Journal of The Electrochemical
Society
OPEN ACCESS
Investigating Capacity Fade Mechanisms in Dual-
Ion Mg-MCl
x
Batteries
To cite this article: Steven H. Stradley
et al
2024
J. Electrochem. Soc.
171 060501
View the
article online
for updates and enhancements.
You may also like
Novel Ambient Temperature Ionic Liquids
Formulated from 1-Methyl-3-
ethylimidazolium Chloride and Anhydrous
Metal Chloride Salts other than Aluminum
Chloride
Towner B. Scheffler and M. Stacey
Thomson
-
Evaluation of the Electrochemical Behavior
of CeCl
3
in Molten LiCl-KCl Eutectic
Utilizing Metallic Ce as an Anode
K. C. Marsden and B. Pesic
-
High Specific Energy Density Aqueous
Lithium-Metal Chloride Rechargeable
Batteries
Yoshinori Morita, Shinya Watanabe, Peng
Zhang et al.
-
This content was downloaded from IP address 131.215.252.175 on 04/06/2024 at 17:54
Investigating Capacity Fade Mechanisms in Dual-Ion Mg-MCl
x
Batteries
Steven H. Stradley,
1
John-Paul Jones,
2
Ratnakumar V. Bugga,
3
and Kimberly A. See
1
,
z
1
Division of Chemistry and Chemical Engineering, California Institute of Technology, Pasadena, California 91125, United
States of America
2
Jet Propulsion Laboratory, California Institute of Technology, Pasadena, California 91109, United States of America
3
Lyten Inc., San Jose, California 95134, United States of America
Mg batteries are a promising alternative to Li-based chemistries due to the high abundance, low cost, and high volumetric capacity
of Mg relative to Li. Mg is also less prone to dendritic plating morphologies, promising safer operation. Mg plating and stripping is
highly ef
fi
cient in chloride-containing electrolytes; however, chloride is incompatible with many candidate cathode materials. In
this work, we capitalize on the positive effect of chloride by using transition metal chloride cathodes with a focus on low cost,
Earth-abundant metals. Both soluble and sparingly soluble chlorides show capacity fade upon cycling. Active material dissolution
and subsequent crossover to the Mg anode are the primary drivers of capacity fade in highly soluble metal chloride cathodes. We
hypothesize that incomplete conversion and chemical reduction by the Grignard-based electrolyte are major promoters of capacity
fade in sparingly soluble metal chlorides. Modi
fi
cations to the electrolyte can improve capacity retention, suggesting that future
work in this system may yield low cost, high retention Mg-MCl
x
batteries.
© 2024 The Author(s). Published on behalf of The Electrochemical Society by IOP Publishing Limited. This is an open access
article distributed under the terms of the Creative Commons Attribution 4.0 License (CC BY,
http://creativecommons.org/licenses/
by/4.0/
), which permits unrestricted reuse of the work in any medium, provided the original work is properly cited. [DOI:
10.1149/
1945-7111/ad4fe4
]
Manuscript submitted April 2, 2024; revised manuscript received April 27, 2024. Published June 4, 2024.
Supplementary material for this article is available
online
The transition to a clean energy economy is hindered by a lack of
safe, reliable, and ef
fi
cient energy storage technologies.
1
Lithium-
ion batteries (LIBs) dominate the modern battery market.
2
,
3
However, LIBs are based on scarce and expensive resources as
reserves of Li, Ni, and Co, key components of modern LIBs, are
concentrated in just a handful of regions.
4
–
6
LIBs are also
approaching the theoretical capacity limit imposed by intercalation
chemistries and are unlikely to undergo more than minor improve-
ments in coming decades.
3
Li-metal batteries are expected to have a
higher storage capacity than LIBs; however, safety concerns related
to dendritic Li plating limit utility and Li is still a major component
that suffers from resource issues.
7
Mg-metal batteries are an appealing alternative to Li-based
chemistries. Based on known reserves, Mg is orders of magnitude
more abundant than Li in the Earth
’
s crust.
8
Mg is also present in
seawater at a concentration of about 1300 ppm. After Na, it is the
most commonly found cation in the oceans.
9
The high relative
abundance makes Mg-metal batteries a more economical and
sustainable choice than batteries that rely on Li.
10
Additionally,
the volumetric capacity of a Mg metal anode is nearly twice that of
Li metal (3830 vs 2060 mAh ml
−
1
). Though some studies have
demonstrated the formation of Mg dendrites under select
conditions,
11
,
12
Mg has been found to plate more smoothly than Li
under comparable current densities due to its low self-diffusion
barrier.
13
,
14
Electrolytes for Mg plating and stripping are dif
fi
cult to develop
and the most successful electrolytes contain Cl.
15
,
16
The electro-
active cationic species identi
fi
ed in most Cl-containing electrolytes
with a Mg:Cl ratio less than unity is the binuclear complex
Mg
y
(
μ
-Cl)
x
+
, where y and x are integers whose values depend on
the solvent, counteranion, and Mg:Cl ratio.
16
–
21
Though the exact
role of Cl in promoting reversible Mg electrodeposition is hard to
determine, evidence suggests that Cl interacts with the Mg-electro-
lyte interface to facilitate Mg
2
+
reduction and prevent passivation.
The formation of a Cl-containing
”
enhancement layer
”
in the all-
phenyl complex (APC) and other Grignard-based electrolytes has
been shown to improve deposition and stripping kinetics and
overpotentials.
22
Computational studies have demonstrated that the
adsorption of cationic (MgCl)
+
monomers is thermodynamically
favorable in the magnesium-aluminum chloride complex electrolyte
and that such complexes have low desolvation energies, promoting
facile Mg
2
+
reduction.
23
Later studies have likewise proposed the
initial adsorption of (MgCl)
+
as a necessary step preceding Mg
electrodeposition in Cl-containing electrolytes.
24
,
25
Jankowski and
coworkers found that both desolvation energy and reductive over-
potential trend inversely with increasing Cl
−
coordination in Mg-Cl
cationic species, although increasing size and complexity can hinder
electron transfer.
26
In addition to the favorable effect of Cl-containing electrolytes on
Mg metal anode electrochemistry, the use of the highly mobile Cl
anion, Cl
−
, as a charge carrier may overcome the issues typically
associated with the sluggish transport of Mg
2
+
in solution. Examples
of Cl transport and conversion in batteries are not uncommon.
Reversible conversion of metal chlorides has been demonstrated in
Na
∣
MCl
x
cells (M
=
Fe
2
+
,Ni
2
+
,Cu
2
+
,Zn
2
+
) at elevated
temperatures.
27
,
28
An Al
∣
Cl
2
cell has been demonstrated with a
room-temperature molten salt electrolyte.
29
Metal chlorides (AgCl,
CuCl
2
, CoCl
2
, VCl
3
, and BiCl
3
), metal oxychlorides (BiOCl, FeOCl,
and VOCl), and layered materials (NiMn-Cl and Ni
2
V
0.9
Al
0.1
–
Cl)
have been paired with Li and room-temperature non-aqueous liquid
electrolytes.
30
–
40
A growing
fi
eld of aqueous chloride-ion batteries
typically combines a Ag cathode with an oxychloride (BiOCl or
Sb
4
O
5
Cl
2
) anode in an NaCl-based electrolyte.
41
–
44
Progress toward
all-solid-state chloride-ion batteries has been made by using
conductive polymer electrolytes in Li
∣
FeOCl and Zn
∣
CuCl
2
·4H
2
O
cells.
45
,
46
To capitalize on the bene
fi
cial effects of Cl on Mg electro-
chemistry, we aim to develop a system that pairs a Mg metal anode
with a Cl-containing electrolyte and a transition metal chloride
cathode of the form MCl
x
(M
=
Cu, Fe, etc.). Such a battery would
operate on a conversion-type mechanism at the cathode. Mg-ion
batteries have historically suffered from poor performance at
relevant rates due to the sluggish transport of Mg
2
+
.
47
This system
aims to alleviate this issue by leveraging the high mobility of Cl
−
to
yield the reduced metal and Cl
−
in the Cl-containing electrolyte
upon discharge:
z
E-mail:
ksee@caltech.edu
Journal of The Electrochemical Society
, 2024
171
060501
+→+
[
]
−−
MCl xe M xCl
1
x
The anode reaction is typical of a Mg metal battery architecture.
The all-phenyl complex (APC) electrolyte demonstrates highly
ef
fi
cient Mg dissolution and deposition with low overpotentials.
48
On discharge, the Mg metal is oxidized to Mg
2
+
in solution:
→+
[]
+−
Mg Mg
e
22
2
We hypothesize that the ionic species will interact in the
electrolyte to form a soluble [Mg
y
Cl
x
]
n
−
complex, the structure of
which depends on the properties of the electrolyte. The reverse
processes occur upon charge. A downside of this chemistry is the
formation of a soluble product during discharge whose solubility
will limit the capacity. However, other successful battery systems
have been developed that yield soluble products, notably the lead
acid battery.
The study of metal chloride cathodes in Mg systems has been
largely con
fi
ned to AgCl. Zhang et al. have investigated AgCl as a
conversion cathode to pair with a Mg anode and the APC
electrolyte.
49
Initial discharge yielded nearly 100% theoretical
conversion with impressive rate performance; however, the chem-
istry suffers from signi
fi
cant capacity fade with cycling, due to the
dissolution of the AgCl active material.
49
Li et al. observed similarly
impressive rate performance, albeit with a lower initial discharge
capacity, by pairing AgCl and Mg with an
Mg(HMDS)
2
–
AlCl
3
–
MgCl
2
electrolyte.
50
Inspired by these initial
studies, we aim to study other metal chlorides that might bypass the
high cost of Ag.
51
,
52
Other Mg
∣
MCl
x
(M
=
Cu
+
,Cu
2
+
,Ni
2
+
) cells
have been discharged, but reversible conversion has not been
demonstrated.
49
In this work, we expand the library of reversible Mg
∣
MCl
x
chemistries to include several metal chlorides based on Earth-
abundant, low cost metals such as Fe and Cu. We establish a cell
geometry that allows for the reversible conversion of these com-
pounds despite the highly corrosive electrolyte environment.
Investigation of several capacity fade mechanisms has led us to
suggest that the primary drivers of capacity fade in Mg
∣
MCl
x
systems are active material dissolution and reactivity with the
Grignard-based electrolyte. Attempts to moderate capacity fade
through modi
fi
cation of the electrolyte shows some improvement,
suggesting that further research in this area may yield low cost, low
fade Mg
∣
MCl
x
cells.
Experimental
Preparation of MCl
x
-C composites.
—
The AgCl-C composite
was prepared according to Zhang et al.
49
Silver nitrate (AgNO
3
,99
+
%, Sigma, 1.69 g) was dissolved in 500 ml of water. Super P carbon
(99
+
%, Fisher, 1.00 g) was then added while stirring. Finally, 20 ml
of concentrated HCl (36.5 to 38.0% w/w, Fisher) was added
dropwise. The precipitate was collected by vacuum
fi
ltration, dried
at 85 °C under vacuum overnight, and immediately brought into a
glove box. CuCl (99
+
%, anhydrous, Fisher), CuCl
2
(99.995%,
anhydrous, Sigma), FeCl
2
(99.99
+
%, anhydrous, Sigma), and
FeCl
3
(99.99
+
%, anhydrous, Sigma) were used as received. These
metal chlorides were combined with Super P carbon in a 1:1 ratio
(w:w) in an Ar-
fi
lled glove box and milled at 300 rpm for 5 h in a
planetary ball mill (MSE supplies) without exposure to air. This high
MCl
x
:C ratio was chosen to ensure suf
fi
cient electronic conductivity
in prototype cathodes to study the electrochemical properties of
these insulating MCl
x
s.
Cathode preparation.
—
The MCl
x
−
C composites were com-
bined with additional Super P carbon and polytetra
fl
uoroethylene
(PTFE, Sigma) to achieve a
fi
nal composition of AgCl:Super P:
PTFE 40:50:10 by mass or MCl
x
:Super P:PTFE (M
≠
Ag) 25:65:10
and ground by hand for 10 minutes. The composite material was
pressed into pellets using an arbor press and a 0.635 cm die and
dried overnight at 120 °C under vacuum.
Electrolyte preparation.
—
The APC electrolyte was prepared
according to Zhang et al.
49
A Peltier plate was used to cool 106.7 mg
AlCl
3
(99.999%, Sigma) to
∼
-20
◦
C. To the cool AlCl
3
, 3.2 ml of
cold tetrahydrofuran (THF, Fisher) was added dropwise to dissolve
the AlCl
3
while stirring. To this solution, 0.8 ml of cold 2.0 M
phenylmagnesium chloride (PhMgCl, Sigma) was added dropwise.
The electrolyte was stored in a foil-wrapped vial to prevent exposure
to light and stirred overnight before use.
The magnesium-aluminum chloride complex (MACC) electro-
lyte was prepared according to Barile et al.
53
THF (2.5 ml) cooled on
a Peltier plate was added dropwise to chilled AlCl
3
and stirred until
dissolved. Another 2.5 ml of THF was added to MgCl
2
(99.9%,
Fisher). The AlCl
3
solution was added to the MgCl
2
solution.
Magnesium hexamethyldisilizide (Mg(HMDS)
2
, 97%, recrystallized
before use, Sigma) was added as a chemical conditioning reagent.
54
The resulting solution, 30 mM AlCl
3
+
60 mM MgCl
2
+
10 mM
Mg(HMDS)
2
, in THF was stirred overnight before use.
The MgCl
2
-AlCl
3
-Mg(HMDS)
2
in tetraglyme electrolyte was
prepared according to Zhao-Karger et al.
55
A Peltier plate was used
to cool 2 ml of tetraglyme (G4,
⩾
99%, Sigma). To the cool G4, 480
mg of chilled AlCl
3
was added dropwise while stirring. Next, 620
mg of Mg(HMDS)
2
was added and stirred overnight. To this
solution, 172 mg of MgCl
2
was added and stirred for 40 hours
before use.
Electrochemical characterization.
—
Galvanostatic cycling ex-
periments were conducted in two electrode, Swagelok type cells.
The cell stack consisted of a Mg foil (99.9%, 0.5 in diameter
×
0.1
mm thick, mechanically cleaned with a razor blade, MTI) anode, a
glass
fi
ber (GFD, dried at 85 °C under vacuum overnight, VWR)
separator, 150
μ
L electrolyte, and the MCl
x
cathode. As constructed,
these Mg
—
MCl
x
cells were limited by the theoretical capacity of the
cathode. Theoretical capacities for each active material are listed in
Table
I
. Mo plungers (Midwest Tungsten Service) served as the
current collectors, as Mo has been shown to resist corrosion by
chloride.
56
All cell body components were made of polytetra
fl
uor-
oethylene (PTFE). Figure S1 illustrates the assembled electroche-
mical cell. Cells were assembled and sealed in a glove box.
Experiments were conducted on a Biologic BCS 805 battery cycler
or a VMP3 potentiostat.
Material characterization.
—
Powder X-ray diffraction (pXRD)
measurements were conducted on a Panalytical X
’
Pert Pro
Diffractometer using a Cu K
α
X-ray source. Rietveld re
fi
nements
were
fi
t using GSAS-II.
57
Scanning electron microscopy was
conducted using a ZEISS 1550VP
fi
eld emission SEM with a 15
kV acceleration voltage. SEM samples were prepared in an air-free
environment but were brie
fl
y exposed to air just before being loaded
into the instrument.
Table I. Initial discharge capacities of MCl
x
cathodes
MCl
x
Discharge
capacity
Theoretical
capacity
% Conversion
a
Solubility
in THF
49
(mAh g
−
1
)
(mAh g
−
1
)
(mg L
−
1
)
AgCl
184
186
99
0.053
CuCl
148
270
55
10.95
FeCl
2
231
327
71
59.7
CuCl
2
128
398
32
384.5
FeCl
3
168
496
34
(no data)
a assuming 100% Faradaic ef
fi
ciency.
Journal of The Electrochemical Society
, 2024
171
060501
Results and Discussion
To determine the effect of active material solubility on the
reversibility of metal chloride electrochemistry, we cycle several
metal chlorides against Mg anodes. Figure
1
shows the
fi
rst
galvanostatic charge/discharge curves for several MCl
x
cathodes,
arranged in order of increasing solubility in THF.
49
The AgCl
cathode achieves an initial discharge capacity of 184 mAh/g, nearly
99% of its theoretical capacity and on par with literature precedent.
49
The discharge pro
fi
le plateaus around 2 V while the charge pro
fi
le
remains similarly
fl
at around 2.2 V vs the Mg anode/reference,
which we take to be near Mg/Mg
2
+
, yielding a relatively low
hysteresis of 200 mV. All subsequent voltages will be given
referenced to the Mg anode/reference. By contrast, the initial
discharge capacities of the other MCl
x
cathodes are signi
fi
cantly
lower than theoretical and the voltage hystereses are larger. CuCl
shows an initial discharge capacity of 150 mAh/g (56% theoretical),
FeCl
2
shows 231 mAh/g (55% theoretical), CuCl
2
shows 128 mAh/g
(33% theoretical), and FeCl
3
shows 170 mAh/g (34% theoretical).
CuCl and CuCl
2
feature sloping discharge pro
fi
les between 1.6 and
0.7 V. In both systems, charge proceeds
fi
rst through a sloping
feature between 1.8 V and 2.2 V followed by a plateau at 2.3 V.
FeCl
2
and FeCl
3
have initial discharge plateaus at 1 V with
corresponding charge plateaus at 1.8 V. To our knowledge, this
represents the
fi
rst demonstrated reversible conversion of CuCl,
CuCl
2
, FeCl
2
, and FeCl
3
. Additionally, the initial discharge capa-
cities of CuCl and CuCl
2
are higher than in previous reports,
potentially due to the difference in cell geometries.
49
Previous work in Mg
∣
MCl
x
systems attributed capacity fade to
the dissolution of the active metal chloride by the electrolyte.
49
This
hypothesis is largely based on the observations that the initial
percent conversion trends inversely with the measured solubility of
the metal chloride and that capacity retention is improved at faster
cycling rates. Though this hypothesis is in line with the percent
conversion and the reported solubilities listed in Table
I
, it fails to
explain the low conversions of more soluble metal chlorides, as the
proposed amount dissolved would vastly supersaturate the small
volume of electrolyte, absent a shuttle effect to facilitate self
discharge. To evaluate this hypothesis in MCl
x
cathodes with a
wide range of measured solubilities in THF, we perform a modi
fi
ed
galvanostatic cycling experiment in which a 24 h open circuit
voltage (OCV) hold is imposed between cycle 1 charge and cycle 2
discharge. Figure
2
compares the discharge capacity of cycle 2 as a
percentage of cycle 1 for cells with and without the 24 h OCV hold.
The metal chlorides are arranged so the measured solubility in THF
increases from left to right.
49
Figure
2
shows the OCV rest causes
Figure 1.
First charge and discharge pro
fi
les of various MCl
x
cathodes at
0.12 C with the APC electrolyte: (a) AgCl, (b) CuCl, (c) FeCl
2
, (d) CuCl
2
,
and (e) FeCl
3
. AgCl has the highest discharge potential and lowest voltage
hysteresis. Charge and discharge proceed through voltage plateaus at
comparatively high voltages relative to the other metal chlorides in this
study. CuCl and CuCl
2
display sloping discharge pro
fi
les and charge pro
fi
les
with a lower potential sloping region and higher potential plateau. FeCl
2
and
FeCl
3
show low voltage discharge plateaus and moderate voltage charge
plateaus.
Figure 2.
(a) Potential vs time traces of an AgCl cell cycled normally and an
AgCl cell with a 24 hour rest after charge. (b) Second cycle capacity
retention of MCl
x
cathodes with and without a 24 hour rest after charge.
AgCl shows essentially no difference in capacity fade. More soluble
chlorides show more rapid fade.
Figure 3.
Initial discharge curve at 0.12C with the APC electrolyte of (a)
AgCl and (b) CuCl
2
. Ex situ pXRD and re
fi
nement for (c) the AgCl cell and
(d) the CuCl
2
cell, with re
fl
ections indicated for Ag and Cu, respectively. Ag
is recovered in the AgCl cell, and Cu is recovered in the CuCl
2
cell.
Journal of The Electrochemical Society
, 2024
171
060501
essentially no difference in capacity fade for AgCl. By contrast, the
more soluble MCl
x
cathodes show a large increase in capacity fade
following a rest after charge. This difference can be explained by the
relatively low solubility of AgCl, which is several orders of
magnitude less soluble than any other MCl
x
, and supports the
hypothesis that active material dissolution is a major driver in
capacity fade.
Though metal chloride dissolution can cause capacity fade, metal
chlorides could also cross over and affect the anode chemistry which
could negatively impact cycling behavior. To explore any effects
arising from crossover, we cycle Mg
∣
APC
∣
AgCl and Mg
∣
APC
∣
CuCl
2
cells
fi
ve times and then replace the cathode with a fresh electrode
and add an additional 75
μ
L of fresh electrolyte. The cells are then
discharged. The initial cycling data along with the post-replacement
discharge curves are shown in Figure S3. Both AgCl and CuCl
2
cells
show essentially full capacity recovery when the cathode is replaced.
The discharge curve in the CuCl
2
cell with the fresh cathode shows a
lower nucleation barrier relative to the freshly assembled cell, likely
due to some conditioning process priming the already cycled anode.
As a control experiment, we perform the same limited cycling
protocol and replace the anode after
fi
ve cycles. Under this
condition, the AgCl cell shows continued capacity fade, and the
CuCl
2
cell shows accelerated capacity fade. These observations
further support our hypothesis that the capacity fade stems from
dissolution of the active material.
To further investigate the primary drivers of capacity fade in
Mg
∣
MCl
x
cells, we choose to focus on AgCl and CuCl
2
cathodes, as
these represent the extremes of measured solubilities in THF, with
CuCl
2
being over 7,000x more soluble in THF than AgCl.
49
Figures
3
a and
3
b show the
fi
rst discharge curves of AgCl and
CuCl
2
cathodes. The AgCl cell discharges to essentially full
theoretical capacity. To probe the reduction product, the cell is
disassembled and pXRD is measured on the cathode. The pXRD and
Rietveld re
fi
nement of the discharged AgCl cathode is shown in
Figure
3
c. The peaks correspond to Ag metal after discharge, which
is the expected discharge product, and no re
fl
ections of the AgCl
starting material are observed. In contrast, the CuCl
2
cell discharges
only to 35% of its theoretical capacity. The pXRD pattern and
Rietveld re
fi
nement of the discharged cathode is shown in Figure
3
d.
The pattern shows only two re
fl
ections, which are re
fi
ned to the
expected discharge product: Cu metal. Despite the low conversion,
no peaks associated with the CuCl
2
starting material are observed
suggesting that the active material is no longer a crystalline solid in
the cathode after discharge.
Upon charging the AgCl cell to 2.75 V, pXRD patterns of the
cathode show re
fl
ections associated with AgCl; however, some
unconverted Ag remains (Fig. S4a). The pattern of the recharged
CuCl
2
cathode, by contrast, does not show any re
fl
ections (Fig. S3b).
We hypothesize that the lack of re
fl
ections for CuCl
2
is due to
dissolution of the compound.
To verify that the discrepancy in initial conversion between AgCl
and the more soluble MCl
x
cathodes is not due to differences in
mixing techniques or kinetic barriers to conversion, galvanostatic
intermittent titration is performed on AgCl and CuCl
2
cells whose
cathodes are prepared with several material processing techniques.
Figure S5 shows the initial discharge and charge GITT curves for
AgCl and CuCl
2
cathodes. The AgCl cathode synthesized in the
presence of C initially shows overpotentials on the order of 100 mV.
These overpotentials grow to
∼
300 mV as the depth of discharge
increases. The corresponding charge curve shows similar over-
potentials. The AgCl cathodes prepared by mechanical milling
show larger overpotentials for both discharge and charge, indicating
that the low overpotentials measured in the cathode synthesized in
the presence of C is at least partially due to the intimate mixing of
the active material with the conductive C matrix. The CuCl
2
cathodes, by contrast, show relatively large conversion overpoten-
tials (
∼
500 mV) for all attempted mixing techniques. Direct
comparison of the high power ball milled and hand ground cathodes
shows that a substantial portion of the overpotential may be due to
intrinsic dif
fi
culties associated with the conversion of CuCl
2
as we
observe little difference in overpotential between the two mixing
conditions.
Though evidence suggests that most of the capacity fade in
highly soluble MCl
x
cathodes is due to dissolution of the active
material by the electrolyte, the volume of electrolyte in our
electrochemical cells should not be able to dissolve a substantial
amount of the active material in the cathode. As assembled, our cells
contain 0.25 ml of electrolyte with a typical active loading of 3.5 mg
Figure 4.
(a) SEM image of a Mg anode from a Mg
∣
APC
∣
CuCl
2
cell after discharge, the corresponding elemental maps for (b) Mg and (c) Cu, and (d) the sum
EDS spectrum for the region. A layer dominated by Mg, O, Cu, and Al is present on the electrode.(e) SEM image of a Mg anode from a Mg
∣
APC
∣
AgCl cell after
discharge, the corresponding elemental maps for (f) Mg and (g) the sum EDS spectrum for the region. A layer dominated by Mg and O is present on the
electrode.
Journal of The Electrochemical Society
, 2024
171
060501
CuCl
2
. A mere 0.096 mg CuCl
2
(approx. 2.75% of the total) should
saturate the electrolyte based on the measured solubility of CuCl
2
in
THF. We hypothesize that a shuttle effect is taking place in which
CuCl
2
is continuously dissolved from the cathode, migrates across
the cell in the electrolyte, and is reduced by the anode. To probe this
hypothesis, we discharged a Mg
∣
CuCl
2
cell to 0.8 V, extracted the
anode, and imaged the surface with a scanning electron microscope.
Figure
4
a shows the surface morphology the Mg anode following
discharge and the corresponding EDS maps for Mg and Cu are
shown in Figures
4
b and
4
c, respectively. Figure
4
a reveals an
uneven, cracked layer coating the Mg anode surface. Elemental
analysis of the imaged area, displayed in Figure
4
d, shows that this
layer is primarily composed of Mg, C, O, and Cu. C likely comes
from reductive decomposition of the THF solvent against the Mg
surface. Decomposition of the ethereal solvent and brief air exposure
prior to the measurement explains the presence of O on the anode
surface. Elemental mapping in Figs.
4
b and
4
c shows that Mg and
Cu are distributed in this layer. The presence of Cu on the anode
after discharge supports the hypothesis that total capacity loss in the
CuCl
2
system is promoted by a continuous shuttle of Cu from the
cathode to the anode, allowing the electrolyte to leach much more
CuCl
2
from the cathode than can be explained by the solubility and
the volume of electrolyte.
An analogous experiment was performed with an AgCl cathode.
Figure
4
e shows the surface morphology of the anode after
discharge. Elemental mapping in Figure
4
f shows Mg distributed
throughout this layer. Analysis of the imaged area in Figure
4
g
shows that surface layer on the anode composed mostly of Mg with a
small amount of O. The layer formed in the AgCl system has a
markedly different surface morphology, dominated by pits rather
than plates. In contrast to the CuCl
2
cell, the layer formed in the
Mg
∣
AgCl cell shows no measurable Ag on the anode. This disparity
may be explained by the stark difference in solubilities between the
metal chlorides. We hypothesize that the electrolyte is saturated with
dissolved AgCl and that a similar shuttle is taking place, but the
concentration of AgCl in the electrolyte and the resultant rate of
leaching is so low that no signi
fi
cant amount of active material is
dissolved before the initial cycling of the cell. The AgCl cell thereby
achieves essentially total conversion in its
fi
rst cycle while the CuCl
2
cell features only 35% conversion. We posit that the more gradual
capacity fade characteristic of AgCl may be driven by incomplete
conversion of the metal Ag to AgCl during charge. This is evident in
the re
fl
ections in the pXRD pattern for a recharged cathode which
can be attributed to Ag (Fig. S4).
To probe the effect of the layer formed on the Mg anode in both
Mg
∣
CuCl
2
and Mg
∣
AgCl cells on Mg plating and stripping at the
anode, we soak Mg foils in saturated solutions of APC
+
AgCl and
APC
+
CuCl
2
for 7 days. The metal foils are collected, rinsed
thoroughly with THF, and assembled into symmetric cells with fresh
APC electrolyte. Chronopotentiometry traces of these cells are
illustrated in Fig. S6. The polarization behavior of the cells soaked
in the two saturated solutions are identical to that of a control soaked
in neat APC for 7 days. This result indicates that any layer formed
on the Mg anode by the reduction of species in the saturated MCl
x
+
APC electrolyte does not impede Mg transport or affect deposition
and stripping processes. We hypothesize that this is due to Mg
2
+
transport through grain and phase boundaries in the heterogeneous
surface layer as has been recently proposed through the SEI formed
on Ca largely composed of the insulating CaO phase.
58
To investigate the chemical reactivity of the active material with
the electrolyte in a Mg
∣
MCl
x
cell, we assemble Mg
∣
APC
∣
AgCl and
Mg
∣
APC
∣
CuCl
2
cells and let them rest in the glove box for 7 days.
The cathodes are then extracted, and their pXRD patterns are
measured (Figure
5
). The patterns associated with the pristine
electrodes in Figs.
5
a and
5
b show re
fl
ections that can be attributed
to the unreacted MCl
x
active material. After the cells have rested for
7 d at OCV, the re
fl
ections attributed to the chlorides are no longer
present even in the case of sparingly soluble AgCl. Instead, the XRD
patterns show new re
fl
ections that can be attributed to the corre-
sponding metals, indicating chemical reduction of the active material
by the electrolyte. Though not quantitative, the low intensity of the
Cu metal re
fl
ections in Figure
5
b suggests that a substantial portion
of the active material is lost, likely deposited on the anode as in
Figure
4
. The APC electrolyte is synthesized from the Grignard
reagent PhMgCl. Hypothesizing that unreacted PhMgCl may che-
mically reduce the active material, we prepared an electrolyte with
20% excess of the Lewis acid AlCl
3
to react with any unreacted
PhMgCl. We perform an analogous soaking experiment with this
modi
fi
ed electrolyte and measure the pXRD pattern of the cathode.
The resulting XRD pattern of the cathode from the Mg
∣
AgCl cell
shown in Figure
5
a shows re
fl
ections for both AgCl and Ag,
suggesting that the excess AlCl
3
has served to slow but not prevent
the chemical reduction of AgCl by the Grignard-containing electro-
lyte. The chemical reduction of AgCl by the APC electrolyte is
Figure 5.
Ex situ pXRD patterns for (a) AgCl and (b) CuCl
2
cathodes,
showing traces for a pristine cathode and cathodes assembled into cells and
rested for 7 days with the APC or APC
+
20 % AlCl
3
electrolyte. APC
chemically reduces the metal chloride to the metal. APC
+
20% AlCl
3
slows
the chemical reduction.
Figure 6.
(a) First discharge of CuCl
2
cells and (b) discharge capacity as
function of cycle number for a CuCl
2
cathode with a variety of electrolytes.
Quenching the Grignard or removing it from the electrolyte while increasing
salt concentration moderate capacity fade but do not prevent it in long-term
cycling.
Journal of The Electrochemical Society
, 2024
171
060501
likely also a major contributor to capacity fade. Results are similar
for CuCl
2
,asre
fl
ections for CuCl are present after soaking in the
APC electrolyte with excess AlCl
3
. This result also suggests that the
chemical reduction of CuCl
2
proceeds through a CuCl intermediate.
We identify three primary drivers of capacity fade in Mg
∣
MCl
x
cells: (1) dissolution of the active material by the electrolyte, (2)
reduction of dissolved MCl
x
by the Mg anode, and (3) chemical
reactivity of MCl
x
with the Grignard electrolyte. We now attempt to
moderate capacity fade in the highly soluble CuCl
2
cathode by
modifying properties of the electrolyte. Figure
6
compares the initial
discharge pro
fi
le and discharge capacities as function of cycle
number of a Mg
∣
CuCl
2
cell with a conventional APC electrolyte
and with modi
fi
ed electrolyte conditions. The reactivity of the APC
electrolyte is reduced by adding additional AlCl
3
to the electrolyte.
The resulting Mg
∣
CuCl
2
cycling behavior, however, is minimally
changed. Figure
6
b illustrates discharge capacities for this electro-
lyte as a function of cycle number, displaying a moderate initial
conversion followed by fade at a rate similar to the APC electrolyte,
which may be attributed to dissolution of the active material by the
electrolyte. To eliminate the Grignard altogether, we use the
magnesium-aluminum chloride complex (MACC) electrolyte, which
is composed entirely of inorganic salts. The cell barely functions,
however, showing minimal redox activity on cycle 1 (Q
=
32 mAh/
g) followed by polarization. We hypothesize that this low conversion
is due to the low conductivity of the relatively low concentration
MACC electrolyte. In an effort to address (1) by reducing the
solubility of CuCl
2
in the electrolyte, we use an electrolyte with
much higher salt concentrations in a more viscous G4 solvent. The
highly concentrated electrolyte composed of 1.8 M AlCl
3
+
0.9 M
MgCl
2
+
0.9 M Mg(HMDS)
2
in G4 shows the highest discharge
capacity and and the highest capacity at cycle 30 of any electrolyte
we investigated. Figure
6
a shows that the discharge also occurs at a
signi
fi
cantly higher voltage than in the cells with Grignard-based
electrolytes, potentially due to decreased chemical reactivity with
the active material. However, severe capacity fade is still observed.
This result suggests that limiting active material dissolution with a
highly concentrated, Grignard-free electrolyte may be a suitable
avenue for moderating both self discharge and long term capacity
fade in Mg
—
MCl
x
cells.
Conclusions
In this work, we have demonstrated the conversion of a variety of
transition metal chlorides based on low-cost, Earth-abundant metals
in chloride-based electrolytes with a Mg metal anode. We enabled
stable electrochemistry by establishing a cell geometry based on
corrosion-resistant Mo current collectors and PTFE body compo-
nents. Initial percent conversion and capacity retention in these cells
are inversely correlated with the measured solubility in THF of the
active materials, suggesting that active material dissolution is a
major contributor to capacity fade in highly soluble metal chlorides.
However, due to the lean electrolyte loading and typical active mass
in prepared CuCl
2
cathodes, the high solubility of certain metal
chlorides is insuf
fi
cient to explain the degree of capacity fade in
highly soluble metal chlorides. We hypothesized that a shuttle effect
is taking place between the MCl
x
cathode and the Mg anode,
enabling the continuous dissolution of the active material far beyond
the quantity necessary to saturate the electrolyte. Resting experi-
ments and subsequent ex situ pXRD characterization revealed that
the active material can also be chemically reduced by the electrolyte.
Reactivity with the electrolyte was moderated by the addition of
20% AlCl
3
to quench unreacted PhMgCl in solution. We hypothe-
size that chemical reduction and incomplete conversion are a major
cause of capacity fade in AgCl cells. Attempts to moderate capacity
fade in CuCl
2
cells through electrolyte modi
fi
cation showed some
success. Increasing salt concentration and the addition of 20% AlCl
3
to the APC electrolyte resulted in modest improvements to capacity
retention but failed to prevent total capacity fade. This effort
demonstrates that future work on this system with a focus on
electrolyte design to prevent active material dissolution may yield
low fade Mg
∣
MCl
x
based on low cost, Earth abundant metals.
Acknowledgments
This research was funded by the President and Director
’
s
Research and Development Fund. Part of the work was carried out
at the Jet Propulsion Laboratory, California Institute of Technology,
under a contract with the National Aeronautics and Space
Administration. S.H.S. acknowledges support from the National
Science Foundation Graduate Research Fellowship under Grant No.
DGE-1745301. K.A.S. acknowledges support from the David and
Lucile Packard Foundation, Alfred P. Sloan Foundation, and
Camille and Henry Dreyfus Foundation. SEM and EDS analyses
were carried out at the Caltech GPS Division Analytical Facility,
which is supported, in part, by NSF Grants EAR-0318518 and
DMR-0080065.
ORCID
Kimberly A. See
https://orcid.org/0000-0002-0133-9693
References
1. S. Chu, Y. Cui, and N. Liu,
“
The path toward sustainable energy.
”
Nat. Mater.
,
16
,
16 (2017).
2. A. Manthiram,
“
Are
fl
ection on lithium-ion battery cathode chemistry.
”
Nat.
Commun.
,
11
, 1550 (2020).
3. J. Cabana, L. Monconduit, D. Larcher, and M. R. Palacín,
“
Beyond intercalation-
based li-ion batteries: the state of the art and challenges of electrode materials
reacting through conversion reactions.
”
Adv Mater
,
22
, E170 (2010).
4. J. A. Ober,
Mineral commodity summaries
,
1
,52
–
53, 100
–
105, 116
–
117 (2016),
USGS IP-072363.
5. P. Greim, A. A. Solomon, and C. Breyer,
“
Assessment of lithium criticality in
theglobal energy transition and addressing policy gaps in transportation.
”
Nat.
Commun.
,
11
, 4570 (2020).
6. A. Elshkaki, B. K. Reck, and T. E. Graedel,
“
Anthropogenic nickel supply, demand,
and associated energy and water use.
”
Resour Conserv Recycl
,
125
, 300 (2017).
7. X.-B. Cheng, R. Zhang, C.-Z. Zhao, and Q. Zhang,
“
Toward safe lithium metal
anode in rechargeable batteries: a review.
”
Chem. Rev.
,
117
, 10403 (2017).
8. K. Hans Wedepohl,
“
The composition of the continental crust.
”
Geochim
Cosmochim Acta
,
59
, 1217 (1995).
9. J. Lyman and R. Fleming,
“
Composition of sea water.
”
J. Mar. Res.
,
3
, 134
–
146
(1940).
10. J. W. Choi and D. Aurbach,
“
Promise and reality of post-lithium-ion batteries with
high energy densities.
”
Nat. Rev. Mater.
,
1
, 1 (2016).
11. R. Davidson et al.,
“
Formation of magnesium dendrites during electrodeposition.
”
ACS Energy Lett.
,
4
, 375 (2019).
12. R. Davidson et al.,
“
Mapping mechanisms and growth regimes of magnesium
electrodeposition at high current densities.
”
Mater. Horiz.
,
7
, 843 (2020).
13. M. Matsui,
“
Study on electrochemically deposited Mg metal.
”
J. Power Sources
,
196
, 7048 (2011).
14. M. Jäckle, K. Helmbrecht, M. Smits, D. Stottmeister, and A. Groß,
“
Self-diffusion
barriers: possible descriptors for dendrite growth in batteries?
”
Energy Environ.
Sci.
,
11
, 3400 (2018).
15. P. Bonnick and J. Muldoon,
“
A trip to Oz and a peak behind the curtain of
magnesium batteries.
”
Adv. Funct. Mater.
,
30
, 1910510 (2020).
16. Q. Wu, K. Shu, L. Sun, and H. Wang,
“
Recent advances in non-nucleophilic Mg
electrolytes.
”
Front. Mater.
,
7
, 612134 (2021).
17. G. Bieker, M. Salama, M. Kolek, Y. Gofer, P. Bieker, D. Aurbach, and M. Winter,
“
The power of stoichiometry: conditioning and speciation of MgCl2/AlCl3 in
tetraethylene glycol dimethyl ether-based electrolytes.
”
ACS Appl. Mater.
Interfaces
,
11
, 24057 (2019).
18. K. A. See, K. W. Chapman, L. Zhu, K. M. Wiaderek, O. J. Borkiewicz, C. J. Barile,
P. J. Chupas, and A. A. Gewirth,
“
The interplay of Al and Mg speciation in
advanced Mg battery electrolyte solutions.
”
J. Am. Chem. Soc.
,
138
, 328 (2016).
19. C. Liao et al.,
“
The unexpected discovery of the Mg (HMDS) 2 /MgCl 2 complex as
a magnesium electrolyte for rechargeable magnesium batteries.
”
J. Mater. Chem. A
,
3
, 6082 (2015).
20. M. Salama, I. Shterenberg, L. J. W. Shimon, K. Keinan-Adamsky, M. Afri,
Y. Gofer, and D. Aurbach,
“
Structural analysis of magnesium chloride complexes
in dimethoxyethane solutions in the context of Mg batteries research.
”
J. Phys.
Chem. C
,
121
, 24909 (2017).
21. N. Sa, B. Pan, A. Saha-Shah, A. A. Hubaud, J. T. Vaughey, L. A. Baker, C. Liao,
and A. K. Burrell,
“
Role of Chloride for a Simple, non-grignard Mg electrolyte in
ether-based solvents.
”
ACS Appl. Mater. Interfaces.
,
8
, 16002 (2016).
22. J. L. Esbenshade, C. J. Barile, T. T. Fister, K. L. Bassett, P. Fenter, R. G. Nuzzo,
and A. A. Gewirth,
“
Improving Electrodeposition of Mg through an Open Circuit
Potential Hold.
”
J. Phys. Chem. C
,
119
, 23366 (2015).
23. P. Canepa, G. S. Gautam, R. Malik, S. Jayaraman, Z. Rong, K. R. Zavadil,
K. Persson, and G. Ceder,
“
Understanding the initial stages of reversible Mg
Journal of The Electrochemical Society
, 2024
171
060501
deposition and stripping in inorganic nonaqueous electrolytes.
”
Chem. Mater.
,
27
,
3317 (2015).
24. A. Benmayza, M. Ramanathan, T. S. Arthur, M. Matsui, F. Mizuno, J. Guo, P.-
A. Glans, and J. Prakash,
“
Effect of electrolytic properties of a magnesium
organohaloaluminate electrolyte on magnesium deposition.
”
J. Phys. Chem. C
,
117
, 26881 (2013).
25. K. Ta, K. A. See, and A. A. Gewirth,
“
Elucidating Zn and Mg electrodeposition
mechanisms in nonaqueous electrolytes for next-generation metal batteries.
”
J. Phys. Chem. C
,
122
, 13790 (2018).
26. P. Jankowski, J. M. G. Lastra, and T. Vegge,
“
Structure of magnesium chloride
complexes in ethereal systems: computational comparison of THF and glymes as
solvents for magnesium battery electrolytes.
”
Batter. Supercaps
,
3
, 1350 (2020).
27. B. V. Ratnakumar, S. D. Stefano, and G. Halpert,
“
Electrochemistry of metal
chloride cathodes in sodium batteries.
”
J. Electrochem. Soc.
,
137
, 2991 (1990).
28. X. Lu, G. Li, J. Y. Kim, J. P. Lemmon, V. L. Sprenkle, and Z. Yang,
“
A novellow-
cost sodium- zinc chloride battery.
”
Energy Environ. Sci.
,
6
, 1837 (2013).
29. P. R. Gifford and J. B. Palmisano,
“
An aluminum/chlorine rechargeable cell
employing a room temperature molten salt electrolyte.
”
J. Electrochem. Soc.
,
135
,
650 (1988).
30. T. Li, Z. X. Chen, Y. L. Cao, X. P. Ai, and H. X. Yang,
“
Transition-metalchlorides
as conversion cathode materials for Li-ion batteries.
”
Electrochim. Acta
,
68
, 202
(2012).
31. J.-l. Liu, W.-j. Cui, C.-x. Wang, and Y.-y. Xia,
“
Electrochemical reaction of lithium
with CoCl2 in nonaqueous electrolyte.
”
Electrochem. Commun.
,
13
, 269 (2011).
32. X. Zhao, S. Ren, M. Bruns, and M. Fichtner,
“
Chloride ion battery: a new member
in the rechargeable battery family.
”
J. Power Sources
,
245
, 706 (2014).
33. X. Zhao, Z. Zhao-Karger, D. Wang, and M. Fichtner,
“
Metal oxychlorides as
cathode materials for chloride ion batteries.
”
Angew.Chem. Int. Ed.
,
52
, 13621
(2013).
34. P. Gao, M. A. Reddy, X. Mu, T. Diemant, L. Zhang, Z. Zhao-Karger, V. S.
K. Chakravadhanula, O. Clemens, R. J. Behm, and M. Fichtner,
“
VOCl as a cathode
for rechargeable chloride ion batteries.
”
Angew. Chem. Int. Ed.
,
55
, 4285 (2016).
35. T. Yu, Q. Li, X. Zhao, H. Xia, L. Ma, J. Wang, Y. S. Meng, and X. Shen,
“
Nanocon
fi
ned iron oxychloride material as a high-performance cathode for
rechargeable chloride ion batteries.
”
ACS Energy Lett.
,
2
, 2341 (2017).
36. R. Yang, T. Yu, and X. Zhao,
“
Polypyrrole-coated iron oxychloride cathode
material with improved cycling stability for chloride ion batteries.
”
J. Alloys and
Compd
,
788
, 407 (2019).
37. Z. Zhao, T. Yu, Y. Miao, and X. Zhao,
“
Chloride ion-doped polyaniline/
carbonnanotube nanocomposite materials as new cathodes for chloride ion battery.
”
Electrochim. Acta
,
270
, 30 (2018).
38. T. Yu, X. Zhao, L. Ma, and X. Shen,
“
Intercalation and electrochemical behaviors
of layered FeOCl cathode material in chloride ion battery.
”
Mater. Res. Bull.
,
96
,
485 (2017).
39. J. Luo, Q. Yin, J. Zhang, S. Zhang, L. Zheng, and J. Han,
“
NiMn-Cl layered double
hydroxide/carbon nanotube networks for high-performance chloride ion batteries.
”
ACS Appl. Energy Mater.
,
3
, 4559 (2020).
40. Q. Yin, J. Luo, J. Zhang, S. Zhang, J. Han, Y. Lin, J. Zhou, L. Zheng, and M. Wei,
“
Ultralong-life chloride ion batteries achieved by the synergistic contribution of
intralayer metals in layered double hydroxides.
”
Adv. Funct. Mater.
,
30
, 1907448
(2020).
41. P. Srimuk, S. Husmann, and V. Presser,
“
Low voltage operation of a silver/
silverchloride battery with high desalination capacity in seawater.
”
RSC Adv.
,
9
,
14849 (2019).
42. X. Hu et al.,
“
Electrochemical performance of Sb4O5Cl2 as a new anode material
in aqueous chloride-ion battery.
”
ACS Appl. Mater. Interfaces
,
11
, 9144 (2019).
43. F. Chen, Z. Y. Leong, and H. Y. Yang,
“
An aqueous rechargeable chloride ion
battery.
”
Energy Storage Mater.
,
7
, 189 (2017).
44. J. Sun, Q. Sun, H. Ye, Y. Wang, J. A. S. Oh, D. Ji, Q. Zeng, K. Zeng, and L. Lu,
“
Ammonium escorted chloride chemistry in stabilizing aqueous chloride ion
battery.
”
Mater. Today Energy
,
26
, 101020 (2022).
45. C. Chen, T. Yu, M. Yang, X. Zhao, and X. Shen,
“
An all-solid-state rechargeable
chloride ion battery.
”
Adv. Sci
,
6
, 1802130 (2019).
46. F. Gschwind, D. Steinle, D. Sandbeck, C. Schmidt, and E. von Hauff,
“
Facile
preparation of chloride-conducting membranes:
fi
rst step towarda room-temperature
solid-state chloride-ion battery.
”
Chemistry Open
,
5
, 525 (2016).
47. E. Levi, M. D. Levi, O. Chasid, and D. Aurbach,
“
A review on the problems of the
solid state ions diffusion in cathodes for rechargeable Mg batteries.
”
J. Electroceram.
,
22
, 13 (2009).
48. D. Aurbach, Z. Lu, A. Schechter, Y. Gofer, H. Gizbar, R. Turgeman, Y. Cohen,
M. Moshkovich, and E. Levi,
“
Prototype systems for rechargeable magnesium
batteries.
”
Nature
,
407
, 724 (2000).
49. R. Zhang, C. Ling, and F. Mizuno,
“
A conceptual magnesium battery with ultrahigh
rate capability.
”
Chem Comm
,
51
, 1487 (2015).
50. X. Li, Y. Zhang, J. Shen, S.-a. Cao, T. Li, and F. Xu,
“
A high-rate rechargeable Mg
battery based on AgCl conversion cathode with fast solid-state Mg2
+
diffusion
kinetics.
”
Energy Technol.
,
7
, 1900454 (2019).
51. Macrotrends. Silver Prices
—
100 Year Historical Chart.
https://www.macrotrends.
net/1470/historical-silver-prices-100-year-chart
(accessed 2024-05-29).
52. Macrotrends. Copper Prices
—
45 Year Historical Chart.
https://www.macrotrends.
net/1476/copper-prices-historical-chart-data
(accessed 2024-05-29).
53. C. J. Barile, E. C. Barile, K. R. Zavadil, R. G. Nuzzo, and A. A. Gewirth,
“
Electrolytic conditioning of a magnesium aluminum chloride complex for
reversible magnesium deposition.
”
J. Phys. Chem. C
,
118
, 27623 (2014).
54. S. S. Kim, S. C. Bevilacqua, and K. A. See,
“
Conditioning-free Mg electrolyte by
the minor addition of Mg(HMDS)2.
”
ACS Appl. Mater. Interfaces
,
12
, 5226
–
5233
(2019).
55. Z. Zhao-Karger, X. Zhao, D. Wang, T. Diemant, R. J. Behm, and M. Fichtner,
“
Performance improvement of magnesium sulfur batteries with modi
fi
ed non-
nucleophilic electrolytes.
”
Adv. Energy Mater.
,
5
, 1401155 (2015).
56. Y. Cheng, T. Liu, Y. Shao, M. H. Engelhard, J. Liu, and G. Li,
“
Electrochemically
stable cathode current collectors for rechargeable magnesium batteries.
”
J. Mater.
Chem. A
,
2
, 2473 (2014).
57. B. H. Toby and R. B. Von Dreele,
“
GSAS-II: the genesis of a modern open-
sourceall purpose crystallography software package.
”
J. Appl. Crystallogr.
,
46
, 544
(2013).
58. S. A. McClary, D. M. Long, A. Sanz-Matias, P. G. Kotula, D. Prendergast, K.
L. Jungjohann, and K. R. Zavadil,
“
A heterogeneous oxide enables reversible
calcium electrodeposition for a calcium battery.
”
ACS Energy Lett.
,
7
, 2792
–
2800
(2022).
Journal of The Electrochemical Society
, 2024
171
060501