Journal of the American Chemical Society
Article
4
7,62
+
one (increase in lithium deposition size), the C.E. approaches or
surpasses 99%, with the PIL 15 coating having one of the highest
size factors and C.E. This indicates that the incorporation of Py+
cation in the polymer is an effective interfacial strategy toward
large lithium deposition structure and improved Coulombic
efficiency.
stable SEI by Li−polymer reaction,
high Li transference
46,48
number interface,
or controlled release of SEI forming salt
We believe that the polymer design concept
reported in this work can be synergistically combined with
previous strategies to develop highly stable electrode−electro-
lyte interfaces to further boost the performance of lithium metal
batteries.
31,53
additives.
To evaluate the long-term cycle stability of the PIL 15 coating
on lithium surfaces, we performed galvanostatic charge and
discharge tests in symmetric lithium cells. We found that the PIL
EXPERIMENTAL SECTION
■
Materials Synthesis and Characterizations. Most chemicals
used for synthesizing ionic liquids and polymers, as well as all salts and
electrolytes used for battery cycling, were acquired from Sigma-Aldrich
except for 1H,1H-heptafluorobutyl acrylate (CAS: 424-64-60) which
was purchased from Oakwood Chemical and 2-isocyanatoethyl acrylate
CAS: 13641-96-8) which was obtained from TCI. NMC532
electrodes were provided by the CAMP facility at Argonne National
Laboratories.
Ionic liquids were synthesized by reacting uncharged molecules with
random copolymers were synthesized by radical polymerization of both
ionic and fluorinated monomers in dimethylformamide (DMF).
1
5 were capable of maintaining stable cycling of the batteries for
2
over 500 cycles both at 1 and 2 mA/cm current densities, while
cells with bare lithium failed in 200 cycles. As shown in Figure
6
a, the zoomed-in window of the voltage profile around the
(
cycles of cell failure exhibit a decrease in operation voltage,
indicative of cell failure by a short-circuit event. We also analyzed
the battery performance using the PIL 0 coating and observed
that the symmetric cells experience close to 0.2 V overpotentials
in the initial stages of cycling, followed by an abrupt drop in the
voltage profile associated with short-circuiting or interfacial
breakdown at 20 cycles. Due to the absence of any ionic groups
in the PIL 0, the lithium transport across the interface is
impeded, resulting in the observed high overpotentials in the
initial cycles and eventual failure.
The chemical structures of the polymers were characterized with
The PIL 15 coating is examined in the Li||NMC full cells,
proton nuclear magnetic resonance spectroscopy (500 MHz, D-
DMSO), fluorine nuclear magnetic resonance spectroscopy (400 MHz,
ethyl acetate), and Fourier-transform infrared spectroscopy (Thermo
Nicolet iS500). Glass transition temperatures (Tgs) of the polymers
were measured with differential scanning calorimetry (TA Q2000)
using heat−cool−heat profile between −50 and 200 °C (10 °C/min
heating, 5 °C/min cooling). Frequency sweeps (0.1−100 Hz) were
performed on each polymer sample at different temperatures (25−
2
where the polymer is applied to a 25 μm (5 mAh/cm ) thin
lithium anode foil and paired against a NMC532 anode with 2.7
2
mAh/cm capacity. The battery is cycled in the voltage range of
2
.7 and 4.3 V. In comparison to cells with bare lithium anodes,
the battery with PIL 15 coated anode shows higher retention of
battery capacity (Figure 6b). The lifetime of the battery can be
determined by the number of cycles taken for the cell to reach
9
5°C) with a TA ARES G2 rheometer to obtain frequency-dependent
8
0% of its original capacity. The cell with PIL 15 coated anode
storage and loss moduli and time−temperature superposition.
To characterize polymers’ electrochemical properties, we coated the
polymers on both Cu and Li surfaces. On Cu foils (1 cm diameter), a
polymer solution (0.1 g/mL ethyl acetate) was spin-coated at 2000 rad/
min speed and dried in a 70 °C vacuum oven for 3 h. Li metal chips or
has a lifetime close to 70 cycles, whereas the uncoated anode’s
lifetime is registered to be about 35 cycles. The voltage profiles
of the coated and uncoated Li||NMC cell at cycle 3 and 35 are
displayed in Figure 6c, showcasing that the cell with PIL 15
polymer coated Li anode has lower overpotential than without.
Similar observation of improved cycle life with coating was
2
20 μm Li foils (1 cm ) were dip-coated in an Ar-filled glovebox with 330
mg/mL ethyl acetate polymer solutions before being dried on a hot
plate at 70 °C for 3 h.
The ionic conductivity of the polymer was measured with biologic
VMP3 system by impedance spectroscopy measurements over a
frequency range from 100 mHz to 7 MHz. The polymers measured
were sandwiched by two steel electrodes. The geometry of the
polymeric samples was controlled as a cylinder with 1 cm cross-
sectional area and 0.36 mm thickness.
The swelling of the polymer was examined by soaking the polymer in
-dioxolane (DOL) 1,2- dimethoxyethane (DME) (volume ratio 1:1)
CONCLUSION
In conclusion, we have shown that organic salts based on
pyrrolidinium cations (Py TFSI ) have reduction potentials
−3.17 V vs SHE) lower than that of Li/Li . This provides a
■
2
+
−
+
(
3
design strategy for a polymer interface that incorporates such
organic salts to resist electrochemical breakdown, while
promoting lithium ion transport and modulating the charge
transfer process at the electrode surface. We simultaneously
incorporate low Tg fluorinated alkyl side chains in the polymer
architecture to provide flowability on the electrode surface and
at the same time to prevent interfacial side reactions with the
bulk electrolyte. The polymer dynamics and the ion transport in
these polymeric materials are shown to be dependent on the
supramolecular ionic interactions that can be tuned by the
tethered ion content or anion chemistry. We further show that
the cationic polymer responds to any external electric field
stimulus that enables a unique “shielding mechanism” of
suppressing morphological instabilities during Li deposition,
resulting in enhanced battery lifetime and Coulombic efficiency.
This mechanism differs from previously reported polymer
interfaces, which were based on strategies like formation of
electrolyte for 2 h. After 2 h of soaking, the polymer was extracted from
the electrolyte and dried for 10 s. The weight difference before and after
the soaking was recorded, and the swelling rate of the polymer is
calculated as the weight increase after soaking divided by the weight of
the polymer before soaking.
Seal-healing measurement of the polymer was conducted on a 1 mm
thick polymer layer formed by solution cast between two parallel
lithium titanate (LTO) electrodes lying on the substrate with a gap of 1
cm. A voltage of 30 V was applied across the electrodes. The cut was
made with a standard scalpel. Pictures/videos of the polymer self-
healing process were taken under a microscope at 2.5× magnification.
The XPS and SEM measurements were performed on the deposited
lithium at the Cu foil electrode. The deposition profile is revealed in the
electrochemical characterization section. After the Li was deposited on
the Cu electrode, the Cu electrode was extracted from the Li||Cu cell in
an Ar-filled glovebox. The foil was then rinsed in DME electrolyte for 10
min to remove excess salts and then dried for 2 min. The XPS profile
was collected with PHI VersaProbe 3 XPS probe with an Al K-alpha
H
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX