REVIEW ARTICLE

PUBLISHED: 24 JULY 2017 | VOLUME: 2 | ARTICLE NUMBER: 17108

A materials perspective on Li-ion batteries at

extreme temperatures
Marco-Tulio F. Rodrigues, Ganguli Babu, Hemtej Gullapalli, Kaushik Kalaga, Farheen N. Sayed,
Keiko Kato, Jarin Joyner and Pulickel M. Ajayan*

With the continuous upsurge in demand for energy storage, batteries are increasingly required to operate under extreme envi -
ronmental conditions. Although they are at the technological forefront, Li-ion batteries have long been limited to room tem-
perature, as internal phenomena during their operation cause thermal fluctuations. This has been the reason for many battery
explosions in recent consumer products. While traditional efforts to address these issues focused on thermal management strat-
egies, the performance and safety of Li-ion batteries at both low (<20°C) and high (>60°C) temperatures are inherently related
to their respective components, such as electrode and electrolyte materials and the so-called solid-electrolyte interphases. This
Review examines recent research that considers thermal tolerance of Li-ion batteries from a materials perspective, spanning a
wide temperature spectrum (–60 °C to 150 °C). The structural stability of promising cathodes, issues with anode passivation,
and the competency of various electrolyte, binder and current collectors are compared for their thermal workability. The pos-
sibilities offered by each of these cell components could extend the environmental frontiers of commercial Li-ion batteries.

E
nergy storage forms the foundation for success of numerous
commercial products. Though many battery chemistries exist,
Li-ion batteries (LIBs) are at the forefront for rechargeable appli-
cations, as the combination of high energy density, light weight, and
low self-discharge makes them highly attractive for portable con-
sumer electronics1. While a large spectrum of consumer applications
operate at room temperature, demand for batteries to survive and
operate under thermal extremes is rising. Military-grade batteries
are expected to operate from –40 °C to 60 °C, and such LIBs are yet
to be fully optimized and developed. Electric vehicles require battery
systems capable of having stable performance in both colder regions
and hot desert conditions. Apart from extending the operability of
conventional batteries, task-specific applications also call for energy
storage at further extremes. Areas like subsurface exploration (such
as for oil and gas), thermal reactors, defensive arsenals and space
vehicles call for operating temperatures over a wide range. On the
other hand, equipment for cold climates and high-altitude drones
can experience temperatures as low as –60 °C. Batteries might also be
required to operate under a predefined temperature regime. Medical
devices, for example, work at room temperature, but are subjected to
120 °C during sterilization. In uses such as energy harvesting, drill-
ing sensors and drones, batteries experience dissimilar temperatures
during charge and discharge. Besides external influences, internal
phenomena while operating LIBs also cause thermal fluctuations2,
sometimes unintentionally swinging to hazardous extremes3. Recent
battery explosions in cases such as the Boeing Dreamliner and
Samsung Note7 have raised already existing safety concerns4.
LIBs are in a unique position to cater to all these applications:
while other cell chemistries have long-lived issues limiting their
operability, the range of electrode and electrolyte materials compat-
ible with LIBs can potentially render these devices active across a
wide temperature range. Among other cell concepts, water-based
technologies, as lead–acid and nickel–metal hydride, are intrinsi-
cally limited by the electrolyte to operate between –50 °C and 50 °C
(ref. 5). Extremely high temperatures are compatible with — and
required by — molten salt batteries, while operation below 90 °C
is impractical. Many applications requiring extreme temperature
windows rely on primary lithium thionyl chloride (Li–SOCl2) bat-
teries, usable from –60 °C to 150 °C (ref. 5). Despite this impres-
sive thermal resilience, peak performance is only achieved between
20 °C and 55 °C, and cell usage is confined to low power conditions.
A LIB with proper formulation can potentially meet and surpass
the temperature window offered by Li–SOCl2 batteries, while also
offering rechargeability, higher power capability and reduced toxic-
ity. Besides serving diversified applications, expanding the tempera-
ture limits of LIBs would also have a rather beneficial side-effect:
improvement of consumer devices’ resistance to thermal abuse.
In designing a battery configuration for a targeted application, it
is critical to understand the workability of the materials being used.
The charge storage mechanism in LIBs involves a synergy of trans-
formations among its components, and temperature plays a deci-
sive role. Many recent efforts have considered thermal effects on the
individual materials used for LIBs in a piecewise approach. Though
it is not guaranteed that a full-cell encompassing each of these frag-
ments would function as desired, it is nevertheless important to
individually understand the limits of these materials.
This Review examines recent reports on thermal characteristics
of battery components and attempts to present a materials perspec-
tive, both at low and high temperature extremes. Reports pertaining
to electrochemical behaviour of electrode materials, development
of unconventional electrolytes and the role of dormant support
components (such as binders and current collectors) in the perfor-
mance at extreme temperatures are discussed in subsequent sec-
tions. Although optimal cell-level performance under demanding
conditions has yet to be fully achieved, the intrinsic properties of
these materials cast favourable prospects over the evolution of this
battery chemistry.
Cathodes
The cathode is the transit centre for both Li + and electrons, and
these materials are susceptible to undesirable phase transitions.
Further, parasitic reactions at the electrode–electrolyte interface in

Department of Material Science and Nanoengineering, Rice University, Houston, Texas 77005, USA. *e-mail: ajayan@rice.edu

NATURE ENERGY 2, 17108 (2017) | DOI: 10.1038/nenergy.2017.108 | www.nature.com/natureenergy 1

©
2
0
1
7
M
a
c
im
lla
n
P
u
b
ils
h
re
s
L
im
it
e
d
,
p
a
rt
o
fp
S
ri
n
g
re
N
a
t
u
r
e
.
A
ll
ri
g
h
ts
r
e
s
e
r
v
e
d
.
REVIEW ARTICLE NATURE ENERGY

a b c d
100
80

retention (%)
Capacity
LFP
60 LCO NMC
LMO
40 Ambient
Ambient (ref. 8) Ambient (ref. 10) Ambient (ref. 32)
20 (our data)
60 ºC (ref. 8) 60 ºC (ref. 10) 55 ºC (ref. 32) 55 ºC (ref. 21)
0
0 25 50 0 25 50 0 25 50 0 25 50
Cycle number (#)

e f Cation 2º phase
Lattice
mixing formation
strain LCO
(Li+–Ni2+)
Metal NMC
dissolution Metal
Oxygen Li+
Li+ dissolution
evolution
Co Slow charge transfer Mn/Ni/Co
Slow charge transfer
and Li+ di¦usivity and Li+ di¦usivity O2–
O2–

g LMO h High
Mn3+ LFP
Mn4+ internal
Li+ impedance
2º phase
formation Mn Li+ Poor e–
O 2- Fe conductivity
Jahn–Teller
distortion P4+ Sluggish
Co-existence of Metal O2– Li transport
two phases at HV dissolution

i j 4.4
260
4.2
240
4.0
220 25 °C
3.8
Temperature (°C)

E vs (Li/Li+)/V

200 Li1–xCoO2 3.6

–20 °C
Li1–xNiO2
180 3.4 4
Li1–xNi0.8Co0.2O2
3.2 1 Pristine
160 Li1–xMn 2O4 2 D500
3.0 3 D1100
Li1–xFePO4 4 D2200 2
140 Mn-rich NMCs 4
2.8 5 D3500 5
Mn-deficient NMCs 1 3
120 2.6 123 5
0.0 0.2 0.4 0.6 0.8 1.0 –20 0 20 40 60 80 100 120 140 160 180 200
Fraction of Li in cathode materials Capacity/mAh g–1

Figure 1 | Temperature-dependent performance constraints in cathode materials. a–h, Cycling plots (a–d) show ambient and elevated temperature
performance while schematics (e–h) show structural instabilities that get aggravated at high (text in red) and low (text in blue) temperature for
representative cathode materials. Layered oxide framework cathodes LiCoO 2 (LCO) (a,e); LiNi1/3Mn1/3Co1/3O2 (NMC) (b,f); spinel LiMn2O4 (c,g); olivine
phosphate LiFePO4 (d,h). The data points for plots a–d have been deduced from previous reports as indicated in the panels. i, Thermal stability as a
function of lithium content in cathode materials. The temperature values indicate the onset of reaction between carbonate-based electrolyte and the
electrode material at the respective state-of-charge. The connected data points are derived from DSC curves, while shaded regions compare influence of
the Mn concentration on the stability of NMC cathode11,110. j, Performance of LiNi0.5Co0.2Mn0.3O2 (here denoted as NCM523) at low and room temperatures.
Substitution of an optimum concentration of Al inhibits impedance rise and accelerates lithium ion diffusion at low temperature, which translates into
improved performance. Panel j reproduced with permission from ref. 26 (Elsevier).

deep-charged conditions (delithiated state) are likely to destabilize
the cathode’s structure.
The reported cyclic stability of well-known cathode materi-
als at high temperature and their source of fragility at temperature
extremes is schematically shown in Figs 1a–h. The listed issues are
minimal or negligible at room temperature but accelerated upon cell
exposure to high temperatures, increasing the extent of metal dis-
solution and oxygen evolution. These processes are generally irre-
versible, leading to performance degradation owing to loss of active
mass and instability in the structural composition. Additionally,
the self-accelerating reactivity of decayed electrodes with flam-
mable electrolyte at elevated temperatures poses safety concerns6
(Fig. 1i). Contrarily, sluggish kinetics limits cathode performance
below 0 °C (Fig. 1j). Although the development of high-energy-
density cathodes is critical for many applications, growing demand
for LIBs to survive/operate under extreme conditions has fostered
considerable research interest in probing the characteristics of cath-
ode materials with temperature as a variable. Table 1 summarizes
representative reports pertaining to known cathode materials and
their modifications.

2 NATURE ENERGY 2, 17108 (2017) | DOI: 10.1038/nenergy.2017.108 | www.nature.com/natureenergy

©
2
0
1
7
M
a
c
im
lla
n
P
u
b
ils
h
re
s
L
im
it
e
d
,
p
a
rt
o
fp
S
ri
n
g
re
N
a
t
u
r
e
.
A
ll
ri
g
h
ts
r
e
s
e
r
v
e
d
.
NATURE ENERGY REVIEW ARTICLE
Table 1 | Electrochemical performance of cathode materials at extreme temperatures.
Cathodes Temperature (°C) Capacity Number of cycles Capacity Electrolyte and anode
(mAhg–1) retention (%)
Layered oxide cathodes
AlPO4/LiCoO2 (ref. 7) 60 180 30 79 Organic and lithium
LiNi0.33Co0.33Mn0.33O2 (ref. 9) 60 175 100 95.3 Organic and Li4Ti5O12
Al doping/ LiNi 0.5Co0.2Mn0.3O2 –20 106 - - Organic and lithium
(ref. 26)

Spinel oxide cathodes

LTO-coated LiMn2O4 (ref. 94) 60 132 100 97 Organic and lithium
LiMn1.5Ni0.5O3.8F0.2 (ref. 18) 60 110 50 40 Organic and lithium
Li1+xMn2–xO4 (ref. 17) 120 80 20 50 RTIL and Li4Ti5O12

Phosphate cathodes
C-LiFePO4 (ref. 29) 60 160 40 93 Organic and lithium
LiFePO4 (ref. 95) 120 160 20 94 PC-LiBOB and lithium
LiFePO4 (ref. 20) 250 - 8 - Molten LiTFSI and lithium
Nano-LiFePO4/C (ref. 96) –20 120 100 99 Organic and lithium
LiFePO4/C (ref. 97) –40 54.7 - 39 Organic and lithium

Layered oxide framework cathodes. Lithium cobalt oxide (LCO),
the commercially established cathode, belongs to the two-dimen-
sional layered materials family (LiMO 2, M = Co,Ni,Mn), wherein
Li+ occupy the sites between two O–M–O layers. Though it shows
satisfactory performance at room temperature with intercalation/
de-intercalation of 0.5 Li per unit formula, structural instability
in the deeply-charged state and fragile Co–O bonding limits its
use to 60 °C (refs 7,8) (Fig. 1a). The intermediate oxidation state
of cobalt (+3.5) in charged Li 1–xCoO2 makes the crystal suscep-
tible to change from the native hexagonal phase to a monoclinic
structure, releasing oxygen due to metal dissolution in the electro-
lyte, in a process that is aggravated beyond 60 °C. Other layered
oxides present similar stability limitations, as the structural breath-
ing during Li-ion intercalation/de-intercalation generates residual
stresses in the crystallographic z-direction. Temperature-provoked
cation exchange and structural transformation to spinel structures
under deeply-charged conditions have written off the commer-
cial use of materials such as LiNiO2 and LiMnO2 at the early stage
of development.
In the case of mixed metal oxides, there is a positive synergy
between their constituting elements. LiNi1/3Mn1/3Co1/3O2 (NMC),
for instance, exhibits high capacity (Ni 2+), minimal cationic dis-
order (Co3+) and extended thermal stability (Mn 4+) (refs 9,10).
Consequently, NMC has demonstrated good capacity retention at
60 °C, presenting high coulombic efficiencies for over 100 cycles9.
These features motivated NMC’s incursion in the commercial mar-
ket, with growing use in cells for portable electronics. At elevated
temperature, challenges for lasting cyclability are metal dissolution
into electrolyte, and minimization of cation mixing, as Li+ and Ni2+
exhibit similar ionic radii (Fig. 1f).
The presence of electrochemically inactive species may exhibit
a pillaring effect, stabilizing the active cations at their respective
crystallographic positions: high manganese concentrations in the
lattice, for instance, stabilize the crystal structure of delithiated
NMC (Fig. 1i; ref. 11). Similarly, LiNi0.8Co0.15Al0.05O2 (NCA) deliv-
ers high stable capacity (~200 mAhg–1), as doping with inert Al
enhances charge retention, facilitating maximum utilization of the
active transition metal. However, capacity fade has been observed
for NCA at 40 °C to 70 °C, due to abrupt surface film growth and
micro-cracks at grain boundaries12. NCA is highly cost-effective,
justifying its adoption in the Panasonic batteries powering the latest
Tesla vehicles13. However, the energetic nature14 of NCA conditions
its use in high-capacity cells to strict state-of-health monitoring and
safety control.
Another approach to improve high-temperature stability is to
minimize the electrode–electrolyte contact through surface engi-
neering. Hydrofluoric acid (HF) originating from electrolyte decay
favours metal dissolution, contributing to oxygen evolution from
the cathode. Coating with electrochemically inert materials, such as
oxides and phosphates, has proven effective as the bulk of the cath-
ode is shielded from interfacial degradations 6. Although structural
modifications have been successful in expanding the temperature
range of layered oxides by a marginal extent, their intrinsic fragility
makes them vulnerable to thermal degradation.
Spinel-structured cathodes. Unlike in layered materials, changes
in unit cell lattice parameters during charge–discharge are iso-
tropic in spinel LiMn2O4 (LMO). This material exhibits good ther-
mal stability and rate capability, owing to its strong edge-shared
(Mn2)O4 octahedral lattice and three-dimensional Li-ion conduc-
tion pathways. Thermogravimetric analysis of deeply charged LMO
(λ-MnO2) showed that oxygen release usually occurs at tempera-
tures higher than for layered materials, indicating the robustness of
spinels. Although λ-MnO2 does not evolve oxygen until ~400 °C,
it converts into inactive MnO2 phases at as low as 190 °C (ref. 14).
Electrolyte oxidation at 4.0 V3+and metal dissolution due to dispro-
portionation reactions (2Mn → Mn4+ + Mn2+ ) are
(solid) (solid) (liquid)
serious concerns for LMO. Moreover, Jahn−Teller distortion due
to the presence of excess Mn 3+ at 3.0 V leads to phase transition
from cubic to tetragonal (Fig. 1g). As these processes become more
prominent at higher temperatures, structural and surface engineer-
ing are required to stabilize the cathode.
Like layered materials, structural stabilization can be attained
using coatings and doping 6,15. Surface protection of LMO using
inert oxides, fluorides and phosphates is an effective approach to
improve performance up to 60 °C, as these materials can act as
HF scavengers. However, insulating coatings inflict structural mis-
match, creating stacking faults that can obstruct Li+ diffusion16. The
surface layer thickness should be in the 1–2 nm range, effectively
blocking solvent molecules while allowing Li-ion conduction.

NATURE ENERGY 2, 17108 (2017) | DOI: 10.1038/nenergy.2017.108 | www.nature.com/natureenergy 3

©
2
0
1
7
M
a
c
im
lla
n
P
u
b
ils
h
re
s
L
im
it
e
d
,
p
a
rt
o
fp
S
ri
n
g
re
N
a
t
u
r
e
.
A
ll
ri
g
h
ts
r
e
s
e
r
v
e
d
.
REVIEW ARTICLE NATURE ENERGY

Li+
E"ects at high temperature Solvent E"ects at low temperature
a c Metal ion e g

–9

LgD (cm2s–1)
SEI
Li dendrite
Reaction with growth –10
metal cations
Delithiated
Lithiated

–11
3 3.4 3.8 4.2
1000/T (k–1)
b d f h
Increase in
charge-transfer 21 °C 5 °C –10 °C –23 °C
and SEI
0.00

(V versus Li/Li+)
Anode polarization
Gas evolution resistances

omposition Ω
change –0.08 Anode active
SEI material
thickening Amorphous carbon
–0.12
Graphite
0.0034 0.0036 0.0038 0.0040
1/T(K–1)

Figure 2 | Challenges imposed by temperature on the performance of graphite anodes. a–f, The many interfacial processes that might take place at
the electrode. Elevated temperatures can damage the SEI (a), promoting further decomposition of the electrolyte to form a thicker and more resistive
layer (b). The passivation film can also be negatively affected by reactive species generated on the cathode (c), which contributes to permanent changes
in SEI composition (d). At the low temperature extreme, lithium plating becomes very problematic (e), and reaction kinetics are negatively affected
by the increased Rct and the SEI’s blocking character. g, Temperature dependence of the apparent chemical diffusion coefficient (D) of Li+ in graphite.
There is a large decrease in D below –20 °C, when the solid-state diffusion also becomes much slower during the lithiation process, with deep effects
on rate capability. h, Anodic polarization for Li+ insertion at different temperatures. The cell overpotential rises quickly at low temperatures, leading to
deposition of metallic lithium at the electrode surface (inset). The low polarization observed for amorphous carbon renders this material much safer at
low temperatures. Panels reproduced with permission from: g, ref. 34 (Elsevier); h ref. 29 (Elsevier).

Doping Mn-sites with suitable ions stabilizes the lattice by expand-
ing the regular MO6 octahedra, improving accommodation of
Mn3+ → Mn4+ transitions, even at 55 °C (ref. 15). For this, several
isovalent (Cr3+, Al3+) and aliovalent (Mg2+, Ti4+) dopants have been
successfully employed. Use of alternative electrolytes can also be
beneficial for applications in more demanding environments,
as cycling at 120 °C has been demonstrated using ionic liquids 17.
Despite its relative thermal stability and historical prominence,
the relatively low capacity (100–120 mAhg–1) of LMO bars it from
widespread commercial use.
Recently, the high-voltage spinel LiMn1.5Ni0.5O4 has been in
the spotlight due to its high energy density. However, introduc-
tion of Ni into the lattice reduces the thermal stability, provok-
ing oxygen release at temperatures much lower than for LMO 18.
Further, structural decomposition is likely to be more prominent
at high temperature during the complete removal of lithium at
high voltages. Although more development is required to confer
thermal stability to high-voltage chemistries, spinels are expected
to be comparatively more thermally resilient than layered oxides,
with the major hurdles being metal dissolution and Jahn–Teller
distortion.
Phosphate cathodes. The limited structural stability of layered
and spinel cathodes arises mostly from the fragility of M–O bonds
constructing the lattice. Polyanion-based phosphates provide more
resilient frameworks, since oxygen atoms are covalently bonded to
phosphorous instead of transition metals, facilitating charge distri-
bution19. Of the known cathode structures, phosphates in general
and LiFePO4 (LFP) specifically are leading choices for applications
involving elevated temperatures, because of features such as: high
thermal and structural stability, as decomposition occurs above
400 °C; oxygen release unlikely from PO43–; and minimal volume
mismatch between LiFePO4/FePO4 phases19,20, improving struc-
tural integrity after complete Li removal. The resilience of LFP in a
cell environment has been highlighted by its use in an LIB at 250 °C
with molten salt electrolyte20.
Despite these favourable properties, LFP’s poor ionic and elec-
tronic conductivities still preclude utilization of micron-sized active
material particles, especially without conductive coating (Fig. 1h).
In addition, LFP is natively susceptible to metal dissolution, requir-
ing conformal coating with carbon, metal oxides or polymers 19,21
to achieve considerable cycle life. Though the low energy density
diminishes the significance of LFP for the consumer electronic
market, its higher power density and long cycle and calendar lives
make this material rather attractive for specific industrial and sta-
tionary applications. LFP has also been employed by Bolloré Group
in solid-state batteries employing polymer electrolyte and lithium
metal anodes for electric vehicles.
Recently, other phosphates, such as VOPO 4 and its lithiated
compounds (LiVOPO4, Li2VOPO4), have been found to exhibit
high thermal stability14. Like LFP, VOPO4 does not evolve O2 upon
heating, and only reacts with electrolyte beyond 200 °C, making it
inherently safe. Li3V2(PO4)3 (LVP) is another candidate for applica-
tions involving a wide range of temperatures, asserting good Li-ion
kinetics. However, the high operating voltages might render LVP
susceptible to metal dissolution at elevated temperatures, and dop-
ing might play an essential role in stabilizing the charge density over
vanadium atoms.

4 NATURE ENERGY 2, 17108 (2017) | DOI: 10.1038/nenergy.2017.108 | www.nature.com/natureenergy

©
2
0
1
7
M
a
c
im
lla
n
P
u
b
ils
h
re
s
L
im
it
e
d
,
p
a
rt
o
fp
S
ri
n
g
re
N
a
t
u
r
e
.
A
ll
ri
g
h
ts
r
e
s
e
r
v
e
d
.
NATURE ENERGY REVIEW ARTICLE
Table 2 | Electrolyte components to extend the temperature range of Li-ion batteries.
Low temperature (>–60°C) High temperature (<60°C)
Compounds Mechanism Compounds Mechanism
Additives Additives
Fluoroethylene carbonate (FEC) (ref. 59); Formation of a more conductive SEI FEC (ref. 102) Formation of a resilient SEI on
butyl sultone (BS) (ref. 60) and decrease in charge-transfer carbon anode
resistance (Rct)
Propane sulfone (PS) (ref. 103); Formation of ionically
propene sulfone (PST) (ref. 103); conducting and thermally
Tris(trimethylsilyl)borate (TMSB) stable layer on cathode
(ref. 104); 1,1’-sulfonyldiimidazole
(SDM) (ref. 105)
1,2-bis(difluoromethylsilyl) Surface protection on
ethane (FSE) (ref. 62); di-(2,2,2 both electrodes
trifluoroethyl)carbonate (DFDEC)
(ref. 106); vinylene carbonate (VC)
(ref. 107)

Lithium salts Lithium salts

Lithium tetrafluoroborate (LiBF4) (ref. 56);
lithium difluoro(oxalate)borate (LiDFOB)
(ref. 98)
Lithium pentafluoroethyltrifluoroborate
(LiFAB) (ref. 99)
High ionic conductivity at low
temperatures, and decrease in Rct
High ionic conductivity below –10 °C
LiBOB (ref. 65); LiDFOB (ref. 98); Formation of a resilient SEI on
Li2B12F9H3 (ref. 63) carbon anode
Lithium tetrafluoro oxalate Surface protection on both
phosphate (LiTFOP) (ref. 61); lithium electrodes
(fluorosulfonyl)(nonafluorobutane)
sulfonimide (LiFNFSI) (ref. 108)

Solvents and co-solvents Solvents and co-solvents

Diethyl carbonate (DEC) (ref. 64);
Ethyl methyl carbonate (EMC) (ref. 64);
Propylene carbonate (PC) (ref. 64)
Ethyl-2,2,2-trifluoroethyl carbonate (ETFEC)
(ref. 100); Propyl-2,2,2-trifluoroethyl
carbonate (PTFEC) (ref. 100)
Ethyl butyrate (EB) (ref. 66);
γ-butyrolactone (γBL) (ref. 101);
Methyl acetate (MA) (ref. 66);
Ethyl acetate (EA)(ref. 66);
Ethyl propionate (EP)(ref. 66)
Decrease in melting point of
electrolyte
Formation of a more conductive
SEI and decrease in Rct
Formation of a more conductive
SEI; decrease in melting point of
electrolyte and Rct
Fluorinated carbonates: FEC(ref. 68); Increase in thermal stability
F-AEC (ref. 69) towards charged electrodes
Ionic liquids (refs 46,70) Decrease in electrolyte
Phosphates: DMMEMP (ref. 109); flammability
DEMEMP (ref. 109)

Effects of low temperature on cathodes. Despite the limited knowl-
edge of the exact mechanisms responsible for the decreased perfor-
mance of LIBs at low temperatures, intrinsic kinetic constraints might
partially account for the technical challenges. The charge–transfer
resistance (Rct) for LFP at –20 °C has been reported to be >300%
higher than at room temperature22, invariably affecting rate perfor-
mance. Additionally, Zhang et al. observed that the Rct for cathode
half-cells at –30 °C is highly dependent on state-of-charge, being
much higher in the fully-lithiated state23. This resonates with the
observation by other groups that, even at moderate rates, charging
batteries at low temperatures is more challenging than discharging24.
Doping has been identified as an alternative to improve kinetics
at low temperature. Cobalt doping in lithium-rich layered materi-
als (Li1.2Ni0.2Mn0.6O2), for instance, contributes to decrease the
acti- vation energy for Ni4+ and Co4+ reduction during cell
discharge, improving capacity retention at –20 °C (ref. 25).
Another example is NCM523 (Fig. 1j), in which surface
substitution with Al, even at ppm level, improved the low
temperature performance. Selective substitution of ions, either at
the Li or transition metal site, have also been successful in
improving both ionic and electronic conductivity at extremely cold
conditions26.
Difficulties in maintaining performance below 0 °C are also asso-
ciated with high grain-boundary resistance and slow Li+ solid-state
diffusion within the active material. The low activation energy for
diffusion exhibited by certain materials can be beneficial for deploy-
ment in sub-zero conditions. LVP, for instance, has an activation
energy of 6.57 kJmol–1, more than seven times lower than LFP,
resulting in almost 300% higher relative capacity retention at –20 °C
(ref. 27). More generally, reducing particle size to the nanoscale,
and conformally coating highly conductive carbon/polymer on the
particles’ surface can be effective in decreasing the diffusional path
length and the inter-grain resistance, respectively. As an example,
Li0.99La0.01Fe0.9Mg0.1PO4/carbon aerogel composites exhibited out-
standing performance at –20 °C, displaying successful operation
from 1C to 50C rate with specific capacities of 120.3 mAhg –1 and
56.6 mAhg–1, respectively28. Although the fabrication of such porous
structures is interesting to probe the intrinsic performance limits of
cathode materials, it drastically reduces the volumetric energy den-
sity of the cell and thus would have limited practical application.
Anode materials
Negative electrodes for LIBs may present lithiation potentials below
that of reductive decomposition of organic electrolytes, giving rise
to a rich and complex interfacial chemistry. Common anode materi-
als, such as graphite and silicon, rely on the formation of effective
passivation layers to operate, and damage to these heterogeneous
REVIEW ARTICLE
NATURE ENERGY 2, 17108 (2017) | DOI: 10.1038/nenergy.2017.108 | www.nature.com/natureenergy
NATURE ENERGY
5

©
2
0
1
7
M
a
c
im
lla
n
P
u
b
ils
h
re
s
L
im
it
e
d
,
p
a
rt
o
fp
S
ri
n
g
re
N
a
t
u
r
e
.
A
ll
ri
g
h
ts
r
e
s
e
r
v
e
d
.
NATURE ENERGY
structures renders the interfaces vulnerable at elevated tempera-
tures. Temperature rise also accelerates side reactions, reducing cou-
lombic efficiency and cycle life1. Conversely, these layers become less
permeable to Li+ at low temperatures, slowing down cell kinetics.
While a significant fraction of the literature on temperature effects
on cathodes focuses on the investigation of intrinsic structural sta-
bility of the charged state, interfacial phenomena seems to dominate
the discussion on the anode side.
Graphite. Many of the problems faced by graphite in extreme
environments can be attributed to the solid electrolyte interphase
(SEI; Fig. 2a–f). Its morphology, composition and passivation prop-
erties are dependent on cell temperature during both formation and
cycling. These observations are in consonance with many reports
on SEI evolution after exposure to high temperatures, exhibiting
increased thickness and, consequently, larger cell impedance, which
can rise by 50% upon aging at 70 °C (ref. 29). Exposition of the cell
to high temperatures for extended times leads to a decrease in the
organic content in the SEI, making it more brittle30.
The SEI might also undergo transformation due to the action of
decomposition products from the positive electrode. For example,
Mn-rich cathodes can suffer from metal dissolution at high tem-
peratures, and these metal ions migrate towards the anode and pen-
etrate the SEI, creating electronically conductive zones31. It has been
proposed that Mn2+ can alter SEI composition by engaging in redox
reactions with some of its native components, forming MnF2 and
MnCO3(ref. 32). In both scenarios, the final effect is detrimental to
the passivation properties of the layer, leading to additional irrevers-
ible reactions with the electrolyte.
While stabilization at elevated temperatures relies essentially
on SEI engineering, the issues arising in colder conditions seem to
be related to intrinsic material properties. At these conditions, cell
impedance rises quickly, with growing contribution from charge
transfer and SEI to the total resistance29,33. Additionally, solid-state
diffusion of Li+ in graphite plummets below 0 °C (ref. 34). Although
there has been discussion on which of these factors becomes rate-
limiting35, the net result is a slow-down of electrode kinetics and a
rapid decrease in rate capability. Zhang et al. have also shown that
the gap between diffusion coefficients of Li+ in the charged and dis-
charged states widens at low temperatures34 (Fig. 2g). Consequently,
ion motion is slower during lithiation, which may be partially due
to the inherent phase instability of lithiated graphite at low tem-
peratures. The combination of all these factors also contributes to
increased cell polarization during lithiation, favouring the deposi-
tion of metallic lithium at the anode29.
Overall, graphite has exhibited fragility at both ends of the
temperature spectrum, and applications aiming specifically for
high or low temperatures could benefit from exploring other car-
bon structures. Hard carbons, for instance, have shown promis-
ing stability at extreme temperatures 29,36 (Fig. 2h) due to their
higher resistance to solvent co-intercalation. Although the amount
of research available for these systems is limited, alternative car-
bon materials hold potential to improve battery performance in
demanding environments.
Spinel lithium titanium oxide. Electrochemical performance
of lithium titanium oxide (Li4Ti5O12, LTO) at high temperature is
dictated by its reactivity with the electrolyte, even with its crystal
structure being stable up to 1000 °C. Despite the lack of consensus
on whether and how a permanent SEI-like layer forms on LTO37,
there is recent evidence of deposition of thick, inorganic-rich films
at high temperatures, triggered by thermally activated side reac-
tions37,38. These chemical processes also result in gas evolution,
which could damage the battery packaging 39. Coating LTO par-
ticles with carbon 39, ceramics40 and polymers41 was shown to be
successful in suppressing these decomposition events, leading to
REVIEW ARTICLE
improved cycle life at 55 °C (ref. 41). LTO’s performance at elevated
temperatures is expected to be solvent-dependent, and reactiv-
ity with polymer electrolytes and ionic liquids has yet to be fully
explored. Reliable cycling properties demonstrated at 150 °C using
quasi-solid electrolytes17,42 suggest that, in these cases, thermal
stability of LTO may surpass the one achieved with conventional
carbonate-based electrolyte.
At lower temperatures, the high lithiation voltage of LTO
(~1.55 V versus Li+/Li) makes it safer than graphite, due to the
unlikeliness of lithium plating. Well-established procedures to pro-
duce LTO in nanoparticle form have succeeded in retaining power
density below 0 °C, by decreasing the Li+ diffusion path. The lower
activation energy for charge transfer in LTO43 is also a potential
advantage at sub-zero conditions: while graphite’s SEI has been
shown to grow increasingly blocking at low temperatures33, the sur-
face layers that may exist on LTO do not seem to exhibit a distinc-
tive electrochemical signature, thus not contributing to the energy
barrier for Li+ transport.
LTO has been the material of choice for commercial batteries
devised for applications with stringent power requirements. Granted
the low energy density is not a decisive factor, the well-known resil-
ience of LTO can also direct its use at temperature extremes: the
level of control on both particle and coating dimensions, along with
the apparent inexistence of a blocking SEI, improves its potential for
applications in demanding environments.
Silicon. Although the onset for phase transition of Li–Si alloys is
low compared to other materials (~175–190 °C)44, it is still compat-
ible with the range expected for most applications. When lithiated
Si is exposed to elevated temperatures, faster diffusion of electro-
lyte trapped within the SEI facilitates further reaction of Li in the
electrode, increasing the thickness of SEI and resulting in capacity
loss45. In contrast to graphite, the interphase layer on Si has higher
inorganic character, and combination of high temperatures and
enormous volume changes during lithiation may lead to cracking of
this brittle layer. Simultaneously, the native oxide on Si surface can
react exothermically with fresh electrolyte, producing a secondary
SEI. Proper electrolyte additives have been found to produce a more
resilient layer and improve the cell performance. Blends of ionic liq-
uids with organic additives have also been shown to stabilize the SEI
at temperatures as high as 100 °C (ref. 46).
Coating carbon on Si is another approach for enhanced capac-
ity retention, since SEI formation is regulated by the outer car-
bon which prevents direct contact of silicon with the electrolyte 47.
Ceramic-based and metal-based coatings were also found to stabi-
lize silicon nanostructures, as the surface layer acts as a barrier that
minimizes side reactions and SEI growth, and assists in maintaining
the electrode’s mechanical integrity upon repeated lithiation/del-
ithiation6. The presence of a 1.5 nm TiO 2 layer within and around
Si nanotubes, for example, has been shown to largely improve cou-
lombic efficiency of the cell, and to extend capacity retention by 50%
in comparison with neat silicon 48. This combination of minimized
SEI formation, decreased surface reactivity and mechanical buffer-
ing granted by ceramic coatings would make it a viable option for
stabilizing silicon anodes at elevated temperatures.
Although a recent report suggests that the impact of low operat-
ing temperature on lithiation kinetics and diffusion in Si is milder
than in graphite49, it is possible that the lower active material uti-
lization may increase anisotropic strain concentration, leading to
particle cracking 50. Use of silicon particles with high surface area
and conductive coatings minimize some of these issues by reducing
mean Li+ diffusion path and enhancing electron transfer. However,
experience with graphite shows that a high surface area fosters
excessive SEI formation, which can exhibit low ionic permittivity
in cold environments. Also, like graphite, deep discharge in silicon
might show susceptibility to lithium plating.

6 NATURE ENERGY 2, 17108 (2017) | DOI: 10.1038/nenergy.2017.108 | www.nature.com/natureenergy

©
2
0
1
7
M
a
c
im
lla
n
P
u
b
ils
h
re
s
L
im
it
e
d
,
p
a
rt
o
fp
S
ri
n
g
re
N
a
t
u
r
e
.
A
ll
ri
g
h
ts
r
e
s
e
r
v
e
d
.
REVIEW ARTICLE NATURE ENERGY

a c d Electrolyte
additives
LiPF6, Rb
LiPF6, Rct
LiBF4, Rb 55 ºC
4 LiBF4, Rct

LMO cathode

Graphite anode
Log(R/K)

Mn4+
Inert SEI
Electrode structural Passivation
0 degradation
–60 –40 –20 0 20
T (oC)
b e
250

4 200
12

Capacity (mAh)
Voltage (V)

150
3

1 LiPF6, 20 ºC 100
3 4 LiPF6
2 2 LiBF4, 20 ºC
3 LiPF6, –30 ºC Li2B12F9H3/LiDFOB
4 LiBF4, –30 ºC 50 LiPF6/LiDFOB
1
0 0.3 0.6 0.9 0
Relative capacity (%) 0 300 600 900 1,200 1,500
Cycle number

Figure 3 | Tailoring salts and electrolyte additives to extend the operating temperature of Li-ion batteries. a,b, The decisive effect of charge transfer.
Although solutions of LiPF6 in alkyl-carbonates exhibit higher ionic conductivities (Rb) than when using LiBF4 (a), electrolytes containing the tetrafluoroborate
anion present lower charge-transfer resistances (Rct) throughout the entire temperature range, translating into larger capacity retention (b) at low
temperatures. c,d, Pictorial representation of dual protection activity by DFDEC and vinylene carbonate (VC) additives in Li1.13Mn0.463Ni0.203Co0.203O2/
graphite full-cells at elevated temperatures: manganese ions originating from the cathode can introduce defects at the SEI (c). The combination of these two
electrolyte additives can produce protective layers on both electrodes, suppressing metal dissolution on the cathode and forming a stable SEI on the anode
(d)106. e, Discharge capacity of Li1.1[Ni1/3Mn1/3Co1/3]0.9O2/graphite cells cycled at C/3 at 55 °C. The cyclic stability is largely enhanced using Li2B12F9H3 as a salt
and LiDFOB as additive. Panels reproduced with permission from: a,b, ref. 56 (Elsevier); e, ref. 63 (Macmillan Publishers Ltd).

The commercial success of Si anodes in bulk LIBs depend on
many factors, including the cost-effective fabrication of surface-
protected nanostructures. While the current industrial trend of
incorporating silicon–graphite composite electrodes would ease
some economical hurdles, it does not mitigate all challenges already
faced by graphite at extreme temperatures.
Lithium metal. Metallic lithium was long forsaken in commercial
devices due to low coulombic efficiency and safety concerns origi-
nating from dendrite growth. Nevertheless, it has been intensively
revisited in the past few years, for its high charge density and the
advent of other Li-based battery chemistries.
Lithium has been largely tested at 60 °C–90 °C in combination
with polymer electrolytes51, showing that operation at elevated
temperatures is not prohibited. More recently, it has been used at
temperatures as high as 150 °C in LIBs containing ionic liquids 17,
highlighting its stability. Elevated temperatures have been shown
to improve plating/stripping efficiency and to reduce the inci-
dence of dendritic deposition52. While the melting point of lithium
(~180 °C) imposes an intrinsic upper temperature limit for cells,
lithium-metal batteries would have more practical challenges in
the low temperature regime. Despite the limited commercial suc-
cess through the years, lithium metal anodes are currently used in
electric vehicles developed by Blue Solutions both in Europe and in
the United States, and the intense research activity on this topic can
be projected to materialize into new ventures in the years to come.
Other materials. Several metallic-based53, intermetallic-based54
and chalcogenide-based55 systems have been investigated as alter-
native high-capacity anodes for LIBs. They all share the feature
of undergoing significant changes in particle morphology upon
lithiation, contributing to a large voltage hysteresis during cycling.
Reaction kinetics are notably slow for such systems, and operation
at higher temperatures can be an effective way of reducing the volt-
age gap between charge and discharge. Conversely, although reports
on temperature-related effects in these materials are scarce, one can
expect cell polarization to be higher at low temperatures, increasing
the already-problematic energy inefficiency during cycling. Among
the explored systems, Sn-based intermetallics, such as Cu6Sn5 and
SnSb, were observed to have greater tolerance for both thermally-
induced and electrochemically-induced lattice volume changes than
pure metals, reducing potential hysteresis53.
Electrolyte
The rich variety of physical- and electro-chemical events in the elec-
trolyte at both temperature extremes raises multifaceted challenges.
At low temperatures, rapid viscosity increase has negative effects on
ion mobility and electrode wettability, reducing rate performance56.
Moreover, reduced conductivity causes Li+ depletion at the electrode
vicinity, which may ultimately favour lithium plating29. At high tem-
peratures, the electrolyte is susceptible to chemical transformations,
including disproportionation of PF6– in the presence of traces of
protic impurities57, and reactions with the charged electrodes58.

NATURE ENERGY 2, 17108 (2017) | DOI: 10.1038/nenergy.2017.108 | www.nature.com/natureenergy 7

©
2
0
1
7
M
a
c
im
lla
n
P
u
b
ils
h
re
s
L
im
it
e
d
,
p
a
rt
o
fp
S
ri
n
g
re
N
a
t
u
r
e
.
A
ll
ri
g
h
ts
r
e
s
e
r
v
e
d
.
NATURE ENERGY REVIEW ARTICLE
a b 100
(i)
10

0.1
100

dT/dt(°C min–1)
10 (ii)
1

0.1
100
(iii)
10

0.1
100 150 200 250 300 350
Temperature (°C)

c d e
1.2 1.2 200
Discharge capacity retention (Cx/C1)

Discharge capacity retention (Cx/C1)

150 °C
1.0 1.0
150

Capacity (mAh g–1)

0.8 0.8

0.6 0.6 100

Delithiation
0.4 0.4 Lithiation
50
25 °C 0 °C 25 °C 60 °C
0.2 0.2 C/2 rate

0.0 0.0 0
0 5 10 15 20 25 0 5 10 15 20 25 0 10 20 30 40
Cycle number Cycle number Cycle number

Figure 4 | Ionic liquids in perspective. a, Imidazolium-based ionic liquid before heating (left), and heated at 250 °C for 12 hours under vacuum (centre)
and N2 atmosphere (right). The effective thermal stability of these solvents can be much lower than suggested by thermogravimetric experiments.
b, Self-heating rate as a function of temperature for lithiated LTO in contact with EMI-FSI (i), EMI-TFSI (ii) and BMMI-TFSI (iii). The dashed lines were
obtained using ethylene carbonate, for comparison. Although certain RTIL compositions indeed increase battery safety, certain combinations can be more
reactive towards charged electrodes than carbonate-based solvents. c–d, High performance RTIL-based electrolytes: using LiTFSI/EMI-FSI (circles) as
electrolyte in graphite half-cells can provide capacities superior to LiPF6/EC+DMC (triangles) both at low (c) and high (d) temperatures. e, Cyclic stability
at 150 °C for a LTO half-cell using a hexagonal boron nitride-based quasi-solid electrolyte. These composites can support stable cell operation over a broad
temperature range, from 24 °C to 150 °C. Panels reproduced with permission from: a, ref. 80 (Elsevier); b, ref. 58 (Elsevier); c, ref. 82 (Elsevier);
d, ref. 41 (ACS); e, ref. 17 (Wiley).

Investigation of multifunctional additives and co-solvents has At high temperatures, additives such as LiTFOP (ref. 61) and
been very promising, extending LIB’s operability to both higher and FSE62 contribute to the construction of a thermally resilient SEI,
lower temperatures. There have also been intense efforts in develop- while also casting a protective coating on the cathode. The outcome
ing alternative electrolytes with enhanced thermal stability, includ- of this dual-protection mechanism (Fig. 3c,d) is the minimization of
ing ceramics and ionic liquids. The electrolyte is the sole element of structural decay on both electrodes, enhancing the cycle life when
the battery in physical contact with all other components of the cell, the battery is stored or tested at 60 °C. Many compounds initially
and the complexity of the processes it might engage in highlights the proposed as new salts, such as LiBOB and LiDFOB, have also been
challenges imposed by modern battery science. tested as additives, since their film-forming properties can benefit
the cell even at low concentrations63 (Fig. 3e). Overall, the screening
Additives and lithium salts. The efficacy of additives to enhance of new salts and additives has been very successful, and it is reason-
the battery’s thermal tolerance mainly rests on their ability to mod- able to expect it to continue having immense impact in managing
ify the surface of cathodes and anodes, and to facilitate the phase temperature-related problems in LIBs.
transfer of Li+. Recent highlights in this area are summarized in
Table 2. Since ion transport through SEI slows down below room Organic solvents and co-solvents. The battery community has long
temperature, the formation of more compact and conductive pas- realized the usefulness of combining different co-solvents to add
sivation films, as provided by FEC 59 and butyl sultone60, is highly functionalities to electrolytes (Table 2). A common approach for low
beneficial to rate performance. Optimizing Rcthas also been shown temperature applications is the addition of low-melting-point, low-
to be effective: although electrolytes containing LiBF4 exhibit lower viscosity solvents, such as EMC and PC64, to extend the electrolyte’s
ionic conductivities than with LiPF6 (Fig. 3a), the lower Rct achieved liquid range and assure good ionic conductivity. Recently, this func-
with tetrafluoroborate anions lead to better capacity retention tion has been further combined with the ability to produce a more
at –30 °C (Fig. 3b)56. conductive SEI and to decrease Rct, for which γ-butyrolactone65 and

8 NATURE ENERGY 2, 17108 (2017) | DOI: 10.1038/nenergy.2017.108 | www.nature.com/natureenergy

©
2
0
1
7
M
a
c
im
lla
n
P
u
b
ils
h
re
s
L
im
it
e
d
,
p
a
rt
o
fp
S
ri
n
g
re
N
a
t
u
r
e
.
A
ll
ri
g
h
ts
r
e
s
e
r
v
e
d
.
REVIEW ARTICLE NATURE ENERGY

a b
LiTFSI LiPF 6 LiBF4
14
0.5 LiTFSI
55 °C 85 °C 85 °C LiBETI
12
Specific current (mA cm–2)

Current density (mA cm–2)

0.0 LiPF6
10
LiBF4
1.5 100 °C 1st 8
2nd
1.0 3rd 6

4
0.5 65 °C
100 °C 2
0.0
0
3 4 5 3 4 5 3 4 5 40 60 80 100 120
Potential (V) versus Li/Li + Temperature (°C)

Figure 5 | Aluminium corrosion in a pyrrolidinium-based ionic liquid at elevated temperatures. a, Cyclic voltammograms of Al/Li cells containing
different electrolytes. The passivation properties are highly dependent on temperature. b, Linear sweep thermometry obtained with the use of different
lithium salts, and voltage held at 5.0 V. The onset for aluminium corrosion occurred at different temperatures, being minimum for LiTFSI and maximum for
LiBF4. Figure 5 reproduced with permission from ref. 93 (Electrochemical Society).

short-chain acetates66 have been quite successful. Smart et al.67 have
also shown that addition of 20vol% methyl propionate or ethyl pro-
pionate to EC:DMC mixtures provides over 70% capacity retention
when the cell operates at –60 °C.
On the high-temperature end, the strategy has been focused
on reducing electrolyte flammability and improving its stability
towards charged electrodes. Much work has been done in explor-
ing fluorinated carbonates as co-solvents, since the addition of fluo-
rine moieties increases thermal resilience and largely improves the
anodic stability of these compounds. Consequently, solvents such
as FEC68 and F-AEC69 have been found to drastically improve safety
and cycle life at 55 °C. Ionic liquids have also been investigated
as co-solvents, usually in combination with PC, to reduce flam-
mability70 and provide additional building blocks for a thermally
resilient SEI46.
Based on the current state of the literature, it is realistic to pro-
ject that these solvents could enable satisfactory battery operation
under very cold environments. However, intrinsic limitations on
their physical properties — such as flammability, high vapour pres-
sure and low flash point — limit the ability of these compositions to
sustain long term storage and cycling beyond 80 °C.
Solid electrolytes. The absence of volatile or flammable compounds
is expected to make solid electrolytes safer than their liquid coun-
terparts at elevated temperatures. Besides thermal stability, solid
polymer electrolytes (SPEs) and ceramic conductors offer attractive
design possibilities, enabling thin film batteries and easy integration
into microdevices.
Although electrochemical testing beyond 60 °C is expedient to
probe properties of SPEs, these materials might present issues at
temperature extremes. For example, slow recrystallization kinetics of
the ubiquitous poly(ethylene oxide) (PEO) renders ionic conductiv-
ity highly dependent on the thermal history of the films71. Recently,
adoption of porous structures of thermally stable polyimide 72 and
poly(tetrafluoroethylene) 73 as skeletons for PEO-succinonitrile-
LiBOB composites provided good results, with conductivities stable
for at least 120 h at 160 °C. Unfortunately, these latter studies did not
present further testing on a cell level, and the actual performance
is unknown. Notwithstanding the uncertainties in SPE behaviour
beyond 100 °C, these materials have been successful in extending
battery operation to 90 °C (refs 51,74). Recent commercialization
of such SPE-based batteries was implemented in the Blue Car/Bus
project by the Bolloré Group, wherein battery packs are heated and
operated between 60–80 °C.
The thermal and chemical stability of ceramics has also spurred
interest in utilizing crystalline ion conductors as electrolytes. Oxides
and phosphates are natural candidates for enabling safe operation
of LIBs in extreme environments, due to their intrinsic structural
stability at high temperatures75. Recently, LiPON has been shown
to be stable up to 200 °C (ref. 76), although the cell exhibited less
capacity than at lower temperatures and long-term performance
was not reported. The sulfide glass 95(0.7Li2S–0.3P2S5)–5Li3PO4 is
also known to provide stable cycling at 100 °C (ref. 77), despite the
questionable environmental inertness of this family of conductors78.
Although the resilience of ceramics is not to be disputed, the ability
of these rather brittle materials to sustain thermal cycling in a bat-
tery has yet to be confirmed.
Reports on the performance of solid electrolytes at low tem-
peratures are scarce in the literature. It can be anticipated that the
large interfacial resistances79 existing in most all-solid-state cells are
a major challenge for retaining power capability in cold environ-
ments. The ionic conductivities, already low at room temperature,
also plunge as temperature decreases. The transition to solid elec-
trolytes is a possible way for LIBs to move forward, but the variety of
challenges to be solved to achieve peak performance, even at room
temperature, makes the complete substitution of liquid electrolytes
a task hard to undertake soon.
Room-temperature ionic liquids. These hold the promise of pro-
ducing safer devices and enabling battery operation at very high
temperatures. The main appeal of these compounds is their non-
flammability and negligible vapour pressure within a wide tempera-
ture range17. Early assessment of their thermal stability neglected
ageing effects on the structural integrity. Albeit not producing
significant amounts of volatile products until ~300 °C, it has been
shown that partial decomposition of the cation can occur after
extended heat treatments at temperatures as low as 140 °C (ref. 80)
(Fig. 4a). Another important point is that their effective thermal sta-
bility when in contact with charged electrodes might divert from the
inertness displayed by the pure solvents. Wang et al. (ref. 58) have
shown that, for certain room-temperature ionic liquids (RTILs),
onset for thermal runaway can actually be lower than for carbonate-
based electrolytes (Fig. 4b). Although this property varies widely
with RTIL structure, TFSI anions and some onium cations are in
fact more stable than organic carbonates.
Besides the elevated cost of high-purity RTILs, they are down-
played for their performance at room temperatures and lower, due
to their high viscosity, large R ct (ref. 46) and low Li+ transference

NATURE ENERGY 2, 17108 (2017) | DOI: 10.1038/nenergy.2017.108 | www.nature.com/natureenergy 9

©
2
0
1
7
M
a
c
im
lla
n
P
u
b
ils
h
re
s
L
im
it
e
d
,
p
a
rt
o
fp
S
ri
n
g
re
N
a
t
u
r
e
.
A
ll
ri
g
h
ts
r
e
s
e
r
v
e
d
.
NATURE ENERGY REVIEW ARTICLE
a
750

700

650

Na-S

Energy density (Wh kg-1)

150
Li-ion

Li-ion polymer

Molten
100
salt
Ni-MH

Li-metal
polymer
50
Lead acid

0
–40 –20 0 20 40 60 80 100 300 400
Temperature (ºC)
b
Consumer electronics Military Liquid Solid ceramic Solid polymer
300

PEO-polyimide (ref. 72)

LiPON (ref. 76)
es (ref. 116)
250
With molten salt (ref. 20)

With RTIL (ref. 17)

PEO-PTFE (ref. 73)

With RTIL (ref. 46)
Melting point
3D carbon decorated (ref. 114)

PP13-TFSI (ref. 17)

Polymer electrolyte (ref. 51)
Organic electrolyte (ref. 97)

luorinated carbonates (ref. 69)

Mixture of carbonates (ref. 64)
With RTIL (ref. 17)

With LiBOB (ref. 94)

With LiPF6 salt (ref. 112)

Carbon-free electrode (ref. 115)

Amorphous monolithic (ref. 49)

Full cell with LMO (ref. 113)

200 With PC-RTIL (ref. 46)

Commercial grade (ref. 111)

77)
Carbon coated (ref. 27)
AlPO4 coated (ref. 7)

With RTIL (ref. 82)

Graphite mixture (ref. 29)

Hard carbon (ref. 36)

Lactones (ref. 101)

oated (ref. 97)
ed (ref. 94)

Ph
With FEC (ref. 68)

30)
With LTO (ref. 113)

With VC (ref. 23)

With LTO (ref. 9)

150 os .

EMI-FSI (ref. 82)

ef

PEO (ref. 51)

Esters (ref. 67)

(r
Temperature (ºC)

ph
oa
tc c PV
at
100 O DF
T
L .
ef
(r
Ca F s
rb
glas
50 on
Sulfide

–50

Cathodes Anodes Electrolytes

Figure 6 | Reported thermal limits for Li-ion battery materials and rechargeable battery systems. a, Typical energy densities and temperature windows of
commercially available rechargeable batteries. b, Summary of the operating temperatures achieved with different cathode, anode and electrolyte materials.
The diversity of LIB-compatible compositions is an asset to the development of thermally resilient devices. Data are from refs 7, 9, 10, 12, 14, 15, 17, 28, 31,
35, 36, 41, 51, 54, 56, 70, 73–75, 80, 89, 90, 93, 94, 100, 111–116 as indicated in the figure.

number17,46,81. Strategies to enable performance at lower tempera- However, practical battery operation in such cold environments has
tures involve optimizing charge-transfer kinetics46,81. Recent devel- yet to be demonstrated.
opment of highly conductive FSI-based RTILs has also improved At the high-temperature end, exploitation of the possibilities ena-
rate performance (Fig. 4c,d), and addition of LiBOB was found bled by RTILs are limited by the lack of state-of-the-art thermally
to be especially effective at improving capacity retention at 0 °C stable separators. Although RTIL-based polymer gels have been
(ref. 82). Although many RTILs freeze below –10 °C, operability at explored85, rapid decay of mechanical properties with temperature
lower temperatures can be explored by forming eutectic mixtures do little to extend the operating range. Performance in more extreme
with suitable combinations of ionic liquids and lithium salts83,84. conditions has been recently demonstrated by immobilizing RTILs

10 NATURE ENERGY 2, 17108 (2017) | DOI: 10.1038/nenergy.2017.108 | www.nature.com/natureenergy

©
2
0
1
7
M
a
c
im
lla
n
P
u
b
ils
h
re
s
L
im
it
e
d
,
p
a
rt
o
fp
S
ri
n
g
re
N
a
t
u
r
e
.
A
ll
ri
g
h
ts
r
e
s
e
r
v
e
d
.
REVIEW ARTICLE NATURE ENERGY

in ceramic materials, as natural clay42 and hexagonal boron nitride17
(h-BN). The combination of h-BN and a piperidinium-based RTIL
enabled LTO half-cells to operate in the 24–150 °C range, with good
capacity retention and rate capability during extended cycling at
high temperatures17 (Fig. 4e). Despite the many exciting develop-
ments during the last decade, RTILs share a common challenge with
SPEs: compositions suitable for deployment at high temperatures
often underperform in milder environments, restricting them to
niche applications.
Separators, binders and current collectors
Although temperature effects are less studied in these electrochemi-
cally inactive components, their state-of-health is essential to sus-
tain cell operation.
Separators. Serving as a physical barrier between electrodes, tradi-
tional polyolefin-based separators also feature a ‘shut-down’ mecha-
nism when exposed to overheating and/or overcharge. A practical
approach to enhance their thermal stability is to coat inorganic nan-
oparticles in the separator films. Several studies encompassed the
use of ceramic materials, such as SiO2 and Al2O3, to enhance surface
morphology, reducing thermal shrinkage at elevated temperatures
and improving ionic conductivities86.
Thermosetting polymers have also been investigated as alterna-
tive separators. The thermal and chemical stability of polyimides,
especially when cross-linked into three-dimensional networks,
makes them especially attractive for high-temperature applications.
Jiang et al. demonstrated that electrospun nano-fibrous polyim-
ide membranes exhibited no thermal shrinkage at 150 °C (ref. 87).
Lee etal. havealsoreported thatcoating polyimide fiberswithalumina
is beneficial for the separator’s wetting and transport properties88.
Binders. Mechanical properties of commonly used poly(vinylene
fluoride) (PVDF) are compromised at elevated temperatures, as
the polymer migrates to the electrode surface at 120 °C, affect-
ing structural integrity 30,37. Yan et al. demonstrated that polyim-
ide can be synthesized via a one-pot solution polycondensation,
and incorporated into a graphite anode 89. Such cell exhibited rate
capability and cycling behaviour is superior to that of conventional
binders. Despite the electrochemical characterization of this work
being carried out at room temperature, the use of this thermally
stable polymer lends itself as an attractive binder candidate for
extreme environments.
Current collector. Copper has been suggested to be susceptible to
oxidation by impurities in the electrolyte90, and this process can be
expected to be accelerated beyond 50 °C, both by enhanced kinet-
ics and decay of LiPF6 into HF57. Although the anode current col-
lector (CC) has not received much attention, aluminium has been
deeply investigated, with the main goal of supporting high voltage
cathodes. Little attention, however, has been paid to temperature-
related effects on the passivation properties of this interface.
A possible problem at elevated temperatures is the loss of pas-
sivation properties due to cracking and dissolution of its com-
ponents. Intrinsic electrolyte limitations have made it difficult to
probe specific effects on the CC, especially since some thermally
stable salts, such as LiTFSI, seem to be unable to protect alumin-
ium beyond 3.7 V versus Li+/Li (ref. 91). Nevertheless, carbonate-
based electrolytes containing LiTFSI-LiBOB mixtures exhibited
promising passivating behaviour up to 60 °C (ref. 92).
The lack of volatility of RTILs make them ideal systems to probe
temperature-related phenomena on CCs. As shown in Fig. 5,
anodic stability of aluminium in a pyrrolidinium-based RTIL
varies widely both with the salt of choice and the operating tem-
perature93. A highlight is LiBF4, which seems able to support 5 V
chemistries beyond 85 °C. Although this knowledge cannot be
directly transferred to alkyl-carbonate systems, it serves the pur-
pose of raising awareness of the fact that passivation behaviour can
be very dynamic.
Outlook
Battery science has always been a game of trade-offs, and this
also holds when considering thermal factors. Figure 6a compares
energy density of commercial rechargeable batteries against their
operable temperature window. Despite LIB’s clear prominence,
the environmental frontiers for these batteries are still modest
compared to the possibilities offered by its core materials. Many
reports focus on extending the thermal usability of individual
components, and Fig. 6b maps some of the standard and prom-
ising materials against their temperature limits, each specifically
discussed in this review.
In designing an optimum battery, materials selection is obvi-
ously made with compatibility in mind. Apart from immediate
reactivity, slow transformations should be considered before suc-
cessful implementation, and being mindful of the interfacial reac-
tions between electrolytes and electrodes at every state-of-charge
is essential. Fluorination of alkyl-carbonates enhances the compat-
ibility between electrolyte and charged electrodes, whereas salts
such as LiBOB and LiDFOB confer various benefits to the thermal
resilience of the cell. Phosphate-based cathodes are particularly
stable at high temperatures, with LiFePO 4 already transitioning
into the consumer’s market. LVP (Li3V2(PO4)3) is another interest-
ing material to be watched, since it inherits the sturdiness from
phosphates while also operating at higher voltages than its ferrous
relative and presenting better low-temperature performance.
On the other hand, reactions at the anode are highly dependent
on the quality and quantity of the SEI formed. The usability of a bat-
tery is dictated by the nature and evolution of this passivation layer
under the operating temperature scenarios. Li + transport through
SEI is one of the major limiting factors at low temperatures, and
eventually favours lithium plating during cell charging. The choice
of cathode could also influence the anode passivation, due to the
presence of minute dissolved ions from its structure. Formation
protocols, which include pretreatments and use of specific addi-
tives, have been a closely guarded know-how of the industry,
and it is uncertain how they will adapt to accommodate thermal
extremes. Diagnostic analyses are essential to rationalize perfor-
mance decay, while systematic studies on the correlation between
thermal history and state-of-health are necessary to project long-
term behaviour. While traditional graphite electrodes have a high
degree of uncertainty at either of the thermal extremes, LTO, hard
carbons and metallic lithium seem to be promising alternatives.
The collective knowledge currently available on the subtleties of
LIBs is astounding, with an ever-growing library of materials being
incorporated to the field. However, commercial adoption of these
materials to engineer battery packs with thermal tolerance has a
long way to go. While most of these reports are impressive and pre-
sent valuable insights, they are usually focused on a single compo-
nent of a battery. In addition to the material properties, cell-level
performance is governed by secondary conditions and interactions
between its components. Industrial know-how clearly surpasses
open-sourced scientific reports in this area, and it is yet to be seen
how technology transitions between them.
Received 2 November 2016; accepted 4 June 2017;
published 24 July 2017
References
1. Tarascon, J. M. & Armand, M. Issues and challenges facing rechargeable
lithium batteries. Nature 414, 359–367 (2001).
2. Forgez, C., Vinh Do, D., Friedrich, G., Morcrette, M. & Delacourt, C.
Thermal modeling of a cylindrical LiFePO4/graphite lithium-ion battery.
J. Power Sour. 195, 2961–2968 (2010).

NATURE ENERGY 2, 17108 (2017) | DOI: 10.1038/nenergy.2017.108 | www.nature.com/natureenergy 11

©
2
0
1
7
M
a
c
im
lla
n
P
u
b
ils
h
re
s
L
im
it
e
d
,
p
a
rt
o
fp
S
ri
n
g
re
N
a
t
u
r
e
.
A
ll
ri
g
h
ts
r
e
s
e
r
v
e
d
.
NATURE ENERGY
3. Feng, X. et al. Thermal runaway features of large format prismatic lithium ion
battery using extended volume accelerating rate calorimetry. J. Power Sour.
255, 294–301 (2014).
4. Williard, N., He, W., Hendricks, C. & Pecht, M. Lessons learned from the 787
Dreamliner issue on lithium-ion battery reliability. Energies
6, 4682–4695 (2013).
5. Linden, D. R. & Thomas B. Handbook of Batteries. 3rd edn
(McGraw-Hill, 2002).
6. Mauger, A. & Julien, C. Surface modifications of electrode materials for
lithium-ion batteries: status and trends. Ionics 20, 751–787 (2014).
7. Cho, Y., Eom, J. & Cho, J. High performance LiCoO2 cathode materials at
60 °C for lithium secondary batteries prepared by the facile nanoscale
dry-coating method. J. Electrochem. Soc. 157, A617–A624 (2010).
8. Aurbach, D. et al. Review on electrode–electrolyte solution interactions,
related to cathode materials for Li-ion batteries. J. Power Sour.
165, 491–499 (2007).
9. Kong, D. P., Ping, P., Wang, Q. S. & Sun, J. H. Study on high
temperature stability of LiNi0.33Co0.33Mn0.33O2/Li4Ti5O12 cells from the safety
perspective. J. Electrochem. Soc. 163, A1697–A1704 (2016).
10. Yan, C., Xu, Y., Xia, J., Gong, C. & Chen, K. Tris(trimethylsilyl) borate
as an electrolyte additive for high-voltage lithium-ion batteries using
LiNi1/3Mn1/3Co1/3O2 cathode. J. Energy Chem. 25, 659–666 (2016).
11. Ma, L., Nie, M., Xia, J. & Dahn, J. A systematic study on the reactivity of
different grades of charged Li[NixMnyCoz]O2 with electrolyte at elevated
temperatures using accelerating rate calorimetry. J. Power Sour.
327, 145–150 (2016).
12. Bloom, I. et al. Effect of cathode composition on capacity fade, impedance rise
and power fade in high-power, lithium-ion cells. J. Power Sour.
124, 538–550 (2003).
13. Nitta, N., Wu, F., Lee, J. T. & Yushin, G. Li-ion battery materials: present and
future. Mater. Today 18, 252–264 (2015).
14. Huang, Y. et al. Thermal stability and reactivity of cathode materials for Li-ion
batteries. ACS Appl. Mater. Interfaces 8, 7013–7021 (2016).
The authors investigate the thermal stability of several cathode materials in
the delithated state, showing that layered cathodes are the least stable, and
phosphates the most.
15. Zhao, H. et al. Enhanced elevated-temperature performance of
LiAlxSi0.05Mg0.05Mn1.90–xO4 (0 ≤ x ≤ 0.08) cathode materials for high-
performance lithium-ion batteries. Electrochimica Acta 199, 18–26 (2016).
16. Sun, Y. K. et al. Nanostructured high-energy cathode materials for advanced
lithium batteries. Nat. Mater. 11, 942–947 (2012).
17. Rodrigues, M.-T. F. et al. Hexagonal boron nitride-based electrolyte composite
for Li-ion battery operation from room temperature to 150 °C.
Adv. Energy Mater. 6, 1600218 (2016).
This work demonstrates that ionic liquids can allow Li-ion batteries to
operate from room temperature to at least 150 °C, with relatively high
coulombic efficiencies.
18. Hagh, N. M. & Amatucci, G. G. Effect of cation and anion doping on
microstructure and electrochemical properties of the LiMn 1.5Ni0.5O4−δ spinel.
J. Power Sour. 256, 457–469 (2014).
19. Maccario, M., Croguennec, L., Le Cras, F. & Delmas, C. Electrochemical
performances in temperature for a C-containing LiFePO4 composite
synthesized at high temperature. J. Power Sour. 183, 411–417 (2008).
20. Muñoz-Rojas, D. et al. Development and implementation of a high
temperature electrochemical cell for lithium batteries. Electrochem. Commun.
9, 708–712 (2007).
21. Chang, H.-H. et al. Effects of TiO2 coating on high-temperature cycle
performance of LiFePO4-based lithium-ion batteries. J. Power Sour.
185, 466–472 (2008).
22. Gao, F. & Tang, Z. Kinetic behavior of LiFePO4/C cathode material for
lithium-ion batteries. Electrochimica Acta 53, 5071–5075 (2008).
23. Zhang, S. S., Xu, K. & Jow, T. R. The low temperature performance of Li-ion
batteries. J. Power Sour. 115, 137–140 (2003).
24. Fan, J. & Tan, S. Studies on charging lithium-ion cells at low temperatures.
J. Electrochem. Soc. 153, A1081–A1092 (2006).
25. Kou, J. et al. Role of cobalt content in improving the low-temperature
performance of layered lithium-rich cathode materials for lithium-ion
batteries. ACS Appl. Mater. Interfaces 7, 17910–17918 (2015).
26. Li, G. et al. Effect of trace Al surface doping on the structure, surface
chemistry and low temperature performance of LiNi0.5Co0.2Mn0.3O2 cathode.
Electrochimica Acta 212, 399–407 (2016).
27. Rui, X. H., Jin, Y., Feng, X. Y., Zhang, L. C. & Chen, C. H. A comparative
study on the low-temperature performance of LiFePO4/C and Li3V2(PO4)3/C
cathodes for lithium-ion batteries. J. Power Sour. 196, 2109–2114 (2011).
This work provides a detailed investigation on the origins of the enhanced
performance of LVP at –20 °C, highlighting the benefit of a low activation
energy for Li+ diffusion in the lattice to capacity retention.
REVIEW ARTICLE
28. Zhang, H., Xu, Y., Zhao, C., Yang, X. & Jiang, Q. Effects of carbon coating and
metal ions doping on low temperature electrochemical properties of LiFePO4
cathode material. Electrochimica Acta 83, 341–347 (2012).
29. Waldmann, T., Wilka, M., Kasper, M., Fleischhammer, M. &
Wohlfahrt-Mehrens, M. Temperature dependent ageing mechanisms in lithium-
ion batteries — a post-mortem study. J. Power Sour.
262, 129–135 (2014).
30. Bodenes, L. et al. Lithium secondary batteries working at very high
temperature: capacity fade and understanding of aging mechanisms.
J. Power Sour. 236, 265–275 (2013).
31. Jalkanen, K. et al. Cycle aging of commercial NMC/graphite pouch cells at
different temperatures. Appl. Energy 154, 160–172 (2015).
32. Shin, H., Park, J., Sastry, A. M. & Lu, W. Degradation of the solid electrolyte
interphase induced by the deposition of manganese ions. J. Power Sour.
284, 416–427 (2015).
33. Zhang, S. S., Xu, K. & Jow, T. R. Electrochemical impedance study on the low
temperature of Li-ion batteries. Electrochimica Acta
49, 1057–1061 (2004).
34. Zhang, S. S., Xu, K. & Jow, T. R. Low temperature performance of graphite
electrode in Li-ion cells. Electrochimica Acta 48, 241–246 (2002).
35. Huang, C. K., Sakamoto, J. S., Wolfenstine, J. & Surampudi, S. The limits of
low-temperature performance of Li-ion cells. J. Electrochem. Soc.
147, 2893–2896 (2000).
36. Zheng, H., Qu, Q., Zhang, L., Liu, G. & Battaglia, V. S. Hard carbon: a
promising lithium-ion battery anode for high temperature applications with
ionic electrolyte. Roy. Soc. Chem. Adv. 2, 4904–4912 (2012).
37. Gieu, J. B., Courrèges, C., El Ouatani, L., Tessier, C. & Martinez, H.
Temperature effects on Li4Ti5O12 electrode/electrolyte interfaces at the first
cycle: a X-ray photoelectron spectroscopy and scanning auger microscopy
study. J. Power Sour. 318, 291–301 (2016).
38. Nordh, T., Younesi, R., Brandell, D. & Edström, K. Depth profiling the solid
electrolyte interphase on lithium titanate (Li4Ti5O12) using synchrotron-based
photoelectron spectroscopy. J. Power Sour. 294, 173–179 (2015).
39. He, Y. B. et al. Gassing in Li4Ti5O12-based batteries and its remedy. Sci. Rep.
2, 913 (2012).
40. Han, C. et al. Suppression of interfacial reactions between Li4Ti5O12 electrode
and electrolyte solution via zinc oxide coating. Electrochimica Acta
157, 266–273 (2015).
41. Lu, Q. et al. A polyimide ion-conductive protection layer to suppress side
reactions on Li4Ti5O12 electrodes at elevated temperature. Roy. Soc. Chem. Adv.
4, 10280 (2014).
42. Kalaga, K. et al. Quasi-solid electrolytes for high temperature lithium ion
batteries. ACS Appl. Mater. Interfaces 7, 25777–25783 (2015).
43. Xu, K. & von Wald Cresce, A. Li+-solvation/desolvation dictates interphasial
processes on graphitic anode in Li ion cells. J. Mater. Res.
27, 2327–2341 (2012).
44. Wang, Y. & Dahn, J. Phase changes in electrochemically lithiated silicon at
elevated temperature. J. Electrochem. Soc. 153, A2314 (2006).
45. Fan, H. et al. Electrochemical properties and thermal stability of silicon
monoxide anode for rechargeable lithium-ion batteries. Electrochemistry
84, 574–577 (2016).
46. Ababtain, K. et al. Ionic liquid-organic carbonate electrolyte blends to stabilize
silicon electrodes for extending lithium ion battery operability to 100 °C.
ACS Appl. Mater. Interfaces 8, 15242–15249 (2016).
47. Profatilova, I. A. et al. Thermally induced reactions between lithiated nano-
silicon electrode and electrolyte for lithium-ion batteries. J. Electrochem. Soc.
159, A657–A663 (2012).
48. Lotfabad, E. M. et al. Si nanotubes ALD coated with TiO2, TiN or Al2O3 as
high performance lithium ion battery anodes. J. Mater. Chem. A
2, 2504–2516 (2014).
49. Markevich, E., Salitra, G. & Aurbach, D. Low temperature performance of
amorphous monolithic silicon anodes: comparative study of silicon and
graphite electrodes. J. Electrochem. Soc. 163, A2407–A2412 (2016).
50. An, Y. et al. Mitigating mechanical failure of crystalline silicon electrodes for
lithium batteries by morphological design. Phys. Chem. Chem. Phys.
17, 17718–17728 (2015).
51. Croce, F., Sacchetti, S. & Scrosati, B. Advanced, lithium batteries based on
high-performance composite polymer electrolytes. J. Power Sour.
162, 685–689 (2006).
52. Akolkar, R. Modeling dendrite growth during lithium electrodeposition
at sub-ambient temperature. J. Power Sour. 246, 84–89 (2014).
53. Winter, M. & Besenhard, J. O. Electrochemical lithiation of tin and tin-based
intermetallics and composites. Electrochimica Acta
45, 31–50 (1999).
54. Jansen, A. N., Clevenger, J. A., Baebler, A. M. & Vaughey, J. T. Variable
temperature performance of intermetallic lithium-ion battery anode materials.
J. Alloys Compounds 509, 4457–4461 (2011).

12 NATURE ENERGY 2, 17108 (2017) | DOI: 10.1038/nenergy.2017.108 | www.nature.com/natureenergy

©
2
0
1
7
M
a
c
im
lla
n
P
u
b
ils
h
re
s
L
im
it
e
d
,
p
a
rt
o
fp
S
ri
n
g
re
N
a
t
u
r
e
.
A
ll
ri
g
h
ts
r
e
s
e
r
v
e
d
.
REVIEW ARTICLE NATURE ENERGY

55. Sen, U. K. & Mitra, S. High-rate and high-energy-density lithium-ion battery
anode containing 2D MoS2 nanowall and cellulose binder.
ACS Appl. Mater. Interfaces 5, 1240–1247 (2013).
56. Zhang, S. S., Xu, K. & Jow, T. R. A new approach toward improved low
temperature performance of Li-ion battery. Electrochem. Commun.
4, 928–932 (2002).
57. Campion, C. L., Li, W. & Lucht, B. L. Thermal decomposition of LiPF6-based
electrolytes for lithium-ion batteries. J. Electrochem. Soc.
152, A2327 (2005).
58. Wang, Y. et al. Accelerating rate calorimetry studies of the reactions
between ionic liquids and charged lithium ion battery electrode materials.
Electrochimica Acta 52, 6346–6352 (2007).
This work disputes some claims of thermal stability of ionic liquids,
showing that some compositions undergo thermal runaway at lower
temperatures than carbonate-based solvents.
59. Liao, L. et al. Fluoroethylene carbonate as electrolyte additive to improve low
temperature performance of LiFePO4 electrode. Electrochimica Acta
87, 466–472 (2013).
60. Liao, L. et al. Enhancement of low-temperature performance of LiFePO4
electrode by butyl sultone as electrolyte additive. Solid State Ionics
254, 27–31 (2014).
61. Qin, Y., Chen, Z., Liu, J. & Amine, K. Lithium tetrafluoro oxalato phosphate as
electrolyte additive for lithium-ion cells. Electrochem. Solid-State Lett.
13, A11 (2010).
62. Yamagiwa, K. et al. Improved high-temperature performance and surface
chemistry of graphite/LiMn2O4 Li-ion cells by fluorosilane-based electrolyte
additive. Electrochimica Acta 160, 347–356 (2015).
63. Chen, Z. et al. New class of nonaqueous electrolytes for long-life and safe
lithium-ion batteries. Nat. Commun. 4, 1513 (2013).
Proposes Li2B12F9H3 as a new Li-salt, exhibiting enhanced stability towards
charged electrodes at elevated temperatures and overcharge protection.
64. Yaakov, D., Gofer, Y., Aurbach, D. & Halalay, I. C. On the study of electrolyte
solutions for Li-ion batteries that can work over a wide temperature range.
J. Electrochem. Soc. 157, A1383 (2010).
65. Xu, K. Tailoring electrolyte composition for LiBOB. J. Electrochem. Soc.
155, A733 (2008).
66. Smart, M. C., Ratnakumar, B. V. & Surampudi, S. Use of organic esters
as cosolvents in electrolytes for lithium-ion batteries with improved low
temperature performance. J. Electrochem. Soc. 149, A361 (2002).
A comprehensive analysis of the implementation of acetates as co-solvents
in electrolytes, extending battery operability to –70 °C.
67. Smart, M. C., Ratnakumar, B. V., Chin, K. B. & Whitcanack, L. D. Lithium-
ion electrolytes containing ester cosolvents for improved low temperature
performance. J. Electrochem. Soc. 157, A1361 (2010).
68. Hu, L., Zhang, Z. & Amine, K. Fluorinated electrolytes for Li-ion battery:
an FEC-based electrolyte for high voltage LiNi0.5Mn1.5O4/graphite couple.
Electrochem. Commun. 35, 76–79 (2013).
69. Zhang, Z. et al. Fluorinated electrolytes for 5 V lithium-ion battery chemistry.
Energy Environ. Sci. 6, 1806 (2013).
70. Kühnel, R.-S. & Balducci, A. Lithium ion transport and solvation in N-Butyl-
N-methylpyrrolidinium Bis(trifluoromethanesulfonyl)imide–Propylene
carbonate mixtures. J. Phys. Chem. C 118, 5742–5748 (2014).
71. Choi, B. K. & Kim, Y. W. Thermal history effects on the ionic conductivity of
PEO-salt electrolytes. Mater. Sci. Eng. B 107, 244–250 (2004).
72. Li, Y.-H. et al. A novel polymer electrolyte with improved high-temperature-
tolerance up to 170 °C for high-temperature lithium-ion batteries.
J. Power Sour. 244, 234–239 (2013).
73. Wu, X.-L. et al. Enhanced working temperature of PEO-based polymer
electrolyte via porous PTFE film as an efficient heat resister. Solid State Ionics
245–246, 1–7 (2013).
74. Bouchet, R. et al. Single-ion BAB triblock copolymers as highly efficient
electrolytes for lithium-metal batteries. Nat. Mater. 12, 452–457 (2013).
75. Wang, Y. et al. Design principles for solid-state lithium superionic conductors.
Nat. Mater. 14, 1026–1031 (2015).
76. Li, D., Ma, Z., Xu, J., Li, Y. & Xie, K. High temperature property of all-solid-
state thin film lithium battery using LiPON electrolyte. Mater. Lett.
134, 237–239 (2014).
77. Mo, S. et al. High-temperature performance of all-solid-state battery
assembled with 95(0.7Li2S-0.3P2S5)-5Li3PO4 glass electrolyte. Solid State Ionics
296, 37–41 (2016).
78. Kato, Y. et al. High-power all-solid-state batteries using sulfide superionic
conductors. Nat. Energy 1, 16030 (2016).
79. Osada, I., de Vries, H., Scrosati, B. & Passerini, S. Ionic-liquid-based polymer
electrolytes for battery applications. Angew. Chem. Int. Edn
55, 500–513 (2016).
80. Del Sesto, R. E. et al. Limited thermal stability of imidazolium and
pyrrolidinium ionic liquids. Thermochimica Acta 491, 118–120 (2009).
81. Rodrigues, M.-T. F., Lin, X., Gullapalli, H., Grinstaff, M. W. & Ajayan, P. M.
Rate limiting activity of charge transfer during lithiation from ionic liquids.
J. Power Sour. 330, 84–91 (2016).
82. Yamagata, M. et al. High-performance graphite negative electrode in a
bis(fluorosulfonyl)imide-based ionic liquid. J. Power Sour.
227, 60–64 (2013).
83. Zhou, Q. et al. Phase behavior of ionic liquid–LiX mixtures: pyrrolidinium
cations and TFSI–anions — linking structure to transport properties.
Chem. Mater. 23, 4331–4337 (2011).
84. Tsai, W.-Y. et al. Outstanding performance of activated graphene based
supercapacitors in ionic liquid electrolyte from −50 to 80 °C. Nano Energy
2, 403–411 (2013).
85. Lee, J. H. et al. Hybrid ionogel electrolytes for high temperature lithium
batteries. J. Mater. Chem. A 3, 2226–2233 (2015).
86. Jeong, H.-S., Hong, S. C. & Lee, S.-Y. Effect of microporous structure on
thermal shrinkage and electrochemical performance of Al2O3/poly(vinylidene
fluoride-hexafluoropropylene) composite separators for lithium-ion batteries.
J. Membrane Sci. 364, 177–182 (2010).
87. Jiang, W. et al. A high temperature operating nanofibrous polyimide separator
in Li-ion battery. Solid State Ionics 232, 44–48 (2013).
88. Lee, J., Lee, C.-L., Park, K. & Kim, I.-D. Synthesis of an Al2O3-coated
polyimide nanofiber mat and its electrochemical characteristics as a separator
for lithium ion batteries. J. Power Sour. 248, 1211–1217 (2014).
89. Yan, X. et al. Polyimide binder by combining with polyimide separator
for enhancing the electrochemical performance of lithium ion batteries.
Electrochimica Acta 216, 1–7 (2016).
90. Zhao, M. et al. Electrochemical stability of copper in lithium-ion battery
electrolytes. J. Electrochem. Soc. 147, 2874 (2000).
91. Morita, M., Shibata, T., Yoshimoto, N. & Ishikawa, M. Anodic behavior
of aluminum current collector in LiTFSI solutions with different solvent
compositions. J. Power Sour. 119–121, 784–788 (2003).
92. Chen, X. et al. Mixed salts of LiTFSI and LiBOB for stable LiFePO4-based
batteries at elevated temperatures. J. Mater. Chem. A 2, 2346 (2014).
93. Mun, J. et al. Linear-sweep thermammetry study on corrosion behavior of Al
current collector in ionic liquid solvent. Electrochem. Solid-State Lett.
13, A109 (2010).
94. Li, J., Zhu, Y., Wang, L. & Cao, C. Lithium titanate epitaxial coating on spinel
lithium manganese oxide surface for improving the performance of lithium
storage capability. ACS Appl. Mater. Interfaces 6, 18742–18750 (2014).
95. Zhang, J. et al. Sustainable, heat-resistant and flame-retardant cellulose-based
composite separator for high-performance lithium ion battery. Sci. Rep.
4, 3935 (2014).
96. Zheng, F. et al. Surfactants assisted synthesis and electrochemical properties of
nano-LiFePO4/C cathode materials for low temperature applications.
J. Power Sour. 288, 337–344 (2015).
97. Liao, L. et al. Effects of temperature on charge/discharge behaviors of LiFePO4
cathode for Li-ion batteries. Electrochimica Acta
60, 269–273 (2012).
98. Shui Zhang, S. An unique lithium salt for the improved electrolyte of Li-ion
battery. Electrochem. Commun. 8, 1423–1428 (2006).
99. Zhou, Z.-B., Takeda, M., Fujii, T. & Ue, M. Li [C2F5BF3] as an electrolyte salt
for 4 V class lithium-ion cells. J. Electrochem. Soc.
152, A351–A356 (2005).
100. Smart, M. C. et al. Improved performance of lithium-ion cells with the use of
fluorinated carbonate-based electrolytes. J. Power Sour.
119–121, 359–367 (2003).
101. Zhu, G. et al. Materials insights into low-temperature performances of
lithium-ion batteries. J. Power Sour. 300, 29–40 (2015).
102. Ryou, M.-H. et al. Effect of fluoroethylene carbonate on high temperature
capacity retention of LiMn2O4/graphite Li-ion cells. Electrochimica Acta
55, 2073–2077 (2010).
103. Kang, K. S. et al. Effect of additives on electrochemical performance of lithium
nickel cobalt manganese oxide at high temperature. J. Power Sour.
253, 48–54 (2014).
104. Liu, Y., Tan, L. & Li, L. Tris(trimethylsilyl) borate as an electrolyte additive to
improve the cyclability of LiMn2O4 cathode for lithium-ion battery.
J. Power Sour. 221, 90–96 (2013).
105. Rong, H. et al. A novel imidazole-based electrolyte additive for improved
electrochemical performance at elevated temperature of high-voltage
LiNi0.5Mn1.5O4 cathodes. J. Power Sour. 329, 586–593 (2016).
106. Pham, H. Q., Hwang, E.-H., Kwon, Y.-G. & Song, S.-W. Understanding the
interfacial phenomena of a 4.7 V and 55 °C Li-ion battery with Li-rich layered
oxide cathode and graphite anode and its correlation to high-energy cycling
performance. J. Power Sour. 323, 220–230 (2016).
107. Aurbach, D. et al. On the use of vinylene carbonate (VC) as an additive to
electrolyte solutions for Li-ion batteries. Electrochimica Acta
47, 1423–1439 (2002).

NATURE ENERGY 2, 17108 (2017) | DOI: 10.1038/nenergy.2017.108 | www.nature.com/natureenergy 13

©
2
0
1
7
M
a
c
im
lla
n
P
u
b
ils
h
re
s
L
im
it
e
d
,
p
a
rt
o
fp
S
ri
n
g
re
N
a
t
u
r
e
.
A
ll
ri
g
h
ts
r
e
s
e
r
v
e
d
.
NATURE ENERGY
108. Han, H. et al. Lithium (fluorosulfonyl)(nonafluorobutanesulfonyl)imide
(LiFNFSI) as conducting salt to improve the high-temperature resilience of
lithium-ion cells. Electrochem. Commun. 13, 265–268 (2011).
109. Jin, Z. et al. A new class of phosphates as co-solvents for nonflammable
lithium ion batteries blectrolytes. ECS Electrochem. Lett.
1, A55–A58 (2012).
110. MacNeil, D., Lu, Z., Chen, Z. & Dahn, J. R. A comparison of the electrode/
electrolyte reaction at elevated temperatures for various Li-ion battery
cathodes. J. Power Sour. 108, 8–14 (2002).
111. Plichta, E. J. & Behl, W. K. A low-temperature electrolyte for lithium and
lithium-ion batteries. J. Power Sour. 88, 192–196 (2000).
112. Yabuuchi, N. & Ohzuku, T. Electrochemical behaviors of
LiCo1/3Ni1/3Mn1/3O2 in lithium batteries at elevated temperatures.
J. Power Sour. 146, 636–639 (2005).
113. Chen, K. et al. Evaluation of the low temperature performance of lithium
manganese oxide/lithium titanate lithium-ion batteries for start/stop
applications. J. Power Sour. 278, 411–419 (2015).
114. Luo, Y. et al. Hierarchical carbon decorated Li3V2(PO4)3 as a bicontinuous
cathode with high-rate capability and broad temperature adaptability.
Adv. Energy Mater. 4, 1400107 (2017).
REVIEW ARTICLE
115. Song, M.-S. et al. Is Li4Ti5O12 a solid-electrolyte-interphase-free electrode
material in Li-ion batteries? Reactivity between the Li4Ti5O12 electrode and
electrolyte. J. Mater. Chem. A 2, 631–636 (2014).
116. Cui, Y., Rohde, M., Mahmoud, M. M., Ziebert, C. & Seifert, H. J.
Phosphate based ceramics as solid electrolyte for high temperature lithium
ion batteries. ECS Meeting Abstracts 120, ma2016-01 (2016).
Additional information
Reprints and permissions information is available at www.nature.com/reprints.
Correspondence should be addressed to P. M. A.
How to cite this article: Rodrigues, M.-T. F. et al. A materials perspective on Li-ion
batteries at extreme temperatures. Nat. Energy 2, 17108 (2017).
Publisher’s note: Springer Nature remains neutral with regard to jurisdictional claims in
published maps and institutional affiliations.
Competing interests
The authors declare no competing financial interests.

14 NATURE ENERGY 2, 17108 (2017) | DOI: 10.1038/nenergy.2017.108 | www.nature.com/natureenergy

©
2
0
1
7
M
a
c
im
lla
n
P
u
b
ils
h
re
s
L
im
it
e
d
,
p
a
rt
o
fp
S
ri
n
g
re
N
a
t
u
r
e
.
A
ll
ri
g
h
ts
r
e
s
e
r
v
e
d
.