
2
and [101]Fe-
̅
II||[110]FeF
2
. Note, [020]Fe-II is along the Li diffusion
direction. While faint and somewhat diffuse at this stage of
the process, they do not have a four-fold symmetry (as in Case-
I). Such a topotactic evolution of Fe over the parent FeF
Cases I & II) was commonly observed in this experiment; see
2
(
Interestingly, one would expect that Fe is least mobile
given its large atomic mass and the strong Fe - F ionic
bonding. However, it has been shown that Fe ions are
quickly reduced to neutral Fe upon lithiation, which
essentially breaks the Fe-F bonds and results in Fe atoms that
are bonded neither to Li nor F . Thus, Fe atoms diffuse freely
out of the mixture and form the metallic phase. To confirm
this, reactive molecular dynamics (MD) simulations were
performed, with the results showing that the diffusion
also another set of diffraction data recorded along [001]
oriented FeF in Supporting Information, Figure S3.
During further lithiation (from t = 17ˈ to 19ˈ), Fe grains
grew with the consumption of the parent FeF , and
correspondingly, those arc-shaped diffraction spots became
stronger, and the spots associated with FeF became weaker.
Figure 2c shows the progression of the spot intensity for
2
+
-
2
2
+
0
34
2
0
+
-
0
2
34
̅
̅
(
110)Fe and (101)Fe as a function of lithiation time, in
comparison to that of the parent (200)FeF
2
spot. As
0
coefficient of Fe is dramatically increased as compared to that
̅
2+
0
demonstrated by the intensity profiles, the (110)Fe spot,
of Fe , and Fe is, indeed, the most mobile species
Supporting Information, Section IV and Table S1). This is
important, as this observation suggests that the growth of the
Fe from FeF is driven by electrochemical lithiation at RT,
which represents the Case I (green arrows) orientation
(
̅
relationship, is dominant over the (101)Fe, which represents
Case II (red arrows). In both Cases I and II, there are large
spikes in the intensities of Fe reflections at the initial state of
lithiation. However, after 17 min, the reflection intensities for
both cases saturate and fade in a gradual manner. The growth
of Fe grain with lithiation introduces lattice distortion,
yielding dislocations (Figure 1g) and coarsening of the Fe
domains, which is likely responsible for deceleration of the
2
2+
which leads to the reduction of Fe . Considering the
difference in diffusivity, the lithium conversion reaction along
the [001] FeF
2
direction must involve the reduction and then
diffusion of Fe as well as the local displacement and ordering
+
of Li within the close-packed F-anion framework since both
+
-
Li and F have a much smaller diffusion coefficient. So, based
on the in situ observations, we attempt to develop a structural
model to explain the observed conversion-driven topotactic
11,39
kinetics of conversion reaction in the bulk.
The overall phase evolution during lithiation was also
examined by obtaining the integrated intensity profiles from
the entire diffraction patterns (full rings); see Figure 2d, e.
phase transformation in FeF
Figure 4a presents a 3D view of the FeF
alternative layers of FeF along [001] direction in the unit cell.
Two large channels exist in the [001] and [110] directions, and
it has been shown that [001] is energetically more favorable for
2
.
2
structure, with
2
Clearly, the reflections associated with the parent FeF are
2
initially intense, and as the lithiation progresses, they become
weaker; while the broad Fe peaks become stronger. We did
not observe an obvious shift in the peak position of the
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 4 of 11
lithium ion diffusion.34,39 Lithium insertion along the [001]
direction (top or bottom surface of FeF particle) would push
II domains with an orientation of [010]Fe-II along the [110]FeF
direction, which is the direction of Li diffusion in the FeF
2
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
2
2
the Fe out either in the [010] or [100] direction, while the
inserted Li ions would likely occupy the center of the
trapezoid constructed by the F atoms, as shown in Figure 4b.
Note that because LiF will expand along the c direction, there
is no room for Fe to move toward the c direction. Because the
bonding length of LiF is about 0.2 nm (based on its bulk
structure), it would shrink along both a and b direction, while
expanding along the c direction to form Li-F bonding, as
particle. The likelihood of forming this kind of domain would
also be low, consistent with the faint diffraction spots (red
arrows) present in Figure 2. As evident from Figure S6, the
distribution of Fe that has the Case I orientation is dominant
when compared to Case II, agreeing well with both the
experimental data and the structural analysis presented
above. These results are also consistent with earlier
computational modeling where the lithiation along the [110]
direction was observed to be very limited as compared to that
2
shown in Figure 4c. Once Fe is out of FeF lattice, it would
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
34
alternately shift about 0.2 nm along the a direction to form
Fe-Fe bonding due to the strong force along the b direction,
as shown in Figure 4b. The (001) Fe layer would then contract
to form a Fe lattice, as shown in Figure 4c. The resulting local
cationic displacement and reordering will result in the
along the [001] direction.
The lithiation process in the plate-like single crystalline
particle proceeds via layer-by-layer, from surface to the bulk
(
as in Figure 1), similar to the observation on small FeF
2
8,9
nanoparticles (10-20nm) despite the different shapes . The
high similarity is due to the fact that FeF is an insulator, the
2
following relationship between the Fe and FeF lattices:
2
̅
[
[
001]Fe-I || [001]FeF
100]FeF
2
, [110]Fe-I || [010]FeF
, as experimentally observed (Case – I) The
2
, [110]Fe-I ||
reaction front propagates, “layer-by-layer”, into the bulk,
during which the Fe percolating network is gradually built up
to provide electronic transport pathway. The plate-like shape
also allows easy access to lithium, similar to that in the small
nanoparticles. In addition, the formation of the percolating Fe
2
expansion/contraction of Fe/LiF is different along two
different directions during the phase transformation of FeF
2
2
+
+
-
Li + 2e → Fe + 2LiF. The phase transformation results in
expansion along the b direction, contraction along the a
direction, and expansion along the c direction for LiF, but
contraction for Fe. Note that due to the four-fold symmetry of
8
network after lithiation resembles the previous ex situ and in
9
situ observations from small nanoparticles. But, by using a
plate-like particle, we were able to reveal more details of the
lithiation process forming checkerboard-like structure
wherein volume change is largely compensated, thus enabling
structural integrity and cycling stability in such conversion-
type electrodes.
2
FeF , Fe can be equivalently pushed out along the [100]
direction. In this case, we get another orientation relationship:
̅
[001]Fe-Iˈ || [001]FeF
010]FeF
2
, [110]Fe-I || [100]FeF
(Fe-Iˈ represents the equivalent Case-I Fe domain).
2
and [110]Fe-Iˈ ||
[
2
The diffraction spots of the Fe-Iˈ domain would overlap with
those of the Fe-I domain. The expansion/contraction of the
Fe-Iˈ domain is the same as that of the Fe-I domain except it
expands along the a direction but contracts along the b
direction. Therefore, the converted LiF and Fe likely form a
checkerboard-like pattern in all three directions, as illustrated
in Figure 4d, in which the expansion/contractions are
compensated to fit the 3D space. The domain size of Fe/LiF
would thus be very small and difficult to grow. Interestingly,
the Fe lattice fringes that are observed along the [110] direction
in the thin area during the lithiation conversion process in
Figures 1c and 1f are consistent with this checkerboard-like
arrangement of Fe/LiF domains. Because of the four-fold
And importantly, the use of single-crystal samples
allowed us to reveal the crystallographic correlation between
the involved intermediates, namely topotactic transformation
process, which was not observed in the previous studies. The
observation in this study also explains the coherent nature of
the converted Fe crystallites reported in the previous
8
,9
studies. On the other hand, such an unexpected topotactic
conversion process resembles the conventional intercalation
2
process, suggesting that conversion in FeF proceeds via co-
+
2+
operative ionic transport of Li and Fe within the close-
packed F-anion array, with the framework retained
throughout the process (so resembling the intercalation
process). In a traditional view, the electrochemical reaction in
electrodes either occurs via intercalation process, namely
2
symmetry of FeF , the Fe fringes can align parallel to [110]
direction. The two orientated Fe fringes may overlap along the
beam direction. This would form a two-dimensional pattern,
but image evidence for this would be less pronounced when
the sample is thick.
+
insertion of Li ions into interstitial sites without breaking the
crystal lattice of the host, or via conversion that involves local
migration/re-ordering of transition metal (TM) ions, and
eventually, the extrusion of metallic TMs. Results from this
study imply that these two types of reaction may be viewed
the same, both through topotactic transformation process
Alternatively, Li insertion along the [110] FeF
from the side surfaces of the FeF particle in Figure 1a) would
push the Fe out along the [110] FeF direction (See Supporting
Information, Figure S5). In this case, LiF would shrink along
2
direction
(
2
2
involving ionic transport and ordering of cations (Li and TM
+
ions) in a close-packed anion framework. These findings may
explain the origin of high cyclability in FeF via topotactic
2
mechanism since the structural relationship between the
parent and converted species are preserved and presumably
provides reaction pathway for further reconversion during
cycling.
̅
both [110] and [110] directions, while expanding along the c
direction to form Li-F bonding, as shown in Figure S5c.
Similarly, the spacing of the Fe layers would expand vertically
to form an Fe lattice. In this case, the relationship between the
Fe and FeF
2
lattice is: [101]Fe-II || [001]FeF
2
, [010]Fe-II ||
̅
̅
[
110]FeF
2
, and [101]Fe-II || [110]FeF
2
. Because the insertion of
Whereas the neutral Fe atoms are the most mobile,
Li along the [110] direction is more difficult than that along the
+
-
charged ions, including Li and F , are expected to have
increased diffusion coefficients due to the presence of the
external field. Formation of intermediate phases during
3
4,39
[001] direction,
a driving force is needed to make this
conversion happen. Thus, it is logical that we only see the Fe-
ACS Paragon Plus Environment

Page 5 of 11
Journal of the American Chemical Society
reconversion is also possible, which is supported by both
SC0001294. This research used resources of the Center for
1
2,14
41
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
experimental
and theoretical evidences. These differences
Functional Nanomaterials, which is a U.S. DOE Office of
Science Facility, at Brookhaven National Laboratory under
Contract No. DE-SC0012704. Lijun Wu and Yimei Zhu were
supported by the U.S. Department of Energy, Office of Basic
Energy Science, Division of Materials Science and
Engineering, under Contract No. DE-SC0012704. MJA and
JDH acknowledge financial support from Science Foundation
Ireland (Project: 07/SRC/I1172). We thank Tiffany Bowman for
the graphic design of the TOC.
could give rise to a distinct reaction mechanism for
reconversion. It would be desirable to track the reconversion
process via in situ TEM method, but difficult with the current
setup. A few studies using single nanowire based batteries and
using LiPON-based electrolyte in the TEM have shown
4
2,43
promising in achieving full charge/discharge cycle.
Nonetheless, it is challenging to accommodate and
manipulate single-crystalline particles as shown in the current
study. In order to observe the re-conversion process, a
completely different in situ TEM platform, incorporated with
robust solid electrolyte and single-crystalline particles with
desired orientation, needs to be developed.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
REFERENCES
(
1)
Li, C.; Chen, K.; Zhou, X.; Maier, J. npj Comput. Mater. 2018,
4 (1), 22.
In summary, through in situ observation of lithiation
process in single-crystalline FeF
2
, we revealed conversion-
(2)
Yu, S.-H.; Feng, X.; Zhang, N.; Seok, J.; Abruña, H. D. Acc.
Chem. Res. 2018, 51 (2), 273.
driven topotactic transformation. During the process, the
converted Fe phase is not randomly oriented as traditionally
believed, but preferentially aligned along specific
crystallographic orientations of the parent FeF . Structural
2
modeling suggests the formation of a checkerboard-like
arrangement of Fe/LiF domains upon topotactic displacement
(
(
3)
4)
Wang, F.; Kim, S.-W.; Seo, D.-H.; Kang, K.; Wang, L.; Su, D.;
Vajo, J. J.; Wang, J.; Graetz, J. Nat Commun 2015, 6.
Badway, F.; Cosandey, F.; Pereira, N.; Amatucci, G. G. J.
Electrochem. Soc. 2003, 150 (10), A1318.
(5)
(6)
Poizot, P.; Laruelle, S.; Grugeon, S.; Dupont, L.; Tarascon, J.-
M. Nature 2000, 407 (6803), 496.
Cabana, J.; Monconduit, L.; Larcher, D.; Palacín, M. R. Adv.
Mater. 2010, 22 (35), E170.
of Fe from the parent FeF
2
. Consequently, a percolating 3D
network of Fe and LiF is formed, which may explain the
previous observation of bi-continuous network, a pathway for
local electron transport. These observations present unique
insights into the conversion reaction mechanisms, and so may
help to pave the way for designing conversion-type electrodes
for high energy density lithium batteries.
(
(8)
7)
Whittingham, M. S. Chem. Rev. 2014, 114 (23), 11414.
Wang, F.; Robert, R.; Chernova, N. A.; Pereira, N.; Omenya,
F.; Badway, F.; Hua, X.; Ruotolo, M.; Zhang, R.; Wu, L.;
Volkov, V.; Su, D.; Key, B.; Whittingham, M. S.; Grey, C. P.;
Amatucci, G. G.; Zhu, Y.; Graetz, J. J. Am. Chem. Soc. 2011, 133
(
46), 18828.
(9)
Wang, F.; Yu, H.-C.; Chen, M.-H.; Wu, L.; Pereira, N.;
Thornton, K.; Van der Ven, A.; Zhu, Y.; Amatucci, G. G.;
Graetz, J. Nat Commun 2012, 3, 1201.
ASSOCIATED CONTENT
(
10)
Zhang, W.; Bock, D. C.; Pelliccione, C. J.; Li, Y.; Wu, L.; Zhu,
Y.; Marschilok, A. C.; Takeuchi, E. S.; Takeuchi, K. J.; Wang,
F. Adv. Energy Mater. 2016, 6 (10), 1502471.
Supporting Information
Materials preparation procedures, X-ray diffraction data and
(11)
Yu, H.-C.; Wang, F.; Amatucci, G. G.; Thornton, K. J. Phase
Equilibria Diffus. 2016, 37 (1), 86.
scanning electron microscopy images of the synthesized FeF
2
particles, in situ TEM experimental methods, electrochemical
discharge/ex situ TEM diffraction, molecular dynamic
simulations are included in the supporting information. This
material is available free of charge via the Internet at
http://pubs.acs.org.
(12)
Li, L.; Jacobs, R.; Gao, P.; Gan, L.; Wang, F.; Morgan, D.; Jin,
S. J. Am. Chem. Soc. 2016, 138 (8), 2838.
(
13)
Doe, R. E.; Persson, K. A.; Meng, Y. S.; Ceder, G. Chem. Mater.
2008, 20 (16), 5274.
(14)
Ko, J. K.; Wiaderek, K. M.; Pereira, N.; Kinnibrugh, T. L.; Kim,
J. R.; Chupas, P. J.; Chapman, K. W.; Amatucci, G. G. ACS
Appl. Mater. Interfaces 2014, 6 (14), 10858.
AUTHOR INFORMATION
(
15)
Chang, D.; Chen, M.-H.; Van der Ven, A. Chem. Mater. 2015,
2
7 (22), 7593.
Corresponding Author
(16)
(17)
Kim, S.; Liu, J.; Sun, K.; Wang, J.; Dillon, S. J.; Braun, P. V.
Adv. Funct. Mater. 2017, 1702783.
Gu, W.; Borodin, O.; Zdyrko, B.; Lin, H.-T.; Kim, H.; Nitta,
N.; Huang, J.; Magasinski, A.; Milicev, Z.; Berdichevsky, G.;
Yushin, G. Adv. Funct. Mater. 2016, 26 (10), 1507.
kkarki@bnl.gov
fwang@bnl.gov
Author Contributions
(18)
Liu, P.; Vajo, J. J.; Wang, J. S.; Li, W.; Liu, J. J. Phys. Chem. C
2
012, 116 (10), 6467.
Ψ
These authors contributed equally.
(
19)
Huang, J. Y.; Zhong, L.; Wang, C. M.; Sullivan, J. P.; Xu, W.;
Zhang, L. Q.; Mao, S. X.; Hudak, N. S.; Liu, X. H.;
Subramanian, A.; Fan, H.; Qi, L.; Kushima, A.; Li, J. Science
(80-. ). 2010, 330 (6010), 1515.
Notes
The authors declare no competing financial interests.
(20)
21)
(22)
(23)
(24)
Karki, K.; Epstein, E.; Cho, J.-H.; Jia, Z.; Li, T.; Picraux, S. T.;
Wang, C.; Cumings, J. Nano Lett. 2012, 12 (3), 1392.
Shannon, R. D.; Rossi, R. C. Nature 1964, 202 (4936), 1000.
Revcolevschi, A.; Dhalenne, G. Nature 1985, 316, 335.
Cahn, R. W. Nature 1985, 316 (6026), 297.
Figlarz, M.; Gérand, B.; Delahaye-Vidal, A.; Dumont, B.;
Harb, F.; Coucou, A.; Fievet, F. Solid State Ionics 1990, 43,
ACKNOWLEDGMENT
(
This work is financially supported by Northeastern Center for
Chemical Energy Storage, an Energy Frontier Research Center
funded by the U.S. DOE, BES under award number DE-
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 6 of 11
1
43.
Söderquist, R.; Dickens, B. J. Phys. Chem. Solids 1967, 28 (5),
23.
Meng, F.; Morin, S. A.; Jin, S. J. Am. Chem. Soc. 2011, 133 (22),
408.
(35)
Thorpe, R.; Rangan, S.; Whitcomb, R.; Basaran, A. C.;
Saerbeck, T.; Schuller, I. K.; Bartynski, R. A. Phys. Chem.
Chem. Phys. 2015, 17 (23), 15218.
Li, L.; Meng, F.; Jin, S. Nano Lett. 2012, 12 (11), 6030.
Gammer, C.; Mangler, C.; Rentenberger, C.; Karnthaler, H. P.
Scr. Mater. 2010, 63 (3), 312.
Tang, L.; Feng, Y. C.; Lee, L.-L.; Laughlin, D. E. J. Appl.
Crystallogr. 1996, 29 (4), 419.
Ma, Y.; Garofalini, S. H. Phys. Chem. Chem. Phys. 2014, 16
(23), 11690.
Lin, F.; Nordlund, D.; Weng, T.-C.; Zhu, Y.; Ban, C.; Richards,
R. M.; Xin, H. L. Nat Commun 2014, 5.
Ma, Y.; Garofalini, S. H. J. Phys. Chem. C 2017, 121 (28), 15002.
Ruzmetov, D.; Oleshko, V. P.; Haney, P. M.; Lezec, H. J.;
Karki, K.; Baloch, K. H.; Agrawal, A. K.; Davydov, A. V;
Krylyuk, S.; Liu, Y.; Huang, J.; Tanase, M.; Cumings, J.; Talin,
A. A. Nano Lett. 2012, 12 (1), 505.
Santhanagopalan, D.; Qian, D.; McGilvray, T.; Wang, Z.;
Wang, F.; Camino, F.; Graetz, J.; Dudney, N.; Meng, Y. S. J.
Phys. Chem. Lett. 2014, 5 (2), 298.
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
(25)
8
(
(
(
26)
27)
28)
(36)
(37)
8
Moon, G. D.; Ko, S.; Min, Y.; Zeng, J.; Xia, Y.; Jeong, U. Nano
Today 2011, 6 (2), 186.
(38)
(39)
(40)
Mairoser, T.; Mundy, J. A.; Melville, A.; Hodash, D.; Cueva,
P.; Held, R.; Glavic, A.; Schubert, J.; Muller, D. A.; Schlom, D.
G.; Schmehl, A. Nat. Commun. 2015, 6, 7716.
He, K.; Zhang, S.; Li, J.; Yu, X.; Meng, Q.; Zhu, Y. Y.; Hu, E.;
Sun, K.; Yun, H.; Yang, X.-Q.; Zhu, Y. Y.; Gan, H.; Mo, Y.;
Stach, E. A.; Murray, C. B.; Su, D. Nat Commun 2016, 7, 11441.
Li, L.; Chen-Wiegart, Y. K.; Wang, J.; Gao, P.; Ding, Q.; Yu,
Y.-S.; Wang, F.; Cabana, J.; Wang, J.; Jin, S. Nat Commun
(
(
29)
30)
(41)
(42)
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
2
015, 6.
(31)
Wiaderek, K. M.; Borkiewicz, O. J.; Castillo-Martínez, E.;
Robert, R.; Pereira, N.; Amatucci, G. G.; Grey, C. P.; Chupas,
P. J.; Chapman, K. W. J. Am. Chem. Soc. 2013, 135 (10), 4070.
Yamakawa, N.; Jiang, M.; Key, B.; Grey, C. P. J. Am. Chem.
Soc. 2009, 131 (30), 10525.
Armstrong, M. J.; Panneerselvam, A.; O’Regan, C.; Morris, M.
A.; Holmes, J. D. J. Mater. Chem. A 2013, 1 (36), 10667.
Ma, Y.; Garofalini, S. H. J. Am. Chem. Soc. 2012, 134 (19), 8205.
(43)
(
(
(
32)
33)
34)
ACS Paragon Plus Environment

Page 7 of 11
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
TOC Graphic
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
7
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 8 of 11
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Figure 1. Lithium conversion and phase propagation within a single crystal of FeF . (a, b) Bright-
2
field TEM image of a typical single-crystalline FeF particle and the corresponding electron diffraction
2
pattern taken from a local region (as labeled by a circle in (a)). Scale bar: 100 nm. (c) Time-lapse TEM
images from a local area (as labeled in a) during lithiation (See Supporting Movie M1). Scale bar: 5 nm.
(
d, e) HRTEM images of pristine FeF
2
(t = 0 s) from the blue framed region in (a), and an inverse FFT
(at t =
image from the selected white framed area. Scale bar: 2 nm. (f) HRTEM image of lithiated FeF
2
30 s), from the same blue framed area in (a). Scale bar: 2 nm. (g) Inverse FFT image from the selected
white framed area showing dislocations (denoted as “T”).
8
ACS Paragon Plus Environment

Page 9 of 11
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Figure 2. Conversion-driven topotactic transformation in FeF tracked by in situ electron
2
diffraction (See also Supporting Movie M2). (a) Representative electron diffraction patterns recorded
from the white circled region in Figure 1(a), and (b) corresponding azimuthal projection images (See
Supporting Movie M2 for the time-resolved diffraction patterns). The blue semi-circle area highlighted
on the image in (a) shows the approximate radius and angles selected for the projection. (c) Comparison
̅
̅
of intensity profiles between 110Fe (Case I) and 101Fe (Case-II) with respect to parent 200FeF
2
. (d)
Radially integrated intensity profile as a function of lithiation time (between 0 to 37ˈ). (e) Zoom-in view
of the patterns from the red framed area in (d) showing the gradual growth of the (110)Fe peak during
the lithiation process.
9
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 10 of 11
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Figure 3. Structural correlation between converted Fe and the parent FeF . (a) Bright-field TEM
2
image taken from an electrochemically lithiated FeF particle. Scale bar: 100 nm. (b) Electron diffraction
2
̅
pattern taken from the white circle area in (a). (c) Intensity histograms comparing 110Fe (Case I) and
̅
(
101)Fe (Case II) spots. (d) HRTEM image and corresponding FFT (inset) from the yellow framed area
in (a). Scale bar: 5 nm.
1
0
ACS Paragon Plus Environment

Page 11 of 11
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Figure 4. Atomistic model of conversion-driven cationic displacement and topotactic
transformation in FeF
2
. (a) Schematic illustration of the structure of FeF with Fe (red) and F
2
(
green). In the 3D view (left), the unit cell is outlined by thick blue lines. The thin black lines
outline a 1×2×2 supercell. Alternative arrangement of Fe-F along [001] direction in the unit cell is
shown in the red and green planes. (b) Li (yellow) insertion along [001] direction. (c)
Expansion/contraction of Fe/LiF along different directions during conversion process. (d)
Perspective view of the checkboard arrangement of the converted Fe domains, illustrated by the
blue and red cubes at different height along [001] FeF direction.
2
1
1
ACS Paragon Plus Environment