Yanbin Li1, Yongming Sun1, Allen Pei1, Kaifeng Chen1, Arturas Vailionis1, Yuzhang Li1, Guangyuan Zheng1, Jie Sun1, Yi Cui1,2. 1. Department of Materials Science and Engineering, Department of Applied Physics, and Stanford Nano Shared Facilities, Stanford University, Stanford, California 94305, United States. 2. Stanford Institute for Materials and Energy Sciences, SLAC National Accelerator Laboratory, 2575 Sand Hill Road, Menlo Park, California 94025, United States.
Abstract
Lithium metal is the ultimate anode choice for high energy density rechargeable lithium batteries. However, it suffers from inferior electrochemical performance and safety issues due to its high reactivity and the growth of lithium dendrites. It has long been desired to develop a materials coating on Li metal, which is pinhole-free, mechanically robust without fracture during Li metal deposition and stripping, and chemically stable against Li metal and liquid electrolytes, all while maintaining adequate ionic conductivity. However, such an ideal material coating has yet to be found. Here we report a novel synthesis method by reacting clean molten lithium foil directly with pure nitrogen gas to generate instantaneously a pinhole-free and ionically conductive α-Li3N film directly bonded onto Li metal foil. The film consists of highly textured large Li3N grains (tens of μm) with (001) crystalline planes parallel to the Li metal surface. The bonding between textured grains is strong, resulting in a mechanically robust film which does not crack even when bent to a 0.8 cm curvature radius and is found to maintain pinhole-free coverage during Li metal deposition and stripping. The measured ionic conductivity is up to 5.2 × 10-4 S cm-1, sufficient for maintaining regular current densities for controllable film thicknesses ranging from 2 to 30 μm. This Li3N coating is chemically stable, isolating the reactive metallic lithium from liquid electrolyte, prevents continuous electrolyte consumption during battery cycling, and promotes dendrite-free uniform lithium plating/stripping underneath. We demonstrated Li|Li4Ti5O12 cells with stable and flat potential profiles for 500 cycles without capacity decay or an increase in potential hysteresis.
Lithium metal is the ultimate anode choice for high energy density rechargeable lithium batteries. However, it suffers from inferior electrochemical performance and safety issues due to its high reactivity and the growth of lithium dendrites. It has long been desired to develop a materials coating on Li metal, which is pinhole-free, mechanically robust without fracture during Li metal deposition and stripping, and chemically stable against Li metal and liquid electrolytes, all while maintaining adequate ionic conductivity. However, such an ideal material coating has yet to be found. Here we report a novel synthesis method by reacting clean molten lithium foil directly with pure nitrogen gas to generate instantaneously a pinhole-free and ionically conductive α-Li3N film directly bonded onto Li metal foil. The film consists of highly textured large Li3N grains (tens of μm) with (001) crystalline planes parallel to the Li metal surface. The bonding between textured grains is strong, resulting in a mechanically robust film which does not crack even when bent to a 0.8 cm curvature radius and is found to maintain pinhole-free coverage during Li metal deposition and stripping. The measured ionic conductivity is up to 5.2 × 10-4 S cm-1, sufficient for maintaining regular current densities for controllable film thicknesses ranging from 2 to 30 μm. This Li3N coating is chemically stable, isolating the reactive metallic lithium from liquid electrolyte, prevents continuous electrolyte consumption during battery cycling, and promotes dendrite-free uniform lithium plating/stripping underneath. We demonstrated Li|Li4Ti5O12 cells with stable and flat potential profiles for 500 cycles without capacity decay or an increase in potential hysteresis.
Li
metal is the most attractive anode for lithium-based rechargeable
batteries because of its high theoretical specific capacity (3861
mAh g–1) and low reduction potential (−3.04
V versus standard hydrogen electrode).[1−4] Although lithium metal has been used in
commercialized primary lithium batteries (e.g., Li-MnO2 battery) featuring high energy density, intrinsic problems such
as low Coulombic efficiency, limited cycle life, and safety concerns
have long hindered its practical application in rechargeable batteries.
To achieve successful Li metal anodes, two fundamental challenges
need to be addressed: (1) the highly reactive nature between metallic
lithium and liquid electrolytes, and (2) the virtually infinite relative
volume change during plating and stripping due to the “hostless”
nature of Li metal, resulting in mossy and dendritic Li deposition.
During battery charging, liquid electrolytes decompose and react with
lithium to form a passivating solid-electrolyte interphase (SEI) on
the anode surface.[5] The drastic volume
change of Li metal during its stripping and plating processes leads
to the significant instability of this SEI. The mechanical fracture
of the formed SEI and its repair take place repetitively during battery
cycling, resulting in the continuous consumption of liquid electrolytes
and lithium, low Coulombic efficiency, large overpotential and short
cycle life, ultimately leading to degradation of battery performance.[6] Another challenging problem is the growth of
Li dendrites, which form due to current inhomogeneities during battery
charging, leading to serious safety problems (Figure a).[7,8] Furthermore, when Li
is stripped away during battery discharging, some dendrites may break
and become electrically isolated from the anode, generating accumulated
“dead lithium”, which accelerates the degradation of
Li batteries and reduces their thermal stability.[9,10]
Figure 1
Design
and structure of various Li metal anodes. (a) A bare Li
metal anode. Upon cycling, Li dendrites grow and liquid electrolytes
are rapidly consumed. (b) A nanofilm modified Li metal anode. The
pores and/or cracks of the film can trigger growth of Li dendrites
and fail in Li metal surface protection. (c) A Li metal anode with
a dense, ionically conductive film with strongly interconnected large
grains. A dense protective layer with good mechanical durability,
high ionic conductivity, and good adhesion to Li metal anodes can
isolate Li metal from corrosive liquid electrolytes and enable stable
electrochemical stripping and plating of Li metal underneath, preventing
dendrite growth and liquid electrolyte consumption.
Design
and structure of various Li metal anodes. (a) A bare Li
metal anode. Upon cycling, Li dendrites grow and liquid electrolytes
are rapidly consumed. (b) A nanofilm modified Li metal anode. The
pores and/or cracks of the film can trigger growth of Li dendrites
and fail in Li metal surface protection. (c) A Li metal anode with
a dense, ionically conductive film with strongly interconnected large
grains. A dense protective layer with good mechanical durability,
high ionic conductivity, and good adhesion to Li metal anodes can
isolate Li metal from corrosive liquid electrolytes and enable stable
electrochemical stripping and plating of Li metal underneath, preventing
dendrite growth and liquid electrolyte consumption.To make Li metal anodes viable, considerable research
effort has
been dedicated to understand the process of Li metal plating and stripping,[11−14] and various nanoscale films have been explored for Li metal surface
protection, including electrolyte additive derived films,[15−17] atomic layer deposited protection layers (e.g., Al2O3),[18] two-dimensional atomic crystal
layers (e.g., BN and graphene),[19] hollow
carbon hemispheres,[20] in situ chemical
reaction formed surface films (e.g., Li3PO4),[21] conformal lithium fluoride protection layers,[22] and grafted polymer layers.[23] However, despite the great progress, these coated Li metal
anodes can still fail during battery fabrication and upon battery
cycling due to the insufficient mechanical durability of the nanofilms
arising from the weak interconnection between the nanoscale particles.
Even with these protective nanofilms, electrolyte can still contact
the Li metal directly through pores and cracks, leading to locally
enhanced Li ion flux and the growth of Li dendrites (Figure b). In this respect, it is
highly desirable to have pinhole-free films with large, strongly interconnected
grains which can provide better mechanical properties, avoid the undesired
side reactions between liquid electrolytes and metallic lithium, and
block the formation of Li dendrites, enabling high-performance lithium
batteries (Figure c). The stable interfacial film would also help stabilizing the three-dimensional
host of Li metal recently demonstrated with hollow carbon spheres,[24] graphene oxide,[25] and nanofibers.[26] Li3N is
an excellent material for Li metal protection. It shows high ionic
conductivity (up to 10–3 S cm–1) and low electronic conductivity (<10–12 S
cm–1).[27] Furthermore,
as uncovered by computation studies, Li3N has unique thermodynamic
stability against Li metal.[28,29] In the past, Li3N has been explored as a coating; however, the grain sizes
are too small (<160 nm), and their interconnection is too weak
to suppress Li metal dendrites.[30,31] In this work, we develop
in situ growth of a stable, mechanically strong and pinhole-free α-Li3N crystalline film comprised of dense, large, and strongly
interconnected grains directly on the truly clean Li metal surface,
and present the excellent electrochemical performance of Li metal
anodes with such a Li3N protection layer.
Results and Discussion
Synthesis
of Pinhole-free-Li3N Solid Electrolyte
on Molten Li Metal
A pinhole-free-Li3N solid electrolyte
layer on Li metal was fabricated by utilizing the nitridation reaction
between pure N2 gas and fresh molten Li metal (Figure a and Movie S1). In a typical procedure, a Li metal
foil was first sandwiched between two Cu foils and heated at 300 °C
under mechanical pressure in Ar atmosphere, producing a Cu/Li/Cu sandwich
architecture with good adhesion between the Cu foils and Li metal.
After transfer to a N2-filled glovebox, the Cu/Li/Cu sandwich
was heated at 450 °C until the Li metal melted. The two Cu foils
with liquid lithium between them were then pulled apart quickly, and
further baked at 450 °C for 1 s to achieve the pinhole-free-Li3N film on Li metal. Right after this, the sample was cooled
down to room temperature and transferred to an Ar-filled glovebox.
Rapidly pulling apart the Cu foils is important to generate truly
fresh and clean liquid Li metal surfaces for reaction. Immediately
upon exposure, the fresh surfaces of the liquid Li metal quickly turned
black, forming a uniform crystalline Li3N layer on the
Li metal surface. The subsequent heat treatment process densified
the Li3N layer and further improved its crystallinity.
Optical and scanning electron microscope (SEM) images show that a
Li3N layer with very large grains is successfully grown
on the Li metal surface (Figure b,c). The surface of the Li3N layer is homogeneous
and shows well-developed Li3N crystal planes, indicating
high crystallinity. Cross-section SEM images show that the Li3N layer is comprised of strongly interconnected grains with
tunable thickness from 2 to 200 μm (Figure d and Figure S1). The dense, pinhole-free structure and reasonable thickness of
the Li3N layer are important for implementation in batteries
to prevent infiltration of liquid electrolytes and to suppress the
growth of Li dendrites.
Figure 2
Fabrication and morphology of the pinhole-free-LiN–Li metal anode. (a) The fabrication
process of the
pinhole-free-Li3N–Li metal anode. A Cu/Li/Cu sandwich
was heated at 450 °C until the Li metal melted. The two Cu foils
with liquid lithium on them were then separated away quickly, exposing
the fresh Li metal surfaces to nitrogen gas, and further baked for
1 s. (b–d) Dark field optical (b), top-view (c), and cross-section
(d) SEM images of a pinhole-free-Li3N–Li metal anode.
(e–g) Digital (e), top-view (f), and cross-section (g) SEM
images of a Li metal foil with a superficial native film after being
placed in a nitrogen filled glovebox for 14 h. A porous Li3N flake was achieved after the nitridation reaction. (h–j)
Digital (h), top-view (i), and cross-section (j) SEM images after
the reaction between a fresh and clean Li metal surface and low-concentration
N2 gas (∼10%) at room temperature. Li3N nanofilms on the Li metal surface were observed.
Fabrication and morphology of the pinhole-free-LiN–Li metal anode. (a) The fabrication
process of the
pinhole-free-Li3N–Li metal anode. A Cu/Li/Cu sandwich
was heated at 450 °C until the Li metal melted. The two Cu foils
with liquid lithium on them were then separated away quickly, exposing
the fresh Li metal surfaces to nitrogen gas, and further baked for
1 s. (b–d) Dark field optical (b), top-view (c), and cross-section
(d) SEM images of a pinhole-free-Li3N–Li metal anode.
(e–g) Digital (e), top-view (f), and cross-section (g) SEM
images of a Li metal foil with a superficial native film after being
placed in a nitrogen filled glovebox for 14 h. A porous Li3N flake was achieved after the nitridation reaction. (h–j)
Digital (h), top-view (i), and cross-section (j) SEM images after
the reaction between a fresh and clean Li metal surface and low-concentration
N2 gas (∼10%) at room temperature. Li3N nanofilms on the Li metal surface were observed.Different from any previous studies, here the two
vital factors
for the successful synthesis of the pinhole-free-Li3N solid
electrolyte are the creation of a truly pristine lithium surface and
utilization of high reaction temperature. Because of its high reactivity,
commercially obtained Li metal is always covered by a native surface
film,[32] and freshly exposed Li metal surfaces
cannot remain clean even in inert atmosphere due to the fast reaction
between Li metal and trace amounts of active gases, such asO2, water, and CO2 (Figure S2). This native film leads to a nonuniform and slow (10–20
h) nitridation reaction (Figure e and Figure S3), where
Li3N flakes with small particle size (∼200 nm) and
porous structure were formed (Figure f,g) instead of a dense pinhole-free-Li3N layer as shown in Figure c,d. To gain insight into the importance of a fresh and clean
Li metal surface for the growth of a uniform Li3N film,
a Cu/Li/Cu sandwich architecture was torn apart in an argon atmosphere
with a low content of N2 gas (∼10% by volume) at
room temperature. Homogeneous nitridation of the fresh Li metal surface
was observed, generating uniform Li3N nanofilms with tunable
thickness of 50–400 nm and corresponding color change during
the reaction (Figure h–j, Figure S4–6 and Movie S2). The size of Li3N particles
for these nanofilms ranges from 50 to 100 nm (Figure i). Hence, using truly pristine Li metal
surfaces avoids the influence of the native film and enables a fast
and homogeneous nitridation reaction over the whole Li metal surface.
The other key factor for the formation of a dense pinhole-free-Li3N film with large grains is the high reaction temperature.
With increased reaction temperature (100 °C, below the melting
point of Li metal at 180.5 °C), a denser Li3N nanofilm
was achieved in comparison to that obtained at room temperature (Figure S7). When we further increased the reaction
temperature to 450 °C, a pinhole-free-Li3N film with
large, strongly interconnected grains was achieved (Figure d). Therefore, the nitridation
of truly pristine Li metal surface at high temperature enables the
fast, uniform growth of a dense pinhole-free-Li3N film
with large grains on the Li metal surface.
Characterization of Pinhole-free-Li3N Solid Electrolyte
One main limitation of nanofilms
is their poor mechanical durability.
They are prone to failure during battery fabrication and cycling processes
due to the weak interconnection between the nanoparticles of the films
(Figures S8 and S9). Until now, little
emphasis has been placed on the development of mechanically robust
protective layers for Li metal. Our pinhole-free-Li3N film
successfully achieves good mechanical durability due to the strongly
interconnected large crystalline grains. The pinhole-free-Li3N film integrated Li metal anode is flexible and pliable. Bending
measurements and SEM investigation show that the pinhole-free-Li3N layer on the Li metal surface does not fail even at a curvature
radius of 0.8 cm, indicating extraordinary mechanical flexibility
and durability compared to the Li3N nanofilm (Figure a and Figure b left). For comparison, bending
measurements were also carried out on a Li3N nanofilm modified
Li metal anode. The Li3N nanofilm shows tiny cracks (∼200
nm) right after its fabrication (Figure S8). Bending the Li3N nanofilm to a curvature radius of
1 cm (Figure b right)
generates a number of cracks with size ∼3 μm, which is
large enough for Li metal dendrites to grow through. Because of the
higher thermal expansion coefficient of Li metal compared to that
of Li3N,[33] built-in compressive
stress is produced when the pinhole-free-Li3N–Li
metal sample cools from 450 °C to room temperature after its
fabrication. The compressive stress in thin film mechanics is known
to counteract the tension during bending and gives rise to good mechanical
robustness.[34]
Figure 3
Mechanical properties
characterizations of the pinhole-free-Li3N–Li metal
anode. (a) Crack size vs curvature plot
for a pinhole-free-Li3N film and a Li3N nanofilm.
The inset in (a) shows a digital image of a pinhole-free-Li3N–Li metal anode under bending. (b) Top-view SEM images of
a pinhole-free-Li3N film (left) and a Li3N nanofilm
(right) under bending at a curvature radius of 1 cm.
Mechanical properties
characterizations of the pinhole-free-Li3N–Li metal
anode. (a) Crack size vs curvature plot
for a pinhole-free-Li3N film and a Li3N nanofilm.
The inset in (a) shows a digital image of a pinhole-free-Li3N–Li metal anode under bending. (b) Top-view SEM images of
a pinhole-free-Li3N film (left) and a Li3N nanofilm
(right) under bending at a curvature radius of 1 cm.X-ray diffraction (XRD) measurements were performed
to determine
the crystalline phase and grain orientation of the product (Figure a). The symmetrical
θ-2θ scan revealed a pure α-phase of Li3N which is corroborated by the powder diffraction file 00-30-0759
(space group: P6/mmm) for the pinhole-free-Li3N. Only (001), (002), and (003) peaks are visible from the
α-phase Li3N, clearly indicating that the crystal
grains of as-grown α-phase Li3N are highly oriented
along the [001] direction perpendicular to the Li substrate surface.
This is quite different from the Li3N nanofilm achieved
at room temperature, which mainly exhibits random grain orientation
shown in Figure a
in blue. To fully characterize the texture of a crystalline Li3N film, the pole figure measurements were carried out.[35] Two pole figures were acquired with the Bragg
angles fixed at 2θ = 22.96° and 70.99° corresponding
to Li3N (001) and Li3N (112) crystallographic
planes. The diffracted intensity was collected by varying the tilt,
χ, and the rotation, ϕ, angles with respect to the normal
direction of a sample surface. Figure b shows the Li3N (001) pole figure from
pinhole-free-Li3N sample. As can be seen from the figure,
all diffracted intensity from (001) planes is concentrated around
χ = 0°, confirming that crystallographic grain orientation
of pinhole-free-Li3N film is exclusively along [001] direction
normal to the film’s surface. As shown in Figure S10, the Li3N (112) pole figure exhibits
a ring of a constant intensity at an angle χ = 46.7° consistent
with the angle between (001) and (112) planes of hexagonal Li3N. The constant intensity of Li3N (112) ring confirms
a random in-plane orientation of the Li3N grains. The distribution
of (001) and (112) pole figure densities indicates that Li3N film is highly textured exhibiting so-called fiber texture with
grains oriented [001] out-of-plane, but having random orientation
in-plane.
Figure 4
Structural characterizations of the pinhole-free-Li3N–Li metal anode. (a) XRD pattern of the pinhole-free-Li3N–Li metal anode and the Li metal anode with a Li3N nanofilm. (b) Pole figure of the (001) plane of the dense
pinhole-free-Li3N layer. (c) 2D XRD pattern of the pinhole-free-Li3N layer within a sample area of 100 × 100 μm. (d)
Schematic of the Li metal anode with an oriented pinhole-free-Li3N layer. XRD results indicate that the as-grown pinhole-free-Li3N layer is highly oriented along the [001] crystalline direction
and is comprised of large grains tens of micrometers in size.
Structural characterizations of the pinhole-free-Li3N–Li metal anode. (a) XRD pattern of the pinhole-free-Li3N–Li metal anode and the Li metal anode with a Li3N nanofilm. (b) Pole figure of the (001) plane of the dense
pinhole-free-Li3N layer. (c) 2D XRD pattern of the pinhole-free-Li3N layer within a sample area of 100 × 100 μm. (d)
Schematic of the Li metal anode with an oriented pinhole-free-Li3N layer. XRD results indicate that the as-grown pinhole-free-Li3N layer is highly oriented along the [001] crystalline direction
and is comprised of large grains tens of micrometers in size.The size of Li3N crystal
grains can be qualitatively
estimated by X-ray diffraction in transmission geometry and using
X-ray source with the spot size of ∼100 μm in diameter.
The two-dimensional (2D) XRD pattern shows only several individual
spots corresponding to Li3N phase within a sample area
of 100 × 100 μm, suggesting that the grain size of Li3N is around few tens of micrometers (Figure c). Meanwhile, the highly oriented crystalline
plane of Li3N crystals also supports the large grain size
of Li3N (Figure c). Thus, the pinhole-free-Li3N layer is vertically
grown on the Li metal substrate along the [001] crystalline direction
of hexagonal α-phase Li3N with a large grain size
(Figure d), in contrast
to the random growth of the Li3N nanofilm achieved at room
temperature. The oriented growth and large grain size of the dense
Li3N film also support its good mechanical properties.Good ambient stability can improve material safety and simplify
requirements for use in industrial battery fabrication environments.
To study the ambient stability of the pinhole-free-Li3N–Li
anode, XRD was performed after exposure to ambient conditions to identify
the formation of any new reaction products. Interestingly, although
Li3N nanofilms usually exhibit high reactivity with ambient
air, the XRD pattern of the pinhole-free-Li3N–Li
metal anode shows clear Li signals and strong Li3N signals
after exposure in ambient air with 40% humidity for 8 h, indicating
the good stability of the pinhole-free-Li3N–Li metal
anode in ambient conditions (Figure S11). Industrial battery fabrication is performed in a dry room with
a relative humidity of less than 1% and an operating period of less
than 8 h. The good ambient stability of the pinhole-free-Li3N–Li metal anode makes it promising in industry battery applications.
Our pinhole-free-Li3N solid electrolyte layer is highly
crystalline and dense, and once it is exposed to ambient air, a protective
passivation layer forms on the surface and limits further reaction
with the surrounding atmosphere.[27] Furthermore,
SEM investigations show that there is no significant change before
and after the storage of a pinhole-free-Li3N–Li
metal anode in a widely used ether-based electrolyte for 30 days,
indicating its good stability and compatibility with liquid electrolytes
(Figure S12).
Electrochemical Performance
of Pinhole-free-Li3N
Integrated Li Metal Anode
To measure the ionic conductivity
of the pinhole-free-Li3N layer, electrochemical impedance
spectroscopy (EIS) measurements were carried out on an all-solid-state
Li|pinhole-free-Li3N|Li symmetric cell (Figure a). Before the construction
of the symmetric cell, a 50 nm thick Au layer was deposited onto a
200 μm thick pinhole-free-Li3N film using thermal
evaporation techniques to serve a buffer layer to improve the contact
between the pinhole-free-Li3N solid electrolyte and Li
metalasAu and Li form an alloy at room temperature.[36,37] The first semicircle at relatively high frequency in the Nyquist
plots can be used to analyze the bulk resistance of the pinhole-free-Li3N film, while the interfacial resistance between the Li metal
and pinhole-free-Li3N layer can be calculated based on
the second semicircle.[36,38] The first semicircle yields a
resistance of 87 Ω cm2, and the ionic conductivity
of the pinhole-free-Li3N film is calculated as 2.3 ×
10–4 S cm–1. Meanwhile, the measured
ionic conductivity is as high as 5.2 × 10–4 S cm–1 for a 20 μm pinhole-free-Li3N film based on symmetric cells made of either two bare Li metal
electrodes or two pinhole-free-Li3N–Li electrodes
using liquid electrolytes (Figure S13).
With the calculated ionic conductivity of 2–5 × 10–4 S cm–1, a 2 μm thick film
will have an ionic resistance of 0.4 to 1 Ohm cm–2, 10 μm thick film will have an ionic resistance of 2 to 5
Ohm cm–2. These resistance values are low enough
for pinhole-free-Li3N to serve as an interfacial coating
layer. Thus, one can expect fast lithium stripping/plating kinetics
and stable cycling behavior of the pinhole-free-Li3N–Li
metal anode due to the excellent mechanical durability, good stability,
and high ionic conductivity of the Li3N film.
Figure 5
Electrochemical
characteristics of the pinhole-free-Li3N–Li metal
anode. (a) Nyquist plots of a Li|pinhole-free-Li3N|Li symmetric
cell. The black dot line denotes the experimental
data, while the red one represents the fitting data using the equivalent
circuit modeling (inset of a). (b) The voltage profiles of the Li|Li4Ti5O12 cells with the pinhole-free-Li3N–Li and the bare Li metal electrodes at 1 C. (c) Cycling
behaviors of the Li|Li4Ti5O12 cells
with the pinhole-free-Li3N–Li and the bare Li metal
electrodes at 1 C. (d, e) Cross-section SEM images
of a pinhole-free-Li3N–Li metal anode (d) and a
bare Li metal anode (e) after 100 cycles.
Electrochemical
characteristics of the pinhole-free-Li3N–Li metal
anode. (a) Nyquist plots of a Li|pinhole-free-Li3N|Li symmetric
cell. The black dot line denotes the experimental
data, while the red one represents the fitting data using the equivalent
circuit modeling (inset of a). (b) The voltage profiles of the Li|Li4Ti5O12 cells with the pinhole-free-Li3N–Li and the bare Li metal electrodes at 1 C. (c) Cycling
behaviors of the Li|Li4Ti5O12 cells
with the pinhole-free-Li3N–Li and the bare Li metal
electrodes at 1 C. (d, e) Cross-section SEM images
of a pinhole-free-Li3N–Li metal anode (d) and a
bare Li metal anode (e) after 100 cycles.To evaluate the electrochemical stability of the pinhole-free-Li3N–Li metal anode, we paired it with a Li4Ti5O12 electrode since Li4Ti5O12 has a flat discharge/charge plateau at ∼1.5
V and very stable cyclability without an increase in potential hysteresis
during cycling. The surface of Li4Ti5O12 particles is free of solid electrolyte interphase (SEI) film, so
it does not consume organic electrolyte and active Li. Therefore,
we can truly evaluate the stability of the pinhole-free-Li3N–Li metal anode without the influence of the counter electrode.
During discharge of a cell with a Li4Ti5O12 counter electrode, lithium is electrochemically stripped
from the Li metal anode, conducted through the pinhole-free-Li3N layer and liquid electrolytes, and intercalated into the
Li4Ti5O12 counter electrode. Then,
upon charging, it is extracted from the Li4Ti5O12 host and deposited back on the Li anode underneath
the pinhole-free-Li3N layer. Figure S14 shows the voltage profiles of a pinhole-free-Li3N–Li|Li4Ti5O12 cell and its
bare Li anode Li|Li4Ti5O12 counterpart
for cycling at 1 C for Li4Ti5O12 (0.8
mA cm–2), respectively. An in-depth comparison of
voltage profiles at each labeled cycle is shown in Figure b and Figure S15. After activation (the first cycle), the pinhole-free-Li3N–Li|Li4Ti5O12 cell
shows similar flat discharge and charge voltage plateaus (∼1.51
V and ∼1.61 V, respectively, Figure b) as the Li|Li4Ti5O12 cell without any potential hysteresis for the second
cycle. The pinhole-free-Li3N–Li|Li4Ti5O12 cell exhibits stable discharge/charge plateaus
and high capacities with ∼99.99% Coulombic efficiency on cycling
with a high capacity of 160 mAh g–1 for the second
cycle and 159 mAh g–1 for the 500th cycle (Figure b,c and Figure S16). In contrast, the Li|Li4Ti5O12 cell displays obviously increased potential
polarization and capacity decay on cycling. Its capacity decays from
159 mAh g–1 for the second cycle to 138 mAh g–1 for the 340th cycle, which quickly drops down to
132 mAh g–1 for the 350th cycle (Figure c). Additionally, the Li|Li4Ti5O12 cell using Li3N nanofilm
coated Li metal anode shows better electrochemical stability than
that of the bare Li|Li4Ti5O12 cell
but worse than the pinhole-free-Li3N–Li|Li4Ti5O12 cell during cycling due to the weak
interconnection between the Li3N nanoparticles and cracking
of the nanofilms (Figure S17). Thus, compared
with the Li3N nanofilm for Li metal anode protection, this
pinhole-free-Li3N material shows much better protection
in terms of suppression of Li metal dendrite growth and improved battery
cycling. Meanwhile, the pinhole-free-Li3N–Li|Li4Ti5O12 cells shows excellent rate capability
with 69% of the 0.2 C (0.19 mA cm–2) capacity at
6 C (5.59 mA cm–2) due to the good ionic conductivity
of the Li3N layer on the Li metal surface (Figure S18). The morphology and structure of
the pinhole-free-Li3N–Li metal anode was observed
after 100 cycles. The pinhole-free-Li3N layer preserves
its structural integrity over battery cycling, suggesting that stable
Li3N-liquid electrolyte and Li metal-Li3N interfaces
are achieved for the cells with the pinhole-free-Li3N–Li
metal anode, providing good protection of the Li metal anode and enabling
good battery performance (Figure d). In comparison, the bare lithium metal anode has
a loose and porous structure with randomly grown dendrites after 100
discharge/charge cycles, suggesting serious corrosion of the anode
(Figure e). The accumulation
of SEI and “dead lithium” at the bare Li metal anode
hurts the kinetics of the control cell, which, combined with the consumption
of liquid electrolytes, eventually leads to the failure of the battery.
Conclusions
In summary, we have fabricated a robust pinhole-free-Li3N solid electrolyte film with large, strongly interconnected
grains
on Li metal surface through the reaction between clean molten lithium
foil and pure nitrogen gas. The pinhole-free-Li3N layer
bonds strongly to Li metal with oriented growth of dense α-Li3N crystals along the [001] direction. The Li3N
coating exhibits excellent mechanical durability, high ionic conductivity,
and high electrochemical stability with liquid electrolytes. By restraining
the growth of lithium dendrites, suppressing parasitic side reactions,
and minimizing the interfacial resistance between the solid electrolyte
and Li metal, the pinhole-free-Li3N film therefore significantly
improves the electrochemical plating and stripping behavior of Li
metal. Our work demonstrates that a pinhole-free-Li3N film
with large and strongly interconnected grains, and high ionic conductivity
opens new prospects to address the challenges that high-energy-density
lithium batteries face.
Methods
Materials Synthesis
To fabricate a Cu/Li/Cu sandwich,
a Li metal foil (99.9%, Alfa Aesar) was first sandwiched between two
Cu foils and then heated on a hot plate at 300 °C in an Ar-filled
glovebox with moisture level and oxygen level below 0.1 ppm. Once
the Li metal melted, a flat aluminum block was pressed on the top
Cu foil, and good adhesion between Li metal and Cu foils was achieved.
During this process, 500 μm thick Si slides were sandwiched
between two Cu foil sheets as spacers to control the thickness of
the final Li layer inside Cu/Li/Cu sandwich structure as 500 μm.
To prepare Li3N nanofilm modified Li metal, a small amount
of N2 gas (∼10% by volume) was introduced into the
pure Ar atmosphere. The cooled Cu/Li/Cu sandwich architecture was
quickly torn apart into two pieces of Cu-foil supported Li metal.
The nitridation reaction took place immediately after the exposure
of the fresh Li metal surface, producing a Li3N nanofilm
on the Li metal surface. To obtain Li3N nanofilms with
certain thickness, the nitridation reaction was terminated by flowing
pure argon gas. Pinhole-free-Li3N -Li metal anodes were
prepared by a similar procedure in a pure N2-filled glovebox.
The Cu/Li/Cu sandwich architecture was heated at 450 °C until
the Li metal melted. The two Cu foils with liquid lithium on them
were then separated away quickly and further baked for 1 s to form
pinhole-free-Li3N–Li metal anodes. After the reaction,
the as-prepared anodes were stored in an Ar atmosphere to avoid further
nitridation reactions. For comparison, the Li metal control sample
was prepared using the same method by pulling apart Cu/Li/Cu sandwich
structures at 450 °C, except that the whole process was done
in an Ar atmosphere. Samples with the thickness between 250–300
μm were chosen for further mechanical and electrochemical tests.
Characterization
The morphology of the samples was
characterized using a FEI XL30 Sirion scanning electron microscope.
Electrodes after electrochemical cycling were rinsed in 1,3-dioxolane
(DOL) and dried in an argon-filled glovebox before their SEM investigation.
All the samples for SEM were sealed in Ar before being transferred
into the SEM chamber. The X-ray diffraction measurements were carried
out employing Panalytical X’Pert Material Research diffractometer
with Cu Kα radiation (λ = 1.5419 Å). In order to
protect specimens from the environmental degradation, the samples
were sealed with Kapton tape before conducting X-ray diffraction experiments.
The pole figure measurements were performed using X-ray lens delivering
a point-focus X-ray beam. The beam size was controlled by a cross-slit
collimator. During the prolonged pole figure measurements, the samples
were additionally sealed inside the Anton-Paar stage with continuous
flow of Ar gas.
Electrochemical Measurements
Battery
performance was
evaluated by galvanostatic cycling of 2032-type coin cells (MTI Corporation)
on a LAND 8-channel battery tester. The electrodes used were pinhole-free-Li3N–Li metal foils or freshly scraped Li metal foils
(99.9%, Alfa Aesar). Li4Ti5O12 electrodes
were used as the counter electrodes. Ether-based electrolyte with
lithium bis(trifluoromethanesulphonyl)imide (LiTFSI) and carbonate-based
electrolyte with LiPF6 are two widely used electrolytes
for Li metal anode studies. In our work, ether-based electrolyte with
LiTFSI was used for the electrochemical characterization. The cells
were assembled in an argon-filled glovebox using 1 M LiTFSI in a mixture
of 1,3-dioxolane/1,2-dimethoxyethane (DME) (1:1 v/v) with 1 wt % lithium
nitrate as the electrolyte and a Celgard 2325 membrane as the separator.
Li4Ti5O12 electrodes were constructed
using a slurry method by mixing commercial Li4Ti5O12 powder, carbon black and polyvinylidene fluoride (PVDF)
binder in the weight ratio of 90:5:5 in N-methyl-2-pyrrolidinone
(NMP) solvent. The typical mass loading of the Li4Ti5O12 electrodes was ∼5 mg cm–2. The cutoff potential range for pinhole-free-Li3N–Li|Li4Ti5O12 and Li|Li4Ti5O12 cells is 1–2.5 V. A Biologic VMP3 system was
employed to carry out the electrochemical impedance spectroscopy measurements.
To fabricate an all-solid-state Li|pinhole-free-Li3N|Li
symmetric cell, a pinhole-free-Li3N/Li/Cu structure with
a ∼ 200 μm pinhole-free-Li3N layer bonded
on a ∼ 50 μm lithium layer was first prepared by tearing
apart a Cu/Li/Cu sandwich architecture at 450 °C in pure N2 atmosphere. Then a 50 nm thick Au layer was deposited on
top of the pinhole-free-Li3N layer using thermal evaporation
techniques. Finally, the all-solid-state Li|pinhole-free-Li3N|Li symmetric cell was assembled by covering the Au deposited pinhole-free-Li3N/Li structure with a 50 μm thick lithium metal foil.
The cell was heated at 80 °C for 1 h to improve the interface
contact between the lithium foil and the pinhole-free-Li3Nelectrolyte.
Authors: Rangeet Bhattacharyya; Baris Key; Hailong Chen; Adam S Best; Anthony F Hollenkamp; Clare P Grey Journal: Nat Mater Date: 2010-05-16 Impact factor: 43.841